Euler

ELEMENTARY
DIFFERENTIAL
EQUATIONS WITH
BOUNDARY VALUES
PROBLEMS
William F. Trench
Trinity University
Trinity University
Elementary Differential Equations with
Boundary Values Problems
William F. Trench
This open text is disseminated via the Open Education Resource (OER) LibreTexts Project (https://LibreTexts.org) and like the
hundreds of other open texts available within this powerful platform, it is licensed to be freely used, adapted, and distributed.
This book is openly licensed which allows you to make changes, save, and print this book as long; the applicable license is
indicated at the bottom of each page.
Instructors can adopt existing LibreTexts texts or Remix them to quickly build course-specific resources to meet the needs of
their students. Unlike traditional textbooks, LibreTexts’ web based origins allow powerful integration of advanced features and
new technologies to support learning.
The LibreTexts mission is to unite students, faculty and scholars in a cooperative effort to develop an easy-to-use online
platform for the construction, customization, and dissemination of OER content to reduce the burdens of unreasonable
textbook costs to our students and society. The LibreTexts project is a multi-institutional collaborative venture to develop the
next generation of open-access texts to improve postsecondary education at all levels of higher learning by developing an
Open Access Resource environment. The project currently consists of 13 independently operating and interconnected libraries
that are constantly being optimized by students, faculty, and outside experts to supplant conventional paper-based books.
These free textbook alternatives are organized within a central environment that is both vertically (from advance to basic level)
and horizontally (across different fields) integrated.
The LibreTexts libraries are Powered by MindTouch® and are supported by the Department of Education Open Textbook Pilot
Project, the UC Davis Office of the Provost, the UC Davis Library, the California State University Affordable Learning
Solutions Program, and Merlot. This material is based upon work supported by the National Science Foundation under Grant
No. 1246120, 1525057, and 1413739. Unless otherwise noted, LibreTexts content is licensed by CC BY-NC-SA 3.0.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not
necessarily reflect the views of the National Science Foundation nor the US Department of Education.
Have questions or comments? For information about adoptions or adaptions contact info@LibreTexts.org. More information
on our activities can be found via Facebook (https://facebook.com/Libretexts), Twitter (https://twitter.com/libretexts), or our
blog (http://Blog.Libretexts.org).
This text was compiled on 09/16/2020

TABLE OF CONTENTS
This text is written for students in science, engineering, and mathematics who have completed calculus through partial differentiation.
This book is intended to serve as the book for students in MAT 266 Differential Equations with Linear Algebra at Community College
of Denver.
1: INTRODUCTION
1.1: APPLICATIONS LEADING TO DIFFERENTIAL EQUATIONS
1.2: BASIC CONCEPTS
1.2E: BASIC CONCEPTS (EXERCISES)

1.3: DIRECTION FIELDS FOR FIRST ORDER EQUATIONS
1.3E: DIRECTION FIELDS FOR FIRST ORDER EQUATIONS (EXERCISES)
2: FIRST ORDER EQUATIONS

In this chapter, we study first-order differential equations for which there are general methods of solution.
2.1: LINEAR FIRST ORDER EQUATIONS

2.1E: LINEAR FIRST ORDER EQUATIONS (EXERCISES)
2.2: SEPARABLE EQUATIONS
2.2E: SEPARABLE EQUATIONS (EXERCISES)
2.3: EXISTENCE AND UNIQUENESS OF SOLUTIONS OF NONLINEAR EQUATIONS
2.3E: EXISTENCE AND UNIQUENESS OF SOLUTIONS OF NONLINEAR EQUATIONS (EXERCISES)

2.4: TRANSFORMATION OF NONLINEAR EQUATIONS INTO SEPARABLE EQUATIONS
2.4E: TRANSFORMATION OF NONLINEAR EQUATIONS INTO SEPARABLE EQUATIONS (EXERCISES)

2.5: EXACT EQUATIONS
2.5E: EXACT EQUATIONS (EXERCISES)

2.6: INTEGRATING FACTORS
2.6E: INTEGRATING FACTORS (EXERCISES)
3: NUMERICAL METHODS
In this chapter we study numerical methods for solving a first-order differential equation
3.1: EULER'S METHOD

3.1E: EULER’S METHOD (EXERCISES)
3.2: THE IMPROVED EULER METHOD AND RELATED METHODS
3.2E: THE IMPROVED EULER METHOD AND RELATED METHODS (EXERCISES)
3.3: THE RUNGE-KUTTA METHOD
3.3E: THE RUNGE-KUTTA METHOD (EXERCISES)
4: APPLICATIONS OF FIRST ORDER EQUATIONS

In this chapter, we consider applications of first order differential equations.
4.1: GROWTH AND DECAY

4.1E: GROWTH AND DECAY (EXERCISES)
4.2: COOLING AND MIXING
4.2E: COOLING AND MIXING (EXERCISES)

4.3: ELEMENTARY MECHANICS
1 9/16/2020
4.3E: ELEMENTARY MECHANICS (EXERCISES)
4.4: AUTONOMOUS SECOND ORDER EQUATIONS
4.4E: AUTONOMOUS SECOND ORDER EQUATIONS (EXERCISES)

4.5: APPLICATIONS TO CURVES
4.5E: APPLICATIONS TO CURVES (EXERCISES)
5: LINEAR SECOND ORDER EQUATIONS

In this Chapter, we study a particularly important class of second order equations. Because of their many applications in science and
engineering, second order differential equation have historically been the most thoroughly studied class of differential equations.
Research on the theory of second order differential equations continues to the present day.
5.1: HOMOGENEOUS LINEAR EQUATIONS

5.1E: HOMOGENEOUS LINEAR EQUATIONS (EXERCISES)
5.2: CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS
5.2E: CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS (EXERCISES)

5.3: NONHOMOGENEOUS LINEAR EQUATIONS
5.3E: NONHOMGENEOUS LINEAR EQUATIONS (EXERCISES)
5.4: THE METHOD OF UNDETERMINED COEFFICIENTS I
5.4E: THE METHOD OF UNDETERMINED COEFFICIENTS I (EXERCISES)
5.5: THE METHOD OF UNDETERMINED COEFFICIENTS II
5.5E: THE METHOD OF UNDETERMINED COEFFICIENTS II (EXERCISES)
5.6: REDUCTION OF ORDER
5.6E: REDUCTION OF ORDER (EXERCISES)
5.7: VARIATION OF PARAMETERS
5.7E: VARIATION OF PARAMETERS (EXERCISES)
6: APPLICATIONS OF LINEAR SECOND ORDER EQUATIONS

In this Chapter, we study applications of linear second order equations.
6.1: SPRING PROBLEMS I

6.1E: SPRING PROBLEMS I (EXERCISES)
6.2: SPRING PROBLEMS II
6.2E: SPRING PROBLEMS II (EXERCISES)
7: SERIES SOLUTIONS OF LINEAR SECOND ORDER EQUATIONS

In this Chapter, we study a class of second order differential equations that occur in many applications, but cannot be solved in closed
form in terms of elementary functions.
7.1: PRELUDE TO SERIES SOLUTIONS OF LINEAR SECOND ORDER EQUATIONS

7.2: REVIEW OF POWER SERIES
7.2E: REVIEW OF POWER SERIES (EXERCISES)
7.3: SERIES SOLUTIONS NEAR AN ORDINARY POINT I
7.3E: SERIES SOLUTIONS NEAR AN ORDINARY POINT I (EXERCISES)
7.4: SERIES SOLUTIONS NEAR AN ORDINARY POINT II
7.4E: SERIES SOLUTIONS NEAR AN ORDINARY POINT II (EXERCISES)
8: LAPLACE TRANSFORMS
8.1: INTRODUCTION TO THE LAPLACE TRANSFORM
8.1E: INTRODUCTION TO THE LAPLACE TRANSFORM (EXERCISES)
2 9/16/2020
8.2: THE INVERSE LAPLACE TRANSFORM
8.2E: THE INVERSE LAPLACE TRANSFORM (EXERCISES)
8.3: SOLUTION OF INITIAL VALUE PROBLEMS
8.3E: SOLUTION OF INITIAL VALUE PROBLEMS (EXERCISES)

8.4: A BRIEF TABLE OF LAPLACE TRANSFORMS
9: LINEAR HIGHER ORDER DIFFERENTIAL EQUATIONS

9.1: INTRODUCTION TO LINEAR HIGHER ORDER EQUATIONS
9.1E: INTRODUCTION TO LINEAR HIGHER ORDER EQUATIONS (EXERCISES)
9.2: HIGHER ORDER CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS
9.2E: HIGHER ORDER CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS (EXERCISES)
9.3: UNDETERMINED COEFFICIENTS FOR HIGHER ORDER EQUATIONS
9.3E: UNDETERMINED COEFFICIENTS FOR HIGHER ORDER EQUATIONS (EXERCISES)

9.4: VARIATION OF PARAMETERS FOR HIGHER ORDER EQUATIONS
9.4E: VARIATION OF PARAMETERS FOR HIGHER ORDER EQUATIONS (EXERCISES)
10: LINEAR SYSTEMS OF DIFFERENTIAL EQUATIONS

10.1: INTRODUCTION TO SYSTEMS OF DIFFERENTIAL EQUATIONS
10.1E: INTRODUCTION TO SYSTEMS OF DIFFERENTIAL EQUATIONS (EXERCISES)
10.2: LINEAR SYSTEMS OF DIFFERENTIAL EQUATIONS
10.2E: LINEAR SYSTEMS OF DIFFERENTIAL EQUATIONS (EXERCISES)
10.3: BASIC THEORY OF HOMOGENEOUS LINEAR SYSTEMS
10.3E: BASIC THEORY OF HOMOGENEOUS LINEAR SYSTEMS (EXERCISES)
10.4: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS I
10.4E: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS I (EXERCISES)

10.5: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS II
10.5E: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS II (EXERCISES)
10.6: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS III
10.6E: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS III (EXERCISES)
11: SYSTEMS OF EQUATIONS

11.1: GEOMETRY
11.2: ELEMENTARY OPERATIONS
11.3: GAUSSIAN ELIMINATION
11.4: EXERCISES
12: MATRICES
12.1: MATRIX ARITHMETIC
12.2: MULTIPLICATION OF MATRICES
12.3: THE IJTH ENTRY OF A PRODUCT
12.4: PROPERTIES OF MATRIX MULTIPLICATION
12.5: THE TRANSPOSE
12.6: THE IDENTITY AND INVERSES
12.7: FINDING THE INVERSE OF A MATRIX
12.E: EXERCISES
13: DETERMINANTS
13.1: BASIC TECHNIQUES
13.2: PROPERTIES OF DETERMINANTS
13.3: FINDING DETERMINANTS USING ROW OPERATIONS
3 9/16/2020
13.4: APPLICATIONS OF THE DETERMINANT
13.E: EXERCISES
14: VECTOR SPACES

14.1: ALGEBRAIC CONSIDERATIONS
14.2: SPANNING SETS
14.3: LINEAR INDEPENDENCE
14.4: SUBSPACES AND BASIS
14.5: SUMS AND INTERSECTIONS
14.6: LINEAR TRANSFORMATIONS
APPENDICES AND ANSWERS TO EXERCISES

A BRIEF TABLE OF INTEGRALS
INDEX
SECTION 1.2 ANSWERS
SECTION 2.1 ANSWERS
SECTION 2.2 ANSWERS
SECTION 2.3 ANSWERS
SECTION 2.4 ANSWERS
SECTION 2.5 ANSWERS
SECTION 2.6 ANSWERS
SECTION 3.1 ANSWERS
SECTION 3.2 ANSWERS
SECTION 3.3 ANSWERS
SECTION 4.1 ANSWERS
SECTION 4.2 ANSWERS
SECTION 4.3 ANSWERS
SECTION 4.4 ANSWERS
SECTION 4.5 ANSWERS
SECTION 5.1 ANSWERS
SECTION 5.2 ANSWERS
SECTION 5.3 ANSWERS
SECTION 5.4 ANSWERS
SECTION 5.5 ANSWERS
SECTION 5.6 ANSWERS
SECTION 5.7 ANSWERS
SECTION 6.1 ANSWERS
SECTION 6.2 ANSWERS
SECTION 6.3 ANSWERS
SECTION 6.4 ANSWERS
SECTION 7.1 ANSWERS
SECTION 7.2 ANSWERS
SECTION 7.3 ANSWERS
SECTION 7.4 ANSWERS
SECTION 7.5 ANSWERS
SECTION 7.6 ANSWERS
SECTION 7.7 ANSWERS
SECTION 8.1 ANSWERS
SECTION 8.2 ANSWERS
SECTION 8.3 ANSWERS
SECTION 8.4 ANSWERS
SECTION 8.5 ANSWERS
SECTION 8.6 ANSWERS
SECTION 8.7 ANSWERS
SECTION 9.1 ANSWERS
SECTION 9.2 ANSWERS
SECTION 9.3 ANSWERS
SECTION 9.4 ANSWERS
SECTION 10.1 ANSWERS
4 9/16/2020
BACK MATTER
INDEX
5 9/16/2020
CHAPTER OVERVIEW
1: INTRODUCTION
In this chapter, we begin our study of differential equations.
1.1: APPLICATIONS LEADING TO DIFFERENTIAL EQUATIONS

This section presents examples of applications that lead to differential equations.
1.2: BASIC CONCEPTS

A differential equation is an equation that contains one or more derivatives of an unknown
function. The order of a differential equation is the order of the highest derivative that it contains.
A differential equation is an ordinary differential equation if it involves an unknown function of
only one variable, or a partial differential equation if it involves partial derivatives of a function of
more than one variable. This section introduces basic concepts and definitions.
1.2E: BASIC CONCEPTS (EXERCISES)

1.3: DIRECTION FIELDS FOR FIRST ORDER EQUATIONS
It's impossible to find explicit formulas for solutions of some differential equations. Even if there are such formulas, they may be so
complicated that they’re useless. In this case we may resort to graphical or numerical methods to get some idea of how the solutions
of the given equation behave. This section presents a geometric method for dealing with differential equations.
1.3E: DIRECTION FIELDS FOR FIRST ORDER EQUATIONS (EXERCISES)
1 9/16/2020
1.1: Applications Leading to Differential Equations
In order to apply mathematical methods to a physical or “real life” problem, we must formulate the problem in mathematical
terms; that is, we must construct a mathematical model for the problem. Many physical problems concern relationships
between changing quantities. Since rates of change are represented mathematically by derivatives, mathematical models often
involve equations relating an unknown function and one or more of its derivatives. Such equations are differential equations.
They are the subject of this book.
Much of calculus is devoted to learning mathematical techniques that are applied in later courses in mathematics and the
sciences; you wouldn’t have time to learn much calculus if you insisted on seeing a specic application of every topic covered
in the course. Similarly, much of this book is devoted to methods that can be applied in later courses. Only a relatively small
part of the book is devoted to the derivation of specific differential equations from mathematical models, or relating the
differential equations that we study to specic applications. In this section we mention a few such applications. The
mathematical model for an applied problem is almost always simpler than the actual situation being studied, since simplifying
assumptions are usually required to obtain a mathematical problem that can be solved. For example, in modeling the motion of
a falling object, we might neglect air resistance and the gravitational pull of celestial bodies other than Earth, or in modeling
population growth we might assume that the population grows continuously rather than in discrete steps.
A good mathematical model has two important properties:
It’s sufficiently simple so that the mathematical problem can be solved.
It represents the actual situation sufficiently well so that the solution to the mathematical problem predicts the outcome of
the real problem to within a useful degree of accuracy. If results predicted by the model don’t agree with physical
observations,the underlying assumptions of the model must be revised until satisfactory agreement is obtained.
We will now give examples of mathematical models involving differential equations. We willreturn to these problems at the
appropriate times, as we learn how to solve the various types of differential equations that occur in the models. All the
examples in this section deal with functions of time, which we denote by t . If y is a function of t , y denotes the derivative of
′
y with respect to t ; thus,
dy
′
y = .
dt
Population Growth and Decay

Although the number of members of a population (people in a given country, bacteria in a laboratory culture, wildowers in a
forest, etc.) at any given time t is necessarily an integer, models that use differential equations to describe the growth and
decay of populations usually rest on the simplifying assumption that the number of members of the population can be regarded
as a differentiable function P = P (t) . In most models it is assumed that the differential equation takes the form
′
P = a(P )P (1.1.1)
where a is a continuous function of P that represents the rate of change of population per unit time per individual. In the
Malthusian model, it is assumed that a(P ) is a constant, so Equation 1.1.1 becomes
′
P = aP . (1.1.2)
(When you see a name in blue italics, just click on it for information about the person.) This model assumes that the numbers
of births and deaths per unit time are both proportional to the population. The constants of proportionality are the birth rate
(births per unit time per individual) and the death rate (deaths per unit time per individual); a is the birth rate minus the death
rate. You learned in calculus that if c is any constant then
at
P = ce (1.1.3)
satises Equation 1.1.2, so Equation 1.1.2 has innitely many solutions. To select the solution of the specic problem that we
are considering, we must know the population P at an initial time, say t = 0 . Setting t = 0 in Equation 1.1.3 yields
0
c = P (0) = P , so the applicable solution is

0
William F. Trench 9/11/2020 1.1.1 CC-BY-NC-SA https://math.libretexts.org/@go/page/44215

at
P (t) = P0 e .
This implies that

∞ if a > 0,
lim P (t) = {
t→∞ 0 if a < 0;
that is, the population approaches infinity if the birth rate exceeds the death rate, or zero if the death rate exceeds the birth rate.
To see the limitations of the Malthusian model, suppose we are modeling the population of a country, starting from a time
t = 0 when the birth rate exceeds the death rate (so a > 0 ), and the country’s resources in terms of space, food supply, and
other necessities of life can support the existing population. Then the prediction P = P e may be reasonably accurate as
0
at
long as it remains within limits that the country’s resources can support. However, the model must inevitably lose validity
when the prediction exceeds these limits. (If nothing else, eventually there will not be enough space for the predicted
population!) This aw in the Malthusian model suggests the need for a model that accounts for limitations of space and
resources that tend to oppose the rate of population growth as the population increases.
Perhaps the most famous model of this kind is the Verhulst model, where Equation 1.1.2 is replaced by
′
P = aP (1 − αP ), (1.1.4)
where α is a positive constant. As long as P is small compared to 1/α, the ratio P /P is approximately equal to a . Therefore
′
the growth is approximately exponential; however, as P increases, the ratio P /P decreases as opposing factors become
′
significant.
Equation 1.1.4 is the logistic equation. You will learn how to solve it in Section 1.2. (See Exercise 2.2.28.) The solution is
P0
P = ,
−at
α P0 + (1 − α P0 )e
where P 0 = P (0) > 0 . Therefore lim P (t) = 1/α , independent of P . 0

t→∞
Figure 1.1.1 shows typical graphs of P versus t for various values of P . 0
Figure 1.1.1 : Solutions of the logistic equation
Newton’s Law of Cooling

According to Newton’s law of cooling, the temperature of a body changes at a rate proportional to the difference between the
temperature of the body and the temperature of the surrounding medium. Thus, if T is the temperature of the medium and
m
T = T (t) is the temperature of the body at time t , then
′
T = −k(T − Tm ) (1.1.5)
where k is a positive constant and the minus sign indicates; that the temperature of the body increases with time if it is less
than the temperature of the medium, or decreases if it is greater. We will see in Section 4.2 that if T is constant then the m

solution of Equation 1.1.5 is
−kt
T = Tm + (T0 − Tm )e (1.1.6)
where T is the temperature of the body when t = 0 . Therefore

0
lim T (t) = Tm
t→∞
independent of T (Common sense suggests this. Why?).

0
Figure 1.1.2 shows typical graphs of T versus t for various values of T . 0
Figure 1.1.2 : Temperature according to Newton’s Law of Cooling

Assuming that the medium remains at constant temperature seems reasonable if we are considering a cup of coffee cooling in a
room, but not if we are cooling a huge cauldron of molten metal in the same room. The difference between the two situations
is that the heat lost by the coffee isn’t likely to raise the temperature of the room appreciably, but the heat lost by the cooling
metal is. In this second situation we must use a model that accounts for the heat exchanged between the object and the
medium. Let T = T (t) and T = T (t) be the temperatures of the object and the medium respectively, and let T and T 0
m m 0 m
be their initial values. Again, we assume that T and Tm are related by Equation 1.1.5. We also assume that the change in heat
of the object as its temperature changes from T to T is a(T − T ) and the change in heat of the medium as its temperature
0 0
changes from T to T is a (T − T ) , where a and am are positive constants depending upon the masses and thermal
m0 m m m m0
properties of the object and medium respectively. If we assume that the total heat of the in the object and the medium remains
constant (that is, energy is conserved), then
a(T − T0 ) + am (Tm − Tm0 ) = 0.
Solving this for T and substituting the result into Equation 1.1.6 yields the differential equation
m
a a
′
T = −k (1 + ) T + k ( Tm0 + T0 )
am am
for the temperature of the object. After learning to solve linear rst order equations, you’ll be able to show (Exercise 4.2.17)
that
aT0 + am Tm0 am (T0 − Tm0 )
−k(1+a/ am )t
T = + e
a + am a + am
Glucose Absorption by the Body

Glucose is absorbed by the body at a rate proportional to the amount of glucose present in the blood stream. Let λ denote the
(positive) constant of proportionality. Suppose there are G units of glucose in the bloodstream when t = 0 , and let G = G(t)
0
be the number of units in the bloodstream at time t > 0 . Then, since the glucose being absorbed by the body is leaving the
bloodstream, G satisfies the equation

′
G = −λG. (1.1.7)
From calculus you know that if c is any constant then

−λt
G = ce (1.1.8)
satises Equation (1.1.7), so Equation 1.1.7 has innitely many solutions. Setting t = 0 in Equation 1.1.8 and requiring that
G(0) = G 0yields c = G , so
0
−λt
G(t) = G0 e .
Now let’s complicate matters by injecting glucose intravenously at a constant rate of r units of glucose per unit of time. Then
the rate of change of the amount of glucose in the bloodstream per unit time is
′
G = −λG + r (1.1.9)
where the rst term on the right is due to the absorption of the glucose by the body and the second term is due to the injection.
After you’ve studied Section 2.1, you’ll be able to show that the solution of Equation 1.1.9 that satisfies G(0) = G is 0
r r −λt
G= + (G0 − )e
λ λ
Graphs of this function are similar to those in Figure 1.1.2. (Why?)
Spread of Epidemics
One model for the spread of epidemics assumes that the number of people infected changes at a rate proportional to the
product of the number of people already infected and the number of people who are susceptible, but not yet infected.
Therefore, if S denotes the total population of susceptible people and I = I (t) denotes the number of infected people at time
t , then S − I is the number of people who are susceptible, but not yet infected. Thus,
′
I = rI (S − I )
, where r is a positive constant. Assuming that I (0) = I , the solution of this equation is
0
SI0
I =
−rSt
I0 + (S − I0 )e
(Exercise 2.2.29). Graphs of this function are similar to those in Figure 1.1.1. (Why?) Since lim I (t) = S , this model predicts
t→∞
that all the susceptible people eventually become infected.
Newton’s Second Law of Motion

According to Newton’s second law of motion, the instantaneous acceleration a of an object with constant mass m is related to
the force F acting on the object by the equation F = ma . For simplicity, let’s assume that m = 1 and the motion of the object
is along a vertical line. Let y be the displacement of the object from some reference point on Earth’s surface, measured
positive upward. In many applications, there are three kinds of forces that may act on the object:
a. A force such as gravity that depends only on the position y, which we write as −p(y), where p(y) > 0 if y ≥ 0 .
b. A force such as atmospheric resistance that depends on the position and velocity of the object, which we write as
−q(y, y )y , where q is a nonnegative function and we’ve put y “outside” to indicate that the resistive force is always in
′ ′ ′
the direction opposite to the velocity.

c. A force f = f (t) , exerted from an external source (such as a towline from a helicopter) that depends only on t .
In this case, Newton’s second law implies that
′′ ′ ′
y = −q(y, y )y − p(y) + f (t),
which is usually rewritten as

′′ ′ ′
y + q(y, y )y + p(y) = f (t).

Since the second (and no higher) order derivative of y occurs in this equation, we say that it is a second order differential
equation.
Interacting Species: Competition

Let P = P (t) and Q = Q(t) be the populations of two species at time t , and assume that each population would grow
exponentially if the other did not exist; that is, in the absence of competition we would have
′ ′
P = aP and Q = bQ, (1.1.10)
where a and b are positive constants. One way to model the effect of competition is to assume that the growth rate per
individual of each population is reduced by an amount proportional to the other population, so Equation 1.1.10 is replaced by
′
P = aP − αQ
′
Q = −βP + bQ,
where α and β are positive constants. (Since negative population doesn’t make sense, this system works only while P and Q
are both positive.) Now suppose P (0) = P > 0 and Q(0) = Q > 0 . It can be shown (Exercise 10.4.42) that there’s a
0 0
positive constant ρ such that if (P , Q ) is above the line L through the origin with slope ρ, then the species with population
0 0
P becomes extinct in finite time, but if (P , Q ) is below L, the species with population Q becomes extinct in finite time.
0 0
Figure 1.1.3 illustrates this. The curves shown there are given parametrically by P = P (t), Q = Q(t), t > 0 . The arrows
indicate direction along the curves with increasing t .
Figure 1.1.3 : Populations of competing species

1.2: Basic Concepts
A differential equation is an equation that contains one or more derivatives of an unknown function. The order of a differential
equation is the order of the highest derivative that it contains. A differential equation is an ordinary differential equation if it
involves an unknown function of only one variable, or a partial differential equation if it involves partial derivatives of a
function of more than one variable. For now we’ll consider only ordinary differential equations, and we’ll just call them
differential equations.
Throughout this text, all variables and constants are real unless it is stated otherwise. We’ll usually use x for the
independent variable unless the independent variable is time; then we’ll use t .
The simplest differential equations are first order equations of the form
dy
= f (x)
dx
or equivalently
′
y = f (x),
where f is a known function of x. We already know from calculus how to find functions that satisfy this kind of equation. For
example, if
′ 3
y =x ,
then
4
x
3
y =∫ x dx = + c,
4
where c is an arbitrary constant. If n > 1 we can find functions y that satisfy equations of the form
(n)
y = f (x) (1.2.1)
by repeated integration. Again, this is a calculus problem.

Except for illustrative purposes in this section, there’s no need to consider differential equations like Equation 1.2.1 . We’ll
usually consider differential equations that can be written as
(n) ′ (n−1)
y = f (x, y, y , … , y ), (1.2.2)
where at least one of the functions y , y , …, y

′ (n−1)
actually appears on the right. Here are some examples:
dy
2
−x = 0 (first order),
dx
dy
2
+ 2x y = −2 (first order),
dx
2
d y dy
2
+2 +y = 2x (second order),
dx dx
′′′ 2
xy +y = sin x (third order),
(n) ′
y + x y + 3y = x (n-th order).
Although none of these equations is written as in Equation 1.2.2, all of them can be written in this form:

′ 2
y = x ,
′ 2
y = −2 − 2x y ,
′′ ′
y = 2x − 2 y − y,
2
sin x − y
′′′
y = ,
x
(n) ′
y = x − x y − 3y.
Solutions of Differential Equations

A solution of a differential equation is a function that satisfies the differential equation on some open interval; thus, y is a
solution of Equation 1.2.2 if y is n times differentiable and
(n) ′ (n−1)
y (x) = f (x, y(x), y (x), … , y (x))
for all x in some open interval (a, b). In this case, we also say that y is a solution of Equation 1.2.2 on (a, b) . Functions that
satisfy a differential equation at isolated points are not interesting. For example, y = x satisfies 2
′ 2
xy + x = 3x
if and only if x = 0 or x = 1 , but it is not a solution of this differential equation because it does not satisfy the equation on an
open interval.
The graph of a solution of a differential equation is a solution curve. More generally, a curve C is said to be an integral curve
of a differential equation if every function y = y(x) whose graph is a segment of C is a solution of the differential equation.
Thus, any solution curve of a differential equation is an integral curve, but an integral curve need not be a solution curve.
Example 1.2.1
If a is any positive constant, the circle
2 2 2
x +y =a (1.2.3)
is an integral curve of
x
′
y =− . (1.2.4)
y
Solution
To see this, note that the only functions whose graphs are segments of Equation 1.2.3 are
− −−−−− − −−−−−
2 2 2 2
y1 = √ a − x and y2 = −√ a − x .
We leave it to you to verify that these functions both satisfy Equation 1.2.4 on the open interval (−a, a) . However,
Equation 1.2.3 is not a solution curve of Equation 1.2.4, since it is not the graph of a function.
Example 1.2.2
Verify that
2
x 1
y = + (1.2.5)
3 x
is a solution of
′ 2
xy + y = x (1.2.6)
on (0, ∞) and on (−∞, 0).

Solution
Substituting Equation 1.2.5 and

′
2x 1
y = −
3 x2
into Equation 1.2.6 yields

2
2x 1 x 1
′ 2
x y (x) + y(x) = x ( − ) +( + ) =x
2
3 x 3 x
for all x ≠ 0 . Therefore y is a solution of Equation 1.2.6 on (−∞, 0) and (0, ∞). However, y isn’t a solution of the
differential equation on any open interval that contains x = 0 , since y is not defined at x = 0 .
Figure 1.2.2 shows the graph of Equation 1.2.5. The part of the graph of Equation 1.2.5 on (0, ∞) is a solution curve of
Equation 1.2.6, as is the part of the graph on (−∞, 0).
Figure 1.2.2 : y = x
3
+
1
Example 1.2.3
Show that if c and c are constants then
1 2
−x
y = (c1 + c2 x)e + 2x − 4 (1.2.7)
is a solution of
′′ ′
y + 2 y + y = 2x (1.2.8)
on (−∞, ∞).
Solution
Differentiating Equation 1.2.7 twice yields
′ −x −x
y = −(c1 + c2 x)e + c2 e +2
and
′′ −x −x
y = (c1 + c2 x)e − 2 c2 e ,
so
′′ ′ −x −x −x −x −x
y + 2y + y = (c1 + c2 x)e − 2 c2 e + 2 [−(c1 + c2 x)e + c2 e + 2] + (c1 + c2 x)e + 2x − 4
−x −x
= (1 − 2 + 1)(c1 + c2 x)e + (−2 + 2)c2 e + 4 + 2x − 4
= 2x
for all values of x. Therefore y is a solution of Equation 1.2.8 on (−∞, ∞).

Example 1.2.4
Find all solutions of
(n) 2x
y =e . (1.2.9)
Solution
Integrating Equation 1.2.9 yields
2x
e
(n−1)
y = + k1 ,
2
where k is a constant. If n ≥ 2 , integrating again yields

1
2x
(n−2)
e
y = + k1 x + k2 .
4
If n ≥ 3 , repeatedly integrating yields

2x n−1 n−2
e x x
y = + k1 + k2 + ⋯ + kn , (1.2.10)
n
2 (n − 1)! (n − 2)!
where k , k , …, k are constants. This shows that every solution of Equation 1.2.9 has the form Equation 1.2.10 for
1 2 n
some choice of the constants k , k , …, k . On the other hand, differentiating Equation 1.2.10 n times shows that if k ,
1 2 n 1
k , …, k
2 are arbitrary constants, then the function y in Equation 1.2.10 satisfies Equation 1.2.9.
n
Since the constants k , k , …, k in Equation 1.2.10 are arbitrary, so are the constants
1 2 n
k1 k2
, , ⋯ , kn .
(n − 1)! (n − 2)!
Therefore Example 1.2.4 actually shows that all solutions of Equation 1.2.9 can be written as
2x
e n−1
y = n
+ c1 + c2 x + ⋯ + cn x ,
2
where we renamed the arbitrary constants in Equation 1.2.10 to obtain a simpler formula. As a general rule, arbitrary constants
appearing in solutions of differential equations should be simplified if possible. You’ll see examples of this throughout the
text.
Initial Value Problems

In Example 1.2.4 we saw that the differential equation y = e has an infinite family of solutions that depend upon the n
(n) 2x
arbitrary constants c , c , …, c . In the absence of additional conditions, there’s no reason to prefer one solution of a
1 2 n
differential equation over another. However, we’ll often be interested in finding a solution of a differential equation that
satisfies one or more specific conditions. The next example illustrates this.
Example 1.2.5
Find a solution of
′ 3
y =x
such that y(1) = 2 .

Solution
At the beginning of this section we saw that the solutions of y ′
=x
3
are
4
x
y = + c.
4

To determine a value of c such that y(1) = 2 , we set x = 1 and y = 2 here to obtain
1
2 = y(1) = +c
4
so
7
c = .
4
Therefore the required solution is

4
x +7
y = .
4
Figure 1.2.2 shows the graph of this solution. Note that imposing the condition y(1) = 2 is equivalent to requiring the
graph of y to pass through the point (1, 2).
Figure 1.2.2 : y =
x +7
We can rewrite the problem considered in Example 1.2.5 more briefly as

′ 3
y =x , y(1) = 2.
We call this an initial value problem. The requirement y(1) = 2 is an initial condition. Initial value problems can also be
posed for higher order differential equations. For example,
′′ ′ x ′
y − 2 y + 3y = e , y(0) = 1, y (0) = 2 (1.2.11)
is an initial value problem for a second order differential equation where y and y are required to have specified values at
′
x = 0 . In general, an initial value problem for an n -th order differential equation requires y and its first n − 1 derivatives to
have specified values at some point x . These requirements are the initial conditions.We’ll denote an initial value problem for
0
a differential equation by writing the initial conditions after the equation, as in Equation 1.2.11. For example, we would write
an initial value problem for Equation 1.2.2 as
(n) ′ (n−1) ′ (n−1)
y = f (x, y, y , … , y ), y(x0 ) = k0 , y (x0 ) = k1 , … , y = kn−1 . (1.2.12)
Consistent with our earlier definition of a solution of the differential equation in Equation 1.2.12, we say that y is a solution of
the initial value problem Equation 1.2.12 if y is n times differentiable and
(n) ′ (n−1)
y (x) = f (x, y(x), y (x), … , y (x))
for all x in some open interval (a, b) that contains x , and y satisfies the initial conditions in Equation 1.2.12. The largest
0
open interval that contains x on which y is defined and satisfies the differential equation is the interval of validity of y .
0
Example 1.2.6

In Example 1.2.5 we saw that
4
x +7
y = (1.2.13)
4
is a solution of the initial value problem

′ 3
y =x , y(1) = 2.
Since the function in Equation 1.2.13 is defined for all x, the interval of validity of this solution is (−∞, ∞).
Example 1.2.7
In Example 1.2.2 we verified that
2
x 1
y = + (1.2.14)
3 x
is a solution of
′ 2
xy + y = x
on (0, ∞) and on (−∞, 0). By evaluating Equation 1.2.14 at x = ±1 , you can see that Equation 1.2.14 is a solution of
the initial value problems
4
′ 2
xy + y = x , y(1) = (1.2.15)
3
and
′ 2
2
xy + y = x , y(−1) = − . (1.2.16)
3
The interval of validity of Equation 1.2.14 as a solution of Equation 1.2.15 is (0, ∞), since this is the largest interval that
contains x = 1 on which Equation 1.2.14 is defined. Similarly, the interval of validity of Equation 1.2.14 as a solution
0
of Equation 1.2.16 is (−∞, 0), since this is the largest interval that contains x = −1 on which Equation 1.2.14 is
0
defined.
Free Fall Under Constant Gravity

The term initial value problem originated in problems of motion where the independent variable is t (representing elapsed
time), and the initial conditions are the position and velocity of an object at the initial (starting) time of an experiment.
Example 1.2.8
An object falls under the influence of gravity near Earth’s surface, where it can be assumed that the magnitude of the
acceleration due to gravity is a constant g .
a. Construct a mathematical model for the motion of the object in the form of an initial value problem for a second order
differential equation, assuming that the altitude and velocity of the object at time t = 0 are known. Assume that
gravity is the only force acting on the object.
b. Solve the initial value problem derived above to obtain the altitude as a function of time.
Solution a
Let y(t) be the altitude of the object at time t . Since the acceleration of the object has constant magnitude g and is in the
downward (negative) direction, y satisfies the second order equation
′′
y = −g,
where the prime now indicates differentiation with respect to t . If y and v denote the altitude and velocity when t = 0 ,
0 0
then y is a solution of the initial value problem

′′ ′
y = −g, y(0) = y0 , y (0) = v0 . (1.2.17)
Solution b
Integrating Equation 1.2.17 twice yields
′
y = −gt + c1 ,
2
gt
y =− + c1 t + c2 .
2
Imposing the initial conditions y(0) = y and y (0) = v in these two equations shows that
0
′
0 c1 = v0 and c2 = y0 .
Therefore the solution of the initial value problem Equation 1.2.17 is
2
gt
y =− + v0 t + y0 .
2

1.2E: Basic Concepts (Exercises)
1. Find the order of the equation.
2 3
d y dy d y
a. dx
2
+2
dx

dx
3
+x = 0
b. y ′′ ′
− 3 y + 2y = x
7
c. y ′
−y
7
=0
d. y ′′
y − (y )
′ 2
=2
2. Verify that the function is a solution of the differential equation on some interval, for any choice of the arbitrary constants
appearing in the function.
a. y = ce 2x
; y
′
= 2y
b. y = x
3
+
c
x
; xy + y = x
′ 2
c. y =
2
1 −x ′
+ ce ; y + 2xy = x
2
2 2
d. y = (1 + ce −x /2
)(1 − c e
−x /2
)
−1
;
′
2 y + x(y
2
− 1) = 0
e. y = tan( x
3
+ c); y
′ 2
= x (1 + y )
2
f. y = (c 1 + c2 x)e
x
+ sin x + x ;
2
y
′′ ′
− 2 y + y = −2 cos x + x
2
− 4x + 2
g. y = c 1e
x
+ c2 x +
2
x
; (1 − x)y
′′ ′
+ x y − y = 4(1 − x − x )x
2 −3
h. y = x −1/2
(c1 sin x + c2 cos x) + 4x + 8 ;x 2
y
′′
+ x y + (x
′ 2
−
1
4
3
)y = 4 x
2
+ 8x + 3x − 2
3. Find all solutions of the equation.

a. y ′
= −x b. y ′
= −x sin x
c. y ′
= x ln x d. y ′′
= x cos x
e. y ′′
= 2x e
x
f. y ′′
= 2x + sin x + e
x
g. y ′′′
= − cos x h. y ′′′
= −x
2
+e
x
i. y ′′′
= 7e
4x
4. Solve the initial value problem.

a. y ′
= −x e ,
x
y(0) = 1
−
−
b. y ′
= x sin x ,
2
y (√
π
2
) =1
c. y ′
= tan x, y(π/4) = 3
d. y ′′
=x ,
4
y(2) = −1,
′
y (2) = −1
e. y ′′
= xe
2x
, y(0) = 7,
′
y (0) = 1
f. y ′′
= −x sin x, y(0) = 1,
′
y (0) = −3
g. y ′′′
=x e ,
2 x
y(0) = 1,
′
y (0) = −2, y
′′
(0) = 3
h. y ′′′
= 2 + sin 2x, y(0) = 1,
′
y (0) = −6, y
′′
(0) = 3
i. y ′′′
= 2x + 1, y(2) = 1,
′
y (2) = −4, y
′′
(2) = 7
5. Verify that the function is a solution of the initial value problem.

a. y = x cos x; y
′
= cos x − y tan x, y(π/4) =
π
4 √2
2 2
x −2 x y+2
b. y = 1+2 ln x
x
2
+
1
2
; y
′
=
x
3
, y(1) =
3
William F. Trench 9/11/2020 1.2E.1 CC-BY-NC-SA https://math.libretexts.org/@go/page/44217

2
c. y = tan( x
2
); y
′
= x(1 + y ),
2
y(0) = 0
−y(y+1)
d. y = x−2
2
; y
′
=
x
, y(1) = −2
6. Verify that the function is a solution of the initial value problem.

′
3xy −4y
a. y = x 2
(1 + ln x); y
′′
=
x
2
, y(e) = 2 e ,
2 ′
y (e) = 5e
2 ′
2
x −xy +y+1
b. y = x
3
+ x − 1; y
′′
=
x2
, y(1) =
1
3
,
′
y (1) =
5
2 2 ′
( x −1)y−x( x +1) y
c. y = (1 + x 2 −1/2
) ; y
′′
=
2
2
, y(0) = 1, y (0) = 0
′
( x +1 )
′
2 2(x+y)(xy −y)
d. y = 1−x
x
; y
′′
= 3
x
, y(1/2) = 1/2,
′
y (1/2) = 3
7. Suppose an object is launched from a point 320 feet above the earth with an initial velocity of 128 ft/sec upward, and the
only force acting on it thereafter is gravity. Take g = 32 ft/sec . 2
a. Find the highest altitude attained by the object.

b. Determine how long it takes for the object to fall to the ground.
8. Let a be a nonzero real number.
a. Verify that if c is an arbitrary constant then
a
y = (x − c) (A)
is a solution of
′ (a−1)/a
y = ay (B)
on (c, ∞).
b. Suppose a < 0 or a > 1 . Can you think of a solution of (B) that isn’t of the form (A)?
9. Verify that
x
e − 1, x ≥ 0,
y ={
−x
1 −e , x < 0,
is a solution of
′
y = |y| + 1
on (−∞, ∞).
10.
(a) Verify that if c is any real number then
2
y =c + cx + 2c + 1 (A)
satisfies
− −−−−−−−− −
2
−(x + 2) + √ x + 4x + 4y
′
y = (B)
2
on some open interval. Identify the open interval.

(b) Verify that
−x(x + 4)
y1 =
4
also satisfies (B) on some open interval, and identify the open interval. (Note that y1 can’t be obtained by selecting a
value of c in (A).)

1.3: Direction Fields for First Order Equations
It’s impossible to find explicit formulas for solutions of some differential equations. Even if there are such formulas, they may
be so complicated that they’re useless. In this case we may resort to graphical or numerical methods to get some idea of how
the solutions of the given equation behave.
In Section 2.3 we’ll take up the question of existence of solutions of a first order equation
′
y = f (x, y). (1.3.1)
In this section we’ll simply assume that Equation 1.3.1 has solutions and discuss a graphical method for approximating them.
In Chapter 3 we discuss numerical methods for obtaining approximate solutions of Equation 1.3.1. Recall that a solution of
Equation 1.3.1 is a function y = y(x) such that
′
y (x) = f (x, y(x))
for all values of x in some interval, and an integral curve is either the graph of a solution or is made up of segments that are
graphs of solutions. Therefore, not being able to solve Equation 1.3.1 is equivalent to not knowing the equations of integral
curves of Equation 1.3.1. However, it is easy to calculate the slopes of these curves. To be specific, the slope of an integral
curve of Equation 1.3.1 through a given point (x , y ) is given by the number f (x , y ). This is the basis of the method of
0 0 0 0
direction fields.
If f is defined on a set R , we can construct a direction field for Equation 1.3.1 in R by drawing a short line segment through
each point (x, y) in R with slope f (x, y). Of course, as a practical matter, we can’t actually draw line segments through every
point in R ; rather, we must select a finite set of points in R . For example, suppose f is defined on the closed rectangular
region
R : {a ≤ x ≤ b, c ≤ y ≤ d}.
Let
a = x0 < x1 < ⋯ < xm = b
be equally spaced points in [a, b] and
c = y0 < y1 < ⋯ < yn = d
form a rectangular grid (Figure 1.3.1). Through each point in the grid we draw a short line segment with slope f (x , y ). The
i j
result is an approximation to a direction field for Equation 1.3.1 in R . If the grid points are sufficiently numerous and close
together, we can draw approximate integral curves of Equation 1.3.1 by drawing curves through points in the grid tangent to
the line segments associated with the points in the grid.

Figure 1.3.1 : A rectangular grid
Unfortunately, approximating a direction field and graphing integral curves in this way is too tedious to be done effectively by
hand. However, there is software for doing this. As you’ll see, the combination of direction fields and integral curves gives
useful insights into the behavior of the solutions of the differential equation even if we can’t obtain exact solutions.
We’ll study numerical methods for solving a single first order equation Equation 1.3.1 in Chapter 3. These methods can be
used to plot solution curves of Equation 1.3.1 in a rectangular region R if f is continuous on R . Figures 1.3.2, 1.3.3, and
1.3.4 show direction fields and solution curves for the differential equations:
2 2
x −y
y
′
=
1+x2 +y 2
,
y
′
= 1 + xy
2
, and
x−y
y
′
=
1+x
2
.
which are all of the form Equation 1.3.1 with f continuous for all (x, y).
2 2
Figure 1.3.2 : A direction and integral curves for y .

′ x −y
=
2 2
1+ x + y

Figure 1.3.3 : A direction and integral curves for y ′
= 1 + xy
2
.
x−y
Figure 1.3.4 : A direction and integral curves for y ′
=
1+x2
.
The methods of Chapter 3 will not work for the equation

′
y = −x/y (1.3.2)
if R contains part of the x-axis, since f (x, y) = −x/y is undefined when y = 0 . Similarly, they will not work for the equation
2
′
x
y = (1.3.3)
1 − x2 − y 2
if R contains any part of the unit circle x + y = 1 , because the right side of Equation
2 2
1.3.3 is undefined if 2
x +y
2
=1 .
However, Equation 1.3.2 and Equation 1.3.3 can written as
A(x, y)
′
y = (1.3.4)
B(x, y)
where A and B are continuous on any rectangle R . Because of this, some differential equation software is based on
numerically solving pairs of equations of the form

dx dy
= B(x, y), = A(x, y) (1.3.5)
dt dt
where x and y are regarded as functions of a parameter t . If x = x(t) and y = y(t) satisfy these equations, then
dy dy dx A(x, y)
′
y = = / = ,
dx dt dt B(x, y)
so y = y(x) satisfies Equation 1.3.4.

Equations 1.3.2 and 1.3.3 can be reformulated as in Equation 1.3.4 with
dx dy
= −y, =x
dt dt
and
dx 2 2
dy 2
= 1 −x −y , =x ,
dt dt
respectively. Even if f is continuous and otherwise “nice” throughout R , your software may require you to reformulate the
equation y = f (x, y) as
′
dx dy
= 1, = f (x, y),
dt dt
which is of the form Equation 1.3.5 with A(x, y) = f (x, y) and B(x, y) = 1 .
Figure 1.3.5 shows a direction field and some integral curves for Equation 1.3.2. As we saw in Example [example:1.2.1} and
will verify again in Section 2.2, the integral curves of Equation 1.3.2 are circles centered at the origin.
Figure 1.3.5 : A direction field and integral curves for y ′

= −
x
Figure 1.3.6 shows a direction field and some integral curves for Equation 1.3.3. The integral curves near the top and bottom
are solution curves. However, the integral curves near the middle are more complicated. For example, Figure 1.3.7 shows the
integral curve through the origin. The vertices of the dashed rectangle are on the circle x + y = 1 ( a ≈ .846 , b ≈ .533),
2 2
where all integral curves of Equation 1.3.3 have infinite slope. There are three solution curves of Equation 1.3.3 on the
integral curve in the figure: the segment above the level y = b is the graph of a solution on (−∞, a) , the segment below the

level y = −b is the graph of a solution on (−a, ∞) , and the segment between these two levels is the graph of a solution on
(−a, a) .
2
x
=
2 2
.
1−x −y

= −
x
Using Technology
As you study from this book, you’ll often be asked to use computer software and graphics. Exercises with this intent are
marked as (computer or calculator required), (computer and/or graphics required), or (laboratory work requiring software
and/or graphics). Often you may not completely understand how the software does what it does. This is similar to the situation
most people are in when they drive automobiles or watch television, and it doesn’t decrease the value of using modern
technology as an aid to learning. Just be careful that you use the technology as a supplement to thought rather than a substitute
for it.

1.3E: Direction Fields for First Order Equations (Exercises)
Exercises for Section 1.3
In Exercises 1–11 a direction field is drawn for the given equation. Sketch some integral curves.
1.
Figure 1.3E. 1 : A direction field for y ′

=
x
2.
2
2xy
=
1+x
2

3.
Figure 1.3E. 3 : A direction field for y ′ 2

= x (1 + y )
2
4.

=
1
2
1+x +y
2

5.

= −(2xy
2
+y )
3
6.

= (x +y )
2 1/2

7.

= sin xy
8.
Figure 1.3E. 8 : A direction field for y ′ xy

=e

9.

= (x − y )(x
2 2
− y)
10.

=x y
2
+ xy
3

11.

= sin(x − 2y)
In Exercises 12 - 22 construct a direction field and plot some integral curves in the indicated rectangular region.
12. y ′
= y(y − 1); {−1 ≤ x ≤ 2, − 2 ≤ y ≤ 2}
13. y ′
= 2 − 3xy; {−1 ≤ x ≤ 4, − 4 ≤ y ≤ 4}
14. y ′
= xy(y − 1); {−2 ≤ x ≤ 2, − 4 ≤ y ≤ 4}
15. y ′
= 3x + y; {−2 ≤ x ≤ 2, 0 ≤ y ≤ 4}
16. y ′
= y −x ;
3
{−2 ≤ x ≤ 2, − 2 ≤ y ≤ 2}
17. y ′
= 1 −x
2
−y ;
2
{−2 ≤ x ≤ 2, − 2 ≤ y ≤ 2}
18. y ′
= x(y
2
− 1); {−3 ≤ x ≤ 3, − 3 ≤ y ≤ 2}
19. y ′
=
y( y
x
2
−1)
; {−2 ≤ x ≤ 2, − 2 ≤ y ≤ 2}
2
xy
20. y ′
=
y−1
; {−2 ≤ x ≤ 2, − 1 ≤ y ≤ 4}
2
x( y −1)
21. y ′
=
y
; {−1 ≤ x ≤ 1, − 2 ≤ y ≤ 2}
2 2
x +y
22. y ′
=−
1−x −y
2 2
; {−2 ≤ x ≤ 2, − 2 ≤ y ≤ 2}
23. By suitably renaming the constants and dependent variables in the equations
′
T = −k(T − Tm ) (A)
and

′
G = −λG + r (B)
discussed in Section 1.2 in connection with Newton’s law of cooling and absorption of glucose in the body, we can write both
as
′
y = −ay + b, (C)
where a is a positive constant and b is an arbitrary constant. Thus, (A) is of the form (C) with y = T , a = k , and b = kT , m
and (B) is of the form (C) with y = G , a = λ , and b = r . We’ll encounter equations of the form (C) in many other
applications in Chapter 2.
Choose a positive a and an arbitrary b . Construct a direction field and plot some integral curves for (C) in a rectangular region
of the form
{0 ≤ t ≤ T , c ≤ y ≤ d} (1.3E.1)
of the ty -plane. Vary T , c , and d until you discover a common property of all the solutions of (C). Repeat this experiment with
various choices of a and b until you can state this property precisely in terms of a and b .
24. By suitably renaming the constants and dependent variables in the equations
′
P = aP (1 − αP ) (A)
and
′
I = rI (S − I ) (B)
discussed in Section 1.1 in connection with Verhulst’s population model and the spread of an epidemic, we can write both in
the form
′ 2
y = ay − b y , (C)
where a and b are positive constants. Thus, (A) is of the form (C) with y = P , a = a , and b = aα , and (B) is of the form (C)
with y = I , a = rS , and b = r . In Chapter 2 we’ll encounter equations of the form (C) in other applications..
Choose positive numbers a and b . Construct a direction field and plot some integral curves for (C) in a rectangular region of
the form
{0 ≤ t ≤ T , 0 ≤ y ≤ d} (1.3E.2)
of the ty -plane. Vary T and d until you discover a common property of all solutions of (C) with y(0) > 0 . Repeat this
experiment with various choices of a and b until you can state this property precisely in terms of a and b .
Choose positive numbers a and b . Construct a direction field and plot some integral curves for (C) in a rectangular region of
the form
{0 ≤ t ≤ T , c ≤ y ≤ 0} (1.3E.3)
of the ty -plane. Vary a , b , T and c until you discover a common property of all solutions of (C) with y(0) < 0 .
You can verify your results later by doing Exercise 2.2.27.

CHAPTER OVERVIEW
2: FIRST ORDER EQUATIONS
In this chapter, we study first-order differential equations for which there are general methods of
solution.
2.1: LINEAR FIRST ORDER EQUATIONS

This section deals with linear equations, the simplest kind of rst order equations. In this section
we introduce the method of variation of parameters. The idea underlying this method will be a
unifying theme for our approach to solving many different kinds of differential equations
throughout the book.
2.1E: LINEAR FIRST ORDER EQUATIONS (EXERCISES)

2.2: SEPARABLE EQUATIONS
The section deals with separable equations, the simplest nonlinear equations. In this section we introduce the idea of implicit and
constant solutions of differential equations, and we point out some differences between the properties of linear and nonlinear
equations.
2.2E: SEPARABLE EQUATIONS (EXERCISES)

2.3: EXISTENCE AND UNIQUENESS OF SOLUTIONS OF NONLINEAR EQUATIONS
Although there are methods for solving some nonlinear equations, it is impossible to find useful formulas for the solutions of most.
Whether we are looking for exact solutions or numerical approximations, it is useful to know conditions that imply the existence and
uniqueness of solutions of initial value problems for nonlinear equations. In this section we state such a condition and illustrate it with
examples.
2.3E: EXISTENCE AND UNIQUENESS OF SOLUTIONS OF NONLINEAR EQUATIONS (EXERCISES)

2.4: TRANSFORMATION OF NONLINEAR EQUATIONS INTO SEPARABLE EQUATIONS
This section deals with nonlinear equations that are not separable, but can be transformed into separable equations by a procedure
similar to variation of parameters.
2.4E: TRANSFORMATION OF NONLINEAR EQUATIONS INTO SEPARABLE EQUATIONS (EXERCISES)

2.5: EXACT EQUATIONS
This section covers exact differential equations, which are given this name because the method for solving them uses the idea of an
exact differential from calculus.
2.5E: EXACT EQUATIONS (EXERCISES)

2.6: INTEGRATING FACTORS
This section deals with equations that are not exact, but can made exact by multiplying them by a function known called integrating
factor.
2.6E: INTEGRATING FACTORS (EXERCISES)
1 9/16/2020
2.1: Linear First Order Equations
A first order differential equation is said to be linear if it can be written as
′
y + p(x)y = f (x). (2.1.1)
A first order differential equation that cannot be written like this is nonlinear. We say that Equation 2.1.1 is homogeneous if
f ≡ 0 ; otherwise it is nonhomogeneous. Since y ≡ 0 is obviously a solution of the homgeneous equation
′
y + p(x)y = 0,
we call it the trivial solution. Any other solution is nontrivial.
Example 2.1.1
The first order equations
2 ′ 2
x y + 3y = x
′ 2
x y − 8 x y = sin x
′
x y + (ln x)y = 0
′ 2
y = x y −2
are not in the form in Equation 2.1.1, but they are linear, since they can be rewritten as
′
3
y + y =1
x2
′
sin x
y − 8xy =
x
′
ln x
y + y =0
x
′ 2
y − x y = −2
Example 2.1.2
Here are some nonlinear first order equations:
′ 2
xy + 3y = 2x (because y is squared)
′ ′
yy =3 (because of the product y y )
′ y y
y + xe = 12 (because of e )
General Solution of a Linear First Order Equation

To motivate a definition that we’ll need, consider the simple linear first order equation
1
′
y = . (2.1.2)
2
x
From calculus we know that y satisfies this equation if and only if

1
y =− + c, (2.1.3)
x
where c is an arbitrary constant. We call c a parameter and say that Equation 2.1.3 defines a one–parameter family of
functions. For each real number c , the function defined by Equation 2.1.3 is a solution of Equation 2.1.2 on (−∞, 0) and

(0, ∞) ; moreover, every solution of Equation 2.1.2 on either of these intervals is of the form Equation 2.1.3 for some choice
of c . We say that Equation 2.1.3 is the general solution of Equation 2.1.2.
We’ll see that a similar situation occurs in connection with any first order linear equation
′
y + p(x)y = f (x); (2.1.4)
that is, if p and f are continuous on some open interval (a, b) then there’s a unique formula y = y(x, c) analogous to Equation
2.1.3 that involves x and a parameter c and has the these properties:
For each fixed value of c , the resulting function of x is a solution of Equation 2.1.4 on (a, b).
If y is a solution of Equation 2.1.4 on (a, b), then y can be obtained from the formula by choosing c appropriately.
We’ll call y = y(x, c) the general solution of Equation 2.1.4.
When this has been established, it will follow that an equation of the form
′
P0 (x)y + P1 (x)y = F (x) (2.1.5)
has a general solution on any open interval (a, b) on which P , P , and F are all continuous and P has no zeros, since in this
0 1 0
case we can rewrite Equation 2.1.5 in the form Equation 2.1.4 with p = P /P and f = F /P , which are both continuous on
1 0 0
(a, b) .
To avoid awkward wording in examples and exercises, we will not specify the interval (a, b) when we ask for the general
solution of a specific linear first order equation. Let’s agree that this always means that we want the general solution on every
open interval on which p and f are continuous if the equation is of the form Equation 2.1.4, or on which P , P , and F are 0 1
continuous and P has no zeros, if the equation is of the form Equation 2.1.5. We leave it to you to identify these intervals in
0
specific examples and exercises.

For completeness, we point out that if P , P , and F are all continuous on an open interval (a, b), but P does have a zero in
0 1 0
(a, b) , then Equation 2.1.5 may fail to have a general solution on (a, b) in the sense just defined. Since this isn’t a major point
that needs to be developed in depth, we will not discuss it further; however, see Exercise 2.1.44 for an example.
Homogeneous Linear First Order Equations

We begin with the problem of finding the general solution of a homogeneous linear first order equation. The next example
recalls a familiar result from calculus.
Example 2.1.3
Let a be a constant.
a. Find the general solution of
′
y − ay = 0. (2.1.6)
b. Solve the initial value problem

′
y − ay = 0, y(x0 ) = y0 .
Solution a
(a) You already know from calculus that if c is any constant, then y = ce satisfies Equation 2.1.6. However, let’s
ax
pretend you’ve forgotten this, and use this problem to illustrate a general method for solving a homogeneous linear first
order equation.
We know that Equation 2.1.6 has the trivial solution y ≡ 0 . Now suppose y is a nontrivial solution of Equation 2.1.6.
Then, since a differentiable function must be continuous, there must be some open interval I on which y has no zeros. We
rewrite Equation 2.1.6 as
′
y
=a
y

Figure 2.1.1 : Solutions of y ′
− ay = 0 , y(0) = 1
for x in I . Integrating this shows that
k ax
ln |y| = ax + k, so |y| = e e ,
where k is an arbitrary constant. Since e ax

can never equal zero, y has no zeros, so y is either always positive or always
negative. Therefore we can rewrite y as
ax
y = ce (2.1.7)
where
k
e ify > 0,
c ={
k
−e ify < 0.
This shows that every nontrivial solution of Equation 2.1.6 is of the form y = ce for some nonzero constant c . Since ax
setting c = 0 yields the trivial solution, all solutions of Equation 2.1.6 are of the form Equation 2.1.7. Conversely,
Equation 2.1.7 is a solution of Equation 2.1.6 for every choice of c , since differentiating Equation 2.1.7 yields
= ay .
′ ax
y = ac e
Solution b
Imposing the initial condition y(x 0) = y0 yields y0 = ce
ax0
, so c = y 0e
−ax0
and
−ax0 ax a(x−x0 )
y = y0 e e = y0 e .
Figure 2.1.1 show the graphs of this function with x 0 =0 ,y 0 =1 , and various values of a .
Example 2.1.4
′
x y + y = 0. (2.1.8)

′
x y + y = 0, y(1) = 3. (2.1.9)
Solution a
We rewrite Equation 2.1.8 as

′
1
y + y = 0, (2.1.10)
x
where x is restricted to either (−∞, 0) or (0, ∞). If y is a nontrivial solution of Equation 2.1.10 , there must be some
open interval I on which y has no zeros. We can rewrite Equation 2.1.10 as
′
y 1
=−
y x
for x in I . Integrating shows that

k
e
ln |y| = − ln |x| + k, so |y| = .
|x|
Since a function that satisfies the last equation can’t change sign on either (−∞, 0) or (0, ∞), we can rewrite this result
more simply as
c
y = (2.1.11)
x
where
k
e ify > 0,
c ={
k
−e ify < 0.
We have now shown that every solution of Equation 2.1.10 is given by Equation 2.1.11 for some choice of c . (Even
though we assumed that y was nontrivial to derive Equation 2.1.11, we can get the trivial solution by setting c = 0 in
Equation 2.1.11.) Conversely, any function of the form Equation 2.1.11 is a solution of Equation 2.1.10, since
differentiating Equation 2.1.11 yields
c
′
y =− ,
2
x
and substituting this and Equation 2.1.11 into Equation 2.1.10 yields
′
1 c 1 c
y + y = − +
x x2 x x
c c
= − + = 0.
2 2
x x
Figure 2.1.2 shows the graphs of some solutions corresponding to various values of c
Solution b
Imposing the initial condition y(1) = 3 in Equation 2.1.11 yields c = 3 . Therefore the solution of Equation 2.1.9 is
3
y = .
x
The interval of validity of this solution is (0, ∞).
The results in Examples 2.1.3a and 2.1.4b are special cases of the next theorem.
Theorem 2.1.2
If p is continuous on (a, b), then the general solution of the homogeneous equation
′
y + p(x)y = 0 (2.1.12)
on (a, b) is
−P (x)
y = ce ,

where
P (x) = ∫ p(x) dx (2.1.13)
is any antiderivative of p on (a, b); that is ,

′
P (x) = p(x), a <x <b (2.1.14)
Figure 2.1.2 : Solutions of xy ′

+y = 0 on (0, ∞) and (−∞, 0)
Proof
If y = ce −P (x)
, differentiating y and using Equation 2.1.14 shows that
′ ′ −P (x) −P (x)
y = −P (x)c e = −p(x)c e = −p(x)y,
so y ′
+ p(x)y = 0 ; that is, y is a solution of Equation 2.1.12, for any choice of c .
Now we’ll show that any solution of Equation 2.1.12 can be written as y = ce for some constant c . The trivial
−P (x)
solution can be written this way, with c = 0 . Now suppose y is a nontrivial solution. Then there’s an open subinterval I of
(a, b) on which y has no zeros. We can rewrite Equation 2.1.12 as
′
y
= −p(x) (2.1.15)
y
for x in I . Integrating Equation 2.1.15 and recalling Equation 2.1.13 yields
ln |y| = −P (x) + k,
where k is a constant. This implies that

k −P (x)
|y| = e e .
Since P is defined for all x in (a, b) and an exponential can never equal zero, we can take I = (a, b) , so y has zeros on
(a, b) (a, b) , so we can rewrite the last equation as y = ce , where −P (x)
k
e if y > 0 on (a, b),
c ={
k
−e if y < 0 on (a, b).

REMARK: Rewriting a first order differential equation so that one side depends only on y and y and the other depends only ′
on x is called separation of variables. We did this in Examples 2.1.3 and 2.1.4, and in rewriting Equation 2.1.12 and Equation
2.1.15. We will apply this method to nonlinear equations in Section 2.2.
Linear Nonhomogeneous First Order Equations

We’ll now solve the nonhomogeneous equation
′
y + p(x)y = f (x). (2.1.16)
When considering this equation we call

′
y + p(x)y = 0
the complementary equation.

We’ll find solutions of Equation 2.1.16 in the form y = uy , where y is a nontrivial solution of the complementary equation
1 1
and u is to be determined. This method of using a solution of the complementary equation to obtain solutions of a
nonhomogeneous equation is a special case of a method called variation of parameters, which you’ll encounter several times
in this book. (Obviously, u can’t be constant, since if it were, the left side of Equation 2.1.16 would be zero. Recognizing this,
the early users of this method viewed u as a “parameter” that varies; hence, the name “variation of parameters.”)
If
′ ′ ′
y = u y1 , then y = u y1 + u y .
1
Substituting these expressions for y and y into Equation 2.1.16 yields

′
′ ′
u y1 + u(y + p(x)y1 ) = f (x),
1
which reduces to
′
u y1 = f (x), (2.1.17)
since y is a solution of the complementary equation; that is,

1
′
y + p(x)y1 = 0.
1
In the proof of Theorem 2.2.1 we saw that y1 has no zeros on an interval where p is continuous. Therefore we can divide
Equation 2.1.17 through by y to obtain
1
′
u = f (x)/ y1 (x).
We can integrate this (introducing a constant of integration), and multiply the result by y to get the general solution of 1
Equation 2.1.16. Before turning to the formal proof of this claim, let’s consider some examples.
Example 2.1.5
Find the general solution of
′ 3 −2x
y + 2y = x e . (2.1.18)
By applying a of Example 2.1.3 with a = −2 , we see that y = e is a solution of the complementary equation
1
−2x
y + 2y = 0 . Therefore we seek solutions of Equation 2.1.18 in the form y = ue , so that

′ −2x
′ ′ −2x −2x ′ ′ −2x −2x −2x ′ −2x

y =u e − 2u e and y + 2y = u e − 2u e + 2u e =u e . (2.1.19)
Therefore y is a solution of Equation 2.1.18 if and only if

′ −2x 3 −2x ′ 3
u e =x e or, equivalently, u =x .
Therefore
4
x
u = + c,
4

Figure 2.1.3 : A direction field and integral curves for y ′ 2
+ 2y = x e
−2x
and
4
−2x −2x
x
y = ue =e ( + c)
4
is the general solution of Equation 2.1.18.

Figure 2.1.3 shows a direction field and some integral curves for Equation 2.1.18.
Example 2.1.6
Find the general solution
′
y + (cot x)y = x csc x. (2.1.20)
Solve the initial value problem

′
y + (cot x)y = x csc x, y(π/2) = 1. (2.1.21)
Here p(x) = cot x and f (x) = x csc x are both continuous except at the points x = rπ, where r is an integer. Therefore
we seek solutions of Equation 2.1.20 on the intervals (rπ, (r + 1)π). We need a nontrival solution y of the 1
complementary equation; thus, y must satisfy y + (cot x)y = 0 , which we rewrite as

1 1
′
1
′
y cos x
1
= − cot x = − . (2.1.22)
y1 sin x
Integrating this yields
ln | y1 | = − ln | sin x|,
where we take the constant of integration to be zero since we need only one function that satisfies Equation 2.1.22.
Clearly y = 1/ sin x is a suitable choice. Therefore we seek solutions of Equation 2.1.20 in the form
1
u
y = ,
sin x
so that
′
u u cos x
′
y = − (2.1.23)
2
sin x sin x
and

′
′ u u cos x u cot x
y + (cot x)y = − +
2
sin x sin x sin x
′
u u cos x u cos x
= − 2
+ 2
(2.1.24)
sin x sin x sin x
′
u
= .
sin x
Therefore y is a solution of Equation 2.1.20 if and only if

′ ′
u / sin x = x csc x = x/ sin x or, equivalently, u = x.

2 2
x u x c
u = + c, and y = = + . (2.1.25)
2 sin x 2 sin x sin x
is the general solution of Equation 2.1.20 on every interval (rπ, (r + 1)π) ( r =integer).
b. Imposing the initial condition y(π/2) = 1 in Equation 2.1.25 yields
2 2
π π
1 = + c orc = 1 − .
8 8
Thus,
2 2
x (1 − π /8)
y = +
2 sin x sin x
is a solution of Equation 2.1.20. The interval of validity of this solution is (0, π); Figure 2.1.4 shows its graph.
Figure 2.1.4 : Solution of y ′

+ (cot x)y = x csc x, y(π/2) = 1
REMARK: It wasn’t necessary to do the computations 2.1.23 and 2.1.24 in Example 2.1.6, since we showed in the
discussion preceding Example 2.1.5 that if y = uy where y + p(x)y = 0 , then y + p(x)y = u y . We did these
1
′
1 1
′ ′
1
computations so you would see this happen in this specific example. We recommend that you include these “unnecesary”
computations in doing exercises, until you’re confident that you really understand the method. After that, omit them.
We summarize the method of variation of parameters for solving
′
y + p(x)y = f (x) (2.1.26)
as follows:
a. Find a function y such that
1
′
y
1
= −p(x).
y1
For convenience, take the constant of integration to be zero.

b. Write
y = uy1 (2.1.27)

to remind yourself of what you’re doing.
c. Write u y′
1 =f and solve for u ; thus, u
′ ′
= f / y1 .
d. Integrate u to obtain u, with an arbitrary constant of integration.
′
e. Substitute u into Equation 2.1.27 to obtain y .

To solve an equation written as
′
P0 (x)y + P1 (x)y = F (x),
we recommend that you divide through by P 0 (x) to obtain an equation of the form Equation 2.1.26 and then follow this
procedure.
Solutions in Integral Form

Sometimes the integrals that arise in solving a linear first order equation can’t be evaluated in terms of elementary functions.
In this case the solution must be left in terms of an integral.
Example 2.1.7
′
y − 2xy = 1.

′
y − 2xy = 1, y(0) = y0 . (2.1.28)
a. To apply variation of parameters, we need a nontrivial solution y1 of the complementary equation; thus,
y − 2x y = 0 , which we rewrite as
′
1 1
′
y
1
= 2x.
y1
Integrating this and taking the constant of integration to be zero yields

2
2 x
ln | y1 | = x , so | y1 | = e .
We choose y and seek solutions of Equation 2.1.28 in the form y = ue , where

2 2
x x
1 =e
2 2
′ x ′ −x
u e = 1, so u =e .
Therefore
2
−x
u = c +∫ e dx,
but we can’t simplify the integral on the right because there’s no elementary function with derivative equal to e
−x
.
Therefore the best available form for the general solution of Equation 2.1.28 is
2 2 2
x x −x
y = ue =e (c + ∫ e dx) . (2.1.29)
b. Since the initial condition in Equation 2.1.28 is imposed at x 0 =0 , it is convenient to rewrite Equation 2.1.29 as
x 0
2 2 2
x −t −t
y =e (c + ∫ e dt) , since ∫ e dt = 0
0 0
Setting x = 0 and y = y here shows that c = y . Therefore the solution of the initial value problem is
0 0
x
2 2
x −t
y =e ( y0 + ∫ e dt) . (2.1.30)
0

For a given value of y and each fixed x, the integral on the right can be evaluated by numerical methods. An alternate
0
procedure is to apply the numerical integration procedures discussed in Chapter 3 directly to the initial value problem
Equation 2.1.28. Figure 2.1.5 shows graphs of of Equation 2.1.30 for several values of y . 0
Figure 2.1.5 : Solution of y ′

− 2xy = 1, y(0) = y0
An Existence and Uniqueness Theorem

The method of variation of parameters leads to this theorem.
Theorem 2.1.2 Existence and Uniqueness Theorem

Suppose p and f are continuous on an open interval (a, b), and let y1 be any nontrivial solution of the complementary
equation
′
y + p(x)y = 0
on (a, b). Then:

a. The general solution of the nonhomogeneous equation
′
y + p(x)y = f (x) (2.1.31)
on (a, b) is
y = y1 (x) (c + ∫ f (x)/ y1 (x) dx) . (2.1.32)
b. If x is an arbitrary point in (a, b) and y is an arbitrary real number , then the initial value problem
0 0
′
y + p(x)y = f (x), y(x0 ) = y0
has the unique solution

x
y0 f (t)
y = y1 (x) ( +∫ dt)
y1 (x0 ) x0 y1 (t)
on (a, b).
Proof
(a) To show that Equation 2.1.32is the general solution of Equation 2.1.31on (a, b) , we must prove that:
i. If c is any constant, the function y in Equation 2.1.32is a solution of Equation 2.1.31on (a, b) .
ii. If y is a solution of Equation 2.1.31on (a, b) then y is of the form Equation 2.1.32for some constant c.
To prove (i), we first observe that any function of the form Equation 2.1.32 is defined on (a, b) , since p and f are
continuous on (a, b) . Differentiating Equation 2.1.32yields

′ ′
y = y (x) (c + ∫ f (x)/ y1 (x) dx) + f (x).
1
Since y 1
′
= −p(x)y1 , this and Equation 2.1.32imply that
′
y = −p(x)y1 (x) (c + ∫ f (x)/ y1 (x) dx) + f (x)
= −p(x)y(x) + f (x),
which implies that y is a solution of Equation 2.1.31.

To prove (ii), suppose y is a solution of Equation 2.1.31on (a, b) . From the proof of Theorem 2.1.1, we know that y 1
has no zeros on (a, b) , so the function u = y/y is defined on (a, b) . Moreover, since
1
′ ′
y = −py + f and y = −p y1 ,
1
′ ′
y1 y − y y
′ 1
u =
2
y
1
y1 (−py + f ) − (−p y1 )y f
= = .
2
y y1
1
Integrating u ′
= f / y1 yields
u = (c + ∫ f (x)/ y1 (x) dx) ,
which implies Equation 2.1.32, since y = uy . 1
(b) We’ve proved (a), where ∫ f (x)/y (x) dx in Equation 2.1.32 is an arbitrary antiderivative of f /y . Now it is
1 1
convenient to choose the antiderivative that equals zero when x = x , and write the general solution of Equation 0
2.1.31as
x
f (t)
y = y1 (x) (c + ∫ dt) .
x0 y1 (t)
Since
x0
f (t)
y(x0 ) = y1 (x0 ) (c + ∫ dt) = c y1 (x0 ),
x0 y1 (t)
we see that y(x 0) = y0 if and only if c = y 0 / y1 (x0 ) .

2.1E: Linear First Order Equations (Exercises)
Q2.1.1
In Exercises 2.1.1-2.1.5 find the general solution.
1. y ′
+ ay = 0 ( a =constant)
2. y ′
+ 3x y = 0
2
3. x y ′
+ (ln x)y = 0
4. x y ′
+ 3y = 0
5. x 2 ′
y +y = 0
Q2.1.2
In Exercises 2.1.6-2.1.11 solve the initial value problem.
1+x
6. y ′
+(
x
) y = 0, y(1) = 1
37 x y ′
+ (1 +
1
ln x
) y = 0, y(e) = 1
8. x y ′
+ (1 + x cot x)y = 0, y(
π
2
) =2
9. y ′
−(
1+x
2x
2
) y = 0, y(0) = 2
10. y ′
+
k
x
y = 0, y(1) = 3 (k = constant)
11. y ′
+ (tan kx)y = 0, y(0) = 2 (k = constant)
Q2.1.3
In Exercises 2.1.12-2.1.15 find the general solution. Also, plot a direction field and some integral curves on the rectangular
region {−2 ≤ x ≤ 2, −2 ≤ y ≤ 2} .
12. y ′
+ 3y = 1
13. y ′
+(
1
x
− 1) y = −
2
14. y
2
′ −x
+ 2xy = x e
−x
15. y ′
+
1+x
2x
2
y =
1+x
e
2
Q2.1.4
16. y ′
+
1
x
y =
7
x
2
+3
17. y ′
+
x−1
4
y =
1
5
+
sin x
4
(x−1) (x−1)
18. x y ′
+ (1 + 2 x )y = x e
2 3 −x
19. x y ′
+ 2y =
x
2
2
+1
20. y ′
+ (tan x)y = cos x
21. (1 + x)y ′
+ 2y =
sin x
1+x
22. (x − 2)(x − 1)y ′

− (4x − 3)y = (x − 2 )
3
23. y ′
+ (2 sin x cos x)y = e
− sin x
24. x 2 ′
y + 3xy = e
x

Q2.1.5
In Exercises 2.1.25-2.1.29 solve the initial value problem and sketch the graph of the solution.
25. y ′
+ 7y = e
3x
, y(0) = 0
26. (1 + x 2 ′
)y + 4xy =
1+x
2
2
, y(0) = 1
27. x y ′
+ 3y =
x(1+x )
2
2
, y(−1) = 0
28. y ′
+ (cot x)y = cos x, y(
π
2
) =1
29. y ′
+
1
x
y =
2
x
2
+ 1, y(−1) = 0
Q2.1.6
In Exercises 2.1.30-2.1.37, solve the initial value problem.
30. (x − 1)y ′
+ 3y =
1
3
+
sin x
2
, y(0) = 1
(x−1) (x−1)
31. x y ′
+ 2y = 8 x ,
2
y(1) = 3
32. x y ′
− 2y = −x ,
2
y(1) = 1
33. y ′
+ 2xy = x, y(0) = 3
2
1+(x−1) sec x
34. (x − 1)y ′
+ 3y = 3
, y(0) = −1
(x−1)
35. (x + 2)y ′
+ 4y =
1+2x
3
, y(−1) = 2
x(x+2)
36. (x 2 ′
− 1)y − 2xy = x(x
2
− 1), y(0) = 4
37. (x 2 ′
− 5)y − 2xy = −2x(x
2
− 5), y(2) = 7
Q2.1.7
In Excercises 2.1.28-2.1.42 solve the initial value problem and leave the answer in a form involving a definite integral. (You
can solve these problems numerically by methods discussed in Chapter 3.)
38. y ′
+ 2xy = x ,
2
y(0) = 3
39. y ′
+
1
x
y =
sin x
x
2
, y(1) = 2
−x
40. y ′
+y =
e tan x
x
, y(1) = 0
41. y ′
+
1+x
2x
2
y =
e
2
2
, y(0) = 1
(1+x )
42. x y ′
+ (x + 1)y = e
x
, y(1) = 2
43. Experiments indicate that glucose is absorbed by the body at a rate proportional to the amount of glucose present in the
bloodstream. Let λ denote the (positive) constant of proportionality. Now suppose glucose is injected into a patient’s
bloodstream at a constant rate of r units per unit of time. Let G = G(t) be the number of units in the patient’s bloodstream at
time t > 0 . Then
′
G = −λG + r, (2.1E.1)
where the first term on the right is due to the absorption of the glucose by the patient’s body and the second term is due to the
injection. Determine G for t > 0 , given that G(0) = G . Also, find lim G(t) . 0 t→∞
44.
(a) Plot a direction field and some integral curves for
′
x y − 2y = −1 (A)
on the rectangular region {−1 ≤ x ≤ 1, −.5 ≤ y ≤ 1.5} . What do all the integral curves have in common?

(b) Show that the general solution of (A) on (−∞, 0) and (0, ∞) is
1 2
y = + cx . (2.1E.2)
2
(c) Show that y is a solution of (A) on (−∞, ∞) if and only if

1 2
⎧ + c1 x , x ≥ 0,
2
y =⎨ (2.1E.3)
⎩ 1 2
+ c2 x , x < 0,
2
where c and c are arbitrary constants.

1 2
(d) Conclude from c that all solutions of (A) on (−∞, ∞) are solutions of the initial value problem
′
1
x y − 2y = −1, y(0) = . (2.1E.4)
2
(e) Use (b) to show that if x 0 ≠0 and y is arbitrary, then the initial value problem
0
′
x y − 2y = −1, y(x0 ) = y0 (2.1E.5)
has infinitely many solutions on ( −∞, ∞). Explain why this doesn't contradict Theorem 2.1.1.
45. Suppose f is continuous on an open interval (a, b) and α is a constant.

(a) Derive a formula for the solution of the initial value problem
′
y + αy = f (x), y(x0 ) = y0 , (A)
where x is in (a, b) and y is an arbitrary real number.

0 0
(b) Suppose (a, b) = (a, ∞) , α > 0 and lim f (x) = L . Show that if y is the solution of (A), then lim y(x) = L/α .
x→∞ x→∞
46. Assume that all functions in this exercise are defined on a common interval (a, b).
(a) Prove: If y and y are solutions of
1 2
′
y + p(x)y = f1 (x) (2.1E.6)
and
′
y + p(x)y = f2 (x) (2.1E.7)
respectively, and c and c are constants, then y = c

1 2 1 y1 + c2 y2 is a solution of
′
y + p(x)y = c1 f1 (x) + c2 f2 (x). (2.1E.8)
(This is the principle of superposition.)

(b) Use (a) to show that if y and y are solutions of the nonhomogeneous equation
1 2
′
y + p(x)y = f (x), (A) (2.1E.9)
then y1 − y2 is a solution of the homogeneous equation

′
y + p(x)y = 0. (B) (2.1E.10)
(c) Use (a) to show that if y is a solution of (A) and y is a solution of (B), then y
1 2 1 + y2 is a solution of (A).
47. Some nonlinear equations can be transformed into linear equations by changing the dependent variable. Show that if
′ ′
g (y)y + p(x)g(y) = f (x) (2.1E.11)
where y is a function of x and g is a function of y , then the new dependent variable z = g(y) satisfies the linear equation
′
z + p(x)z = f (x). (2.1E.12)
48. Solve by the method discussed in Exercise 47.

(a) (sec 2 ′
y)y − 3 tan y = −1
(b) e
2
y ′ 2 1
(2y y + ) =
x 2
x
′
xy
(c) y
+ 2 ln y = 4 x
2
′
y
(d) 2
−
1
x(1+y)
=−
3
2
x
(1+y)
49. We’ve shown that if p and f are continuous on (a, b) then every solution of
′
y + p(x)y = f (x) (A)
on (a, b) can be written as y = uy , where y is a nontrivial solution of the complementary equation for (A) and u
1 1
′
= f / y1 .
Now suppose f , f , …, f and p, p , …, p
′
are continuous on (a, b), where m is a positive integer, and define
(m) ′ (m−1)
f0 = f ,
′
fj =f + p fj−1 , 1 ≤ j ≤ m.
j−1
Show that
fj
(j+1)
u = , 0 ≤ j ≤ m. (2.1E.13)
y1

2.2: Separable Equations
A first order differential equation is separable if it can be written as
′
h(y)y = g(x), (2.2.1)
where the left side is a product of y and a function of y and the right side is a function of x. Rewriting a separable differential
′
equation in this form is called separation of variables. In Section 2.1, we used separation of variables to solve homogeneous
linear equations. In this section we’ll apply this method to nonlinear equations.
To see how to solve Equation 2.2.1, let’s first assume that y is a solution. Let G(x) and H (y) be antiderivatives of g(x) and
h(y); that is,
′ ′
H (y) = h(y) and G (x) = g(x). (2.2.2)
Then, from the chain rule,

d
′ ′ ′
H (y(x)) = H (y(x))y (x) = h(y)y (x).
dx
Therefore Equation 2.2.1 is equivalent to

d d
H (y(x)) = G(x).
dx dx
Integrating both sides of this equation and combining the constants of integration yields
H (y(x)) = G(x) + c. (2.2.3)
Although we derived this equation on the assumption that y is a solution of Equation 2.2.1, we can now view it differently:
Any differentiable function y that satisfies Equation 2.2.3 for some constant c is a solution of Equation 2.2.1. To see this, we
differentiate both sides of Equation 2.2.3, using the chain rule on the left, to obtain
′ ′ ′
H (y(x))y (x) = G (x),
which is equivalent to
′
h(y(x))y (x) = g(x)
because of Equation 2.2.2.

In conclusion, to solve Equation 2.2.1 it suffices to find functions G = G(x) and H = H (y) that satisfy Equation 2.2.2. Then
any differentiable function y = y(x) that satisfies Equation 2.2.3 is a solution of Equation 2.2.1.
Example 2.2.1
Solve the equation
′ 2
y = x(1 + y ).
Solution
Separating variables yields
′
y
= x.
2
1 +y
Integrating yields
2
x
−1
tan y = +c
2
Therefore

2
x
y = tan( + c).
2
Example 2.2.2
a. Solve the equation
x
′
y =− . (2.2.4)
y

′
x
y =− , y(1) = 1. (2.2.5)
y
c. Solve the initial value problem

′
x
y =− , y(1) = −2. (2.2.6)
y
Solution a
Separating variables in Equation 2.2.4 yields
′
yy = −x.
Integrating yields
2 2
y x
2 2
=− + c, or equivalently x +y = 2c.
2 2
The last equation shows that c must be positive if y is to be a solution of Equation 2.2.4 on an open interval. Therefore
we let 2c = a (with a > 0 ) and rewrite the last equation as
2
2 2 2
x +y =a . (2.2.7)
This equation has two differentiable solutions for y in terms of x:

−−−−−−
2 2
y = √a − x , −a < x < a, (2.2.8)
and
− −−−−−
2 2
y = −√ a − x , −a < x < a. (2.2.9)
The solution curves defined by Equation 2.2.8 are semicircles above the x-axis and those defined by Equation 2.2.9 are
semicircles below the x-axis (Figure 2.2.1).
Solution b
The solution of Equation 2.2.5 is positive when x = 1 ; hence, it is of the form Equation 2.2.8. Substituting x = 1 and
y = 1 into Equation 2.2.7 to satisfy the initial condition yields a = 2 ; hence, the solution of Equation 2.2.5 is
2
−−−− − – –
2
y = √2 − x , −√2 < x < √2.
Solution c
The solution of Equation 2.2.6 is negative when x = 1 and is therefore of the form Equation 2.2.9. Substituting x = 1
and y = −2 into Equation 2.2.7 to satisfy the initial condition yields a = 5 . Hence, the solution of Equation 2.2.6 is
2
− −−− − – –
2
y = −√ 5 − x , −√5 < x < √5.

−−−−− – – −−−−− – –
Figure 2.2.1 : (a) y = √2 − x ,2
−√2 < x < √2 ; (b) y = −√5 − x , 2
−√5 < x < √5
Implicit Solutions of Separable Equations

In Examples 2.2.1 and 2.2.2 we were able to solve the equation H (y) = G(x) + c to obtain explicit formulas for solutions of
the given separable differential equations. As we’ll see in the next example, this isn’t always possible. In this situation we must
broaden our definition of a solution of a separable equation. The next theorem provides the basis for this modification. We
omit the proof, which requires a result from advanced calculus called as the implicit function theorem.
Theorem 2.2.1 : implicit function theorem

Suppose g = g(x) is continous on (a, b) and h = h(y) are continuous on (c, d). Let G be an antiderivative of g on (a, b)
and let H be an antiderivative of h on (c, d). Let x be an arbitrary point in (a, b), let y be a point in (c, d) such that
0 0
h(y ) ≠ 0, and define

0
c = H (y0 ) − G(x0 ). (2.2.10)
Then there’s a function y = y(x) defined on some open interval (a1 , b1 ), where a ≤ a1 < x0 < b1 ≤ b, such that
y(x ) = y
0 and 0
H (y) = G(x) + c (2.2.11)
for a 1 < x < b1 . Therefore y is a solution of the initial value problem

′
h(y)y = g(x), y(x0 ) = x0 . (2.2.12)
It’s convenient to say that Equation 2.2.11 with c arbitrary is an implicit solution of h(y)y = g(x) . Curves defined by ′
Equation 2.2.11 are integral curves of h(y)y = g(x) . If c satisfies Equation 2.2.10, we’ll say that Equation 2.2.11 is an
′
implicit solution of the initial value problem Equation 2.2.12. However, keep these points in mind:
For some choices of c there may not be any differentiable functions y that satisfy Equation 2.2.11.
The function y in Equation 2.2.11 (not Equation 2.2.11 itself) is a solution of h(y)y = g(x) . ′
Example 2.2.3
a. Find implicit solutions of
2x + 1
′
y = . (2.2.13)
4
5y +1
b. Find an implicit solution of

′
2x + 1
y = , y(2) = 1. (2.2.14)
4
5y +1
Solution a
4 ′
(5 y + 1)y = 2x + 1.
Integrating yields the implicit solution

5 2
y +y = x + x + c. (2.2.15)
of Equation 2.2.13.
Solution b
Imposing the initial condition y(2) = 1 in Equation 2.2.15 yields 1 + 1 = 4 + 2 + c , so c = −4 . Therefore
5 2
y +y = x +x −4
is an implicit solution of the initial value problem Equation 2.2.14. Although more than one differentiable function
y = y(x) satisfies Equation 2.2.13 near x = 1 , it can be shown that there is only one such function that satisfies the initial
condition y(1) = 2 . Figure 2.2.2 shows a direction field and some integral curves for Equation 2.2.13.
Figure 2.2.2 : A direction field and integral curves for y ′ 2x+1

=
4
5y +1
Constant Solutions of Separable Equations

An equation of the form
′
y = g(x)p(y)
is separable, since it can be rewritten as

1
′
y = g(x).
p(y)
However, the division by p(y) is not legitimate if p(y) = 0 for some values of y . The next two examples show how to deal
with this problem.
Example 2.2.4

′ 2
y = 2x y . (2.2.16)
Solution
Here we must divide by p(y) = y to separate variables. This isn’t legitimate if y is a solution of Equation 2.2.16 that
2
equals zero for some value of x. One such solution can be found by inspection: y ≡ 0 . Now suppose y is a solution of
Equation 2.2.16 that isn’t identically zero. Since y is continuous there must be an interval on which y is never zero. Since
division by y is legitimate for x in this interval, we can separate variables in Equation 2.2.16 to obtain
2
′
y
= 2x.
2
y

1
2
− =x + c,
y
1
y =− . (2.2.17)
2
x +c
We’ve now shown that if y is a solution of Equation 2.2.16 that is not identically zero, then y must be of the form
Equation 2.2.17. By substituting Equation 2.2.17 into Equation 2.2.16, you can verify that Equation 2.2.17 is a solution
of Equation 2.2.16. Thus, solutions of Equation 2.2.16 are y ≡ 0 and the functions of the form Equation 2.2.17. Note that
the solution y ≡ 0 isn’t of the form Equation 2.2.17 for any value of c .
Figure 2.2.3 shows a direction field and some integral curves for Equation 2.2.16

= 2xy
2
Example 2.2.5
′
1 2
y = x(1 − y ). (2.2.18)
2
Here we must divide by p(y) = 1 − y to separate variables. This isn’t legitimate if y is a solution of Equation 2.2.18
2
that equals ±1 for some value of x. Two such solutions can be found by inspection: y ≡ 1 and y ≡ −1 . Now suppose y is
a solution of Equation 2.2.18 such that 1 − y isn’t identically zero. Since 1 − y is continuous there must be an interval
2 2
on which 1 − y is never zero. Since division by 1 − y is legitimate for x in this interval, we can separate variables in
2 2
Equation 2.2.18 to obtain

′
2y
= −x.
2
y −1
A partial fraction expansion on the left yields

1 1
′
[ − ]y = −x,
y −1 y +1
and integrating yields

2
∣ y −1 ∣ x
ln∣ ∣ =− + k;
∣ y +1 ∣ 2
hence,
∣ y −1 ∣ k
2
−x /2
∣ ∣ =e e .
∣ y +1 ∣
Since y(x) ≠ ±1 for x on the interval under discussion, the quantity (y − 1)/(y + 1) cannot change sign in this interval.
Therefore we can rewrite the last equation as
y −1 2
−x /2
= ce ,
y +1
where c = ±e , depending upon the sign of (y − 1)/(y + 1) on the interval. Solving for y yields
k
2
−x /2
1 + ce
y = . (2.2.19)
2
−x /2
1 − ce
We’ve now shown that if y is a solution of Equation 2.2.18 that is not identically equal to ±1, then y must be as in
Equation 2.2.19. By substituting Equation 2.2.19 into Equation 2.2.18 you can verify that Equation 2.2.19 is a solution
of Equation 2.2.18. Thus, the solutions of Equation 2.2.18 are y ≡ 1 , y ≡ −1 and the functions of the form Equation
2.2.19. Note that the constant solution y ≡ 1 can be obtained from this formula by taking c = 0 ; however, the other
constant solution, y ≡ −1 , cannot be obtained in this way.

Figure 2.2.4 shows a direction field and some integrals for Equation 2.2.18.
2
x(1− y )
=
2
Differences Between Linear and Nonlinear Equations

Theorem 2.1.2 states that if p and f are continuous on (a, b) then every solution of

′
y + p(x)y = f (x)
on (a, b) can be obtained by choosing a value for the constant c in the general solution, and if x is any point in (a, b) and y
0 0
is arbitrary, then the initial value problem

′
y + p(x)y = f (x), y(x0 ) = y0
has a solution on (a, b).

The not true for nonlinear equations. First, we saw in Examples 2.2.4 and 2.2.5 that a nonlinear equation may have solutions
that cannot be obtained by choosing a specific value of a constant appearing in a one-parameter family of solutions. Second, it
is in general impossible to determine the interval of validity of a solution to an initial value problem for a nonlinear equation
by simply examining the equation, since the interval of validity may depend on the initial condition. For instance, in Example
2.2.2 we saw that the solution of
dy x
=− , y(x0 ) = y0
dx y
−−−−−−
is valid on (−a, a) , where a = √x 2
0
+y
0
2
.
Example 2.2.6
′ 2
y = 2x y , y(0) = y0
and determine the interval of validity of the solution.

Solution
First suppose y 0 ≠0 . From Example 2.2.4, we know that y must be of the form
1
y =− . (2.2.20)
2
x +c
Imposing the initial condition shows that c = −1/y . Substituting this into Equation 2.2.20 and rearranging terms yields
0
the solution
y0
y = .
2
1 − y0 x
This is also the solution if y = 0 . If y < 0 , the denominator isn’t zero for any value of x, so the the solution is valid on
0 0
(−∞, ∞) . If y > 0 , the solution is valid only on (−1/ √y , 1/ √y ).

−− − −
0 0 0

2.2E: Separable Equations (Exercises)
Q2.2.1
In Exercises 2.2.1-2.2.6 find all solutions.
2
1. y ′
=
3 x +2x+1
y−2
&
2. (sin x)(sin y) + (cos y)y ′

=0
3. x y ′
+y
2
+y = 0 &
4. y ′
ln |y| + x y = 0
2
(2x+1)y
5. (3y 3
+ 3y cos y + 1)y +
′
1+x
2
=0
6. x 2
yy
′
= (y
2
− 1)
3/2
Q2.2.2
In Exercises 2.2.7-2.2.10 find all solutions. Also, plot a direction field and some integral curves on the indicated rectangular
region.
7. y ′ 2 2
= x (1 + y ); {−1 ≤ x ≤ 1, − 1 ≤ y ≤ 1}
8. y ′ 2
(1 + x ) + xy = 0; {−2 ≤ x ≤ 2, − 1 ≤ y ≤ 1}
9. y ′
= (x − 1)(y − 1)(y − 2); {−2 ≤ x ≤ 2, − 3 ≤ y ≤ 3}
10. (y − 1) 2
y
′
= 2x + 3; {−2 ≤ x ≤ 2, − 2 ≤ y ≤ 5}
Q2.2.3
In Exercises 2.2.11 and 2.2.12 solve the initial value problem.
2
11. y ′
=
x +3x+2
y−2
, y(1) = 4
12. y ′
+ x(y
2
+ y) = 0, y(2) = 1
Q2.2.4
In Exercises 2.2.13-2.2.16 solve the initial value problem and graph the solution.
13. (3y 2
+ 4y)y + 2x + cos x = 0,
′
y(0) = 1
(y+1)(y−1)(y−2)
14. y ′
+
x+1
= 0, y(1) = 0
15. y ′
+ 2x(y + 1) = 0, y(0) = 2
16. y ′
= 2xy(1 + y ),
2
y(0) = 1
Q2.2.5
In Exercises 2.2.17-2.2.23 solve the initial value problem and find the interval of validity of the solution.
17. y ′
(x
2
+ 2) + 4x(y
2
+ 2y + 1) = 0, y(1) = −1
18. y ′
= −2x(y
2
− 3y + 2), y(0) = 3
19. y ′
=
1+2y
2x
, y(2) = 0 &
20. y ′
= 2y − y ,
2
y(0) = 1
21. x + y y ′
= 0, y(3) = −4
22. y ′ 2
+ x (y + 1)(y − 2 )
2
= 0, y(4) = 2
23. (x + 1)(x − 2)y ′

+ y = 0, y(1) = −3

Q2.2.6
2
(1+y )
24. Solve y ′
=
(1+x )
2
explicitly.
−−−− − −−−− −
25. Solve ′
y √1 − x2 + √1 − y 2 = 0 explicitly.
26. Solve y ′
=
cos x
sin y
, y(π) =
π
2
explicitly.
27. Solve the initial value problem

′ 2
y = ay − b y , y(0) = y0 . (2.2E.1)
Discuss the behavior of the solution if a y 0 ≥0 ;by 0 <0 .

28. The population P = P (t) of a species satisfies the logistic equation
′
P = aP (1 − αP ) (2.2E.2)
and P (0) = P 0 >0 . Find P for t > 0 , and find lim t→∞ .
P (t)
29. An epidemic spreads through a population at a rate proportional to the product of the number of people already infected
and the number of people susceptible, but not yet infected. Therefore, if S denotes the total population of susceptible people
and I = I (t) denotes the number of infected people at time t , then
′
I = rI (S − I ), (2.2E.3)
where r is a positive constant. Assuming that I (0) = I , find I (t) for t > 0 , and show that lim
0 t→∞ I (t) = S .
30. The result of Exercise 2.2.29 is discouraging: if any susceptible member of the group is initially infected, then in the long
run all susceptible members are infected! On a more hopeful note, suppose the disease spreads according to the model of
Exercise 2.2.29, but there’s a medication that cures the infected population at a rate proportional to the number of infected
individuals. Now the equation for the number of infected individuals becomes
′
I = rI (S − I ) − qI (A)
where q is a positive constant.

a. Choose r and S positive. By plotting direction fields and solutions of (A) on suitable rectangular grids
R = {0 ≤ t ≤ T , 0 ≤ I ≤ d} (2.2E.4)
in the (t, I )-plane, verify that if I is any solution of (A) such that I (0) > 0 , then lim I (t) = S − q/r if q < rS and t→∞
lim t→∞ I (t) = 0 if q ≥ rS .
b. To verify the experimental results of (a), use separation of variables to solve (A) with initial condition I (0) = I > 0 , and 0
find lim t→∞ I (t) .
31. Consider the differential equation

′ 2
y = ay − b y − q, (A)
where a , b are positive constants, and q is an arbitrary constant. Suppose y denotes a solution of this equation that satisfies the
initial condition y(0) = y . 0
a. Choose a and b positive and q < a 2

/4b . By plotting direction fields and solutions of (A) on suitable rectangular grids
R = {0 ≤ t ≤ T , c ≤ y ≤ d} (B)
in the (t, y)-plane, discover that there are numbers y and y with y < y such that if y > y then lim
1 2 y(t) = y , 1 2 0 1 t→∞ 2
and if y < y then y(t) = −∞ for some finite value of t . (What happens if y = y ?)
0 1 0 1
b. Choose a and b positive and q = a /4b . By plotting direction fields and solutions of (A) on suitable rectangular grids of
2
the form (B), discover that there’s a number y such that if y ≥ y then lim 1 y(t) = y , while if y < y 0then 1 t→∞ 1 0 1
y(t) = −∞ for some finite value of t .
c. Choose positive a , b and q > a /4b . By plotting direction fields and solutions of (A) on suitable rectangular grids of the
2
form (B), discover that no matter what y is, y(t) = −∞ for some finite value of t .
0

d. Verify your results experiments analytically. Start by separating variables in (A) to obtain
′
y
= 1. (2.2E.5)
2
ay − b y −q
To decide what to do next you’ll have to use the quadratic formula. This should lead you to see why there are three cases.
Take it from there! Because of its role in the transition between these three cases, q = a /4b is called a bifurcation value0
2
of q. In general, if q is a parameter in any differential equation, q is said to be a bifurcation value of q if the nature of the
0
solutions of the equation with q < q is qualitatively different from the nature of the solutions with q > q .
0 0
32. By plotting direction fields and solutions of

′ 3
y = qy − y , (2.2E.6)
convince yourself that q 0 =0 is a bifurcation value of q for this equation. Explain what makes you draw this conclusion.
33. Suppose a disease spreads according to the model of Exercise 2.2.29, but there’s a medication that cures the infected
population at a constant rate of q individuals per unit time, where q > 0 . Then the equation for the number of infected
individuals becomes
′
I = rI (S − I ) − q. (2.2E.7)
Assuming that I (0) = I 0 >0 , use the results of Exercise 2.2.31 to describe what happens as t → ∞ .
34. Assuming that p ≢ 0 , state conditions under which the linear equation
′
y + p(x)y = f (x) (2.2E.8)
is separable. If the equation satisfies these conditions, solve it by separation of variables and by the method developed in
Section 2.1.
Q2.2.7
Solve the equations in Exercises 2.2.35-2.2.38 using variation of parameters followed by separation of variables.
−x
35. y ′
+y =
2xe
1+yex
&
6
36. x y ′
− 2y =
y+x
x
2
4x
(x+1)e
37. y ′
−y =
x
2
&
(y+e )
2x
38. y ′
− 2y =
xe
1−ye−2x
39. Use variation of parameters to show that the solutions of the following equations are of the form y = uy1 , where u
satisfies a separable equation u = g(x)p(u) . Find y and g for each equation.

′
1
a. x y + y = h(x)p(xy)
′
y
b. x y − y = h(x)p ( )
′
c. y + y = h(x)p(e y)
′ x
d. x y + ry = h(x)p(x y)
′ r
′
v (x)
e. ′
y +
v(x)
y = h(x)p (v(x)y)

2.3: Existence and Uniqueness of Solutions of Nonlinear Equations
Although there are methods for solving some nonlinear equations, it is impossible to find useful formulas for the solutions of
most. Whether we are looking for exact solutions or numerical approximations, it is useful to know conditions that imply the
existence and uniqueness of solutions of initial value problems for nonlinear equations. In this section we state such a
condition and illustrate it with examples.
Figure 2.3.1 : An open rectangle

Some terminology: an open rectangle R is a set of points (x, y) such that
a <x <b and c <y <d
(Figure 2.3.1). We’ll denote this set by R : {a < x < b, c < y < d} . “Open” means that the boundary rectangle (indicated by
the dashed lines in Figure 2.3.1) is not included in R .
The next theorem gives sufficient conditions for existence and uniqueness of solutions of initial value problems for first order
nonlinear differential equations. We omit the proof, which is beyond the scope of this book.
Theorem 2.3.1 : existence and uniqueness

a. If f is continuous on an open rectangle
R : {a < x < b, c < y < d}
that contains (x 0, y0 ) then the initial value problem

′
y = f (x, y), y(x0 ) = y0 (2.3.1)
has at least one solution on some open subinterval of (a, b) that contains x . 0
b. If both f and f are continuous on R then Equation 2.3.1 has a unique solution on some open subinterval of
y (a, b)
that contains x 0
It’s important to understand exactly what Theorem 2.3.1 says.

(a) is an existence theorem. It guarantees that a solution exists on some open interval that contains x , but provides no 0
information on how to find the solution, or to determine the open interval on which it exists. Moreover, (a) provides no
information on the number of solutions that Equation 2.3.1 may have. It leaves open the possibility that Equation 2.3.1
may have two or more solutions that differ for values of x arbitrarily close to x . We will see in Example 2.3.6 that this
0
can happen.
(b) is a uniqueness theorem. It guarantees that Equation 2.3.1 has a unique solution on some open interval (a,b) that
contains x . However, if (a, b) ≠ (−∞, ∞) , Equation 2.3.1 may have more than one solution on a larger interval that
0
contains (a, b). For example, it may happen that b < ∞ and all solutions have the same values on (a, b), but two solutions
y and y are defined on some interval (a, b ) with b > b , and have different values for b < x < b ; thus, the graphs of
1 2 1 1 1
the y and y “branch off” in different directions at x = b . (See Example 2.3.7 and Figure 2.3.3). In this case, continuity
1 2
implies that y (b) = y (b) (call their common value y ), and y and y are both solutions of the initial value problem
1 2 1 2

¯¯
y = f (x, y), y(b) = ȳ (2.3.2)
that differ on every open interval that contains b . Therefore f or f must have a discontinuity at some point in each open
y
rectangle that contains (b, y), since if this were not so, 2.3.2 would have a unique solution on some open interval that contains
b . We leave it to you to give a similar analysis of the case where a > −∞ .
Example 2.3.1
Consider the initial value problem
2 2
x −y
′
y = , y(x0 ) = y0 . (2.3.3)
2 2
1 +x +y
Since
2 2 2
x −y 2y(1 + 2 x )
f (x, y) = and fy (x, y) = −
1 + x2 + y 2 (1 + x2 + y 2 )2
are continuous for all (x, y), Theorem 2.3.1 implies that if (x 0, y0 ) is arbitrary, then Equation 2.3.3 has a unique solution
on some open interval that contains x .
0
Example 2.3.2
2 2
x −y
′
y = , y(x0 ) = y0 . (2.3.4)
2 2
x +y
Here
2 2 2
x −y 4x y
f (x, y) = and fy (x, y) = −
2 2 2 2 2
x +y (x +y )
are continuous everywhere except at (0, 0). If (x , y ) ≠ (0, 0) , there’s an open rectangle R that contains (x , y ) that
0 0 0 0
does not contain (0, 0). Since f and f are continuous on R , Theorem 2.3.1 implies that if (x , y ) ≠ (0, 0) then
y 0 0
Equation 2.3.4 has a unique solution on some open interval that contains x . 0
Example 2.3.3
x +y
′
y = , y(x0 ) = y0 . (2.3.5)
x −y
Here
x +y 2x
f (x, y) = and fy (x, y) =
2
x −y (x − y)
are continuous everywhere except on the line y = x . If y ≠ x , there’s an open rectangle R that contains (x , y ) that
0 0 0 0
does not intersect the line y = x . Since f and f are continuous on R , Theorem 2.3.1 implies that if y ≠ x , Equation
y 0 0
2.3.5 has a unique solution on some open interval that contains x . 0
Example 2.3.4
In Example 2.2.4, we saw that the solutions of
′ 2
y = 2x y (2.3.6)

are
1
y ≡0 and y =− ,
2
x +c
where c is an arbitrary constant. In particular, this implies that no solution of Equation 2.3.6 other than y ≡0 can equal
zero for any value of x. Show that Theorem 2.3.1b implies this.
We’ll obtain a contradiction by assuming that Equation 2.3.6 has a solution y that equals zero for some value of x, but is 1
not identically zero. If y has this property, there’s a point x such that y (x ) = 0 , but y (x) ≠ 0 for some value of x in
1 0 1 0 1
every open interval that contains x . This means that the initial value problem
0
′ 2
y = 2x y , y(x0 ) = 0 (2.3.7)
has two solutions y ≡ 0 and y = y that differ for some value of

1 x on every open interval that contains x0 . This
contradicts Theorem 2.3.1(b), since in Equation 2.3.6 the functions
2
f (x, y) = 2x y and fy (x, y) = 4xy.
are both continuous for all , which implies that Equation

(x, y) 2.3.7 has a unique solution on some open interval that
contains x .0
Example 2.3.5
10
′ 2/5
y = xy , y(x0 ) = y0 . (2.3.8)
3
a. For what points (x 0, y0 ) does Theorem 2.3.1a imply that Equation 2.3.8 has a solution?
b. For what points (x 0, y0 ) does Theorem 2.3.1b imply that Equation 2.3.8 has a unique solution on some open interval
that contains x ?0
Solution a
Since
10 2/5
f (x, y) = xy
3
is continuous for all (x, y), Theorem 2.3.1 implies that Equation 2.3.8 has a solution for every (x 0, y0 ) .
Solution b
Here
4
−3/5
fy (x, y) = xy
3
is continuous for all (x, y) with y ≠ 0 . Therefore, if y ≠ 0 there’s an open rectangle on which both f and f are
0 y
continuous, and Theorem 2.3.1 implies that Equation 2.3.8 has a unique solution on some open interval that contains x . 0
If y = 0 then f y (x, y) is undefined, and therefore discontinuous; hence, Theorem 2.3.1 does not apply to Equation 2.3.8
if y = 0 .
0
Example 2.3.6
Example 2.3.5 leaves open the possibility that the initial value problem
10
′ 2/5
y = xy , y(0) = 0 (2.3.9)
3
has more than one solution on every open interval that contains x 0 =0 . Show that this is true.

Solution
By inspection, y ≡ 0 is a solution of the differential equation
′
10 2/5
y = xy . (2.3.10)
3
Since y ≡ 0 satisfies the initial condition y(0) = 0 , it is a solution of Equation 2.3.9.

Now suppose y is a solution of Equation 2.3.10 that is not identically zero. Separating variables in Equation 2.3.10 yields
10
−2/5 ′
y y = x
3
on any open interval where y has no zeros. Integrating this and rewriting the arbitrary constant as 5c/3 yields
5 3/5
5 2
y = (x + c).
3 3
Therefore
2 5/3
y = (x + c) . (2.3.11)
Since we divided by y to separate variables in Equation 2.3.10, our derivation of Equation 2.3.11 is legitimate only on
open intervals where y has no zeros. However, Equation 2.3.11 actually defines y for all x, and differentiating Equation
2.3.11 shows that
10 10
′ 2 2/3 2/5
y = x(x + c) = xy , −∞
3 3
−− −−
Therefore Equation 2.3.11 satisfies Equation 2.3.10 on (−∞, ∞) even if c ≤0 , so that y(√|c|) = y(−√|c|) = 0 . In
particular, taking c = 0 in Equation 2.3.11 yields
10/3
y =x
as a second solution of Equation 2.3.9. Both solutions are defined on (−∞, ∞), and they differ on every open interval
that contains x = 0 (Figure 2.3.2). In fact, there are four distinct solutions of Equation 2.3.9 defined on (−∞, ∞) that
0
differ from each other on every open interval that contains x = 0 . Can you identify the other two?
0
Figure 2.3.2 : Two solutions ( y = 0 and y = x 1/2

) of Equation 2.3.9 that differ on every interval containing x 0 = 0
Example 2.3.7
From Example 2.3.5, the initial value problem

′
10 2/5
y = xy , y(0) = −1 (2.3.12)
3
has a unique solution on some open interval that contains x 0 =0 . Find a solution and determine the largest open interval
(a, b) on which it is unique.
Solution
Let y be any solution of Equation 2.3.12. Because of the initial condition y(0) = −1 and the continuity of y , there’s an
open interval I that contains x = 0 on which y has no zeros, and is consequently of the form Equation 2.3.11. Setting
0
x = 0 and y = −1 in Equation 2.3.11 yields c = −1 , so
2 5/3
y = (x − 1) (2.3.13)
for x in I . Therefore every solution of Equation 2.3.12 differs from zero and is given by Equation 2.3.13 on (−1, 1); that
is, Equation 2.3.13 is the unique solution of Equation 2.3.12 on (−1, 1). This is the largest open interval on which
Equation 2.3.12 has a unique solution. To see this, note that Equation 2.3.13 is a solution of Equation 2.3.12 on
(−∞, ∞) . From Exercise 2.2.15, there are infinitely many other solutions of Equation 2.3.12 that differ from Equation
2.3.13 on every open interval larger than (−1, 1). One such solution is
2 5/3
(x − 1) , −1 ≤ x ≤ 1,
y ={
0, |x| > 1.
Figure 2.3.3 : Two solutions of Equation 2.3.12 on (−1,1) that coincide on (−1, 1), but on no larger open interval. (right)
Example 2.3.8
From Example 2.3.5), the initial value problem
10
′ 2/5
y = xy , y(0) = 1 (2.3.14)
3
has a unique solution on some open interval that contains x0 = 0 . Find the solution and determine the largest open
interval on which it is unique.
Solution
Let y be any solution of Equation 2.3.14. Because of the initial condition y(0) = 1 and the continuity of y , there’s an
open interval I that contains x = 0 on which y has no zeros, and is consequently of the form Equation 2.3.11. Setting
0
x = 0 and y = 1 in Equation 2.3.11 yields c = 1 , so
2 5/3
y = (x + 1) (2.3.15)
for x in I . Therefore every solution of Equation 2.3.14 differs from zero and is given by Equation 2.3.15 on (−∞, ∞);
that is, Equation 2.3.15 is the unique solution of Equation 2.3.14 on (−∞, ∞). Figure 2.3.4) shows the graph of this
solution.

Figure 2.3.4 : The unique solution of Equation 2.3.14 .

2.3E: Existence and Uniqueness of Solutions of Nonlinear Equations (Exercises)
Q2.3.1
In Exercises 2.3.1-2.3.13, find all (x , y ) for which Theorem 2.3.1 implies that the initial value problem
0 0
′
y = f (x, y), y(x ) = y has (a) a solution and (b) a unique solution on some open interval that contains x .
0 0 0
2 2
x +y
1. y ′
=
sin x
x
e +y
2. y ′
= 2
x +y
2
3. y ′
= tan xy
2 2
x +y
4. y ′
=
ln xy
5. y ′
= (x
2
+ y )y
2 1/3
6. y ′
= 2xy
7. y ′
= ln(1 + x
2
+y )
2
2x+3y
8. y ′
=
x−4y
9. y ′
= (x
2
+y )
2 1/2
10. y ′
= x(y
2
− 1)
2/3
11. y ′
= (x
2
+y )
2 2
12. y ′
= (x + y )
1/2
tan y
13. y ′
=
x−1
Q2.3.2
14. Apply Theorem 2.3.1 to the initial value problem
′
y + p(x)y = q(x), y(x0 ) = y0 (2.3E.1)
for a linear equation, and compare the conclusions that can be drawn from it to those that follow from Theorem 2.1.2.
15.
a. Verify that the function
2 5/3
(x − 1) , −1 < x < 1,
y ={ (2.3E.2)
0, |x| ≥ 1,
is a solution of the initial value problem

10
′ 2/5
y = xy , y(0) = −1 (2.3E.3)
3
on (−∞, ∞). HINT: You'll need the definition

¯¯
y(x) − y(x̄)
¯′
¯
y (x̄) = lim (2.3E.4)
¯¯
x→x
¯¯
¯
x − x̄
to verify that y satisfies the differential equation at x̄ = ±1 . ¯

¯
b. Verify that if ϵ = 0 or 1 for i = 1 , 2 and a , b > 1 , then the function

i

2 2 5/3
⎧ ϵ1 (x
⎪ −a ) , −∞ < x < −a,
⎪
⎪
⎪
⎪
⎪
⎪
⎪ 0, −a ≤ x ≤ −1,
⎪
⎪
⎪
2 5/3
y =⎨ (x − 1) , −1 < x < 1, (2.3E.5)
⎪
⎪
⎪
⎪
⎪
⎪ 0, 1 ≤ x ≤ b,
⎪
⎪
⎪
⎪
⎩
⎪ 2 2 5/3
ϵ2 (x −b ) , b < x < ∞,
is a solution of the initial value problem of a on (−∞, ∞).

16. Use the ideas developed in Exercise 2.3.15 to find infinitely many solutions of the initial value problem
′ 2/5
y =y , y(0) = 1 (2.3E.6)
on (−∞, ∞).
17. Consider the initial value problem
′ 1/3
y = 3x(y − 1 ) , y(x0 ) = y0 . (A)
a. For what points (x 0, y0 ) does Theorem 2.3.1 imply that (A) has a solution?
b. For what points (x 0, y0 ) does Theorem 2.3.1 imply that (A) has a unique solution on some open interval that contains x ? 0
18. Find nine solutions of the initial value problem

′ 1/3
y = 3x(y − 1 ) , y(0) = 1 (2.3E.7)
that are all defined on (−∞, ∞) and differ from each other for values of x in every open interval that contains x 0 =0 .
19. From Theorem 2.3.1, the initial value problem
′ 1/3
y = 3x(y − 1 ) , y(0) = 9 (2.3E.8)
has a unique solution on an open interval that contains x0 = 0 . Find the solution and determine the largest open interval on
which it is unique.
20.
a. From Theorem 2.3.1, the initial value problem
′ 1/3
y = 3x(y − 1 ) , y(3) = −7 (A)
has a unique solution on some open interval that contains x = 3 . Determine the largest such open interval, and find the
0
solution on this interval.

b. Find infinitely many solutions of (A), all defined on (−∞, ∞).
21. Prove:
a. If
f (x, y0 ) = 0, a < x < b, (A)
andx is in (a, b), then y ≡ y is a solution of

0 0
′
y = f (x, y), y(x0 ) = y0
on (a, b).
b. If f and f are continuous on an open rectangle that contains (x
y 0, y0 ) and (A) holds, no solution of y ′
= f (x, y) other than
y ≡y 0can equal y at any point in (a, b).
0

2.4: Transformation of Nonlinear Equations into Separable Equations
In Section 2.1, we found that the solutions of a linear nonhomogeneous equation
′
y + p(x)y = f (x)
are of the form y = uy , where y is a nontrivial solution of the complementary equation

1 1
′
y + p(x)y = 0 (2.4.1)
and u is a solution of
′
u y1 (x) = f (x).
Note that this last equation is separable, since it can be rewritten as

f (x)
′
u = .
y1 (x)
In this section we’ll consider nonlinear differential equations that are not separable to begin with, but can be solved in a similar
fashion by writing their solutions in the form y = uy , where y is a suitably chosen known function and u satisfies a
1 1
separable equation. We’llsay in this case that we transformed the given equation into a separable equation.
Bernoulli Equations
A Bernoulli equation is an equation of the form
′ r
y + p(x)y = f (x)y , (2.4.2)
where r can be any real number other than 0 or 1. (Note that Equation 2.4.2 is linear if and only if r = 0 or r = 1 .) We can
transform Equation 2.4.2 into a separable equation by variation of parameters: if y is a nontrivial solution of Equation 2.4.1, 1
substituting y = uy into Equation 2.4.2 yields

1
′ ′ r
u y1 + u(y + p(x)y1 ) = f (x)(u y1 ) ,
1
which is equivalent to the separable equation

′
′ r r
u r−1
u y1 (x) = f (x)(y1 (x)) u or = f (x)(y1 (x)) ,
r
u
since y1
′
+ p(x)y1 = 0 .
Example 2.4.1 :
Solve the Bernoulli equation
′ 2
y − y = xy . (2.4.3)
Since y 1 =e
x
is a solution of y ′
−y = 0 , we look for solutions of Equation 2.4.3 in the form y = ue , where x
′ x 2 2x ′ 2 x
u e = xu e or equivalently u = xu e .

′
u
x
= xe ,
2
u

1
x
− = (x − 1)e + c.
u
Hence,

1
u =−
x
(x − 1)e +c
and
1
y =− .
−x
x − 1 + ce
Figure 2.4.1 shows direction field and some integral curves of Equation 2.4.3.

− y = xy
2
Other Nonlinear Equations That Can be Transformed Into Separable Equations

We’ve seen that the nonlinear Bernoulli equation can be transformed into a separable equation by the substitution y = uy if 1
y is suitably chosen. Now let’s discover a sufficient condition for a nonlinear first order differential equation
1
′
y = f (x, y) (2.4.4)
to be transformable into a separable equation in the same way. Substituting y = uy into Equation 2.4.4 yields 1
′ ′
u y1 (x) + u y (x) = f (x, u y1 (x)),
1
′ ′
u y1 (x) = f (x, u y1 (x)) − u y (x). (2.4.5)
1
If
′
f (x, u y1 (x)) = q(u)y (x)
1
for some function q, then Equation 2.4.5 becomes

′ ′
u y1 (x) = (q(u) − u)y (x), (2.4.6)
1
which is separable. After checking for constant solutions u ≡ u such that q(u
0 0) = u0 , we can separate variables to obtain
′ ′
u y (x)
1
= .
q(u) − u y1 (x)
Homogeneous Nonlinear Equations

In the text we will consider only the most widely studied class of equations for which the method of the preceding paragraph
works. Other types of equations appear in Exercises 2.4.44- 2.4.51.

The differential equation Equation 2.4.4 is said to be homogeneous if x and y occur in f in such a way that f (x, y) depends
only on the ratio y/x; that is, Equation 2.4.4 can be written as
′
y = q(y/x), (2.4.7)
where q = q(u) is a function of a single variable. For example,

−y/x
y + xe y
′ −y/x
y = = +e
x x
and
2 2
y + xy − x y 2 y
′
y = =( ) + −1
2
x x x
are of the form Equation 2.4.7, with

−u 2
q(u) = u + e and q(u) = u + u − 1,
respectively. The general method discussed above can be applied to Equation 2.4.7 with y 1 =x (and therefore y
1
′
= 1) . Thus,
substituting y = ux in Equation 2.4.7 yields
′
u x + u = q(u),
and separation of variables (after checking for constant solutions u ≡ u such that q(u 0 0) = u0 ) yields
′
u 1
= .
q(u) − u x
Before turning to examples, we point out something that you may’ve have already noticed: the definition of homogeneous
equation given here is not the same as the definition given in Section 2.1, where we said that a linear equation of the form
′
y + p(x)y = 0
is homogeneous. We make no apology for this inconsistency, since we didn’t create it historically, homogeneous has been used
in these two inconsistent ways. The one having to do with linear equations is the most important. This is the only section of the
book where the meaning defined here will apply.
Since y/x is in general undefined if x = 0 , we’ll consider solutions of nonhomogeneous equations only on open intervals that
do not contain the point x = 0 .
Example 2.4.2
Solve
−y/x
y + xe
′
y = . (2.4.8)
x
Substituting y = ux into Equation 2.4.8 yields

−ux/x
′
ux + xe −u
u x +u = = u +e .
x
Simplifying and separating variables yields

1
u ′
e u = .
x
Integrating yields e u
= ln |x| + c . Therefore u = ln(ln |x| + c) and y = ux = x ln(ln |x| + c) .
Figure 2.4.2 shows a direction field and integral curves for Equation 2.4.8.

−y/x
y + xe
Figure 2.4.2 : A direction field and some integral curves for y ′
=
x
Example 2.4.3
a. Solve
2 ′ 2 2
x y =y + xy − x . (2.4.9)

2 ′ 2 2
x y =y + xy − x , y(1) = 2. (2.4.10)
Solution a
We first find solutions of Equation 2.4.9 on open intervals that don’t contain x = 0 . We can rewrite Equation 2.4.9 as
2 2
y + xy − x
′
y =
2
x
for x in any such interval. Substituting y = ux yields

2 2
(ux ) + x(ux) − x
′ 2
u x +u = =u + u − 1,
2
x
so
′ 2
u x =u − 1. (2.4.11)
By inspection this equation has the constant solutions u ≡ 1 and u ≡ −1 . Therefore y = x and y = −x are solutions of
Equation 2.4.9. If u is a solution of Equation 2.4.11 that does not assume the values ±1 on some interval, separating
variables yields
′
u 1
= ,
2
u −1 x
or, after a partial fraction expansion,

1 1 1 1
′
[ − ]u = .
2 u −1 u +1 x
Multiplying by 2 and integrating yields

∣ u −1 ∣
ln∣ ∣ = 2 ln |x| + k,
∣ u +1 ∣

or
∣ u −1 ∣ k 2
∣ ∣ =e x ,
∣ u +1 ∣
which holds if
u −1 2
= cx (2.4.12)
u +1
where c is an arbitrary constant. Solving for u yields

2
1 + cx
u = .
2
1 − cx
Figure 2.4.3 : A direction field and integral curves for x 2

y
′
= y
2
+ xy − x
2
Therefore
2
x(1 + c x )
y = ux = (2.4.13)
1 − cx2
is a solution of Equation 2.4.10 for any choice of the constant c . Setting c = 0 in Equation 2.4.13 yields the solution
y = x . However, the solution y = −x can’t be obtained from Equation 2.4.13. Thus, the solutions of Equation 2.4.9 on
intervals that don’t contain x = 0 are y = −x and functions of the form Equation 2.4.13.
The situation is more complicated if x = 0 is the open interval. First, note that y = −x satisfies Equation 2.4.9 on
(−∞, ∞) . If c and c are arbitrary constants, the function
1 2
2
x(1+c1 x )
⎧
⎪
2
, a <x <0
1−c1 x
y =⎨ (2.4.14)
2
x(1+c2 x
⎩
⎪ , 0 ≤x <b
1−c2 x2
is a solution of Equation 2.4.9 on (a, b), where

1 1
− if c1 > 0, if c2 > 0,
√c1 √c2
a ={ and b ={
−∞ if c1 ≤ 0, ∞ if c2 ≤ 0.
We leave it to you to verify this. To do so, note that if y is any function of the form Equation 2.4.13 then y(0) = 0 and
y (0) = 1 .
′
Solution b

We could obtain c by imposing the initial condition y(1) = 2 in Equation 2.4.13, and then solving for c . However, it is
easier to use Equation 2.4.12. Since u = y/x, the initial condition y(1) = 2 implies that u(1) = 2 . Substituting this into
Equation 2.4.12 yields c = 1/3 . Hence, the solution of Equation 2.4.10 is
2
x(1 + x /3)
y = .
2
1 − x /3
– –
The interval of validity of this solution is (−√3, √3) . However, the largest interval on which Equation 2.4.10 has a
–
unique solution is (0, √3). To see this, note from Equation 2.4.14 that any function of the form
2
x(1+cx )
⎧
⎪
2
, a <x ≤0
1−cx
y =⎨ 2
(2.4.15)
x(1+x /3) –
⎩
⎪ 0 ≤ x < √3
1−x2 /3
–
is a solution of Equation 2.4.10 on (a, √3) , where a = −1/ √c if c >0 or a = −∞ if c ≤0 . Why does this not
contradict Theorem 2.3.1?
Figure 2.4.4 shows several solutions of the initial value problem Equation 2.4.10 . Note that these solutions coincide on
–
(0, √3).
Figure 2.4.4 : Solutions of x 2

y
′
= y
2 2
+ xy − x , y(1) = 2
In the last two examples we were able to solve the given equations explicitly. However, this is not always possible, as you’ll
see in the exercises.

2.4E: Transformation of Nonlinear Equations into Separable Equations
(Exercises)
Q2.4.1
In Exercises 2.4.1-2.4.4 solve the given Bernoulli equation.
1. y ′
+y = y
2
2. 7x y ′
− 2y = −
x
y
6
3. x 2 ′
y + 2y = 2 e
1/x
y
1/2
4. (1 + x 2 ′
)y + 2xy =
1
(1+x2 )y
Q2.4.2
In Exercises 2.4.5 and 2.4.6 find all solutions. Also, plot a direction field and some integral curves on the indicated rectangular
region.
5. y ′
− xy = x y ;
3 3
{−3 ≤ x ≤ 3, 2 ≤ y ≥ 2}
6. y ′
−
1+x
3x
y =y ;
4
{−2 ≤ x ≤ 2, −2 ≤ y ≤ 2}
Q2.4.3
–
7. y ′
− 2y = x y ,
3
y(0) = 2 √2
8. y ′
− xy = x y
3/2
, y(1) = 4
9. x y ′
+y = x y ,
4 4
y(1) = 1/2
10. y ′
− 2y = 2 y
1/2
, y(0) = 1
11. y ′
− 4y =
48x
y2
, y(0) = 1
Q2.4.4
In Exercises 2.4.12 and 2.4.13 solve the initial value problem and graph the solution.
–
12. x 2 ′
y + 2xy = y ,
3
y(1) = 1/ √2
13. y ′
− y = xy
1/2
, y(0) = 4
Q2.4.5
14. You may have noticed that the logistic equation
′
P = aP (1 − αP ) (2.4E.1)
from Verhulst’s model for population growth can be written in Bernoulli form as
′ 2
P − aP = −aα P . (2.4E.2)
This isn’t particularly interesting, since the logistic equation is separable, and therefore solvable by the method studied in
Section 2.2. So let’s consider a more complicated model, where a is a positive constant and α is a positive continuous function
of t on [0, ∞). The equation for this model is
′ 2
P − aP = −aα(t)P , (2.4E.3)
a non-separable Bernoulli equation.

a. Assuming that P (0) = P 0 >0 , find P for t > 0 .

b. Verify that your result reduces to the known results for the Malthusian model where α = 0 , and the Verhulst model where
α is a nonzero constant.
c. Assuming that
t
−at aτ
lim e ∫ α(τ )e dτ = L (2.4E.4)
t→∞
0
exists (finite or infinite), find lim t→∞ .

P (t)
Q2.4.6
In Exercises 2.4.15-2.4.18 solve the equation explicitly.
y+x
15. y ′
=
x
2
y +2xy
16. y ′
=
x
2
17. x y 3
y
′
=y
4
+x
4
y y
18. y ′
=
x
+ sec
x
Q2.4.7
In Exercises 2.4.19-2.4.21 solve the equation explicitly. Also, plot a direction field and some integral curves on the indicated
rectangular region.
19. x 2
y
′
= xy + x
2
+y ;
2
{−8 ≤ x ≤ 8, −8 ≤ y ≤ 8}
20. xy y ′
=x
2
+ 2y ;
2
{−4 ≤ x ≤ 4, −4 ≤ y ≤ 4}
2
2 2 −( y/x)
2y +x e
21. y ′
=
2xy
; {−8 ≤ x ≤ 8, −8 ≤ y ≤ 8}
Q2.4.8
2
xy+y
22. y ′
=
x
2
, y(−1) = 2
3 3
x +y
23. y ′
=
xy
2
, y(1) = 3
24. xy y ′
+x
2
+y
2
= 0, y(1) = 2
2 2
y −3xy−5 x
25. y ′
=
x
2
, y(1) = −1
26. x 2
y
′
= 2x
2
+y
2
+ 4xy, y(1) = 1
–
27. xy y ′
= 3x
2
+ 4y ,
2
y(1) = √3
Q2.4.9
In Exercises 2.4.28-2.4.34 solve the given homogeneous equation implicitly.
x+y
28. y ′
=
x−y
29. (y ′
x − y)(ln |y| − ln |x|) = x
3 2 2 3
y +2xy +x y+x
30. y ′
= 2
x(y+x)
x+2y
31. y ′
=
2x+y
y
32. y ′
=
y−2x
2 3
xy +2 y
33. y ′
= 3
x +x y+xy
2 2
3 2 3
x +x y+3 y
34. y ′
=
x +3xy
3 2

Q2.4.10
35.
a. Find a solution of the initial value problem
2 ′ 2 2
x y =y + xy − 4 x , y(−1) = 0 (A)
on the interval (−∞, 0). Verify that this solution is actually valid on (−∞, ∞).
b. Use Theorem 2.3.1 to show that (A) has a unique solution on (−∞, 0).
c. Plot a direction field for the differential equation in (A) on a square
{−r ≤ x ≤ r, −r ≤ y ≤ r}, (2.4E.5)
where r is any positive number. Graph the solution you obtained in (a) on this field.
d. Graph other solutions of (A) that are defined on (−∞, ∞).
e. Graph other solutions of (A) that are defined only on intervals of the form (−∞, a) , where is a finite positive number.
36.
′ 2 2
xy y =x − xy + y (A)
implicitly.
b. Plot a direction field for (A) on a square
{0 ≤ x ≤ r, 0 ≤ y ≤ r} (2.4E.6)
where r is any positive number.

c. Let K be a positive integer. (You may have to try several choices for K .) Graph solutions of the initial value problems
kr
′ 2 2
xy y =x − xy + y , y(r/2) = , (2.4E.7)
K
for k = 1 , 2, …, K . Based on your observations, find conditions on the positive numbers x0 and y0 such that the initial
value problem
′ 2 2
xy y =x − xy + y , y(x0 ) = y0 , (B)
has a unique solution (i) on (0, ∞) or (ii) only on an interval (a, ∞), where a > 0 ?
d. What can you say about the graph of the solution of (B) as x → ∞ ? (Again, assume that x 0 >0 and y 0 >0 .)
37.
2 2
′
2y − xy + 2 x
y = (A)
xy + 2x2
implicitly.
b. Plot a direction field for (A) on a square
{−r ≤ x ≤ r, −r ≤ y ≤ r} (2.4E.8)
where r is any positive number. By graphing solutions of (A), determine necessary and sufficient conditions on (x0 , y0 )
such that (A) has a solution on (i) (−∞, 0) or (ii) (0, ∞) such that y(x ) = y . 0 0
38. Follow the instructions of Exercise 2.4.37 for the equation

2 2
′
xy + x +y
y = . (2.4E.9)
xy
39. Pick any nonlinear homogeneous equation y = q(y/x) you like, and plot direction fields on the square
′
{−r ≤ x ≤ r, − r ≤ y ≤ r} , where r > 0 . What happens to the direction field as you vary r ? Why?

40. Prove: If ad − bc ≠ 0 , the equation
ax + by + α
′
y = (2.4E.10)
cx + dy + β
can be transformed into the homogeneous nonlinear equation

dY aX + bY
= (2.4E.11)
dX cX + dY
by the substitution x = X − X 0, y = Y − Y0 , where X and Y are suitably chosen constants.

0 0
Q2.4.11
In Exercises 2.4.21-2.4.43 use a method suggested by Exercise 2.4.40 to solve the given equation implicitly.
−6x+y−3
41. y ′
=
2x−y−1
2x+y+1
42. y ′
=
x+2y−4
−x+3y−14
43. y ′
=
x+y−2
Q2.4.12
In Exercises 2.4.44-2.4.51 find a function y such that the substitution y = uy transforms the given equation into a separable
1 1
equation of the form (2.4.6). Then solve the given equation explicitly.
44. 3x y 2
y
′
=y
3
+x
45. xy y ′
= 3x
6
+ 6y
2
46. x 3
y
′
= 2(y
2
+x y −x )
2 4
47. y ′
=y e
2 −x
+ 4y + 2 e
x
2 2
y +y tan x+tan x
48. y ′
=
2
sin x
49. x(ln x ) 2
y
′
= −4(ln x )
2
+ y ln x + y
2
50. 2x(y + 2√−

x )y
′ − 2
= (y + √x )
51. (y + e
2 2 2
x ′ 2 x 2x
)y = 2x(y + ye +e
Q2.4.13
2 2
2 3x y + 6xy + 2
′
y + y = , y(2) = 2. (2.4E.12)
2
x x (2xy + 3)

4 2 2
3 3x y + 10 x y + 6
′
y + y = , y(1) = 1. (2.4E.13)
3 2
x x (2 x y + 5)
54. Prove: If y is a solution of a homogeneous nonlinear equation y

′
= q(y/x) , so is y1 = y(ax)/a , where a is any nonzero
constant.
55. A generalized Riccati equation is of the form
′ 2
y = P (x) + Q(x)y + R(x)y . (A)
(If R ≡ −1 , (A) is a Riccati equation.) Let y be a known solution and 1 y an arbitrary solution of (A). Let z = y − y1 . Show
that z is a solution of a Bernoulli equation with n = 2 .
Q2.4.14

In Exercises 2.4.56-2.4.59, given that y1 is a solution of the given equation, use the method suggested by Exercise 2.4.55 to
find other solutions.
56. y ′
= 1 + x − (1 + 2x)y + x y
2
;y 1 =1
57. y ′
=e
2x x
+ (1 − 2 e )y + y
2
;y 1 =e
x
58. x y ′
= 2 − x + (2x − 2)y − x y
2
;y 1 =1
59. x y ′
=x
3 2
+ (1 − 2 x )y + x y
2
;y 1 =x

2.5: Exact Equations
In this section it is convenient to write first order differential equations in the form
M (x, y) dx + N (x, y) dy = 0. (2.5.1)
This equation can be interpreted as

dy
M (x, y) + N (x, y) = 0, (2.5.2)
dx
where x is the independent variable and y is the dependent variable, or as

dx
M (x, y) + N (x, y) = 0, (2.5.3)
dy
where y is the independent variable and x is the dependent variable. Since the solutions of Equation 2.5.2 and Equation 2.5.3
will often have to be left in implicit form we will say that F (x, y) = c is an implicit solution of Equation 2.5.1 if every
differentiable function y = y(x) that satisfies F (x, y) = c is a solution of Equation 2.5.2 and every differentiable function
x = x(y) that satisfies F (x, y) = c is a solution of Equation 2.5.3
Here are some examples:

Table 2.5.1 : Examples of Exact Differential Equations in three forms
Equation 2.5.1 Equation 2.5.2 Equation 2.5.3
2 2 3 2 2 3 dy 2 2 dx 3
3x y dx + 2 x y dy = 0 3x y + 2x y = 0 3x y + 2x y = 0
dx dy
2 2 2 2 dy 2 2 dx
(x + y ) dx + 2xy dy = 0 (x + y ) + 2xy = 0 (x +y ) + 2xy = 0
dx dy
dy dx
3y sin x dx − 2xy cosx dy = 0 3y sin x − 2xy cosx = 0 3y sin x − 2xy cosx = 0
dx dy
Note that a separable equation can be written as Equation 2.5.1 as
M (x) dx + N (y) dy = 0.
we will develop a method for solving Equation 2.5.1 under appropriate assumptions on M and N . This method is an
extension of the method of separation of variables. Before stating it we consider an example.
Example 2.5.1
Show that
4 3 2 5
x y +x y + 2xy = c (2.5.4)
is an implicit solution of
3 3 5 4 2 2 4
(4 x y + 2x y + 2y) dx + (3 x y + 5x y + 2x) dy = 0. (2.5.5)
Solution
Regarding y as a function of x and differentiating Equation 2.5.4 implicitly with respect to x yields
dy
3 3 5 4 2 2 4
(4 x y + 2x y + 2y) + (3 x y + 5x y + 2x) = 0.
dx
Similarly, regarding x as a function of y and differentiating Equation 2.5.4 implicitly with respect to y yields
dx
3 3 5 4 2 2 4
(4 x y + 2x y + 2y) + (3 x y + 5x y + 2x) = 0.
dy
Therefore Equation 2.5.4 is an implicit solution of Equation 2.5.5 in either of its two possible interpretations.

You may think Example 2.5.1 is pointless, since concocting a differential equation that has a given implicit solution is not
particularly interesting. However, it illustrates the next important theorem, which we will prove by using implicit
differentiation, as in Example 2.5.1.
Theorem 2.5.1
If F = F (x, y) has continuous partial derivatives F and F , then
x y
F (x, y) = c (2.5.6)
(with c as a constant) is an implicit solution of the differential equation
Fx (x, y) dx + Fy (x, y) dy = 0. (2.5.7)
Proof
Regarding y as a function of x and differentiating Equation 2.5.6implicitly with respect to x yields
dy
Fx (x, y) + Fy (x, y) = 0.
dx
On the other hand, regarding x as a function of y and differentiating Equation 2.5.6implicitly with respect to y yields
dx
Fx (x, y) + Fy (x, y) = 0.
dy
Thus, Equation 2.5.6is an implicit solution of Equation 2.5.7in either of its two possible interpretations.
We will say that the equation
M (x, y) dx + N (x, y) dy = 0 (2.5.8)
is exact on an an open rectangle R if there’s a function F = F (x, y) such F and F are continuous, and
x y
Fx (x, y) = M (x, y) and Fy (x, y) = N (x, y) (2.5.9)
for all (x, y) in R . This usage of “exact” is related to its usage in calculus, where the expression
Fx (x, y) dx + Fy (x, y) dy
(obtained by substituting Equation 2.5.9 into the left side of Equation 2.5.8) is the exact differential of F .
Example 2.5.1 shows that it is easy to solve Equation 2.5.8 if it is exact and we know a function F that satisfies Equation
2.5.9. The important questions are:
Question 1. Given an equation Equation 2.5.8, how can we determine whether it is exact?
Question 2. If Equation 2.5.8 is exact, how do we find a function F satisfying Equation 2.5.9?
To discover the answer to Question 1, assume that there’s a function F that satisfies Equation 2.5.9 on some open rectangle R ,
and in addition that F has continuous mixed partial derivatives F and F . Then a theorem from calculus implies that
xy yx
Fxy = Fyx . (2.5.10)
If Fx =M and F y =N , differentiating the first of these equations with respect to y and the second with respect to x yields
Fxy = My and Fyx = Nx . (2.5.11)
From Equation 2.5.10 and Equation 2.5.11, we conclude that a necessary condition for exactness is that My = Nx . This
motivates the next theorem, which we state without proof.
Theorem 2.5.2 : The Exactness Condition

Suppose M and N are continuous and have continuous partial derivatives M and N on an open rectangle R. Then
y x

M (x, y) dx + N (x, y) dy = 0
is exact on R if and only if

My (x, y) = Nx (x, y) (2.5.12)
for all (x, y) in R..
To help you remember the exactness condition, observe that the coefficients of dx and dy are differentiated in Equation 2.5.12
with respect to the “opposite” variables; that is, the coefficient of dx is differentiated with respect to y , while the coefficient of
dy is differentiated with respect to x.
Example 2.5.2
Show that the equation
2 3
3 x y dx + 4 x dy = 0
is not exact on any open rectangle.

Solution
Here
2 3
M (x, y) = 3 x y and N (x, y) = 4 x
so
2 2
My (x, y) = 3 x andNx (x, y) = 12 x .
Therefore My = Nx on the line x = 0 , but not on any open rectangle, so there’s no function F such that
Fx (x, y) = M (x, y) and F (x, y) = N (x, y) for all (x, y) on any open rectangle.
y
The next example illustrates two possible methods for finding a function F that satisfies the condition F x =M and F y =N
if M dx + N dy = 0 is exact.
Example 2.5.3
Solve
3 3 2 4 2 2
(4 x y + 3 x ) dx + (3 x y + 6 y ) dy = 0. (2.5.13)
Solution (Method 1)
Here
3 3 2 4 2 2
M (x, y) = 4 x y + 3x , N (x, y) = 3 x y + 6y ,
and
3 2
My (x, y) = Nx (x, y) = 12 x y
for all (x, y). Therefore Theorem 2.5.2 implies that there’s a function F such that
3 3 2
Fx (x, y) = M (x, y) = 4 x y + 3x (2.5.14)
and
4 2 2
Fy (x, y) = N (x, y) = 3 x y + 6y (2.5.15)
for all (x, y). To find F , we integrate Equation 2.5.14 with respect to x to obtain
4 3 3
F (x, y) = x y +x + ϕ(y), (2.5.16)

where ϕ(y) is the “constant” of integration. (Here ϕ is “constant” in that it is independent of x, the variable of
integration.) If ϕ is any differentiable function of y then F satisfies Equation 2.5.14. To determine ϕ so that F also
satisfies Equation 2.5.15, assume that ϕ is differentiable and differentiate F with respect to y . This yields
4 2 ′
Fy (x, y) = 3 x y + ϕ (y).
Comparing this with Equation 2.5.15 shows that

′ 2
ϕ (y) = 6 y .
We integrate this with respect to y and take the constant of integration to be zero because we are interested only in finding
some F that satisfies Equation 2.5.14 and Equation 2.5.15. This yields
3
ϕ(y) = 2 y .
Substituting this into Equation 2.5.16 yields

4 3 3 3
F (x, y) = x y +x + 2y . (2.5.17)
Now Theorem 2.5.1 implies that

4 3 3 3
x y +x + 2y =c
is an implicit solution of Equation 2.5.13. Solving this for y yields the explicit solution
1/3
3
c −x
y =( ) .
4
2 +x
Solution (Method 2)
Instead of first integrating Equation 2.5.14 with respect to x, we could begin by integrating Equation 2.5.15 with respect
to y to obtain
4 3 3
F (x, y) = x y + 2y + ψ(x), (2.5.18)
where ψ is an arbitrary function of x. To determine ψ , we assume that ψ is differentiable and differentiate F with respect
to x, which yields
3 3 ′
Fx (x, y) = 4 x y + ψ (x).

′ 2
ψ (x) = 3 x .
Integrating this and again taking the constant of integration to be zero yields
3
ψ(x) = x .

Figure 2.5.1 : A direction field and integral curves for (4x 3
y
3 2 4
+ 3x ) dx + (3x y
2 2
+ 6y ) dy = 0
Substituting this into Equation 2.5.18 yields Equation 2.5.17.

Figure 2.5.1 shows a direction field and some integral curves of Equation 2.5.13.
Here’s a summary of the procedure used in Method 1 of this example. You should summarize procedure used in Method 2.
HOWTO: Procedure For Solving An Exact Equation (Method 1)

Step 1. Check that the equation
M (x, y) dx + N (x, y) dy = 0 (2.5.19)
satisfies the exactness condition M y = Nx . If not, don’t go further with this procedure.
Step 2. Integrate
∂F (x, y)
= M (x, y)
∂x
with respect to x to obtain

F (x, y) = G(x, y) + ϕ(y), (2.5.20)
where G is an antiderivative of M with respect to x, and ϕ is an unknown function of y .

Step 3. Differentiate Equation 2.5.20 with respect to y to obtain
∂F (x, y) ∂G(x, y)
′
= + ϕ (y).
∂y ∂y
Step 4. Equate the right side of this equation to N and solve for ϕ ; thus, ′
∂G(x, y) ∂G(x, y)
′ ′
+ ϕ (y) = N (x, y), so ϕ (y) = N (x, y) − .
∂y ∂y
Step 5. Integrate ϕ with respect to y , taking the constant of integration to be zero, and substitute the result in Equation
′
2.5.20 to obtain F (x, y).
Step 6. Set F (x, y) = c to obtain an implicit solution of Equation 2.5.19. If possible, solve for y explicitly as a
function of x.
It’s a common mistake to omit Step 6 in the procedure above. However, it is important to include this step, since F isn’t itself a
solution of Equation 2.5.19. Many equations can be conveniently solved by either of the two methods used in Example 2.5.3.

However, sometimes the integration required in one approach is more difficult than in the other. In such cases we choose the
approach that requires the easier integration.
Example 2.5.4
Solve the equation
xy xy 2 xy
(y e tan x + e sec x) dx + x e tan x dy = 0 (2.5.21)
Solution
We leave it to you to check that M y = Nx on any open rectangle where tan x and sec x are defined. Here we must find a
function F such that
xy xy 2
Fx (x, y) = y e tan x + e sec x (2.5.22)
and
xy
Fy (x, y) = x e tan x. (2.5.23)
It’s difficult to integrate Equation 2.5.22 with respect to x, but easy to integrate Equation 2.5.23 with respect to y . This
yields
xy
F (x, y) = e tan x + ψ(x). (2.5.24)
Differentiating this with respect to x yields

xy xy 2 ′
Fx (x, y) = y e tan x + e sec x + ψ (x).
Comparing this with Equation 2.5.22 shows that ′

ψ (x) = 0 . Hence, ψ is a constant, which we can take to be zero in
Equation 2.5.24, and
xy
e tan x = c,
is an implicit solution of Equation 2.5.21.
Attempting to apply our procedure to an differential equation that is not exact will lead to failure in Step 4, since the function
∂G
N −
∂y
will not be independent of x if My ≠ Nx , and therefore cannot be the derivative of a function of y alone. Example 2.5.5
illustrates this.
Example 2.5.5
Verify that the equation
2 2 3
3x y dx + 6 x y dy = 0 (2.5.25)
is not exact, and show that the procedure for solving exact equations fails when applied to Equation 2.5.25.
Solution
Here
2 2
My (x, y) = 6 x y and Nx (x, y) = 18 x y,
so Equation 2.5.25 is not exact. Nevertheless, let us try to find a function F such that
2 2
Fx (x, y) = 3 x y (2.5.26)
and

3
Fy (x, y) = 6 x y. (2.5.27)
Integrating Equation 2.5.26 with respect to x yields

3 2
F (x, y) = x y + ϕ(y),
and differentiating this with respect to y yields

3 ′
Fy (x, y) = 2 x y + ϕ (y).
For this equation to be consistent with Equation 2.5.27,

3 3 ′
6 x y = 2 x y + ϕ (y),
or
′ 3
ϕ (y) = 4 x y.
This is a contradiction, since ϕ must be independent of x. Therefore the procedure fails.

′

2.5E: Exact Equations (Exercises)
Q2.5.1
In Exercises 2.5.1-2.5.17 determine which equations are exact and solve them.
1. 6x 2
y
2
dx + 4 x y dy = 0
3
2. (3y cos x + 4x e x 2 x
+ 2 x e ) dx + (3 sin x + 3) dy = 0
3. 14x 2
y
3
dx + 21 x y
2 2
dy = 0
4. (2x − 2y 2
) dx + (12 y
2
− 4xy) dy = 0
5. (x + y ) 2
dx + (x + y )
2
dy = 0
6. (4x + 7y) dx + (3x + 4y) dy = 0

7. (−2y 2
sin x + 3 y
3
− 2x) dx + (4y cos x + 9x y ) dy = 0
2
8. (2x + y) dx + (2y + 2x) dy = 0

9. (3x 2
+ 2xy + 4 y ) dx + (x
2 2
+ 8xy + 18y) dy = 0
10. (2x 2
+ 8xy + y ) dx + (2 x
2 2 3
+ x y /3) dy = 0
11. ( 1
x
+ 2x) dx + (
1
y
+ 2y) dy = 0
12. (y sin xy + x y 2
cos xy) dx + (x sin xy + x y
2
cos xy) dy = 0
y dy
13. x dx
3/2
+
3/2
=0
( x2 +y 2 ) ( x2 +y 2 )
14. (e x
(x y
2 2 2
+ 2x y ) + 6x) dx + (2 x y e
2 x
+ 2) dy = 0
15. (x
2 2
2 x +y 2 3 x +y 2
e (2 x + 3) + 4x) dx + (x e − 12 y ) dy = 0
16. (e xy 4 3
(x y + 4 x ) + 3y) dx + (x e
5 xy
+ 3x) dy = 0
17. (3x 2 3
cos xy − x y sin xy + 4x) dx + (8y − x
4
sin xy) dy = 0
Q2.5.2
18. (4x 3
y
2 2
− 6 x y − 2x − 3) dx + (2 x y − 2 x ) dy = 0,
4 3
y(1) = 3
19. (−4y cos x + 4 sin x cos x + sec 2

x) dx + (4y − 4 sin x) dy = 0, y(π/4) = 0
20. (y 3
− 1)e
x
dx + 3 y (e
2 x
+ 1) dy = 0, y(0) = 0
21. (sin x − y sin x − 2 cos x) dx + cos x dy = 0, y(0) = 1
22. (2x − 1)(y − 1) dx + (x + 2)(x − 3) dy = 0, y(1) = −1
Q2.5.3
23. Solve the exact equation
(7x + 4y) dx + (4x + 3y) dy = 0.
Plot a direction field and some integral curves for this equation on the rectangle
{−1 ≤ x ≤ 1, −1 ≤ y ≤ 1}.
24. Solve the exact equation

x 4 2 3 2 4 x
e (x y + 4x y + 1) dx + (2 x y e + 2y) dy = 0.

Plot a direction field and some integral curves for this equation on the rectangle
{−2 ≤ x ≤ 2, −1 ≤ y ≤ 1}.
25. Plot a direction field and some integral curves for the exact equation
3 4 4 3
(x y + x) dx + (x y + y) dy = 0
on the rectangle {−1 ≤ x ≤ 1, −1 ≤ y ≤ 1} . (See Exercise 2.5.37(a)).

26. Plot a direction field and some integral curves for the exact equation
2
(3 x + 2y) dx + (2y + 2x) dy = 0
on the rectangle {−2 ≤ x ≤ 2, −2 ≤ y ≤ 2} . (See Exercise 2.5.37(b)).

27.
a. Solve the exact equation
3 4 4 3
(x y + 2x) dx + (x y + 3y) dy = 0 (A)
implicitly.
b. For what choices of (x 0, y0 ) does Theorem 2.3.1 imply that the initial value problem
3 4 4 3
(x y + 2x) dx + (x y + 3y) dy = 0, y(x0 ) = y0 , (B)
has a unique solution on an open interval (a, b) that contains x ? 0
c. Plot a direction field and some integral curves for (A) on a rectangular region centered at the origin. What is the interval of
validity of the solution of (B)?
28.
a. Solve the exact equation
2 2
(x + y ) dx + 2xy dy = 0 (A)
implicitly.
b. For what choices of (x 0, y0 ) does Theorem 2.3.1 imply that the initial value problem
2 2
(x + y ) dx + 2xy dy = 0, y(x0 ) = y0 , (B)
has a unique solution y = y(x) on some open interval (a, b) that contains x ? 0
c. Plot a direction field and some integral curves for (A). From the plot determine, the interval (a, b) of b, the monotonicity
properties (if any) of the solution of (B), and lim y(x) and lim y(x).
x→a+ x→b−
29. Find all functions M such that the equation is exact.

a. M (x, y) dx + (x − y ) dy = 0
2 2
b. M (x, y) dx + 2xy sin x cos y dy = 0

c. M (x, y) dx + (e − e sin x) dy = 0
x y
30. Find all functions N such that the equation is exact.

a. (x y + 2xy + 3y ) dx + N (x, y) dy = 0
3 2 2
b. (ln xy + 2y sin x) dx + N (x, y) dy = 0

c. (x sin x + y sin y) dx + N (x, y) dy = 0
31. Suppose M , N , and their partial derivatives are continuous on an open rectangle R , and G is an antiderivative of M with
respect to x; that is,
∂G
= M.
∂x
Show that if M y ≠ Nx in R then the function

∂G
N −
∂y
is not independent of x.
32. Prove: If the equations M 1 dx + N1 dy = 0 and M 2 dx + N2 dy = 0 are exact on an open rectangle R , so is the equation
(M1 + M2 ) dx + (N1 + N2 ) dy = 0.
33. Find conditions on the constants A , B , C , and D such that the equation
(Ax + By) dx + (C x + Dy) dy = 0
is exact.
34. Find conditions on the constants A , B , C , D, E , and F such that the equation
2 2 2 2
(Ax + Bxy + C y ) dx + (Dx + Exy + F y ) dy = 0
is exact.
35. Suppose M and N are continuous and have continuous partial derivatives M and N that satisfy the exactness condition y x
M =N
y on an open rectangle R . Show that if (x, y) is in R and
x
x y
F (x, y) = ∫ M (s, y0 ) ds + ∫ N (x, t) dt,

x0 y0
then F x =M and F
y =N .
36. Under the assumptions of Exercise 2.5.35, show that
y x
F (x, y) = ∫ N (x0 , s) ds + ∫ M (t, y) dt.

y x0
0
37. Use the method suggested by Exercise 2.5.35, with (x 0, y0 ) = (0, 0) , to solve the these exact equations:
a. (x y + x) dx + (x y + y) dy = 0
3 4 4 3
b. (x + y ) dx + 2xy dy = 0
2 2
c. (3x + 2y) dx + (2y + 2x) dy = 0

2
′
2 2xy
y + y =− , y(1) = −2.
2 2
x x + 2x y + 1

4 3
3 2 x (4 x − 3y)
′
y − y = , y(1) = 1.
5 3
x 3x + 3x + 2y

2
x
′ −x
2 3x + 2ye
y + 2xy = −e ( ), y(0) = −1.
2
2x + 3yex
41. Rewrite the separable equation

′
h(y)y = g(x) (A)
as an exact equation
M (x, y) dx + N (x, y) dy = 0. (B)
Show that applying the method of this section to (B) yields the same solutions that would be obtained by applying the method
of separation of variables to (A)

42. Suppose all second partial derivatives of M = M (x, y) and N = N (x, y) are continuous and M dx + N dy = 0 and
−N dx + M dy = 0 are exact on an open rectangle R . Show that M +M xx =N yy +N xx =0 on R .
yy
43. Suppose all second partial derivatives of F = F (x, y) are continuous and F + F = 0 on an open rectangle R . (A
xx yy
function with these properties is said to be harmonic; see also Exercise 2.5.42.) Show that −F dx + F dy = 0 is exact on R ,
y x
and therefore there’s a function G such that G = −F and G = F in R . (A function G with this property is said to be a
x y y x
harmonic conjugate of F .)
44. Verify that the following functions are harmonic, and find all their harmonic conjugates. (See Exercise 2.5.43.)
a. x − y
2 2
b. e cos y
x
c. x − 3x y
3 2
d. cos x cosh y
e. sin x cosh y

2.6: Integrating Factors
In Section 2.5 we saw that if M , N , M and N are continuous and M
y x y = Nx on an open rectangle R then
M (x, y) dx + N (x, y) dy = 0 (2.6.1)
is exact on R . Sometimes an equation that isn’t exact can be made exact by multiplying it by an appropriate function. For
example,
2
(3x + 2 y ) dx + 2xy dy = 0 (2.6.2)
is not exact, since M y (x, y) = 4y ≠ Nx (x, y) = 2y in Equation 2.6.2. However, multiplying Equation 2.6.2 by x yields
2 2 2
(3 x + 2x y ) dx + 2 x y dy = 0, (2.6.3)
which is exact, since M (x, y) = N (x, y) = 4xy in Equation

y x 2.6.3 . Solving Equation 2.6.3 by the procedure given in
Section 2.5 yields the implicit solution
3 2 2
x +x y = c.
A function μ = μ(x, y) is an integrating factor for Equation 2.6.1 if

μ(x, y)M (x, y) dx + μ(x, y)N (x, y) dy = 0 (2.6.4)
is exact. If we know an integrating factor μ for Equation 2.6.1, we can solve the exact equation Equation 2.6.4 by the method
of Section 2.5. It would be nice if we could say that Equation 2.6.1 and Equation 2.6.4 always have the same solutions, but
this isn’t so. For example, a solution y = y(x) of Equation 2.6.4 such that μ(x, y(x)) = 0 on some interval a < x < b could
fail to be a solution of 2.6.1 (Exercise 2.6.1), while Equation 2.6.1 may have a solution y = y(x) such that μ(x, y(x)) isn’t
even defined (Exercise 2.6.2). Similar comments apply if y is the independent variable and x is the dependent variable in
Equation 2.6.1 and Equation 2.6.4. However, if μ(x, y) is defined and nonzero for all (x, y), Equation 2.6.1 and Equation
2.6.4 are equivalent; that is, they have the same solutions.
Finding Integrating Factors

By applying Theorem 2.5.2 (with M and N replaced by μM and μN ), we see that Equation 2.6.4 is exact on an open
rectangle R if μM , μN , (μM ) , and (μN ) are continuous and
y x
∂ ∂
(μM ) = (μN ) or, equivalently, μy M + μMy = μx N + μNx
∂y ∂x
on R . It’s better to rewrite the last equation as
μ(My − Nx ) = μx N − μy M , (2.6.5)
which reduces to the known result for exact equations; that is, if M y = Nx then Equation 2.6.5 holds with μ = 1 , so Equation
2.6.1 is exact.
You may think Equation 2.6.5 is of little value, since it involves partial derivatives of the unknown integrating factor μ , and
we haven’t studied methods for solving such equations. However, we’ll now show that Equation 2.6.5 is useful if we restrict
our search to integrating factors that are products of a function of x and a function of y ; that is, μ(x, y) = P (x)Q(y). We’re
not saying that every equation M dx + N dy = 0 has an integrating factor of this form; rather, we are saying that some
equations have such integrating factors.We’llnow develop a way to determine whether a given equation has such an integrating
factor, and a method for finding the integrating factor in this case.
If μ(x, y) = P (x)Q(y), then μ x (x,
′
y) = P (x)Q(y) and μ y (x, y) = P (x)Q (y)
′
, so Equation 2.6.5 becomes
′ ′
P (x)Q(y)(My − Nx ) = P (x)Q(y)N − P (x)Q (y)M , (2.6.6)
or, after dividing through by P (x)Q(y),

′ ′
P (x) Q (y)
My − Nx = N − M. (2.6.7)
P (x) Q(y)
Now let
′ ′
P (x) Q (y)
p(x) = and q(y) = ,
P (x) Q(y)
so Equation 2.6.7 becomes

My − Nx = p(x)N − q(y)M . (2.6.8)
We obtained Equation 2.6.8 by assuming that M dx + N dy = 0 has an integrating factor μ(x, y) = P (x)Q(y). However,
we can now view Equation 2.6.7 differently: If there are functions p = p(x) and q = q(y) that satisfy Equation 2.6.8 and we
define
∫ p(x) dx ∫ q(y) dy
P (x) = ±e and Q(y) = ±e , (2.6.9)
then reversing the steps that led from Equation 2.6.6 to Equation 2.6.8 shows that μ(x, y) = P (x)Q(y) is an integrating
factor for M dx + N dy = 0 . In using this result, we take the constants of integration in Equation 2.6.9 to be zero and choose
the signs conveniently so the integrating factor has the simplest form.
There’s no simple general method for ascertaining whether functions p = p(x) and q = q(y) satisfying Equation 2.6.8 exist.
However, the next theorem gives simple sufficient conditions for the given equation to have an integrating factor that depends
on only one of the independent variables x and y , and for finding an integrating factor in this case.
Theorem 2.6.1
Let M , N , M y, and N be continuous on an open rectangle R. Then:
x
(a) If (M y − Nx )/N is independent of y on R and we define

My − Nx
p(x) =
N
then
∫ p(x) dx
μ(x) = ±e (2.6.10)
is an integrating factor for

M (x, y) dx + N (x, y) dy = 0 (2.6.11)
on R.
(b) If (N x − My )/M is independent of x on R and we define
Nx − My
q(y) = ,
M
then
∫ q(y) dy
μ(y) = ±e (2.6.12)
is an integrating factor for Equation 2.6.11 on R.
Proof
(a) If (M y − Nx )/N is independent of y, then Equation 2.6.8holds with p = (M y − Nx )/N and q ≡ 0 . Therefore
∫ p(x) dx ∫ q(y) dy 0
P (x) = ±e and Q(y) = ±e = ±e = ±1,
so Equation 2.6.10is an integrating factor for Equation 2.6.11on R.

(b) If (N − M )/M is independent of x then eqrefeq:2.6.8 holds with p ≡ 0 and q = (N
x y x − My )/M , and a
similar argument shows that Equation 2.6.12is an integrating factor for Equation 2.6.11on R.
The next two examples show how to apply Theorem 2.6.1.
Example 2.6.1
Find an integrating factor for the equation
3 3 3 2 2 2
(2x y − 2x y − 4x y + 2x) dx + (3 x y + 4y) dy = 0 (2.6.13)
and solve the equation.

Solution
In Equation 2.6.13
3 3 3 2 2 2
M = 2x y − 2x y − 4x y + 2x, N = 3 x y + 4y,
and
2 3 2 2 3 2
My − Nx = (6x y − 6x y − 8xy) − 6x y = −6 x y − 8xy,
so Equation 2.6.13 isn’t exact. However,

3 2
My − Nx 6x y + 8xy
=− = −2x
2 2
N 3x y + 4y
is independent of y , so Theorem 2.6.1 (a) applies with p(x) = −2x. Since
2
∫ p(x) dx = − ∫ 2x dx = −x ,
is an integrating factor. Multiplying Equation 2.6.13 by μ yields the exact equation

2
−x
μ(x) = e
2 2
−x 3 3 3 2 −x 2 2
e (2x y − 2x y − 4x y + 2x) dx + e (3 x y + 4y) dy = 0. (2.6.14)
To solve this equation, we must find a function F such that

2
−x 3 3 3 2
Fx (x, y) = e (2x y − 2x y − 4x y + 2x) (2.6.15)
and
2
−x 2 2
Fy (x, y) = e (3 x y + 4y). (2.6.16)
Integrating Equation 2.6.16 with respect to y yields

2
−x 2 3 2
F (x, y) = e (x y + 2 y ) + ψ(x). (2.6.17)
Differentiating this with respect to x yields

2
−x 3 3 3 2 ′
Fx (x, y) = e (2x y − 2x y − 4x y ) + ψ (x).
Comparing this with Equation shows that ; therefore, we can let in Equation
2 2
′ −x −x
2.6.15 ψ (x) = 2x e ψ(x) = −e
2.6.17 and conclude that
2
−x 2 2
e (y (x y + 2) − 1) = c
is an implicit solution of Equation 2.6.14. It is also an implicit solution of Equation 2.6.13.

Figure 2.6.1 shows a direction field and some integral curves for Equation 2.6.13

Figure 2.6.1 : A direction field and integral curves for (2xy 3
− 2x y
3 3
− 4xy
2 2
+ 2x) dx + (3x y
2
+ 4y) dy = 0
Example 2.6.2
Find an integrating factor for
3 2 2 2 3
2x y dx + (3 x y +x y + 1)dy = 0 (2.6.18)

Solution
In Equation 2.6.18,
3 2 2 2 3
M = 2x y , N = 3x y +x y + 1,
and
2 2 3 3
My − Nx = 6 x − (6x y + 2x y ) = −2x y ,
so Equation 2.6.18 isn't exact. Moreover,

3
My − Nx 2xy
=−
N 3 x2 y 2 + x2 y 2 + 1
is not independent of y , so Theorem 2.6.1(a) does not apply. However, Theorem 2.6.1(b) does apply, since
3
Nx − My 2xy
= =1
M 2xy 3
is not independent of x, so we can take q(y) = 1 . Since
∫ q(y)dy = ∫ dy = y,
μ(y) = e
y
is an integrating factor. Multiplying Equation 2.6.18 by μ yields the exact equation.
3 y 2 2 2 3 y
2x y e dx + (3 x y +x y + 1)e dy = 0. (2.6.19)
To solve this equation, we must find a function F such that

3 y
Fx (x, y) = 2x y e (2.6.20)
and
2 2 2 3 y
Fy (x, y) = (3 x y +x y + 1)e . (2.6.21)
Integrating Equation 2.6.20 with respect to x yields

2 3 y
F (x, y) = x y e + ϕ(y) (2.6.22)
Differentiating this with respect to y yields

2 2 2 3 y ′
Fy = (3 x y + x y )e + ϕ (y)
and comparing this with Equation 2.6.21 shows that φ 0 (y) = e y . Therefore we set φ(y) = e y in Equation 2.6.22 and
conclude that
2 3 y
(x y + 1)e =c
is an implicit solution of 2.6.19. It is also an implicit solution of 2.6.18. Figure 2.6.2 shows a direction field and some
integral curves for 2.6.18.
Figure 2.6.2 : A direction field and integral curves for (2xy 3 y

e dx + (3x y
2 2 2
+x y
3 y
+ 1) e dy = 0
Theorem 2.6.1 does not apply in the next example, but the more general argument that led to Theorem 2.6.1 provides an
integrating factor.
Example 2.6.3
Find an integrating factor for
2 2
(3xy + 6 y )dx + (2 x + 9xy)dy = 0 (2.6.23)

Solution
In Equation 2.6.23
2 2
M = 3xy + 6 y , N = 2x + 9xy,
and
My − Nx = (3x + 12y) − (4x + 9y) = −x + 3y.
Therefore
My − Nx −x + 3y Nx − My x − 3y
= and =
M 3xy + 6y 2 N 2 x2 + 9xy
so Theorem 2.6.1 does not apply. Following the more general argument that led to Theorem 2.6.1, we look for functions
p = p(x) and q = q(y) such that
My − Nx = p(x)N − q(y)M ;
that is,

2 2
−x + 3y = p(x)(2 x + 9xy) − q(y)(3xy + 6 y ).
Since the left side contains only first degree terms in x and y , we rewrite this equation as
xp(x)(2x + 9y) − yq(y)(3x + 6y) = −x + 3y.
This will be an identity if
xp(x) = A andyq(y) = B, (2.6.24)
where A and B are constants such that
−x + 3y = A(2x + 9y) − B(3x + 6y),
or, equivalently,
−x + 3y = (2A − 3B)x + (9A − 6B)y.
Equating the coefficients of x and y on both sides shows that the last equation holds for all (x, y) if
2A − 3B = −1
9A − 6B = 3
which has the solution A = 1, B = 1. Therefore Equation 2.6.24 implies that

1 1
p(x) = and q(y) = .
x y
Since
∫ p(x)dx = ln |x| and ∫ q(y)dy = ln |y|,
we can let P (x) = x and Q(y) = y ; hence, µ(x, y) = xy is an integrating factor. Multiplying Equation 2.6.23 by µ yields
the exact equation
2 2 3 3 2 2
(3 x y + 6x y )dx + (2 x y + 9 x y )dy = 0.
Figure 2.6.3 : A direction field and integral curves for (3xy + 6y 2

)dx + (2x
2
+ 9xy)dy = 0
We leave it to you to use the method of Section 2.5 to show that this equation has the implicit solution
3 2 2 3
x y + 3x y = c. (2.6.25)
This is also an implicit solution of Equation 2.6.23. Since x ≡ 0 and y ≡ 0 satisfy Equation 2.6.25, you should check to
see that x ≡ 0 and y ≡ 0 are also solutions of Equation 2.6.23. (Why is it necesary to check this?) Figure 2.6.3 shows a
direction field and integral curves for Equation 2.6.23. See Exercise 2.6.28 for a general discussion of equations like
Equation 2.6.23.

Example 2.6.4
The separable equation
6
−ydx + (x + x )dy = 0 (2.6.26)
can be converted to the exact equation

dx dy
− + =0 (2.6.27)
6
x +x y
by multiplying through by the integrating factor

1
μ(x, y) = .
6
y(x + x )
However, to solve Equation 2.6.27 by the method of Section 2.5 we would have to evaluate the nasty integral
dx
∫ .
6
x +x
Instead, we solve Equation 2.6.26 explicitly for y by finding an integrating factor of the form µ(x, y) = x a
y
b
.
Figure 2.6.4 : A direction field and integral curves for −ydx + (x + x 6

)dy = 0
Solution
In Equation 2.6.26
6
M = −y, N = x + x ,
and
5 5
My − Nx = −1 − (1 + 6 x ) = −2 − 6 x .
We look for functions p = p(x) and q = q(y) such that
My − Nx = p(x)N − q(y)M ;
that is,
5 6
−2 − 6 x = p(x)(x + x ) + q(y)y. (2.6.28)
The right side will contain the term −6x if p(x) = −6/x. Then Equation 2.6.28 becomes
5
5 5
−2 − 6 x = −6 − 6 x + q(y)y,
so q(y) = 4/y . Since

6 1
∫ p(x) dx = − ∫ dx = −6 ln |x| = ln ,
6
x x
and
4 4
∫ q(y) dy = ∫ dy = 4 ln |y| = ln y ,
y
we can take P (x) = x and Q(y) = y , which yields the integrating factor
−6 4
μ(x, y) = x
−6
y
4
. Multiplying Equation
2.6.26 by μ yields the exact equation
5 4
y y
4
− dx + ( + y ) dy = 0.
6 5
x x
We leave it to you to use the method of the Section 2.5 to show that this equation has the implicit solution
y 5
5
( ) +y = k.
x
Solving for y yields

1/5 5 −1/5
y =k x(1 + x ) ,
which we rewrite as
5 −1/5
y = cx(1 + x )
by renaming the arbitrary constant. This is also a solution of Equation 2.6.26.


2.6E: Integrating Factors (Exercises)
Q2.6.1
1.
a. Verify that μ(x, y) = y is an integrating factor for
1
ydx + (2x + ) dy = 0 (A)
y
on any open rectangle that does not intersect the x axis or, equivalently, that
2
y dx + (2xy + 1)dy = 0 (B)
is exact on any such rectangle.

b. Verify that y ≡ 0 is a solution of (B), but not of (A).
c. Show that
y(xy + 1) = c (C)
is an implicit solution of (B), and explain why every differentiable function y = y(x) other than y ≡ 0 that satisfies (C) is
also a solution of (A).
2.
a. Verify that μ(x, y) = 1/(x − y) is an integrating factor for
2
2 2
−y dx + x dy = 0 (A)
on any open rectangle that does not intersect the line y = x or, equivalently, that
2 2
y x
− dx + dy = 0 (B)
(x − y)2 (x − y)2
is exact on any such rectangle.

b. Use Theorem 2.2.1 to show that
xy
=c (C)
(x − y)
is an implicit solution of (B), and explain why it is also an implicit solution of (A)
c. Verify that y = x is a solution of (A), even though it can’t be obtained from (C).
Q2.6.2
In Exercises 2.6.3-2.6.16 find an integrating factor; that is a function of only one variable, and solve the given equation.
3. ydx − xdy = 0
4. 3x 2
ydx + 2 x dy = 0
3
5. 2y 3 2
dx + 3 y dy = 0
6. (5xy + 2y + 5)dx + 2xdy = 0

7. (xy + x + 2y + 1) dx + (x + 1) dy = 0
8. (27x y 2 3 2
+ 8 y ) dx + (18 x y + 12x y ) dy = 0
2
9. (6x y 2 2
+ 2y) dx + (12 x y + 6x + 3) dy = 0
10. y 2
dx + (x y
2
+ 3xy +
1
y
) dy = 0
11. (12x 3 2 2
y + 24 x y ) dx + (9 x
4 3
+ 32 x y + 4y) dy = 0

12. (x 2
y + 4xy + 2y) dx + (x
2
+ x) dy = 0
13. −y dx + (x 4
− x) dy = 0
14. cos x cos y dx + (sin x cos y − sin x sin y + y) dy = 0

15. (2xy + y 2
) dx + (2xy + x
2 2
− 2x y
2 3
− 2x y ) dy = 0
16. y sin y dx + x(sin y − y cos y) dy = 0
Q2.6.3
In Exercises 2.6.17-2.6.23 find an integrating factor of the form μ(x, y) = P (x)Q(y) and solve the given equation.
17. y(1 + 5 ln |x|) dx + 4x ln |x| dy = 0
18. (αy + γxy) dx + (βx + δxy) dy = 0
19. (3x 2
y
3
−y
2
+ y) dx + (−xy + 2x) dy = 0
20. 2y dx + 3(x 2 2
+ x y ) dy = 0
3
21. (a cos xy − y sin xy) dx + (b cos xy − x sin xy) dy = 0

22. x4
y
4
dx + x y
5 3
dy = 0
23. y(x cos x + 2 sin x) dx + x(y + 1) sin x dy = 0
Q2.6.4
In Exercises 2.6.24-2.6.27 find an integrating factor and solve the equation. Plot a direction field and some integral curves for
the equation in the indicated rectangular region.
24. (x 4
y
3
+ y) dx + (x y
5 2
− x) dy = 0; {−1 ≤ x ≤ 1, −1 ≤ y ≤ 1}
25. (3xy + 2y 2
+ y) dx + (x
2
+ 2xy + x + 2y) dy = 0; {−2 ≤ x ≤ 2, −2 ≤ y ≤ 2}
26. (12xy + 6y 3
) dx + (9 x
2 2
+ 10x y ) dy = 0; {−2 ≤ x ≤ 2, −2 ≤ y ≤ 2}
27. (3x 2
y
2
+ 2y) dx + 2x dy = 0; {−4 ≤ x ≤ 4, −4 ≤ y ≤ 4}
Q2.6.5
28. Suppose a , b , c , and d are constants such that ad − bc ≠ 0 , and let m and n be arbitrary real numbers. Show that
m n+1 m+1 n
(ax y + by ) dx + (c x + dx y ) dy = 0 (2.6E.1)
has an integrating factor μ(x, y) = x α

y
β
.
29. Suppose M , N , M , and N are continuous for all (x, y), and μ = μ(x, y) is an integrating factor for
x y
M (x, y)dx + N (x, y)dy = 0. (A)
Assume that and μx μy are continuous for all (x, y), and suppose y = y(x) is a differentiable function such that
μ(x, y(x)) = 0 and μ x (x, y(x)) ≠ 0 for all x in some interval I . Show that y is a solution of (A) on I .
30. According to Theorem 2.1.2, the general solution of the linear nonhomogeneous equation
′
y + p(x)y = f (x) (A)
is
y = y1 x (c + ∫ f (x)/ y1 (x)dx) , (B)
where y is any nontrivial solution of the complementary equation y + p(x)y = 0 . In this exercise we obtain this conclusion
1
′
in a different way. You may find it instructive to apply the method suggested here to solve some of the exercises in Section 2.1.
a. Rewrite (A) as
[p(x)y − f (x)]dx + dy = 0, (C)

and show that μ = ±e is an integrating factor for (C).
∫ p(x) dx
b. Multiply (A) through by μ = ±e and verify that the resulting equation can be rewritten as
∫ p(x) dx
′
(μ(x)y ) = μ(x)f (x). (2.6E.2)
Then integrate both sides of this equation and solve for y to show that the general solution of (A) is
1
y = (c + ∫ f (x)μ(x) dx) . (2.6E.3)
μ(x)
Why is this form of the general solution equivalent to (B)?

CHAPTER OVERVIEW
3: NUMERICAL METHODS
In this chapter we study numerical methods for solving a first-order differential equation
3.1: EULER'S METHOD

This section deals with Euler's method, which is really too crude to be of much use in practical
applications. However, its simplicity allows for an introduction to the ideas required to understand
the better methods discussed in the other two sections.
3.1E: EULER’S METHOD (EXERCISES)

3.2: THE IMPROVED EULER METHOD AND RELATED METHODS
Euler’s method implies that we can achieve arbitrarily accurate results with Euler’s method by
simply choosing the step size sufficiently small. However, this isn’t a good idea, for two reasons. (1) After a certain point decreasing
the step size will increase roundoff errors to the point where the accuracy will deteriorate rather than improve. (2)The expensive part
of the computation is the evaluation of the solution. This section discusses improvements on Euler’s method.
3.2E: THE IMPROVED EULER METHOD AND RELATED METHODS (EXERCISES)

3.3: THE RUNGE-KUTTA METHOD
This section deals with the Runge-Kutta method, perhaps the most widely used method for numerical solution of differential
equations.
3.3E: THE RUNGE-KUTTA METHOD (EXERCISES)
1 9/16/2020
3.1: Euler's Method
If an initial value problem
′
y = f (x, y), y(x0 ) = y0 (3.1.1)
cannot be solved analytically, it is necessary to resort to numerical methods to obtain useful approximations to a solution of
Equation 3.1.1. we will consider such methods in this chapter.
We’re interested in computing approximate values of the solution of Equation 3.1.1 at equally spaced points x0 , x1 , …,
x = b in an interval [ x , b]. Thus,
n 0
xi = x0 + ih, i = 0, 1, … , n,
where
b − x0
h = .
n
we will denote the approximate values of the solution at these points by y , y , …, y ; thus, y is an approximation to y(x ).
0 1 n i i
we will call
ei = y(xi ) − yi
the error at the ith step. Because of the initial condition y(x 0) = y0 , we will always have e 0 =0 . However, in general e i ≠0
if i > 0 .
We encounter two sources of error in applying a numerical method to solve an initial value problem:
The formulas defining the method are based on some sort of approximation. Errors due to the inaccuracy of the
approximation are called truncation errors.
Computers do arithmetic with a fixed number of digits, and therefore make errors in evaluating the formulas defining the
numerical methods. Errors due to the computer’s inability to do exact arithmetic are called roundoff errors.
Since a careful analysis of roundoff error is beyond the scope of this book, we will consider only truncation errors.
Euler’s Method
The simplest numerical method for solving Equation 3.1.1 is Euler’s method. This method is so crude that it is seldom used in
practice; however, its simplicity makes it useful for illustrative purposes. Euler’s method is based on the assumption that the
tangent line to the integral curve of Equation 3.1.1 at (x , y(x )) approximates the integral curve over the interval [x , x ].
i i i i+1
Since the slope of the integral curve of Equation 3.1.1 at (x , y(x )) is y (x ) = f (x , y(x )) , the equation of the tangent line
i i
′
i i i
to the integral curve at (x , y(x )) is

i i
y = y(xi ) + f (xi , y(xi ))(x − xi ). (3.1.2)
Setting x = x i+1 = xi + h in Equation 3.1.2 yields
yi+1 = y(xi ) + hf (xi , y(xi )) (3.1.3)
as an approximation to y(x i+1 ) . Since y(x

0) = y0 is known, we can use Equation 3.1.3 with i = 0 to compute
y1 = y0 + hf (x0 , y0 ).
However, setting i = 1 in Equation 3.1.3 yields
y2 = y(x1 ) + hf (x1 , y(x1 )),
which isn’t useful, since we don’t know y(x ). Therefore we replace y(x
1 1) by its approximate value y and redefine1
y2 = y1 + hf (x1 , y1 ).
Having computed y , we can compute

2

y3 = y2 + hf (x2 , y2 ).
In general, Euler’s method starts with the known value y(x0 ) = y0 and computes y1 , y2 , …, yn successively by with the
formula
yi+1 = yi + hf (xi , yi ), 0 ≤ i ≤ n − 1. (3.1.4)
The next example illustrates the computational procedure indicated in Euler’s method.
Example 3.1.1
Use Euler’s method with h = 0.1 to find approximate values for the solution of the initial value problem
′ 3 −2x
y + 2y = x e , y(0) = 1 (3.1.5)
at x = 0.1, 0.2, 0.3.

Solution
We rewrite Equation 3.1.5 as
′ 3 −2x
y = −2y + x e , y(0) = 1,
which is of the form Equation 3.1.1, with

3 −2x
f (x, y) = −2y + x e , x0 = 0, and y0 = 1.
Euler’s method yields

y1 = y0 + hf (x0 , y0 )
= 1 + (0.1)f (0, 1) = 1 + (0.1)(−2) = 0.8,
y2 = y1 + hf (x1 , y1 )
3 −0.2
= 0.8 + (0.1)f (0.1, 0.8) = 0.8 + (0.1) (−2(0.8) + (0.1 ) e ) = 0.640081873,
y3 = y2 + hf (x2 , y2 )
3 −0.4
= 0.640081873 + (0.1) (−2(0.640081873) + (0.2 ) e ) = 0.512601754.
We’ve written the details of these computations to ensure that you understand the procedure. However, in the rest of the
examples as well as the exercises in this chapter, we will assume that you can use a programmable calculator or a computer to
carry out the necessary computations.
Examples Illustrating The Error in Euler’s Method
Example 3.1.2
Use Euler’s method with step sizes h = 0.1 , h = 0.05, and h = 0.025 to find approximate values of the solution of the
initial value problem
′ 3 −2x
y + 2y = x e , y(0) = 1
at x = 0 , 0.1, 0.2, 0.3, …, 1.0. Compare these approximate values with the values of the exact solution
−2x
e
4
y = (x + 4), (3.1.6)
4
which can be obtained by the method of Section 2.1. (Verify.)

Table 3.1.1 shows the values of the exact solution Equation 3.1.6 at the specified points, and the approximate values of
the solution at these points obtained by Euler’s method with step sizes h = 0.1 , h = 0.05, and h = 0.025. In examining
this table, keep in mind that the approximate values in the column corresponding to h = 0.05 are actually the results of 20
steps with Euler’s method. We haven’t listed the estimates of the solution obtained for x = 0.05, 0.15, …, since there’s

nothing to compare them with in the column corresponding to h = 0.1 . Similarly, the approximate values in the column
corresponding to h = 0.025 are actually the results of 40 steps with Euler’s method.
x h = 0.1 h = 0.05 h = 0.025 Exact
0.0 1.000000000 1.000000000 1.000000000 1.000000000
0.1 0.800000000 0.810005655 0.814518349 0.818751221

0.2 0.640081873 0.656266437 0.663635953 0.670588174
0.3 0.512601754 0.532290981 0.541339495 0.549922980
0.4 0.411563195 0.432887056 0.442774766 0.452204669
0.5 0.332126261 0.353785015 0.363915597 0.373627557
0.6 0.270299502 0.291404256 0.301359885 0.310952904
0.7 0.222745397 0.242707257 0.252202935 0.261398947
0.8 0.186654593 0.205105754 0.213956311 0.222570721
0.9 0.159660776 0.176396883 0.184492463 0.192412038
1.0 0.139778910 0.154715925 0.162003293 0.169169104
Table 3.1.1 : Numerical solution of y ′

+ 2y = x e
3 −2x
, y(0) = 1 , by Euler’s method.
You can see from Table 3.1.1 that decreasing the step size improves the accuracy of Euler’s method. For example,
⎧ 0.0293withh = 0.1,
⎪
yexact (1) − yapprox (1) ≈ ⎨ 0.0144 with h = 0.05,

⎩
⎪
0.0071 with h = 0.025.
Based on this scanty evidence, you might guess that the error in approximating the exact solution at a fixed value of x by
Euler’s method is roughly halved when the step size is halved. You can find more evidence to support this conjecture by
examining Table 3.1.2, which lists the approximate values of y −y at x = 0.1, 0.2, …, 1.0.
exact approx
x h = 0.1 h = 0.05 h = 0.0.25
0.1 0.0187 0.0087 0.0042
0.2 0.0305 0.0143 0.0069

0.3 0.0373 0.0176 0.0085
0.4 0.0406 0.0193 0.0094
0.5 0.0415 0.0198 0.0097
0.6 0.0406 0.0195 0.0095
0.7 0.0386 0.0186 0.0091
0.8 0.0359 0.0174 0.0086
0.9 0.0327 0.0160 0.0079
1.0 0.0293 0.0144 0.0071
Table 3.1.2 : Errors in approximate solutions of y ′

+ 2y = x e
3 −2x
, y(0) = 1 , obtained by Euler’s method.
Example 3.1.3
Tables 3.1.2 and 3.1.4 show analogous results for the nonlinear initial value problem
′ 2 2
y = −2 y + xy + x , y(0) = 1, (3.1.7)
except in this case we cannot solve Equation 3.1.7 exactly. The results in the “Exact” column were obtained by using a
more accurate numerical method known as the Runge-Kutta method with a small step size. They are exact to eight

decimal places.
x h = 0.1 h = 0.05 h = 0.025 Exact
0.0 1.000000000 1.000000000 1.000000000 1.000000000
0.1 0.800000000 0.821375000 0.829977007 0.837584494

0.2 0.681000000 0.707795377 0.719226253 0.729641890
0.3 0.605867800 0.633776590 0.646115227 0.657580377
0.4 0.559628676 0.587454526 0.600045701 0.611901791
0.5 0.535376972 0.562906169 0.575556391 0.587575491
0.6 0.529820120 0.557143535 0.569824171 0.581942225
0.7 0.541467455 0.568716935 0.581435423 0.593629526
0.8 0.569732776 0.596951988 0.609684903 0.621907458
0.9 0.614392311 0.641457729 0.654110862 0.666250842
1.0 0.675192037 0.701764495 0.714151626 0.726015790

= −2 y
2 2
+ xy + x , y(0) = 1 , by Euler’s method.
Since we think it is important in evaluating the accuracy of the numerical methods that we will be studying in this chapter,
we often include a column listing values of the exact solution of the initial value problem, even if the directions in the
example or exercise don’t specifically call for it. If quotation marks are included in the heading, the values were obtained
by applying the Runge-Kutta method in a way that’s explained in Section 3.3. If quotation marks are not included, the
values were obtained from a known formula for the solution. In either case, the values are exact to eight places to the right
of the decimal point.
x h = 0.1 h = 0.05 h = 0.025
0.1 0.0376 0.0162 0.0076
0.2 0.0486 0.0218 0.0104

0.3 0.0517 0.0238 0.0115
0.4 0.0523 0.0244 0.0119
0.5 0.0522 0.0247 0.0121
0.6 0.0521 0.0248 0.0121
0.7 0.0522 0.0249 0.0122
0.8 0.0522 0.0250 0.0122
0.9 0.0519 0.0248 0.0121
1.0 0.0508 0.0243 0.0119
Table 3.1.4 : Errors in approximate solutions of y ′

= −2 y
2 2
+ xy + x , y(0) = 1 , obtained by Euler’s method.
Truncation Error in Euler’s Method

Consistent with the results indicated in Tables 3.1.1 - 3.1.4, we will now show that under reasonable assumptions on f there’s
a constant K such that the error in approximating the solution of the initial value problem
′
y = f (x, y), y(x0 ) = y0 ,
at a given point b > x by Euler’s method with step size h = (b − x

0 0 )/n satisfies the inequality
|y(b) − yn | ≤ Kh,
where K is a constant independent of n .

There are two sources of error (not counting roundoff) in Euler’s method:
1. The error committed in approximating the integral curve by the tangent line Equation 3.1.2 over the interval [x , x ]. i i+1
2. The error committed in replacing y(x ) by y in Equation 3.1.2 and using Equation 3.1.4 rather than Equation 3.1.2 to
i i
compute y . i+1
Euler’s method assumes that y defined in Equation 3.1.2 is an approximation to

i+1 y(xi+1 ) . We call the error in this
approximation the local truncation error at the ith step, and denote it by T ; thus, i
Ti = y(xi+1 ) − y(xi ) − hf (xi , y(xi )). (3.1.8)
we will now use Taylor’s theorem to estimate T , assuming for simplicity that f , f , and f are continuous and bounded for all
i x y
(x, y). Then y exists and is bounded on [x , b]. To see this, we differentiate
′′
0
′
y (x) = f (x, y(x))
to obtain
′′ ′
y (x) = fx (x, y(x)) + fy (x, y(x))y (x)
= fx (x, y(x)) + fy (x, y(x))f (x, y(x)).
Since we assumed that f , f and f are bounded, there’s a constant M such that
x y
′
| fx (x, y(x)) + fy (x, y(x))y (x)| ≤ M x0 < x < b
which implies that

′′
|y (x)| ≤ M , x0 < x < b (3.1.9)
Since x i+1 = xi + h , Taylor’s theorem implies that

2
h
′ ′′ ~
y(xi+1 ) = y(xi ) + h y (xi ) + y (xi ),
2
where x
~
is some number between x and x
i i i+1 . Since y ′
(xi ) = f (xi , y(xi )) this can be written as
2
h
′′ ~
y(xi+1 ) = y(xi ) + hf (xi , y(xi )) + y (xi ),
2
or, equivalently,
2
h
′′ ~
y(xi+1 ) − y(xi ) − hf (xi , y(xi )) = y (xi ).
2

2
h ′′ ~
Ti = y (xi ).
2
Recalling Equation 3.1.9, we can establish the bound

2
Mh
| Ti | ≤ , 1 ≤ i ≤ n. (3.1.10)
2
Although it may be difficult to determine the constant M , what is important is that there’s an M such that Equation 3.1.10
holds. We say that the local truncation error of Euler’s method is of order h , which we write as O(h ). 2 2
Note that the magnitude of the local truncation error in Euler’s method is determined by the second derivative y of the ′′
solution of the initial value problem. Therefore the local truncation error will be larger where |y | is large, or smaller where ′′
| y | is small.
′′
Since the local truncation error for Euler’s method is O(h ), it is reasonable to expect that halving h reduces the local
2
truncation error by a factor of 4. This is true, but halving the step size also requires twice as many steps to approximate the
solution at a given point. To analyze the overall effect of truncation error in Euler’s method, it is useful to derive an equation
relating the errors

ei+1 = y(xi+1 ) − yi+1 and ei = y(xi ) − yi .
To this end, recall that

y(xi+1 ) = y(xi ) + hf (xi , y(xi )) + Ti (3.1.11)
and
yi+1 = yi + hf (xi , yi ). (3.1.12)
Subtracting Equation 3.1.12 from Equation 3.1.11 yields

ei+1 = ei + h [f (xi , y(xi )) − f (xi , yi )] + Ti . (3.1.13)
The last term on the right is the local truncation error at the ith step. The other terms reflect the way errors made at previous
steps affect e . Since |T | ≤ M h /2 , we see from Equation 3.1.13 that
i+1 i
2
2
Mh
| ei+1 | ≤ | ei | + h|f (xi , y(xi )) − f (xi , yi )| + . (3.1.14)
2
Since we assumed that f is continuous and bounded, the mean value theorem implies that
y
∗ ∗
f (xi , y(xi )) − f (xi , yi ) = fy (xi , y )(y(xi ) − yi ) = fy (xi , y )ei ,
i i
where y is between y and y(x ). Therefore

∗
i
i i
|f (xi , y(xi )) − f (xi , yi )| ≤ R| ei |
for some constant R . From this and Equation 3.1.14,

2
Mh
| ei+1 | ≤ (1 + Rh)| ei | + , 0 ≤ i ≤ n − 1. (3.1.15)
2
For convenience, let C = 1 + Rh . Since e 0 = y(x0 ) − y0 = 0 , applying Equation 3.1.15 repeatedly yields
2
Mh
| e1 | ≤
2
2 2
Mh Mh
| e2 | ≤ C | e1 | + ≤ (1 + C )
2 2
2 2
Mh 2
Mh
| e3 | ≤ C | e2 | + ≤ (1 + C + C )
2 2
⋮
2 2
Mh n−1
Mh
| en | ≤ C | en−1 | + ≤ (1 + C + ⋯ + C ) . (3.1.16)
2 2
Recalling the formula for the sum of a geometric series, we see that
n n
1 −C (1 + Rh ) −1
n−1
1 +C +⋯ +C = =
1 −C Rh
(since C = 1 + Rh ). From this and Equation 3.1.16,

n
(1 + Rh ) −1 Mh
|y(b) − yn | = | en | ≤ . (3.1.17)
R 2
Since Taylor’s theorem implies that

rh
1 + Rh < e
(verify),
n nRh R(b−x0 )
(1 + Rh ) <e =e (since nh = b − x0 ).
This and Equation 3.1.17 imply that

|y(b) − yn | ≤ Kh, (3.1.18)
with
R(b−x0 )
e −1
K =M .
2R
Because of Equation 3.1.18 we say that the global truncation error of Euler’s method is of order h , which we write as O(h) .
Semilinear Equations and Variation of Parameters

An equation that can be written in the form
′
y + p(x)y = h(x, y) (3.1.19)
with p ≢ 0 is said to be semilinear. (Of course, Equation 3.1.19 is linear if h is independent of y .) One way to apply Euler’s
method to an initial value problem
′
y + p(x)y = h(x, y), y(x0 ) = y0 (3.1.20)
for Equation 3.1.19 is to think of it as

′
y = f (x, y), y(x0 ) = y0 ,
where
f (x, y) = −p(x)y + h(x, y).
However, we can also start by applying variation of parameters to Equation 3.1.20, as in Sections 2.1 and 2.4; thus, we write
the solution of Equation 3.1.20 as y = uy , where y is a nontrivial solution of the complementary equation y + p(x)y = 0 .
1 1
′
Then y = uy is a solution of Equation 3.1.20 if and only if u is a solution of the initial value problem
1
′
u = h(x, u y1 (x))/ y1 (x), u(x0 ) = y(x0 )/ y1 (x0 ). (3.1.21)
We can apply Euler’s method to obtain approximate values u , u , …, u of this initial value problem, and then take
0 1 n
yi = ui y1 (xi )
as approximate values of the solution of Equation 3.1.20. we will call this procedure the Euler semilinear method.
The next two examples show that the Euler and Euler semilinear methods may yield drastically different results.
Example 3.1.2
In Example 3.1.7 we had to leave the solution of the initial value problem
′
y − 2xy = 1, y(0) = 3 (3.1.22)
in the form
x
2 2
x −t
y =e (3 + ∫ e dt) (3.1.23)
0
because it was impossible to evaluate this integral exactly in terms of elementary functions. Use step sizes h = 0.2 ,
h = 0.1 , and h = 0.05 to find approximate values of the solution of Equation 3.1.22 at x = 0 , 0.2, 0.4, 0.6, …, 2.0 by
(a) Euler’s method; (b) the Euler semilinear method.

Solution a
Rewriting Equation 3.1.22 as
′
y = 1 + 2xy, y(0) = 3 (3.1.24)
and applying Euler’s method with f (x, y) = 1 + 2xy yields the results shown in Table 3.1.5. Because of the large
differences between the estimates obtained for the three values of h , it would be clear that these results are useless even if
the “exact” values were not included in the table.

x h = 0.2 h = 0.1 h = 0.05 Exact
0.0 3.000000000 3.000000000 3.000000000 3.000000000
0.2 3.200000000 3.262000000 3.294348537 3.327851973

0.4 3.656000000 3.802028800 3.881421103 3.966059348
0.6 4.440960000 4.726810214 4.888870783 5.067039535
0.8 5.706790400 6.249191282 6.570796235 6.936700945
1.0 7.732963328 8.771893026 9.419105620 10.184923955
1.2 11.026148659 13.064051391 14.405772067 16.067111677
1.4 16.518700016 20.637273893 23.522935872 27.289392347
1.6 25.969172024 34.570423758 41.033441257 50.000377775
1.8 42.789442120 61.382165543 76.491018246 98.982969504
2.0 73.797840446 115.440048291 152.363866569 211.954462214

− 2xy = 1, y(0) = 3 , with Euler’s method.
It’s easy to see why Euler’s method yields such poor results. Recall that the constant M in Equation 3.1.10 – which plays
an important role in determining the local truncation error in Euler’s method – must be an upper bound for the values of
the second derivative y of the solution of the initial value problem Equation 3.1.22 on (0, 2). The problem is that y
′′ ′′
assumes very large values on this interval. To see this, we differentiate Equation 3.1.24 to obtain
′′ ′ 2
y (x) = 2y(x) + 2x y (x) = 2y(x) + 2x(1 + 2xy(x)) = 2(1 + 2 x )y(x) + 2x,
where the second equality follows again from Equation 3.1.24. Since Equation 3.1.23 implies that y(x) > 3e x
if x > 0 ,
2
′′ 2 x
y (x) > 6(1 + 2 x )e + 2x, x > 0.
For example, letting x = 2 shows that y ′′

(2) > 2952 .
Solution b
Since y = e is a solution of the complementary equation y , we can apply the Euler semilinear method to
2
x ′
1 − 2xy = 0
Equation 3.1.22, with

2 2
x ′ −x
y = ue and u =e , u(0) = 3.
The results listed in Table 3.1.6 are clearly better than those obtained by Euler’s method.
x h = 0.2 h = 0.1 h = 0.05 Exact
0.0 3.000000000 3.000000000 3.000000000 3.000000000
0.2 3.330594477 3.329558853 3.328788889 3.327851973

0.4 3.980734157 3.974067628 3.970230415 3.966059348
0.6 5.106360231 5.087705244 5.077622723 5.067039535
0.8 7.021003417 6.980190891 6.958779586 6.936700945
1.0 10.350076600 10.269170824 10.227464299 10.184923955
1.2 16.381180092 16.226146390 16.147129067 16.067111677
1.4 27.890003380 27.592026085 27.441292235 27.289392347
1.6 51.183323262 50.594503863 50.298106659 50.000377775
1.8 101.424397595 100.206659076 99.595562766 98.982969504
2.0 217.301032800 214.631041938 213.293582978 211.954462214

− 2xy = 1, y(0) = 3 , by the Euler semilinear method.

We cannot give a general procedure for determining in advance whether Euler’s method or the semilinear Euler method will
produce better results for a given semilinear initial value problem Equation 3.1.19. As a rule of thumb, the Euler semilinear
method will yield better results than Euler’s method if |u | is small on [x , b], while Euler’s method yields better results if |u |
′′
0
′′
is large on [x , b]. In many cases the results obtained by the two methods don’t differ appreciably. However, we propose the an
0
intuitive way to decide which is the better method: Try both methods with multiple step sizes, as we did in Example
[example:3.1.4}, and accept the results obtained by the method for which the approximations change less as the step size
decreases.
Example 3.1.5
Applying Euler’s method with step sizes h = 0.1 , h = 0.05, and h = 0.025 to the initial value problem
′
x
y − 2y = , y(1) = 7 (3.1.25)
2
1 +y
on [1, 2] yields the results in Table 3.1.7.
x h = 0.1 h = 0.05 h = 0.025 Exact
1.0 7.000000000 7.000000000 7.000000000 7.000000000
1.1 8.402000000 8.471970569 8.510493955 8.551744786

1.2 10.083936450 10.252570169 10.346014101 10.446546230
1.3 12.101892354 12.406719381 12.576720827 12.760480158
1.4 14.523152445 15.012952416 15.287872104 15.586440425
1.5 17.428443554 18.166277405 18.583079406 19.037865752
1.6 20.914624471 21.981638487 22.588266217 23.253292359
1.7 25.097914310 26.598105180 27.456479695 28.401914416
1.8 30.117766627 32.183941340 33.373738944 34.690375086
1.9 36.141518172 38.942738252 40.566143158 42.371060528
2.0 43.369967155 47.120835251 49.308511126 51.752229656

− 2y = x/(1 + y ), y(1) = 7
2
, by Euler’s method.
Applying the Euler semilinear method with
−2x
xe
2x ′ −2
y = ue and u = , u(1) = 7 e
2 4x
1 +u e
yields the results in Table 3.1.8. Since the latter are clearly less dependent on step size than the former, we conclude that
the Euler semilinear method is better than Euler’s method for Equation 3.1.25. This conclusion is supported by comparing
the approximate results obtained by the two methods with the “exact” values of the solution.
x h = 0.1 h = 0.05 h = 0.025 Exact
1.0 7.000000000 7.000000000 7.000000000 7.000000000
1.1 8.552262113 8.551993978 8.551867007 8.551744786

1.2 10.447568674 10.447038547 10.446787646 10.446546230
1.3 12.762019799 12.761221313 12.760843543 12.760480158
1.4 15.588535141 15.587448600 15.586934680 15.586440425
1.5 19.040580614 19.039172241 19.038506211 19.037865752
1.6 23.256721636 23.254942517 23.254101253 23.253292359
1.7 28.406184597 28.403969107 28.402921581 28.401914416
1.8 34.695649222 34.692912768 34.691618979 34.690375086

x h = 0.1 h = 0.05 h = 0.025 Exact
1.9 42.377544138 42.374180090 42.372589624 42.371060528

2.0 51.760178446 51.756054133 51.754104262 51.752229656
Table 3.1.8 : Numerical solution of y ′ 2

− 2y = x/(1 + y ), y(1) = 7 , by the Euler semilinear method.
Example 3.1.6
Applying Euler’s method with step sizes h = 0.1 , h = 0.05, and h = 0.025 to the initial value problem
′ 2 2
y + 3x y = 1 + y , y(2) = 2 (3.1.26)
on [2, 3] yields the results in Table 3.1.9. Applying the Euler semilinear method with
3 3 3
−x ′ x 2 −2x 8
y = ue and u =e (1 + u e ), u(2) = 2 e
yields the results in Table 3.1.10. Noting the close agreement among the three columns of Table 3.1.9 (at least for larger
values of x) and the lack of any such agreement among the columns of Table 3.1.10, we conclude that Euler’s method is
better than the Euler semilinear method for Equation 3.1.26. Comparing the results with the exact values supports this
conclusion.
x h = 0.1 h = 0.05 h = 0.025 Exact
2.0 2.000000000 2.000000000 2.000000000 2.000000000
2.1 0.100000000 0.493231250 0.609611171 0.701162906

2.2 0.068700000 0.122879586 0.180113445 0.236986800
2.3 0.069419569 0.070670890 0.083934459 0.103815729
2.4 0.059732621 0.061338956 0.063337561 0.068390786
2.5 0.056871451 0.056002363 0.056249670 0.057281091
2.6 0.050560917 0.051465256 0.051517501 0.051711676
2.7 0.048279018 0.047484716 0.047514202 0.047564141
2.8 0.042925892 0.043967002 0.043989239 0.044014438
2.9 0.042148458 0.040839683 0.040857109 0.040875333
3.0 0.035985548 0.038044692 0.038058536 0.038072838:
Table 3.1.9 Numerical solution of y ′ 2

+ 3x y = 1 + y ,
2
y(2) = 2 , by Euler’s method.
x h = 0.1 h = 0.05 h = 0.025 h = .0125
2.0 2.000000000 2.000000000 2.000000000 2.000000000
2.1 0.708426286 0.702568171 0.701214274 0.701162906

2.2 0.214501852 0.222599468 0.228942240 0.236986800
2.3 0.069861436 0.083620494 0.092852806 0.103815729
2.4 0.032487396 0.047079261 0.056825805 0.068390786
2.5 0.021895559 0.036030018 0.045683801 0.057281091
2.6 0.017332058 0.030750181 0.040189920 0.051711676
2.7 0.014271492 0.026931911 0.036134674 0.047564141
2.8 0.011819555 0.023720670 0.032679767 0.044014438
2.9 0.009776792 0.020925522 0.029636506 0.040875333
3.0 0.008065020 0.018472302 0.026931099 0.038072838

Table 3.1.10 : Numerical solution of y ′ 2 2
+ 3x y = 1 + y , y(2) = 2 , by the Euler semilinear method.
In the next two sections we will study other numerical methods for solving initial value problems, called the improved Euler
method, the midpoint method, Heun’s method and the Runge- Kutta method. If the initial value problem is semilinear as in
Equation 3.1.19, we also have the option of using variation of parameters and then applying the given numerical method to the
initial value problem Equation 3.1.21 for u. By analogy with the terminology used here, we will call the resulting procedure
the improved Euler semilinear method, the midpoint semilinear method, Heun’s semilinear method or the Runge- Kutta
semilinear method, as the case may be.

3.1E: Euler’s Method (Exercises)
Q3.1.1
You may want to save the results of these exercises, since we will revisit in the next two sections. In Exercises 3.1.1-3.1.5 use
Euler’s method to find approximate values of the solution of the given initial value problem at the points x = x + ih , where i 0
x is the point where the initial condition is imposed and i = 1 , 2 , 3 . The purpose of these exercises is to familiarize you with
0
the computational procedure of Euler’s method.

1. y ′
= 2x
2
+ 3y
2
− 2, y(2) = 1; h = 0.05
−−−−− −
2. y ′
= y + √x2 + y 2 , y(0) = 1; h = 0.1
3. y ′
+ 3y = x
2
− 3xy + y ,
2
y(0) = 2; h = 0.05
4. y ′
=
1+x
1−y 2
, y(2) = 3; h = 0.1
5. y ′ 2
+ x y = sin xy, y(1) = π; h = 0.2
Q3.1.2
6. Use Euler’s method with step sizes h = 0.1 , h = 0.05 , and h = 0.025 to find approximate values of the solution of the
′ 4x
y + 3y = 7 e , y(0) = 2 (3.1E.1)
at x = 0 , 0.1, 0.2, 0.3, …, 1.0. Compare these approximate values with the values of the exact solution y =e
4x
+e
−3x
,
which can be obtained by the method of Section 2.1. Present your results in a table like Table 3.1.1.
′
2 3
y + y = + 1, y(1) = 1 (3.1E.2)
x x3
at x = 1.0, 1.1, 1.2, 1.3, …, 2.0. Compare these approximate values with the values of the exact solution
1 3
y = (9 ln x + x + 2), (3.1E.3)
3x2
8. Use Euler’s method with step sizes h = 0.05, h = 0.025, and h = 0.0125 to find approximate values of the solution of the
2 2
y + xy − x
′
y = , y(1) = 2 (3.1E.4)
2
x
2
x(1 + x /3)
y = (3.1E.5)
2
1 − x /3
obtained in Example [example:2.4.3}. Present your results in a table like Table 3.1.1.
9. In Example [example:2.2.3} it was shown that
5 2
y +y = x +x −4 (3.1E.6)
is an implicit solution of the initial value problem

2x + 1
′
y = , y(2) = 1. (A)
4
5y +1

Use Euler’s method with step sizes h = 0.1 , h = 0.05, and h = 0.025 to find approximate values of the solution of (A) at
x = 2.0 , 2.1, 2.2, 2.3, …, 3.0. Present your results in tabular form. To check the error in these approximate values, construct
another table of values of the residual

5 2
R(x, y) = y +y −x −x +4 (3.1E.7)
for each value of (x, y) appearing in the first table.

10. You can see from Example 2.5.1 that
4 3 2 5
x y +x y + 2xy = 4 (3.1E.8)

3 3 5
4x y + 2x y + 2y
′
y =− , y(1) = 1. (A)
4 2 2 4
3x y + 5x y + 2x
Use Euler’s method with step sizes h = 0.1 , h = 0.05, and h = 0.025 to find approximate values of the solution of (A) at
x = 1.0 , 1.1, 1.2, 1.3, …, 2.0. Present your results in tabular form. To check the error in these approximate values, construct
another table of values of the residual

4 3 2 5
R(x, y) = x y +x y + 2xy − 4 (3.1E.9)

2 ′
(3 y + 4y)y + 2x + cos x = 0, y(0) = 1; (Exercise 2.2.13) (3.1E.10)
at x = 0 , 0.1, 0.2, 0.3, …, 1.0.

(y + 1)(y − 1)(y − 2)
′
y + = 0, y(1) = 0 (Exercise 2.2.14) (3.1E.11)
x +1
at x = 1.0, 1.1, 1.2, 1.3, …, 2.0.

13. Use Euler’s method and the Euler semilinear method with step sizes h = 0.1 , h = 0.05 , and h = 0.025 to find
approximate values of the solution of the initial value problem
′ −3x
y + 3y = 7 e , y(0) = 6 (3.1E.12)
at x = 0 , 0.1, 0.2, 0.3, …, 1.0. Compare these approximate values with the values of the exact solution y = e −3x
(7x + 6) ,
which can be obtained by the method of Section 2.1. Do you notice anything special about the results? Explain.
Q3.1.3
The linear initial value problems in Exercises 3.1.14–3.1.19 can’t be solved exactly in terms of known elementary functions. In
each exercise, use Euler’s method and the Euler semilinear methods with the indicated step sizes to find approximate values of
the solution of the given initial value problem at 11 equally spaced points (including the endpoints) in the interval.
14. y ′
− 2y =
1+x
1
2
, y(2) = 2 ; h = 0.1, 0.05, 0.025on [2, 3]
15. y ′
+ 2xy = x ,
2
y(0) = 3 (Exercise 2.1.38); h = 0.2, 0.1, 0.05 on [0, 2]
16. y ′
+
1
x
y =
sin x
x
2
, y(1) = 2; (Exercise 2.1.39); h = 0.2, 0.1, 0.05 on [1, 3]
−x
17. y ′
+y =
e tan x
x
, y(1) = 0; (Exercise 2.1.40); h = 0.05, 0.025, 0.0125 on [1, 1.5]
x
18. y ′
+
1+x2
2x
y =
e
2 2
, y(0) = 1; (Exercise 2.1.41); h = 0.2, 0.1, 0.05 on [0, 2]
(1+x )
19. x y on [1, 1.5]

2
′ x
+ (x + 1)y = e , y(1) = 2; (Exercise 2.1.42); h = 0.05, 0.025, 0.0125

Q3.1.4
In Exercises 3.1.20-3.1.22, use Euler’s method and the Euler semilinear method with the indicated step sizes to find
approximate values of the solution of the given initial value problem at 11 equally spaced points (including the endpoints) in
the interval.
20. y ′ 2
+ 3y = x y (y + 1), y(0) = 1 ; h = 0.1, 0.05, 0.025on [0, 1]
21. y ′
− 4y =
y
2
x
(y+1)
, y(0) = 1 ; h = 0.1, 0.05, 0.025on [0, 1]
2
22. y ′
+ 2y =
1+y
x
2
, y(2) = 1 ; h = 0.1, 0.05, 0.025on [2, 3]
Q3.1.5
23. Numerical Quadrature. The fundamental theorem of calculus says that if f is continuous on a closed interval [a, b] then it
has an antiderivative F such that F (x) = f (x) on [a, b] and
′
∫ f (x) dx = F (b) − F (a). (A)

a
This solves the problem of evaluating a definite integral if the integrand f has an antiderivative that can be found and
evaluated easily. However, if f doesn’t have this property, (A) doesn’t provide a useful way to evaluate the definite integral. In
this case we must resort to approximate methods. There’s a class of such methods called numerical quadrature, where the
approximation takes the form
b n
∫ f (x) dx ≈ ∑ ci f (xi ), (B)

a
i=0
where a = x < x < ⋯ < x

0 1 n =b are suitably chosen points and c , c , …, c are suitably chosen constants. We call (B) a
0 1 n
quadrature formula.
a. Derive the quadrature formula
b n−1
∫ f (x) dx ≈ h ∑ f (a + ih) (C)

a i=0
where h = (b − a)/n) by applying Euler’s method to the initial value problem

′
y = f (x), y(a) = 0. (3.1E.13)
b. The quadrature formula (C) is sometimes called the left rectangle rule. Draw a figure that justifies this terminology.
c. For several choices of a , b , and A , apply (C) to f (x) = A with n = 10, 20, 40, 80, 160, 320. Compare your results with
the exact answers and explain what you find.
d. For several choices of a , b , A , and B , apply (C) to f (x) = A + Bx with n = 10 , 20, 40, 80, 160, 320. Compare your
results with the exact answers and explain what you find.

3.2: The Improved Euler Method and Related Methods
In Section 3.1, we saw that the global truncation error of Euler’s method is O(h) , which would seem to imply that we can
achieve arbitrarily accurate results with Euler’s method by simply choosing the step size sufficiently small. However, this is
not a good idea, for two reasons. First, after a certain point decreasing the step size will increase roundoff errors to the point
where the accuracy will deteriorate rather than improve. The second and more important reason is that in most applications of
numerical methods to an initial value problem
′
y = f (x, y), y(x0 ) = y0 , (3.2.1)
the expensive part of the computation is the evaluation of f . Therefore we want methods that give good results for a given
number of such evaluations. This is what motivates us to look for numerical methods better than Euler’s.
To clarify this point, suppose we want to approximate the value of e by applying Euler’s method to the initial value problem
′
y = y, y(0) = 1
(with solution y = e ) on [0, 1], with h = 1/12, 1/24, and 1/48, respectively. Since each step in Euler’s method requires one
x
evaluation of f , the number of evaluations of f in each of these attempts is n = 12 , 24, and 48, respectively. In each case we
accept y as an approximation to e . The second column of Table 3.2.1 shows the results. The first column of the table
n
indicates the number of evaluations of f required to obtain the approximation, and the last column contains the value of e
rounded to ten significant figures.
In this section we will study the improved Euler method, which requires two evaluations of f at each step. We’ve used this
method with h = 1/6 , 1/12, and 1/24. The required number of evaluations of f were 12, 24, and 48, as in the three
applications of Euler’s method; however, you can see from the third column of Table 3.2.1 that the approximation to e
obtained by the improved Euler method with only 12 evaluations of f is better than the approximation obtained by Euler’s
method with 48 evaluations.
In Section 3.3, we will study the Runge- Kutta method, which requires four evaluations of f at each step. We’ve used this
method with h = 1/3 , 1/6, and 1/12. The required number of evaluations of f were again 12, 24, and 48, as in the three
applications of Euler’s method and the improved Euler method; however, you can see from the fourth column of Table 3.2.1
that the approximation to e obtained by the Runge-Kutta method with only 12 evaluations of f is better than the
approximation obtained by the improved Euler method with 48 evaluations.
n Euler Improved Euler Runge-Kutta Exact
12 2.613035290 2.707188994 2.718069764 2.718281828
24 2.663731258 2.715327371 2.718266612 2.718281828

48 2.690496599 2.717519565 2.718280809 2.718281828
Table 3.2.1 : Approximations to e obtained by three numerical methods.
The Improved Euler Method

The improved Euler method for solving the initial value problem Equation 3.2.1 is based on approximating the integral curve
of Equation 3.2.1 at (x , y(x )) by the line through (x , y(x )) with slope
i i i i
f (xi , y(xi )) + f (xi+1 , y(xi+1 ))

mi = ;
2
that is, m is the average of the slopes of the tangents to the integral curve at the endpoints of [x
i i, . The equation of the
xi+1 ]
approximating line is therefore

f (xi , y(xi )) + f (xi+1 , y(xi+1 ))
y = y(xi ) + (x − xi ). (3.2.2)
2

h
yi+1 = y(xi ) + (f (xi , y(xi )) + f (xi+1 , y(xi+1 ))) (3.2.3)
2
as an approximation to y(x ). As in our derivation of Euler’s method, we replace

i+1 y(xi ) (unknown if i >0 ) by its
approximate value y ; then Equation 3.2.3 becomes
i
h
yi+1 = yi + (f (xi , yi ) + f (xi+1 , y(xi+1 )) .
2
However, this still will not work, because we do not know y(x ), which appears on the right. We overcome this by replacing
i+1
y(x i+1) by y + hf (x , y ) , the value that the Euler method would assign to y
i i i . Thus, the improved Euler method starts i+1
with the known value y(x ) = y and computes y , y , …, y successively with the formula
0 0 1 2 n
h
yi+1 = yi + (f (xi , yi ) + f (xi+1 , yi + hf (xi , yi ))) . (3.2.4)
2
The computation indicated here can be conveniently organized as follows: given y , compute i
k1i = f (xi , yi ),
k2i = f (xi + h, yi + h k1i ) ,
h
yi+1 = yi + (k1i + k2i ).
2
The improved Euler method requires two evaluations of f (x, y) per step, while Euler’s method requires only one. However,
we will see at the end of this section that if f satisfies appropriate assumptions, the local truncation error with the improved
Euler method is O(h ), rather than O(h ) as with Euler’s method. Therefore the global truncation error with the improved
3 2
Euler method is O(h ); however, we will not prove this.

2
We note that the magnitude of the local truncation error in the improved Euler method and other methods discussed in this
section is determined by the third derivative y of the solution of the initial value problem. Therefore the local truncation error
′′′
will be larger where |y | is large, or smaller where |y | is small.

′′′ ′′′
The next example, which deals with the initial value problem considered in Example 3.2.1, illustrates the computational
procedure indicated in the improved Euler method.
Example 3.2.1
Use the improved Euler method with h = 0.1 to find approximate values of the solution of the initial value problem
′ 3 −2x
y + 2y = x e , y(0) = 1 (3.2.5)
at x = 0.1, 0.2, 0.3.

Solution
As in Example 3.1.1, we rewrite Equation 3.2.5 as
′ 3 −2x
y = −2y + x e , y(0) = 1,

3 −2x
f (x, y) = −2y + x e , x0 = 0, and y0 = 1.
The improved Euler method yields

k10 = f (x0 , y0 ) = f (0, 1) = −2,
k20 = f (x1 , y0 + h k10 ) = f (0.1, 1 + (0.1)(−2))
3 −0.2
= f (0.1, 0.8) = −2(0.8) + (0.1 ) e = −1.599181269,
h
y1 = y0 + (k10 + k20 ),
2
= 1 + (0.05)(−2 − 1.599181269) = 0.820040937,
3 −0.2
k11 = f (x1 , y1 ) = f (0.1, 0.820040937) = −2(0.820040937) + (0.1 ) e = −1.639263142,
k21 = f (x2 , y1 + h k11 ) = f (0.2, 0.820040937 + 0.1(−1.639263142)),
3 −0.4
= f (0.2, 0.656114622) = −2(0.656114622) + (.2 ) e = −1.306866684,
h
y2 = y1 + (k11 + k21 ),
2
= .820040937 + (.05)(−1.639263142 − 1.306866684) = 0.672734445,
3 −.4
k12 = f (x2 , y2 ) = f (.2, .672734445) = −2(.672734445) + (.2 ) e = −1.340106330,
k22 = f (x3 , y2 + h k12 ) = f (.3, .672734445 + .1(−1.340106330)),
3 −.6
= f (.3, .538723812) = −2(.538723812) + (.3 ) e = −1.062629710,
h
y3 = y2 + (k12 + k22 )
2
= .672734445 + (.05)(−1.340106330 − 1.062629710) = 0.552597643.
Example 3.2.2
Table 3.2.2 shows results of using the improved Euler method with step sizes h = 0.1 and h = 0.05 to find approximate
values of the solution of the initial value problem
′ 3 −2x
y + 2y = x e , y(0) = 1
at x = 0 , 0.1, 0.2, 0.3, …, 1.0. For comparison, it also shows the corresponding approximate values obtained with Euler’s
method in [example:3.1.2}, and the values of the exact solution
−2x
e
4
y = (x + 4).
4
The results obtained by the improved Euler method with h = 0.1 are better than those obtained by Euler’s method with
h = 0.05 .
Euler Improved Euler Exact
x h = 0.1 h = 0.05 h = 0.1 h = 0.05 Exact
0.0 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000
0.1 0.800000000 0.810005655 0.820040937 0.819050572 0.818751221

0.2 0.640081873 0.656266437 0.672734445 0.671086455 0.670588174
0.3 0.512601754 0.532290981 0.552597643 0.550543878 0.549922980
0.4 0.411563195 0.432887056 0.455160637 0.452890616 0.452204669
0.5 0.332126261 0.353785015 0.376681251 0.374335747 0.373627557
0.6 0.270299502 0.291404256 0.313970920 0.311652239 0.310952904
0.7 0.222745397 0.242707257 0.264287611 0.262067624 0.261398947
0.8 0.186654593 0.205105754 0.225267702 0.223194281 0.222570721
0.9 0.159660776 0.176396883 0.194879501 0.192981757 0.192412038
1.0 0.139778910 0.154715925 0.171388070 0.169680673 0.169169104
Table 3.2.2: Numerical solution of y ′

+ 2y = x e
3 −2x
, y(0) = 1 , by Euler’s method and the improved Euler method.

Example 3.2.3
Table 3.2.3 shows analogous results for the nonlinear initial value problem
′ 2 2
y = −2 y + xy + x , y(0) = 1.
We applied Euler’s method to this problem in Example 3.2.3.
Euler Improved Euler Exact
x h = 0.1 h = 0.05 h = 0.1 h = 0.05 Exact
0.0 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000
0.1 0.800000000 0.821375000 0.840500000 0.838288371 0.837584494

0.2 0.681000000 0.707795377 0.733430846 0.730556677 0.729641890
0.3 0.605867800 0.633776590 0.661600806 0.658552190 0.657580377
0.4 0.559628676 0.587454526 0.615961841 0.612884493 0.611901791
0.5 0.535376972 0.562906169 0.591634742 0.588558952 0.587575491
0.6 0.529820120 0.557143535 0.586006935 0.582927224 0.581942225
0.7 0.541467455 0.568716935 0.597712120 0.594618012 0.593629526
0.8 0.569732776 0.596951988 0.626008824 0.622898279 0.621907458
0.9 0.614392311 0.641457729 0.670351225 0.667237617 0.666250842
1.0 0.675192037 0.701764495 0.730069610 0.726985837 0.726015790

= −2 y
2 2
+ xy + x , y(0) = 1 , by Euler’s method and the improved Euler
method.
Example 3.2.4
Use step sizes h = 0.2 , h = 0.1 , and h = 0.05 to find approximate values of the solution of
′
y − 2xy = 1, y(0) = 3 (3.2.6)
at x = 0 , 0.2, 0.4, 0.6, …, 2.0 by (a) the improved Euler method; (b) the improved Euler semilinear method. (We used
Euler’s method and the Euler semilinear method on this problem in [example:3.1.4}.)
Solution
a. Rewriting Equation 3.2.6 as
′
y = 1 + 2xy, y(0) = 3
and applying the improved Euler method with f (x, y) = 1 + 2xy yields the results shown in Table 3.2.4.
b. Since y = e is a solution of the complementary equation y , we can apply the improved Euler semilinear
2
x ′
1 − 2xy = 0
method to Equation 3.2.6, with

2 2
x ′ −x
y = ue and u =e , u(0) = 3.
The results listed in Table 3.2.5 are clearly better than those obtained by the improved Euler method.
x h = 0.2 h = 0.1 h = 0.05 Exact
0.0 3.000000000 3.000000000 3.000000000 3.000000000
0.2 3.328000000 3.328182400 3.327973600 3.327851973

0.4 3.964659200 3.966340117 3.966216690 3.966059348

x h = 0.2 h = 0.1 h = 0.05 Exact
0.6 5.057712497 5.065700515 5.066848381 5.067039535

0.8 6.900088156 6.928648973 6.934862367 6.936700945
1.0 10.065725534 10.154872547 10.177430736 10.184923955
1.2 15.708954420 15.970033261 16.041904862 16.067111677
1.4 26.244894192 26.991620960 27.210001715 27.289392347
1.6 46.958915746 49.096125524 49.754131060 50.000377775
1.8 89.982312641 96.200506218 98.210577385 98.982969504
2.0 184.563776288 203.151922739 209.464744495 211.954462214

− 2xy = 1, y(0) = 3 , by the improved Euler method.
x h = 0.2 h = 0.1 h = 0.05 Exact
0.0 3.000000000 3.000000000 3.000000000 3.000000000
0.2 3.326513400 3.327518315 3.327768620 3.327851973

0.4 3.963383070 3.965392084 3.965892644 3.966059348
0.6 5.063027290 5.066038774 5.066789487 5.067039535
0.8 6.931355329 6.935366847 6.936367564 6.936700945
1.0 10.178248417 10.183256733 10.184507253 10.184923955
1.2 16.059110511 16.065111599 16.066611672 16.067111677
1.4 27.280070674 27.287059732 27.288809058 27.289392347
1.6 49.989741531 49.997712997 49.999711226 50.000377775
1.8 98.971025420 98.979972988 98.982219722 98.982969504
2.0 211.941217796 211.951134436 211.953629228 211.954462214

− 2xy = 1, y(0) = 3 , by the improved Euler semilinear method.
A Family of Methods with O(h³) Local Truncation Error

We will now derive a class of methods with O(h ) local truncation error for solving Equation 3.2.1. For simplicity, we assume
3
that f , f , f , f , f , and f are continuous and bounded for all (x, y). This implies that if y is the solution of Equation
x y xx yy xy
3.2.1 then y and y are bounded (Exercise 3.2.31).

′′ ′′′
We begin by approximating the integral curve of Equation 3.2.1 at (x i, y(xi )) by the line through (xi, y(xi )) with slope
′ ′
mi = σ y (xi ) + ρy (xi + θh),
where σ, ρ, and θ are constants that we will soon specify; however, we insist at the outset that 0 < θ ≤ 1 , so that
xi < xi + θh ≤ xi+1 .
The equation of the approximating line is
y = y(xi ) + mi (x − xi )
(3.2.7)
′ ′
= y(xi ) + [σ y (xi ) + ρy (xi + θh)] (x − xi ).

′ ′
y
î+1 = y(xi ) + h [σ y (xi ) + ρy (xi + θh)]
as an approximation to y(x i+1 ) .

To determine σ, ρ, and θ so that the error

Ei = y(xi+1 ) − y
^
i+1
(3.2.8)
′ ′
= y(xi+1 ) − y(xi ) − h [σ y (xi ) + ρy (xi + θh)]
in this approximation is O(h ), we begin by recalling from Taylor’s theorem that

3
2 3
′
h ′′
h ′′′
y(xi+1 ) = y(xi ) + h y (xi ) + y (xi ) + y î ),
(x
2 6
where x
^ is in (x
i i, . Since y is bounded this implies that
xi+1 )
′′′
2
′
h ′′ 3
y(xi+1 ) − y(xi ) − h y (xi ) − y (xi ) = O(h ).
2
Comparing this with Equation 3.2.8 shows that E i = O(h )

3
if
h
′ ′ ′ ′′ 2
σ y (xi ) + ρy (xi + θh) = y (xi ) + y (xi ) + O(h ). (3.2.9)
2
However, applying Taylor’s theorem to y shows that ′
2
(θh)
′ ′ ′′ ′′′ ¯¯
¯
y (xi + θh) = y (xi ) + θh y (xi ) + y (xi ),
2
where x is in (x
¯¯
¯
i i, xi + θh) . Since y is bounded, this implies that
′′′
′ ′ ′′ 2
y (xi + θh) = y (xi ) + θh y (xi ) + O(h ).
Substituting this into Equation 3.2.9 and noting that the sum of two O(h 2
) terms is again O(h 2
) shows that E
i
3
= O(h ) if
′ ′′ ′
h ′′
(σ + ρ)y (xi ) + ρθh y (xi ) = y (xi ) + y (xi ),
2
which is true if
1
σ +ρ = 1 and ρθ = . (3.2.10)
2
Since y ′
= f (x, y) , we can now conclude from Equation 3.2.8 that
3
y(xi+1 ) = y(xi ) + h [σf (xi , yi ) + ρf (xi + θh, y(xi + θh))] + O(h ) (3.2.11)
if σ, ρ, and θ satisfy Equation 3.2.10. However, this formula would not be useful even if we knew y(x ) exactly (as we would i
for i = 0 ), since we still wouldn’t know y(x + θh) exactly. To overcome this difficulty, we again use Taylor’s theorem to
i
write
2
h
′ ′′ ~
y(xi + θh) = y(xi ) + θh y (xi ) + y (xi ),
2
where x
~
is in (x
i i, xi + θh) . Since y ′
(xi ) = f (xi , y(xi )) and y is bounded, this implies that
′′
2
|y(xi + θh) − y(xi ) − θhf (xi , y(xi ))| ≤ K h (3.2.12)
for some constant K . Since f is bounded, the mean value theorem implies that
y
|f (xi + θh, u) − f (xi + θh, v)| ≤ M |u − v|
for some constant M . Letting
u = y(xi + θh) and v = y(xi ) + θhf (xi , y(xi ))
and recalling Equation 3.2.12 shows that

2
f (xi + θh, y(xi + θh)) = f (xi + θh, y(xi ) + θhf (xi , y(xi ))) + O(h ).

y(xi+1 ) = y(xi ) + h [σf (xi , y(xi ))+
3
ρf (xi + θh, y(xi ) + θhf (xi , y(xi )))] + O(h ).
This implies that the formula
yi+1 = yi + h [σf (xi , yi ) + ρf (xi + θh, yi + θhf (xi , yi ))]
has O(h 3
) local truncation error if σ, ρ, and θ satisfy Equation 3.2.10. Substituting σ = 1 − ρ and θ = 1/2ρ here yields
h h
yi+1 = yi + h [(1 − ρ)f (xi , yi ) + ρf ( xi + , yi + f (xi , yi ))] . (3.2.13)
2ρ 2ρ
The computation indicated here can be conveniently organized as follows: given y , compute i
k1i = f (xi , yi ),
h h
k2i = f ( xi + , yi + k1i ) ,
2ρ 2ρ
yi+1 = yi + h[(1 − ρ)k1i + ρk2i ].
Consistent with our requirement that 0 < θ < 1 , we require that ρ ≥ 1/2 . Letting ρ = 1/2 in Equation 3.2.13 yields the
improved Euler method Equation 3.2.4. Letting ρ = 3/4 yields Heun’s method,
1 3 2 2
yi+1 = yi + h [ f (xi , yi ) + f ( xi + h, yi + hf (xi , yi ))] ,
4 4 3 3
which can be organized as

k1i = f (xi , yi ),
2h 2h
k2i = f ( xi + , yi + k1i ) ,
3 3
h
yi+1 = yi + (k1i + 3 k2i ).
4
Letting ρ = 1 yields the midpoint method,

h h
yi+1 = yi + hf ( xi + , yi + f (xi , yi )) ,
2 2
which can be organized as

k1i = f (xi , yi ),
h h
k2i = f ( xi + , yi + k1i ) ,
2 2
yi+1 = yi + h k2i .
Examples involving the midpoint method and Heun’s method are given in Exercises 3.2.23 - 3.3.30.

3.2E: The Improved Euler Method and Related Methods (Exercises)
Most of the following numerical exercises involve initial value problems considered in the exercises in Section 3.1. You’ll find
it instructive to compare the results that you obtain here with the corresponding results that you obtained in Section 3.1.
Q3.2.1
In Exercises 3.2.1–3.2.5 use the improved Euler method to find approximate values of the solution of the given initial value
problem at the points x = x + ih , where x is the point where the initial condition is imposed and i = 1 , 2, 3.
i 0 0
1. y ′
= 2x
2
+ 3y
2
− 2, y(2) = 1; h = 0.05
−−−−−−
2. y ′ 2 2
= y + √x + y , y(0) = 1; h = 0.1
3. y ′
+ 3y = x
2
− 3xy + y ,
2
y(0) = 2; h = 0.05
4. y ′
=
1+x
1−y
2
, y(2) = 3; h = 0.1
5. y ′ 2
+ x y = sin xy, y(1) = π; h = 0.2
Q3.2.2
6. Use the improved Euler method with step sizes h = 0.1 , h = 0.05 , and h = 0.025 to find approximate values of the
solution of the initial value problem
′ 4x
y + 3y = 7 e , y(0) = 2
4x
+e
−3x
,
′
2 3
y + y = + 1, y(1) = 1
x x3
1 3
y = (9 ln x + x + 2)
2
3x
2 2
′
y + xy − x
y = , y(1) = 2,
2
x
2
x(1 + x /3)
y =
2
1 − x /3
obtained in Example [example:2.4.3}. Present your results in a table like Table 3.2.2.
9. In Example [example:3.2.2} it was shown that
5 2
y +y = x +x −4

2x + 1
′
y = , y(2) = 1. (A)
4
5y +1

Use the improved Euler method with step sizes h = 0.1 , h = 0.05, and h = 0.025 to find approximate values of the solution
of (A) at x = 2.0, 2.1, 2.2, 2.3, …, 3.0. Present your results in tabular form. To check the error in these approximate values,
construct another table of values of the residual
5 2
R(x, y) = y +y −x −x +4

10. You can see from Example 2.5.1 that
4 3 2 5
x y +x y + 2xy = 4

3 3 5
4x y + 2x y + 2y
′
y =− , y(1) = 1. (A)
4 2 2 4
3x y + 5x y + 2x
Use the improved Euler method with step sizes h = 0.1 , h = 0.05, and h = 0.025 to find approximate values of the solution
of (A) at x = 1.0, 1.14, 1.2, 1.3, …, 2.0. Present your results in tabular form. To check the error in these approximate values,
4 3 2 5
R(x, y) = x y +x y + 2xy − 4

2 ′
(3 y + 4y)y + 2x + cos x = 0, y(0) = 1 (Exercise 2.2.13) (3.2E.1)
at x = 0 , 0.1, 0.2, 0.3, …, 1.0.

(y + 1)(y − 1)(y − 2)
′
y + = 0, y(1) = 0 (Exercise 2.2.14) (3.2E.2)
x +1
at x = 1.0, 1.1, 1.2, 1.3, …, 2.0.

13. Use the improved Euler method and the improved Euler semilinear method with step sizes h = 0.1 , h = 0.05 , and
h = 0.025 to find approximate values of the solution of the initial value problem
′ −3x
y + 3y = e (1 − 2x), y(0) = 2,
at x =0 , 0.1 , 0.2 , 0.3, …, 1.0. Compare these approximate values with the values of the exact solution
y =e
−3x
(2 + x − x )
2
, which can be obtained by the method of Section 2.1. Do you notice anything special about the results?
Explain.
Q3.2.3
The linear initial value problems in Exercises 3.2.14-3.2.19 can’t be solved exactly in terms of known elementary functions. In
each exercise use the improved Euler and improved Euler semilinear methods with the indicated step sizes to find approximate
values of the solution of the given initial value problem at 11 equally spaced points (including the endpoints) in the interval.
14. y ′
− 2y =
1+x2
1
, y(2) = 2 ; h = 0.1, 0.05, 0.025on [2, 3]
15. y ′
+ 2xy = x ,
2
y(0) = 3 ; h = 0.2, 0.1, 0.05on [0, 2] (Exercise 2.1.38)
16. y ′
+
x
1
y =
sin x
x
2
, y(1) = 2 , h = 0.2, 0.1, 0.05on [1, 3] (Exercise 2.1.39)
−x
17. y ′
+y =
e tan x
x
, y(1) = 0 ; h = 0.05, 0.025, 0.0125on [1, 1.5] (Exercise 2.1.40),
x
18. y ′
+
1+x2
2x
y =
e
2
, y(0) = 1 ; h = 0.2, 0.1, 0.05on [0, 2] (Exercise 2.1.41)
(1+x2 )

19. x y ; h = 0.05, 0.025, 0.0125on [1, 1.5] (Exercise 2.1.42)
2
′ x
+ (x + 1)y = e , y(1) = 2
Q3.2.4
In Exercises 3.2.20-3.2.22 use the improved Euler method and the improved Euler semilinear method with the indicated step
sizes to find approximate values of the solution of the given initial value problem at 11 equally spaced points (including the
endpoints) in the interval.
20. y ′
+ 3y = x y (y + 1),
2
y(0) = 1 ; h = 0.1, 0.05, 0.025on [0, 1]
21. y ′
− 4y =
y
2
x
(y+1)
, y(0) = 1 ; h = 0.1, 0.05, 0.025on [0, 1]
2
22. y ′
+ 2y =
1+y
x
2
, y(2) = 1 ; h = 0.1, 0.05, 0.025on [2, 3]
Q3.2.5
23. Do Exercise 3.2E. 7 with “improved Euler method” replaced by “midpoint method.”
24. Do Exercise 3.2E. 7 with “improved Euler method” replaced by “Heun’s method.”
31. Show that if f , f , f , f , f , and f x y xx yy xy are continuous and bounded for all (x, y) and y is the solution of the initial value
problem
′
y = f (x, y), y(x0 ) = y0 ,
then y and y are bounded.

′′ ′′′
32. Numerical Quadrature (see Exercise 3.1.23).

b n−1
∫ f (x) dx ≈ 0.5h(f (a) + f (b)) + h ∑ f (a + ih)

a i=1
(where h = (b − a)/n) by applying the improved Euler method to the initial value problem
′
y = f (x), y(a) = 0.
b. The quadrature formula (A) is called the trapezoid rule. Draw a figure that justifies this terminology.
c. For several choices of a , b , A , and B , apply (A) to f (x) = A + Bx , with n = 10, 20, 40, 80, 160, 320. Compare your
results with the exact answers and explain what you find.
d. For several choices of a , b , A , B , and C , apply (A) to f (x) = A + Bx + C x , with n = 10 , 20, 40, 80, 160, 320. 2
Compare your results with the exact answers and explain what you find.

3.3: The Runge-Kutta Method
In general, if k is any positive integer and f satisfies appropriate assumptions, there are numerical methods with local
truncation error O(h ) for solving an initial value problem
k+1
′
y = f (x, y), y(x0 ) = y0 . (3.3.1)
Moreover, it can be shown that a method with local truncation error O(h ) has global truncation error O(h ) . In Sections
k+1 k
3.1 and 3.2 we studied numerical methods where k = 1 and k = 2 . We’ll skip methods for which k = 3 and proceed to the
Runge-Kutta method, the most widely used method, for which k = 4 . The magnitude of the local truncation error is
determined by the fifth derivative y of the solution of the initial value problem. Therefore the local truncation error will be
(5)
larger where |y | is large, or smaller where |y | is small. The Runge-Kutta method computes approximate values y , y ,
(5) (5)
1 2
…, y of the solution of Equation 3.3.1 at x , x + h , …, x + nh as follows: Given y , compute

n 0 0 0 i
k1i = f (xi , yi ),
h h
k2i = f ( xi + , yi + k1i ) ,
2 2
h h
k3i = f ( xi + , yi + k2i ) ,
2 2
k4i = f (xi + h, yi + h k3i ),
and
h
yi+1 = yi + (k1i + 2 k2i + 2 k3i + k4i ).
6
The next example, which deals with the initial value problem considered in Examples and Example 3.3.1, illustrates the
computational procedure indicated in the Runge-Kutta method.
Example 3.3.1
Use the Runge-Kutta method with h = 0.1 to find approximate values for the solution of the initial value problem
′ 3 −2x
y + 2y = x e , y(0) = 1, (3.3.2)
at x = 0.1, 0.2.
Solution
Again we rewrite Equation 3.3.2 as
′ 3 −2x
y = −2y + x e , y(0) = 1,

3 −2x
f (x, y) = −2y + x e , x0 = 0, and y0 = 1.
The Runge-Kutta method yields

k10 = f (x0 , y0 ) = f (0, 1) = −2,
k20 = f (x0 + h/2, y0 + h k10 /2) = f (.05, 1 + (.05)(−2))
3 −.1
= f (.05, .9) = −2(.9) + (.05 ) e = −1.799886895,
k30 = f (x0 + h/2, y0 + h k20 /2) = f (.05, 1 + (.05)(−1.799886895))
3 −.1
= f (.05, .910005655) = −2(.910005655) + (.05 ) e = −1.819898206,
k40 = f (x0 + h, y0 + h k30 ) = f (.1, 1 + (.1)(−1.819898206))
3 −.2
= f (.1, .818010179) = −2(.818010179) + (.1 ) e = −1.635201628,
h
y1 = y0 + (k10 + 2 k20 + 2 k30 + k40 ),
6
.1
=1+ (−2 + 2(−1.799886895) + 2(−1.819898206) − 1.635201628) = .818753803,
6
3 −.2
k11 = f (x1 , y1 ) = f (.1, .818753803) = −2(.818753803)) + (.1 ) e = −1.636688875,
k21 = f (x1 + h/2, y1 + h k11 /2) = f (.15, .818753803 + (.05)(−1.636688875))

3 −.3
= f (.15, .736919359) = −2(.736919359) + (.15 ) e = −1.471338457,
k31 = f (x1 + h/2, y1 + h k21 /2) = f (.15, .818753803 + (.05)(−1.471338457))
3 −.3
= f (.15, .745186880) = −2(.745186880) + (.15 ) e = −1.487873498,
k41 = f (x1 + h, y1 + h k31 ) = f (.2, .818753803 + (.1)(−1.487873498))
3 −.4
= f (.2, .669966453) = −2(.669966453) + (.2 ) e = −1.334570346,
h
y2 = y1 + (k11 + 2 k21 + 2 k31 + k41 ),
6
.1
= .818753803 + (−1.636688875 + 2(−1.471338457) + 2(−1.487873498) − 1.334570346)
6
= .670592417.
The Runge-Kutta method is sufficiently accurate for most applications.
Example 3.3.2
Table 3.3.1 shows results of using the Runge-Kutta method with step sizes h = 0.1 and h = 0.05 to find approximate
values of the solution of the initial value problem
′ 3 −2x
y + 2y = x e , y(0) = 1
at x = 0 , 0.1, 0.2, 0.3, …, 1.0. For comparison, it also shows the corresponding approximate values obtained with the
improved Euler method in Example 3.2.2:
Add text here. For the automatic number to work, you need to add the “AutoNum” template (preferably at3.2.2}, and the
values of the exact solution
−2x
e 4
y = (x + 4).
4
The results obtained by the Runge-Kutta method are clearly better than those obtained by the improved Euler method in
fact; the results obtained by the Runge-Kutta method with h = 0.1 are better than those obtained by the improved Euler
method with h = 0.05.
Improved Euler Runge-Kutta
x h=0.1 h=0.05 h=0.1 h-0.05 "exact"
0.0 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000
0.1 0.820040937 0.819050572 0.818753803 0.818751370 0.818751221

0.2 0.672734445 0.671086455 0.670592417 0.670588418 0.670588174
0.3 0.552597643 0.550543878 0.549928221 0.549923281 0.549922980
0.4 0.455160637 0.452890616 0.452210430 0.452205001 0.452204669

x h=0.1 h=0.05 h=0.1 h-0.05 "exact"
0.5 0.376681251 0.374335747 0.373633492 0.373627899 0.373627557

0.6 0.313970920 0.311652239 0.310958768 0.310953242 0.310952904
0.7 0.264287611 0.262067624 0.261404568 0.261399270 0.261398947
0.8 0.225267702 0.223194281 0.222575989 0.222571024 0.222570721
0.9 0.194879501 0.192981757 0.192416882 0.192412317 0.192412038
1.0 0.171388070 0.169680673 0.169173489 0.169169356 0.169169104

+ 2y = x e
3 −2x
, y(0) = 1 , by the Runge-Kuttta method and the improved Euler
method.
Example 3.3.3
Table 3.3.2 shows analogous results for the nonlinear initial value problem
′ 2 2
y = −2 y + xy + x , y(0) = 1.
We applied the improved Euler method to this problem in Example 3.2.3.

x h=0.1 h=0.05 h=0.1 h-0.05 "exact"
0.0 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000
0.1 0.840500000 0.838288371 0.837587192 0.837584759 0.837584494

0.2 0.733430846 0.730556677 0.729644487 0.729642155 0.729641890
0.3 0.661600806 0.658552190 0.657582449 0.657580598 0.657580377
0.4 0.615961841 0.612884493 0.611903380 0.611901969 0.611901791
0.5 0.591634742 0.588558952 0.587576716 0.587575635 0.587575491
0.6 0.586006935 0.582927224 0.581943210 0.581942342 0.581942225
0.7 0.597712120 0.594618012 0.593630403 0.593629627 0.593629526
0.8 0.626008824 0.622898279 0.621908378 0.621907553 0.621907458
0.9 0.670351225 0.667237617 0.666251988 0.666250942 0.666250842
1.0 0.730069610 0.726985837 0.726017378 0.726015908 0.726015790

= −2 y
2 2
+ xy + x , y(0) = 1 , by the Runge-Kuttta method and the improved
Euler method.
Example 3.3.4
Tables 3.3.3 and 3.3.4 show results obtained by applying the Runge-Kutta and Runge-Kutta semilinear methods to to the
′
y − 2xy = 1, y(0) = 3,
which we considered in Examples 3.3.3 and 3.3.4
x h = 0.2 h = 0.1 h = 0.05 "Exact"
0.0 3.000000000 3.000000000 3.000000000 3.000000000
0.2 3.327846400 3.327851633 3.327851952 3.327851973

x h = 0.2 h = 0.1 h = 0.05 "Exact"
0.4 3.966044973 3.966058535 3.966059300 3.966059348

0.6 5.066996754 5.067037123 5.067039396 5.067039535
0.8 6.936534178 6.936690679 6.936700320 6.936700945
1.0 10.184232252 10.184877733 10.184920997 10.184923955
1.2 16.064344805 16.066915583 16.067098699 16.067111677
1.4 27.278771833 27.288605217 27.289338955 27.289392347
1.6 49.960553660 49.997313966 50.000165744 50.000377775
1.8 98.834337815 98.971146146 98.982136702 98.982969504
2.0 211.393800152 211.908445283 211.951167637 211.954462214

− 2xy = 1, y(0) = 3 , by the Runge-Kutta method.
x h = 0.2 h = 0.1 h = 0.05 "Exact"
0.0 3.000000000 3.000000000 3.000000000 3.000000000
0.2 3.327853286 3.327852055 3.327851978 3.327851973

0.4 3.966061755 3.966059497 3.966059357 3.966059348
0.6 5.067042602 5.067039725 5.067039547 5.067039535
0.8 6.936704019 6.936701137 6.936700957 6.936700945
1.0 10.184926171 10.184924093 10.184923963 10.184923955
1.2 16.067111961 16.067111696 16.067111678 16.067111677
1.4 27.289389418 27.289392167 27.289392335 27.289392347
1.6 50.000370152 50.000377302 50.000377745 50.000377775
1.8 98.982955511 98.982968633 98.982969450 98.982969504
2.0 211.954439983 211.954460825 211.954462127 211.954462214

− 2xy = 1, y(0) = 3 , by the Runge-Kutta semilinear method.
The Case Where xₒ Isn’t The Left Endpoint

So far in this chapter we’ve considered numerical methods for solving an initial value problem
′
y = f (x, y), y(x0 ) = y0 (3.3.3)
on an interval [x , b], for which x is the left endpoint. We haven’t discussed numerical methods for solving Equation 3.3.3on
0 0
an interval [a, x ], for which x is the right endpoint. To be specific, how can we obtain approximate values y , y , …, y
0 0 −1 −2 −n
of the solution of Equation 3.3.3at x − h, … , x − nh , where h = (x − a)/n ? Here’s the answer to this question:
0 0 0
Question
′
z = −f (−x, z), z(−x0 ) = y0 , (3.3.4)
on the interval [−x , −a], for which −x is the left endpoint. Use a numerical method to obtain approximate values z ,
0 0 1
z , …, z
2 n of the solution of (3.3.4) at −x + h , −x + 2h , …, −x + nh = −a . Then y = z , y = z , …,
0 0 0 −1 1 −2 2
y−n =z n are approximate values of the solution of (3.3.3) at x − h , x − 2h , …, x − nh = a .

0 0 0
The justification for this answer is sketched in Exercise 3.3.23. Note how easy it is to make the change the given problem
Equation 3.3.3 to the modified problem Equation 3.3.4 : first replace f by −f and then replace x, x , and y by −x, −x , and 0 0
z , respectively.

Example 3.3.5
Use the Runge-Kutta method with step size h = 0.1 to find approximate values of the solution of
2 ′
(y − 1 ) y = 2x + 3, y(1) = 4 (3.3.5)
at x = 0 , 0.1, 0.2, …, 1.
Solution
We first rewrite Equation 3.3.5 in the form Equation 3.3.3 as
2x + 3
′
y = , y(1) = 4. (3.3.6)
2
(y − 1)
Since the initial condition y(1) = 4 is imposed at the right endpoint of the interval [0, 1] , we apply the Runge-Kutta
method to the initial value problem
2x − 3
′
z = , z(−1) = 4 (3.3.7)
2
(z − 1)
on the interval [−1, 0]. (You should verify that Equation 3.3.7 is related to Equation 3.3.6 as Equation 3.3.4 is related to
Equation 3.3.3.) Table [table:3.3.5} shows the results. Reversing the order of the rows in Table [table:3.3.5} and changing
the signs of the values of x yields the first two columns of Table [table:3.3.6}. The last column of Table [table:3.3.6}
shows the exact values of y , which are given by
2 1/3
y = 1 + (3 x + 9x + 15 ) .
(Since the differential equation in Equation 3.3.6is separable, this formula can be obtained by the method of Section 2.2.)
x z
-1.0 4.000000000
-0.9 3.944536474
-0.8 3.889298649
-0.7 3.834355648
-0.6 3.779786399
-0.5 3.725680888
-0.4 3.672141529
-0.3 3.619284615
-0.2 3.567241862
-0.1 3.516161955
0.0 3.466212070
Table 3.3.5 : Numerical solution of z ′

=
2x−3
2
, z(−1) = 4 , on [−1, 0].
(z−1)
x y Exact
0.00 3.466212070 3.466212074
0.10 3.516161955 3.516161958

0.20 3.567241862 3.567241864
0.30 3.619284615 3.619284617
0.40 3.672141529 3.672141530
0.50 3.725680888 3.725680889
0.60 3.779786399 3.779786399
0.70 3.834355648 3.834355648

x y Exact
0.80 3.889298649 3.889298649

0.90 3.944536474 3.944536474
1.00 4.000000000 4.000000000
Table 3.3.6 : Numerical solution of (y − 1)

2
y
′
= 2x + 3, y(1) = 4 , on [0, 1].
We leave it to you to develop a procedure for handling the numerical solution of Equation 3.3.3 on an interval [a, b] such
that a < x < b (Exercises 3.3.26 and 3.3.27).
0

3.3E: The Runge-Kutta Method (Exercises)
Most of the following numerical exercises involve initial value problems considered in the exercises in Sections 3.2. You’ll
find it instructive to compare the results that you obtain here with the corresponding results that you obtained in those sections.
Q3.3.1
In Exercises 3.3.1 -3.3.5 use the Runge-Kutta method to find approximate values of the solution of the given initial value
problem at the points x = x + ih, where x is the point where the initial condition is imposed and i = 1 , 2.
i 0 0
1. y ′
= 2x
2
+ 3y
2
− 2, y(2) = 1; h = 0.05
−−−−−−
2. y ′ 2 2
= y + √x + y , y(0) = 1; h = 0.1
3. y ′
+ 3y = x
2
− 3xy + y ,
2
y(0) = 2; h = 0.05
4. y ′
=
1+x
1−y
2
, y(2) = 3; h = 0.1
5. y ′ 2
+ x y = sin xy, y(1) = π; h = 0.2
Q3.3.2
6. Use the Runge-Kutta method with step sizes h = 0.1 , h = 0.05, and h = 0.025 to find approximate values of the solution
of the initial value problem
′ 4x
y + 3y = 7 e , y(0) = 2, (3.3E.1)
4x
+e
−3x
,
2 3
′
y + y = + 1, y(1) = 1 (3.3E.2)
3
x x
1
3
y = (9 ln x + x + 2), (3.3E.3)
2
3x
8. Use the Runge-Kutta method with step sizes h = 0.05 , h = 0.025 , and h = 0.0125 to find approximate values of the
2 2
′
y + xy − x
y = , y(1) = 2 (3.3E.4)
x2
at x = 1.0, 1.05, 1.10, 1.15 …, 1.5. Compare these approximate values with the values of the exact solution
2
x(1 + x /3)
y = , (3.3E.5)
2
1 − x /3
which was obtained in Example

Example 3.3E. 1:
Add text here. For the automatic number to work, you need to add the “AutoNum” template (preferably at 2.2.3}. Present your
results in a table like Table 3.3.1.
9. In Example
Example 3.3E. 1:

Add text here. For the automatic number to work, you need to add the “AutoNum” template (preferably at 2.2.3} it was shown
that
5 2
y +y = x +x −4 (3.3E.6)

2x + 1
′
y = , y(2) = 1. (A)
5y 4 + 1
Use the Runge-Kutta method with step sizes h = 0.1, h = 0.05, and h = 0.025 to find approximate values of the solution of
(A) at x = 2.0, 2.1, 2.2, 2.3, …, 3.0. Present your results in tabular form. To check the error in these approximate values,
5 2
R(x, y) = y +y −x −x +4 (3.3E.7)

10. You can see from Example
Example 3.3E. 1:
Add text here. For the automatic number to work, you need to add the “AutoNum” template (preferably at 2.5.1} that
4 3 2 5
x y +x y + 2xy = 4 (3.3E.8)

3 3 5
′
4x y + 2x y + 2y
y =− , y(1) = 1. (A)
4 2 2 4
3x y + 5x y + 2x
Use the Runge-Kutta method with step sizes h = 0.1, h = 0.05, and h = 0.025 to find approximate values of the solution of
(A) at x = 1.0, 1.1, 1.2, 1.3, …, 2.0. Present your results in tabular form. To check the error in these approximate values,
4 3 2 5
R(x, y) = x y +x y + 2xy − 4 (3.3E.9)

2 ′
(3 y + 4y)y + 2x + cos x = 0, y(0) = 1 (Exercise 2.2.13) (3.3E.10)
at x = 0 , 0.1, 0.2, 0.3, …, 1.0.

(y + 1)(y − 1)(y − 2)
′
y + = 0, y(1) = 0 (Exercise 2.2.14) (3.3E.11)
x +1
at x = 1.0, 1.1, 1.2, 1.3, …, 2.0.

13. Use the Runge-Kutta method and the Runge-Kutta semilinear method with step sizes h = 0.1 , h = 0.05 , and h = 0.025
to find approximate values of the solution of the initial value problem

′ −3x 2 3
y + 3y = e (1 − 4x + 3 x − 4 x ), y(0) = −3 (3.3E.12)
at x =0 , 0.1 , ,
0.2 0.3, …, 1.0 . Compare these approximate values with the values of the exact solution
y = −e
−3x
(3 − x + 2 x
2
−x
3
+x )
4
, which can be obtained by the method of Section 2.1. Do you notice anything special
about the results? Explain.
Q3.3.3
The linear initial value problems in Exercises 3.3.14–3.3.19 can’t be solved exactly in terms of known elementary functions. In
each exercise use the Runge-Kutta and the Runge-Kutta semilinear methods with the indicated step sizes to find approximate

values of the solution of the given initial value problem at 11 equally spaced points (including the endpoints) in the interval.
14. y ′
− 2y =
1+x
1
2
, y(2) = 2 ; h = 0.1, 0.05, 0.025 on [2, 3]
15. y ′
+ 2xy = x ,
2
y(0) = 3 ; h = 0.2, 0.1, 0.05on [0, 2] (Exercise 2.1.38)
16. y ′
+
1
x
y =
sin x
x
2
, y(1) = 2; h = 0.2, 0.1, 0.05 on [1, 3] (Exercise 2.1.39)
−x
17. y ′
+y =
e tan x
x
, y(1) = 0; h = 0.05, 0.025, 0.0125 on [1, 1.5] (Exercise 2.1.40)
x
18. y ′
+
1+x
2x
2
y =
e
2
2
, y(0) = 1; h = 0.2, 0.1, 0.05 on [0, 2] (Exercise 2.1.41)
(1+x )
19. x y ′
+ (x + 1)y = e
x
, y(1) = 2 ; h = 0.05, 0.025, 0.0125on [1, 1.5] (Exercise 2.1.42)
Q3.3.4
In Exercises 3.3.20–3.3.22 use the Runge-Kutta method and the Runge-Kutta semilinear method with the indicated step sizes
to find approximate values of the solution of the given initial value problem at 11 equally spaced points (including the
endpoints) in the interval.
20. y ′
+ 3y = x y (y + 1),
2
y(0) = 1 ; h = 0.1, 0.05, 0.025 on [0, 1]
21. y ′
− 4y =
y
2
x
(y+1)
, y(0) = 1 ; h = 0.1, 0.05, 0.025 on [0, 1]
2
22. y ′
+ 2y =
x
1+y 2
, y(2) = 1 ; h = 0.1, 0.05, 0.025 on [2, 3]
Q3.3.5
23. Suppose a < x , so that −x 0 0 < −a . Use the chain rule to show that if z is a solution of
′
z = −f (−x, z), z(−x0 ) = y0 , (3.3E.13)
on [−x 0, −a] , then y = z(−x) is a solution of

′
y = f (x, y), y(x0 ) = y0 , (3.3E.14)
on [a, x ]. 0
of
2 2
′
y + xy − x
y = , y(2) = −1 (3.3E.15)
2
x
at x = 1.1, 1.2, 1.3, …2.0. Compare these approximate values with the values of the exact solution
2
x(4 − 3 x )
y = , (3.3E.16)
2
4 + 3x
which can be obtained by referring to Example

Example 3.3E. 1:
Add text here. For the automatic number to work, you need to add the “AutoNum” template (preferably at 2.4.3}.
of
′ 2 2
y = −x y − x y , y(1) = 1 (3.3E.17)
at x = 0 , 0.1, 0.2, …, 1.
of
1 7 3
′
y + y = + 3, y(1) = (3.3E.18)
2
x x 2

at x = 0.5, 0.6,…, 1.5. Compare these approximate values with the values of the exact solution
7 ln x 3x
y = + , (3.3E.19)
x 2
which can be obtained by the method discussed in Section 2.1.

of
′ 2
x y + 2y = 8 x , y(2) = 5 (3.3E.20)
at x = 1.0, 1.1, 1.2, …, 3.0. Compare these approximate values with the values of the exact solution
2
12
y = 2x − , (3.3E.21)
2
x
which can be obtained by the method discussed in Section 2.1.

28. Numerical Quadrature (see Exercise 3.1.23).
b n−1 n
h h 2h
∫ f (x) dx ≈ (f (a) + f (b)) + ∑ f (a + ih) + ∑ f (a + (2i − 1)h/2) (A)
a 6 3 3
i=1 i=1
(where h = (b − a)/n) by applying the Runge-Kutta method to the initial value problem
′
y = f (x), y(a) = 0. (3.3E.22)
This quadrature formula is called Simpson’s Rule.

b. For several choices of a , b , A , B , C , and D apply (A) to f (x) = A + Bx + C x + Dx , with 3
n = 10 , ,
20 40, 80 , 160,
320. Compare your results with the exact answers and explain what you find.
c. For several choices of , b , A , B , C , D, and E apply (A) to f (x) = A + Bx + C x + Dx

a
2 3
+ Ex
4
, with
. Compare your results with the exact answers and explain what you find.
n = 10, 20, 40, 80, 160, 320

CHAPTER OVERVIEW
4: APPLICATIONS OF FIRST ORDER EQUATIONS
In this chapter, we consider applications of first order differential equations.
4.1: GROWTH AND DECAY

This section begins with a discussion of exponential growth and decay, which you have probably
already seen in calculus. We consider applications to radioactive decay, carbon dating, and
compound interest. We also consider more complicated problems where the rate of change of a
quantity is in part proportional to the magnitude of the quantity, but is also influenced by other
other factors for example, a radioactive susbstance is manufactured at a certain rate, but decays at
a rate proportional
4.1E: GROWTH AND DECAY (EXERCISES)

4.2: COOLING AND MIXING
This section deals with applications of Newton's law of cooling and with mixing problems.
4.2E: COOLING AND MIXING (EXERCISES)

4.3: ELEMENTARY MECHANICS
This section discusses applications to elementary mechanics involving Newton's second law of motion. The problems considered
include motion under the influence of gravity in a resistive medium, and determining the initial velocity required to launch a satellite.
4.3E: ELEMENTARY MECHANICS (EXERCISES)

4.4: AUTONOMOUS SECOND ORDER EQUATIONS
This section deals with methods for dealing with a type of second order equation that often arises in applications of Newton's second
law of motion, by reformulating it as first order equation with a different independent variable. Although the method doesn't usually
lead to an explicit solution of the given equation, it does provide valuable insights into the behavior of the solutions.
4.4E: AUTONOMOUS SECOND ORDER EQUATIONS (EXERCISES)

4.5: APPLICATIONS TO CURVES
This section deals with applications of differential equations to curves.
4.5E: APPLICATIONS TO CURVES (EXERCISES)
1 9/16/2020
4.1: Growth and Decay
This section begins with a discussion of exponential growth and decay, which you have probably already seen in calculus. We
consider applications to radioactive decay, carbon dating, and compound interest. We also consider more complicated
problems where the rate of change of a quantity is in part proportional to the magnitude of the quantity, but is also influenced
by other other factors for example, a radioactive substance is manufactured at a certain rate, but decays at a rate proportional to
its mass, or a saver makes regular deposits in a savings account that draws compound interest.
Since the applications in this section deal with functions of time, we’ll denote the independent variable by t . If Q is a function
of t , Q will denote the derivative of Q with respect to t ; thus,
′
dQ
′
Q = .
dt
Exponential Growth and Decay

One of the most common mathematical models for a physical process is the exponential model, where it is assumed that the
rate of change of a quantity Q is proportional to Q; thus
′
Q = aQ, (4.1.1)
where a is the constant of proportionality.

From Example 2.1.3, the general solution of Equation 4.1.1 is
at
Q = ce
and the solution of the initial value problem

′
Q = aQ, Q(t0 ) = Q0
is
a(t−t0 )
Q = Q0 e . (4.1.2)
Since the solutions of Q = aQ are exponential functions, we say that a quantity

′
Q that satisfies this equation grows
exponentially if a > 0 , or decays exponentially if a < 0 (Figure 4.1.1).
Figure 4.1.1 : Exponential growth and decay
Radioactive Decay
Experimental evidence shows that radioactive material decays at a rate proportional to the mass of the material present.
According to this model the mass Q(t) of a radioactive material present at time t satisfies Equation 4.1.1, where a is a
negative constant whose value for any given material must be determined by experimental observation. For simplicity, we’ll

replace the negative constant a by −k , where k is a positive number that we’ll call the decay constant of the material. Thus,
Equation 4.1.1 becomes
′
Q = −kQ.
If the mass of the material present at t = t is Q , the mass present at time t is the solution of
0 0
′
Q = −kQ, Q(t0 ) = Q0 .
From Equation 4.1.2 with a = −k , the solution of this initial value problem is
−k(t−t0 )
Q = Q0 e . (4.1.3)
The half–life τ of a radioactive material is defined to be the time required for half of its mass to decay; that is, if Q(t 0) = Q0 ,
then
Q0
Q(τ + t0 ) = . (4.1.4)
2
From Equation 4.1.3 with t = τ + t , Equation 4.1.4 is equivalent to

0
−kτ
Q0
Q0 e = ,
2
so
1
−kτ
e = .
2
Taking logarithms yields

1
−kτ = ln = − ln 2,
2
so the half-life is
1
τ = ln 2. (4.1.5)
k
(Figure 4.1.2). The half-life is independent of t and 0 Q0 , since it is determined by the properties of material, not by the
amount of the material present at any particular time.
Example 4.1.1
A radioactive substance has a half-life of 1620 years.
a. If its mass is now 4 g (grams), how much will be left 810 years from now?
b. Find the time t when 1.5 g of the substance remain.
1
Solution a
From Equation 4.1.3 with t 0 =0 and Q
0 =4 ,
−kt
Q = 4e , (4.1.6)
where we determine k from Equation 4.1.5, with τ = 1620 years:

ln 2 ln 2
k = = .
τ 1620
Substituting this in Equation 4.1.6 yields

−(t ln 2)/1620
Q = 4e . (4.1.7)

Figure 4.1.2 : Half-life of a radioactive substance
Therefore the mass left after 810 years will be
−(810 ln 2)/1620 −(ln 2)/2
Q(810) = 4e = 4e
–
= 2 √2 g.
Solution b
Setting t = t in Equation 4.1.7 and requiring that Q(t
1 1) = 1.5 yields
3
(−t1 ln 2)/1620
= 4e .
2
Dividing by 4 and taking logarithms yields

3 t1 ln 2
ln =− .
8 1620
Since ln 3/8 = − ln 8/3,

ln 8/3
t1 = 1620 ≈ 2292.4 years.
ln 2
Interest Compounded Continuously

Suppose we deposit an amount of money Q in an interest-bearing account and make no further deposits or withdrawals for t
0
years, during which the account bears interest at a constant annual rate r. To calculate the value of the account at the end of t
years, we need one more piece of information: how the interest is added to the account, or—as the bankers say—how it is
compounded. If the interest is compounded annually, the value of the account is multiplied by 1 + r at the end of each year.
This means that after t years the value of the account is
t
Q(t) = Q0 (1 + r) .
If interest is compounded semiannually, the value of the account is multiplied by (1 + r/2) every 6 months. Since this occurs
twice annually, the value of the account after t years is
2t
r
Q(t) = Q0 (1 + ) .
2
In general, if interest is compounded n times per year, the value of the account is multiplied n times per year by (1 + r/n) ;
therefore, the value of the account after t years is
nt
r
Q(t) = Q0 (1 + ) . (4.1.8)
n

Thus, increasing the frequency of compounding increases the value of the account after a fixed period of time. Table 4.1.1
shows the effect of increasing the number of compoundings over t = 5 years on an initial deposit of Q = 100 (dollars), at an 0
annual interest rate of 6%.

Table 4.1.1 : The effect of compound interest
5n
n (number of compoundings per year) $100(1 +
. 06
n
) (value in dollars after 5 years)
1 $133.82
2 $134.39
4 $134.68
8 $134.83
364 $134.98
You can see from Table 4.1.1 that the value of the account after 5 years is an increasing function of n . Now suppose the
maximum allowable rate of interest on savings accounts is restricted by law, but the time intervals between successive
compoundings isn’t; then competing banks can attract savers by compounding often. The ultimate step in this direction is to
compound continuously, by which we mean that n → ∞ in Equation 4.1.8. Since we know from calculus that
r n
r
lim (1 + ) =e ,
n→∞ n
this yields
nt n t
r r
Q(t) = limn→∞ Q0 (1 + ) = Q0 [ limn→∞ (1 + ) ]
n n
rt
= Q0 e .
Observe that Q = Q 0e
rt
is the solution of the initial value problem
′
Q = rQ, Q(0) = Q0 ;
that is, with continuous compounding the value of the account grows exponentially.
Example 4.1.2
If $150 is deposited in a bank that pays 5
1
2
% annual interest compounded continuously, the value of the account after t
years is
.055t
Q(t) = 150e
dollars. (Note that it is necessary to write the interest rate as a decimal; thus, r = .055.) Therefore, after t = 10 years the
value of the account is
.55
Q(10) = 150 e ≈ $259.99.
Example 4.1.3
We wish to accumulate $10,000 in 10 years by making a single deposit in a savings account bearing 5 % annual interest 1
compounded continuously. How much must we deposit in the account?

Solution
The value of the account at time t is
.055t
Q(t) = Q0 e . (4.1.9)
Since we want Q(10) to be $10,000, the initial deposit Q must satisfy the equation
0
.55
10000 = Q0 e , (4.1.10)

obtained by setting t = 10 and Q(10) = 10000 in Equation 4.1.9. Solving Equation 4.1.10 for Q yields 0
−.55
Q0 = 10000 e ≈ $5769.50.
Mixed Growth and Decay
Example 4.1.4
A radioactive substance with decay constant k is produced at a constant rate of a units of mass per unit time.
a. Assuming that Q(0) = Q , find the mass Q(t) of the substance present at time t .
0
b. Find lim Q(t).

t→∞
Solution a:
Here
′
Q = rate of increase of Q − rate of decrease of Q.
The rate of increase is the constant a . Since Q is radioactive with decay constant k , the rate of decrease is kQ. Therefore
′
Q = a − kQ.
This is a linear first order differential equation. Rewriting it and imposing the initial condition shows that Q is the solution
′
Q + kQ = a, Q(0) = Q0 . (4.1.11)
Since e is a solution of the complementary equation, the solutions of Equation

−kt
4.1.11 are of the form Q = ue
−kt
,
where u e′
= a , so u = ae
−kt ′
. Hence,
kt
a kt
u = e +c
k
Figure 4.1.3 : Q(t) approaches the steady state value a
k
as t → ∞
and
−kt
a −kt
Q = ue = + ce .
k
Since Q(0) = Q , setting t = 0 here yields

0
a a
Q0 = +c or c = Q0 − .
k k
Therefore

a a
−kt
Q = + (Q0 − )e . (4.1.12)
k k
b. Since k > 0 , limt→∞ e

−kt
=0 , so from Equation 4.1.12
a
lim Q(t) = .
t→∞ k
This limit depends only on a and k , and not on Q . We say that a/k is the steady state value of Q. From Equation 4.1.12
0
we also see that Q approaches its steady state value from above if Q > a/k , or from below if Q < a/k . If Q = a/k ,
0 0 0
then Q remains constant (Figure 4.1.3).
Carbon Dating
The fact that Q approaches a steady state value in the situation discussed in Example 4 underlies the method of carbon dating,
devised by the American chemist and Nobel Prize Winner W.S. Libby.
Carbon 12 is stable, but carbon-14, which is produced by cosmic bombardment of nitrogen in the upper atmosphere, is
radioactive with a half-life of about 5570 years. Libby assumed that the quantity of carbon-12 in the atmosphere has been
constant throughout time, and that the quantity of radioactive carbon-14 achieved its steady state value long ago as a result of
its creation and decomposition over millions of years. These assumptions led Libby to conclude that the ratio of carbon-14 to
carbon-12 has been nearly constant for a long time. This constant, which we denote by R , has been determined experimentally.
Living cells absorb both carbon-12 and carbon-14 in the proportion in which they are present in the environment. Therefore
the ratio of carbon-14 to carbon-12 in a living cell is always R . However, when the cell dies it ceases to absorb carbon, and the
ratio of carbon-14 to carbon-12 decreases exponentially as the radioactive carbon-14 decays. This is the basis for the method
of carbon dating, as illustrated in the next example.
Example 4.1.5
An archaeologist investigating the site of an ancient village finds a burial ground where the amount of carbon-14 present
in individual remains is between 42 and 44% of the amount present in live individuals. Estimate the age of the village and
the length of time for which it survived.
Solution
Let Q = Q(t) be the quantity of carbon-14 in an individual set of remains t years after death, and let Q be the quantity 0
that would be present in live individuals. Since carbon-14 decays exponentially with half-life 5570 years, its decay
constant is
ln 2
k = .
5570
Therefore
−t(ln 2)/5570
Q = Q0 e
if we choose our time scale so that t = 0 is the time of death. If we know the present value of
0 Q we can solve this
equation for t , the number of years since death occurred. This yields
ln Q/Q0
t = −5570 .
ln 2
It is given that Q = .42Q in the remains of individuals who died first. Therefore these deaths occurred about
0
ln .42
t1 = −5570 ≈ 6971
ln 2
years ago. For the most recent deaths, Q = .44Q ; hence, these deaths occurred about
0
ln .44
t2 = −5570 ≈ 6597
ln 2

years ago. Therefore it is reasonable to conclude that the village was founded about 7000 years ago, and lasted for about
400 years.
A Savings Program
Example 4.1.6 :
A person opens a savings account with an initial deposit of $1000 and subsequently deposits $50 per week. Find the value
Q(t) of the account at time t > 0 , assuming that the bank pays 6% interest compounded continuously.
Solution
Observe that Q isn’t continuous, since there are 52 discrete deposits per year of $50 each. To construct a mathematical
model for this problem in the form of a differential equation, we make the simplifying assumption that the deposits are
made continuously at a rate of $2600 per year. This is essential, since solutions of differential equations are continuous
functions. With this assumption, Q increases continuously at the rate
′
Q = 2600 + 0.06Q
and therefore Q satisfies the differential equation

′
Q − .06Q = 2600. (4.1.13)
(Of course, we must recognize that the solution of this equation is an approximation to the true value of Q at any given
time. We’ll discuss this further below.) Since e is a solution of the complementary equation, the solutions of Equation
.06t
4.1.13 are of the form Q = ue , where u e = 2600 . Hence, u = 2600 e ,

.06t ′ .06t ′ −.06t
2600
−0.06t
u =− e +c
.06
and
.06t
2600 .06t
Q = ue =− + ce . (4.1.14)
.06
Setting t = 0 and Q = 1000 here yields

2600
c = 1000 + ,
0.06
and substituting this into Equation 4.1.14 yields

2600
.06t .06t
Q = 1000 e + (e − 1) (4.1.15)
.06
where the first term is the value due to the initial deposit and the second is due to the subsequent weekly deposits.
Mathematical models must be tested for validity by comparing predictions based on them with the actual outcome of
experiments. Example 6 is unusual in that we can compute the exact value of the account at any specified time and compare it
with the approximate value predicted by Equation 4.1.15 (See Exercise 4.1.21). The following table gives a comparison for a
ten year period. Each exact answer corresponds to the time of the year-end deposit, and each year is assumed to have exactly
52 weeks.
Table 4.1.3
Approximate Value of Q Exact Value of P Percentage Error
Year Error Q − P
(Example 4.1.6) (Exercise 4.1.21) (Q − P )/P
1 $3741.42 $3739.87 $1.55 .0413
2 6652.36 6649.17 3.19 .0479
3 9743.30 9738.37 4.93 .0506

Approximate Value of Q Exact Value of P Percentage Error
Year Error Q − P
(Example 4.1.6) (Exercise 4.1.21) (Q − P )/P
4 13, 025.38 13, 018.60 6.78 .0521
5 16, 510.41 16, 501.66 8.75 .0530
6 20, 210.94 20, 200.11 10.83 .0536
7 24, 140.30 24, 127.25 13.05 .0541
8 28, 312.63 28, 297.23 15.40 .0544
9 32, 742.97 32, 725.07 17.90 .0547
10 37, 447.27 37, 426.72 20.55 .0549

4.1E: Growth and Decay (Exercises)
Q4.1.1
1. The half-life of a radioactive substance is 3200 years. Find the quantity Q(t) of the substance left at time t >0 if
Q(0) = 20 g.
2. The half-life of a radioactive substance is 2 days. Find the time required for a given amount of the material to decay to 1/10
of its original mass.
3. A radioactive material loses 25% of its mass in 10 minutes. What is its half-life?
4. A tree contains a known percentage p of a radioactive substance with half-life τ . When the tree dies the substance decays
0
and isn’t replaced. If the percentage of the substance in the fossilized remains of such a tree is found to be p , how long has the
1
tree been dead?

5. If t and t are the times required for a radioactive material to decay to 1/p and 1/q times its original mass (respectively),
p q
how are t and t related?

p q
6. Find the decay constant k for a radioactive substance, given that the mass of the substance is Q at time t and Q at time
1 1 2
t .
2
7. A process creates a radioactive substance at the rate of 2 g/hr and the substance decays at a rate proportional to its mass,
with constant of proportionality k = .1(hr) . If Q(t) is the mass of the substance at time t , find lim
−1
Q(t).
t→∞
8. A bank pays interest continuously at the rate of 6%. How long does it take for a deposit of Q to grow in value to 2Q ?
0 0
9. At what rate of interest, compounded continuously, will a bank deposit double in value in 8 years?
10. A savings account pays 5% per annum interest compounded continuously. The initial deposit is Q0 dollars. Assume that
there are no subsequent withdrawals or deposits.
a. How long will it take for the value of the account to triple?
b. What is Q if the value of the account after 10 years is $100,000 dollars?
0
11. A candymaker makes 500 pounds of candy per week, while his large family eats the candy at a rate equal to Q(t)/10
pounds per week, where Q(t) is the amount of candy present at time t .
a. Find Q(t) for t > 0 if the candymaker has 250 pounds of candy at t = 0 .
b. Find lim Q(t).
t→∞
12. Suppose a substance decays at a yearly rate equal to half the square of the mass of the substance present. If we start with
50 g of the substance, how long will it be until only 25 g remain?
13. A super bread dough increases in volume at a rate proportional to the volume V present. If V increases by a factor of 10 in
2 hours and V (0) = V , find V at any time t . How long will it take for V to increase to 100V ?
0 0
14. A radioactive substance decays at a rate proportional to the amount present, and half the original quantity Q is left after
0
1500 years. In how many years would the original amount be reduced to 3Q /4? How much will be left after 2000 years?
0
15. A wizard creates gold continuously at the rate of 1 ounce per hour, but an assistant steals it continuously at the rate of 5%
of however much is there per hour. Let W (t) be the number of ounces that the wizard has at time t . Find W (t) and
lim t→∞W (t) if W (0) = 1 .
16. A process creates a radioactive substance at the rate of 1 g/hr, and the substance decays at an hourly rate equal to 1/10 of
the mass present (expressed in grams). Assuming that there are initially 20 g, find the mass S(t) of the substance present at
time t , and find lim S(t) .
t→∞
17. A tank is empty at t = 0 . Water is added to the tank at the rate of 10 gal/min, but it leaks out at a rate (in gallons per
minute) equal to the number of gallons in the tank. What is the smallest capacity the tank can have if this process is to continue
forever?

18. A person deposits $25,000 in a bank that pays 5% per year interest, compounded continuously. The person continuously
withdraws from the account at the rate of $750 per year. Find V (t), the value of the account at time t after the initial deposit.
19. A person has a fortune that grows at rate proportional to the square root of its worth. Find the worth W of the fortune as a
function of t if it was $1 million 6 months ago and is $4 million today.
20. Let p = p(t) be the quantity of a product present at time t . The product is manufactured continuously at a rate proportional
to p, with proportionality constant 1/2, and it is consumed continuously at a rate proportional to p , with proportionality
2
constant 1/8. Find p(t) if p(0) = 100.

21.
a. In the situation of Example 4.1.6 find the exact value P(t) of the person’s account after t years, where t is an integer. Assume
that each year has exactly 52 weeks, and include the year-end deposit in the computation.
HINT: At time t the initial $1000 has been on deposit for t years. There have been 52t deposits of $50 each. The first $50 has
been on deposit for t − 1/52 years, the second for t − 2/52 years ... in general, the j th $50 has been on deposit for t − j/52
years (1 ≤ j ≤ 52t ). Find the present value of each $50 deposit assuming 6% interest compounded continuously, and use the
formula
n+1
2 n
1 −x
1 + x + x +. . . +x = (x ≠ 1) (4.1E.1)
1 −x
to find their total value.

b. Let
Q(t) − P (t)
p(t) = (4.1E.2)
P (t)
be the relative error after t years. Find

p(∞) = lim p(t). (4.1E.3)
t→∞
22. A homebuyer borrows P dollars at an annual interest rate r, agreeing to repay the loan with equal monthly payments of
0
M dollars per month over N years.
a. Derive a differential equation for the loan principal (amount that the homebuyer owes) P (t) at time t > 0 , making the
simplifying assumption that the homebuyer repays the loan continuously rather than in discrete steps. (See Example 4.1.6.)
b. Solve the equation derived in (a).
c. Use the result of (b) to determine an approximate value for M assuming that each year has exactly 12 months of equal
length.
d. It can be shown that the exact value of M is given by
rP0 −1
−12N
M = (1 − (1 + r/12 ) ) . (4.1E.4)
12
Compare the value of M obtained from the answer in (c) to the exact value if (i) P 0 = $50, 000 , $r=7{1\over2}$%, N = 20
(ii) P = $150, 000, r = 9.0 %, N = 30 .

0
23. Assume that the homebuyer of Exercise 4.1.22 elects to repay the loan continuously at the rate of αM dollars per month,
where α is a constant greater than 1. (This is called accelerated payment.)
a. Determine the time T (α) when the loan will be paid off and the amount S(α) that the homebuyer will save.
b. Suppose P = $50, 000, r = 8 %, and N = 15 . Compute the savings realized by accelerated payments with
0
α = 1.05, 1.10, and 1.15.
24. A benefactor wishes to establish a trust fund to pay a researcher’s salary for T years. The salary is to start at S dollars per0
year and increase at a fractional rate of a per year. Find the amount of money P that the benefactor must deposit in a trust
0
fund paying interest at a rate r per year. Assume that the researcher’s salary is paid continuously, the interest is compounded
continuously, and the salary increases are granted continuously.

25. A radioactive substance with decay constant k is produced at the rate of
at
(4.1E.5)
1 + btQ(t)
units of mass per unit time, where a and b are positive constants and Q(t) is the mass of the substance present at time t; thus,
the rate of production is small at the start and tends to slow when Q is large.
a. Set up a differential equation for Q.
b. Choose your own positive values for a , b , k , and Q = Q(0) . Use a numerical method to discover what happens to
0 Q(t)
as t → ∞ . (Be precise, expressing your conclusions in terms of a , b , k . However, no proof is required.)

26. Follow the instructions of Exercise 4.1.25, assuming that the substance is produced at the rate of at/(1 + bt(Q(t)) 2
) units
of mass per unit of time.
27. Follow the instructions of Exercise 4.1.25, assuming that the substance is produced at the rate of at/(1 + bt) units of mass
per unit of time.

4.2: Cooling and Mixing
Newton’s law of cooling states that if an object with temperature T (t) at time t is in a medium with temperature Tm (t) , the
rate of change of T at time t is proportional to T (t) − T (t) ; thus, T satisfies a differential equation of the form
m
′
T = −k(T − Tm ). (4.2.1)
Here k > 0 , since the temperature of the object must decrease if T > Tm , or increase if T < Tm . We’ll call k the temperature
decay constant of the medium.
For simplicity, in this section we’ll assume that the medium is maintained at a constant temperature T . This is another m
example of building a simple mathematical model for a physical phenomenon. Like most mathematical models it has its
limitations. For example, it is reasonable to assume that the temperature of a room remains approximately constant if the
cooling object is a cup of coffee, but perhaps not if it is a huge cauldron of molten metal. (For more on this see Exercise
4.2.17.)
To solve Equation 4.2.1, we rewrite it as
′
T + kT = kTm .
Since e is a solution of the complementary equation, the solutions of this equation are of the form
−kt
T = ue
−kt
, where
′
u e
−kt
= kT , so u = kT e . Hence,
m
′
m
kt
kt
u = Tm e + c,
so
−kt −kt
T = ue = Tm + c e .
If T (0) = T , setting t = 0 here yields c = T

0 0 − Tm , so
−kt
T = Tm + (T0 − Tm )e . (4.2.2)
Note that T − T m decays exponentially, with decay constant k .
Example 4.2.1
A ceramic insulator is baked at 400 C and cooled in a room in which the temperature is
∘
25
∘
C. After 4 minutes the
temperature of the insulator is 200 C. What is its temperature after 8 minutes?
∘
Solution
Here T 0 = 400 and Tm = 25 , so Equation 4.2.2 becomes
−kt
T = 25 + 375 e . (4.2.3)
We determine k from the stated condition that T (4) = 200 ; that is,
−4k
200 = 25 + 375 e ;
hence,
175 7
−4k
e = = .
375 15
Taking logarithms and solving for k yields

1 7 1 15
k =− ln = ln .
4 15 4 7

t 15
− ln
T = 25 + 375e 4 7
(Figure 4.2.1). Therefore the temperature of the insulator after 8 minutes is

15
−2 ln
T (8) = 25 + 375e 7
2
7 ∘
= 25 + 375 ( ) ≈ 107 C.
15
Example 4.2.2
An object with temperature 72 F is placed outside, where the temperature is −20 F. At 11:05 the temperature of the
∘ ∘
object is 60 F and at 11:07 its temperature is 50 F. At what time was the object placed outside?
∘ ∘
Solution
Let T (t) be the temperature of the object at time t . For convenience, we choose the origin t 0 =0 of the time scale to be
11:05 so that T = 60 . We must determine the time τ when T (τ ) = 72 . Substituting T
0 0 = 60 and T = −20 into
m
Equation 4.2.2 yields

−kt
T = −20 + (60 − (−20))e
or
−kt
T = −20 + 80 e . (4.2.4)
Figure 4.2.1 : T = 25 + 375e

−(t/4) ln 15/7
We obtain k from the stated condition that the temperature of the object is 50 F at 11:07. Since 11:07 is t = 2 on our time
∘
scale, we can determine k by substituting T = 50 and t = 2 into Equation 4.2.4 to obtain

−2k
50 = −20 + 80e
This is shown in Figure 4.2.2.

Hence,
70 7
−2k
e = = .
80 8
Taking logarithms and solving for k yields

1 7 1 8
k =− ln = ln .
2 8 2 7

t 8
− ln
T = −20 + 80 e 2 7
,
and the condition T (τ ) = 72 implies that

τ 8
− ln
72 = −20 + 80 e 2 7 ;
hence,
−
τ
ln
8
92 23
e 2 7
= = .
80 20
Taking logarithms and solving for τ yields

23
2 ln
20
τ =− ≈ −2.09 min.
8
ln
7
t 8
Figure 4.2.2 : −20 + 80e −

2
ln
7
Therefore the object was placed outside about 2 minutes and 5 seconds before 11:05; that is, at 11:02:55.
Mixing Problems
In the next two examples a saltwater solution with a given concentration (weight of salt per unit volume of solution) is added
at a specified rate to a tank that initially contains saltwater with a different concentration. The problem is to determine the
quantity of salt in the tank as a function of time. This is an example of a mixing problem. To construct a tractable mathematical
model for mixing problems we assume in our examples (and most exercises) that the mixture is stirred instantly so that the salt
is always uniformly distributed throughout the mixture. Exercises 4.2.22 and 4.2.23 deal with situations where this isn’t so, but
the distribution of salt becomes approximately uniform as t → ∞ .
Example 4.2.3
A tank initially contains 40 pounds of salt dissolved in 600 gallons of water. Starting at t = 0 , water that contains 1/2
0
pound of salt per gallon is poured into the tank at the rate of 4 gal/min and the mixture is drained from the tank at the
same rate (Figure 4.2.3).
a. Find a differential equation for the quantity Q(t) of salt in the tank at time t > 0 , and solve the equation to determine
Q(t).
b. Find lim t→∞ Q(t).
Solution a
To find a differential equation for Q, we must use the given information to derive an expression for Q . But Q is the rate ′ ′
of change of the quantity of salt in the tank changes with respect to time; thus, if rate in denotes the rate at which salt

enters the tank and rate out denotes the rate by which it leaves, then
′
Q = rate in − rate out. (4.2.5)
Figure 4.2.3 : A mixing problem

The rate in is
1
( lb/gal) × (4 gal/min) = 2 lb/min.
2
Determining the rate out requires a little more thought. We’re removing 4 gallons of the mixture per minute, and there are
always 600 gallons in the tank; that is, we are removing 1/150 of the mixture per minute. Since the salt is evenly
distributed in the mixture, we are also removing 1/150 of the salt per minute. Therefore, if there are Q(t) pounds of salt
in the tank at time t , the rate out at any time t is Q(t)/150. Alternatively, we can arrive at this conclusion by arguing that
rate out = (concentration) × (rate of flow out)
= (lb/gal) × (gal/min)
Q(t) Q(t)
= ×4 = .
600 150
We can now write Equation 4.2.5 as

Q
′
Q =2− .
150
This first order equation can be rewritten as

Q
′
Q + = 2.
150
Since e −t/150
is a solution of the complementary equation, the solutions of this equation are of the form Q = ue
−t/150
,
where u e′ −t/150
= 2 , so u = 2 e
′ t/150
. Hence,
t/150
u = 300 e + c,

Figure 4.2.4 : Q = 300 − 260e −t/150
so
−t/150 −t/150
Q = ue = 300 + c e (4.2.6)
(Figure 4.2.4). Since Q(0) = 40, c = −260 ; therefore,

−t/150
Q = 300 − 260 e .
Solution b
From Equation 4.2.6, we see that that lim Q(t) = 300 for any value of Q(0). This is intuitively reasonable, since the
t→∞
incoming solution contains 1/2 pound of salt per gallon and there are always 600 gallons of water in the tank.
Example 4.2.4
A 500-liter tank initially contains 10 g of salt dissolved in 200 liters of water. Starting at t = 0 , water that contains 1/4 g
0
of salt per liter is poured into the tank at the rate of 4 liters/min and the mixture is drained from the tank at the rate of 2
liters/min (Figure [figure:4.2.5}). Find a differential equation for the quantity Q(t) of salt in the tank at time t prior to the
time when the tank overflows and find the concentration K(t) (g/liter) of salt in the tank at any such time.
Solution
We first determine the amount W (t) of solution in the tank at any time t prior to overflow. Since W (0) = 200 and we are
adding 4 liters/min while removing only 2 liters/min, there’s a net gain of 2 liters/min in the tank; therefore,
W (t) = 2t + 200.
Since W (150) = 500 liters (capacity of the tank), this formula is valid for 0 ≤ t ≤ 150 .
Now let Q(t) be the number of grams of salt in the tank at time t , where 0 ≤ t ≤ 150 . As in Example 4.2.3
′
Q = rate in − rate out. (4.2.7)
Figure 4.2.5 : Another mixing problem

The rate in is
1
( g/liter ) × (4 liters/min ) = 1 g/min. (4.2.8)
4
To determine the rate out, we observe that since the mixture is being removed from the tank at the constant rate of 2
liters/min and there are 2t + 200 liters in the tank at time t, the fraction of the mixture being removed per minute at time t
is
2 1
= .
2t + 200 t + 100
We’re removing this same fraction of the salt per minute. Therefore, since there are Q(t) grams of salt in the tank at time
t,
Q(t)
rate out = . (4.2.9)
t + 100
Alternatively, we can arrive at this conclusion by arguing that

rate out = (concentration) × (rate of flow out) = (g/liter) × (liters/min)
Q(t) Q(t)
= ×2 = .
2t+200 t+100
Substituting Equation 4.2.8and Equation 4.2.9into Equation 4.2.7yields

Q 1
′ ′
Q =1− , so Q + Q = 1. (4.2.10)
t + 100 t + 100
By separation of variables, 1/(t + 100) is a solution of the complementary equation, so the solutions of Equation 4.2.10
are of the form
′
u u
′
Q = , where , so u = t + 100.
t + 100 t + 100 = 1
Hence,
2
(t + 100)
u = + c. (4.2.11)
2
Since Q(0) = 10 and u = (t + 100)Q , Equation 4.2.11implies that

2
(100)
(100)(10) = + c,
2
so
2
(100)
c = 100(10) − = −4000
2
and therefore
2
(t + 100)
u = − 4000.
2
Hence,
u t + 100 4000
Q = = − .
t + 200 2 t + 100
Now let K(t) be the concentration of salt at time t. Then

1 2000
K(t) = −
2
4 (t + 100)
This is shown in Figure 4.2.6.

Figure 4.2.6 : K(t) = 1
4
−
2000
2
(t+100)

4.2E: Cooling and Mixing (Exercises)
Q4.2.1
1. A thermometer is moved from a room where the temperature is 70 F to a freezer where the temperature is 12
∘ ∘
F . After 30
seconds the thermometer reads 40 F. What does it read after 2 minutes?
∘
2. A fluid initially at 100 C is placed outside on a day when the temperature is −10 C, and the temperature of the fluid drops
∘ ∘
20 C in one minute. Find the temperature T (t) of the fluid for t > 0 .
∘
3. At 12:00 pm a thermometer reading 10 F is placed in a room where the temperature is 70 F. It reads 56 when it is placed
∘ ∘ ∘
outside, where the temperature is 5 F, at 12:03. What does it read at 12:05 pm?
∘
4. A thermometer initially reading 212 F is placed in a room where the temperature is 70 F. After 2 minutes the thermometer
∘ ∘
reads 125 F.
∘
a. What does the thermometer read after 4 minutes?

b. When will the thermometer read 72 F? ∘
c. When will the thermometer read 69 F? ∘
5. An object with initial temperature 150 C is placed outside, where the temperature is
∘
35
∘
C. Its temperatures at 12:15 and
12:20 are 120 C and 90 C, respectively.
∘ ∘
a. At what time was the object placed outside?

b. When will its temperature be 40 C? ∘
6. An object is placed in a room where the temperature is 20 C. The temperature of the object drops by 5 C in 4 minutes and
∘ ∘
by 7 C in 8 minutes. What was the temperature of the object when it was initially placed in the room?
∘
7. A cup of boiling water is placed outside at 1:00 pm. One minute later the temperature of the water is 152
∘
F. After another
minute its temperature is 112 F. What is the outside temperature?
∘
8. A tank initially contains 40 gallons of pure water. A solution with 1 gram of salt per gallon of water is added to the tank at 3
gal/min, and the resulting solution drains out at the same rate. Find the quantity Q(t) of salt in the tank at time t > 0 .
9. A tank initially contains a solution of 10 pounds of salt in 60 gallons of water. Water with 1/2 pound of salt per gallon is
added to the tank at 6 gal/min, and the resulting solution leaves at the same rate. Find the quantity Q(t) of salt in the tank at
time t > 0 .
10. A tank initially contains 100 liters of a salt solution with a concentration of .1 g/liter. A solution with a salt concentration
of .3 g/liter is added to the tank at 5 liters/min, and the resulting mixture is drained out at the same rate. Find the concentration
K(t) of salt in the tank as a function of t .
11. A 200 gallon tank initially contains 100 gallons of water with 20 pounds of salt. A salt solution with 1/4 pound of salt per
gallon is added to the tank at 4 gal/min, and the resulting mixture is drained out at 2 gal/min. Find the quantity of salt in the
tank as it is about to overflow.
12. Suppose water is added to a tank at 10 gal/min, but leaks out at the rate of 1/5 gal/min for each gallon in the tank. What is
the smallest capacity the tank can have if the process is to continue indefinitely?
13. A chemical reaction in a laboratory with volume V (in ft ) produces q ft /min of a noxious gas as a byproduct. The gas is
3
1
3
dangerous at concentrations greater than c̄, but harmless at concentrations ≤ c̄ . Intake fans at one end of the laboratory pull in
¯ ¯
fresh air at the rate of q ft /min and exhaust fans at the other end exhaust the mixture of gas and air from the laboratory at the
2
3
same rate. Assuming that the gas is always uniformly distributed in the room and its initial concentration c is at a safe level,
0
find the smallest value of q required to maintain safe conditions in the laboratory for all time.
2
14. A 1200-gallon tank initially contains 40 pounds of salt dissolved in 600 gallons of water. Starting at t = 0 , water that
0
contains 1/2 pound of salt per gallon is added to the tank at the rate of 6 gal/min and the resulting mixture is drained from the
tank at 4 gal/min. Find the quantity Q(t) of salt in the tank at any time t > 0 prior to overflow.

15. Tank T initially contain 50 gallons of pure water. Starting at t = 0 , water that contains 1 pound of salt per gallon is
1 0
poured into T at the rate of 2 gal/min. The mixture is drained from T at the same rate into a second tank T , which initially
1 1 2
contains 50 gallons of pure water. Also starting at t = 0 , a mixture from another source that contains 2 pounds of salt per
0
gallon is poured into T at the rate of 2 gal/min. The mixture is drained from T at the rate of 4 gal/min.
2 2
a. Find a differential equation for the quantity Q(t) of salt in tank T at time t > 0 . 2
b. Solve the equation derived in (a) to determine Q(t).

c. Find lim Q(t).
t→∞
16. Suppose an object with initial temperature T is placed in a sealed container, which is in turn placed in a medium with
0
temperature T . Let the initial temperature of the container be S . Assume that the temperature of the object does not affect
m 0
the temperature of the container, which in turn does not affect the temperature of the medium. (These assumptions are
reasonable, for example, if the object is a cup of coffee, the container is a house, and the medium is the atmosphere.)
a. Assuming that the container and the medium have distinct temperature decay constants k and k respectively, use m
Newton’s law of cooling to find the temperatures S(t) and T (t) of the container and object at time t .
b. Assuming that the container and the medium have the same temperature decay constant k , use Newton’s law of cooling to
find the temperatures S(t) and T (t) of the container and object at time t .
c. Find lim. S(t) and lim
t→∞ T (t) . t→∞
17. In our previous examples and exercises concerning Newton’s law of cooling we assumed that the temperature of the
medium remains constant. This model is adequate if the heat lost or gained by the object is insignificant compared to the heat
required to cause an appreciable change in the temperature of the medium. If this isn’t so, we must use a model that accounts
for the heat exchanged between the object and the medium. Let T = T (t) and T = T (t) be the temperatures of the object m m
and the medium, respectively, and let T and T be their initial values. Again, we assume that T and T are related by
0 m0 m
Newton’s law of cooling,

′
T = −k(T − Tm ). (A)
We also assume that the change in heat of the object as its temperature changes from T to T is a(T − T ) and that the change 0 0
in heat of the medium as its temperature changes from T to T is a (T − T ) , where a and a are positive constants
m0 m m m m0 m
depending upon the masses and thermal properties of the object and medium, respectively. If we assume that the total heat of
the system consisting of the object and the medium remains constant (that is, energy is conserved), then
a(T − T0 ) + am (Tm − Tm0 ) = 0. (B)
a. Equation (A) involves two unknown functions T and T . Use (A) and (B) to derive a differential equation involving only
m
T.
b. Find T (t) and T (t) for t > 0 .

m
c. Find lim T (t) and lim

t→∞ T (t) . t→∞ m
18. Control mechanisms allow fluid to flow into a tank at a rate proportional to the volume V of fluid in the tank, and to flow
out at a rate proportional to V . Suppose V (0) = V and the constants of proportionality are a and b , respectively. Find V (t)
2
0
for t > 0 and find lim V (t) . t→∞
19. Identical tanks T and T initially contain W gallons each of pure water. Starting at t = 0 , a salt solution with constant
1 2 0
concentration c is pumped into T at r gal/min and drained from T into T at the same rate. The resulting mixture in T is
1 1 2 2
also drained at the same rate. Find the concentrations c (t) and c (t) in tanks T and T for t > 0 .
1 2 1 2
20. An infinite sequence of identical tanks T , T , …, T , …, initially contain W gallons each of pure water. They are hooked
1 2 n
together so that fluid drains from T into T (n = 1, 2, ⋯). A salt solution is circulated through the tanks so that it enters
n n+1
and leaves each tank at the constant rate of r gal/min. The solution has a concentration of c pounds of salt per gallon when it
enters T .
1
a. Find the concentration c (t) in tank T for t > 0 .

n n
b. Find lim c (t) for each n .

t→∞ n
21. Tanks T and T have capacities W and W liters, respectively. Initially they are both full of dye solutions with
1 2 1 2
concentrations c and c grams per liter. Starting at t = 0 , the solution from T is pumped into T at a rate of r liters per
1 2 0 1 2

minute, and the solution from T is pumped into T at the same rate.
2 1
a. Find the concentrations c (t) and c (t) of the dye in T and T for t > 0 .
1 2 1 2
b. Find lim t→∞ c (t) and lim

1 c (t) .
t→∞ 2
22. Consider the mixing problem of Example 4.2.3, but without the assumption that the mixture is stirred instantly so that the
salt is always uniformly distributed throughout the mixture. Assume instead that the distribution approaches uniformity as
t → ∞ . In this case the differential equation for Q is of the form
a(t)
′
Q + Q =2 (4.2E.1)
150
where lim t→∞ a(t) = 1 .

a. Assuming that Q(0) = Q , can you guess the value of lim
0 Q(t)?. t→∞
b. Use numerical methods to confirm your guess in the these cases:

2
−t −t
(i) a(t) = t/(1 + t) (ii) a(t) = 1 − e (iii) a(t) = 1 − sin(e ). (4.2E.2)
23. Consider the mixing problem of Example 4.2.4 in a tank with infinite capacity, but without the assumption that the mixture
is stirred instantly so that the salt is always uniformly distributed throughout the mixture. Assume instead that the distribution
approaches uniformity as t → ∞ . In this case the differential equation for Q is of the form
a(t)
′
Q + Q =1 (4.2E.3)
t + 100
where lim t→∞ a(t) = 1 .

a. Let K(t) be the concentration of salt at time t . Assuming that Q(0) = Q , can you guess the value of lim
0 t→∞ K(t) ?
b. Use numerical methods to confirm your guess in the these cases:
2
−t −t
(i) a(t) = t/(1 + t) (ii) a(t) = 1 − e (iii) a(t) = 1 + sin(e ). (4.2E.4)

4.3: Elementary Mechanics
Newton's Second Law of Motion
In this section we consider an object with constant mass m moving along a line under a force F . Let y = y(t) be the
displacement of the object from a reference point on the line at time t , and let v = v(t) and a = a(t) be the velocity and
acceleration of the object at time t . Thus, v = y and a = v = y , where the prime denotes differentiation with respect to t .
′ ′ ′′
Newton’s second law of motion asserts that the force F and the acceleration a are related by the equation
F = ma. (4.3.1)
Units
In applications there are three main sets of units in use for length, mass, force, and time: the cgs, mks, and British systems. All
three use the second as the unit of time. Table 1 shows the other units. Consistent with Equation 4.3.1, the unit of force in each
system is defined to be the force required to impart an acceleration of (one unit of length)/s to one unit of mass. 2
Table 4.3.1
Length Force Mass
cgs centimeter (cm) dyne (d) gram (g)
mks meter (m) newton (N) kilogram (kg)

British foot (ft) pound (lb) slug (sl)
If we assume that Earth is a perfect sphere with constant mass density, Newton’s law of gravitation (discussed later in this
section) asserts that the force exerted on an object by Earth’s gravitational field is proportional to the mass of the object and
inversely proportional to the square of its distance from the center of Earth. However, if the object remains sufficiently close to
Earth’s surface, we may assume that the gravitational force is constant and equal to its value at the surface. The magnitude of
this force is mg, where g is called the acceleration due to gravity. (To be completely accurate, g should be called the
magnitude of the acceleration due to gravity at Earth’s surface.) This quantity has been determined experimentally.
Approximate values of g are
2
g = 980 cm/s (cgs)
2
g = 9.8 m/s (mks)
2
g = 32 ft/s (British).
In general, the force F in Equation 4.3.1 may depend upon t , y , and y . Since ′
a =y
′′
, Equation 4.3.1 can be written in the
form
′′ ′
my = F (t, y, y ), (4.3.2)
which is a second order equation. We’ll consider this equation with restrictions on F later; however, since Chapter 2 dealt only
with first order equations, we consider here only problems in which Equation 4.3.2 can be recast as a first order equation. This
is possible if F does not depend on y , so Equation 4.3.2 is of the form
′′ ′
my = F (t, y ).
Letting v = y and v
′ ′
=y
′′
yields a first order equation for v :
′
m v = F (t, v). (4.3.3)
Solving this equation yields v as a function of t . If we know y(t0 ) for some time t0 , we can integrate v to obtain y as a
function of t .
Equations of the form Equation 4.3.3 occur in problems involving motion through a resisting medium.
Motion Through a Resisting Medium Under Constant Gravitational Force

Now we consider an object moving vertically in some medium. We assume that the only forces acting on the object are gravity
and resistance from the medium. We also assume that the motion takes place close to Earth’s surface and take the upward
direction to be positive, so the gravitational force can be assumed to have the constant value −mg. We’ll see that, under
reasonable assumptions on the resisting force, the velocity approaches a limit as t → ∞ . We call this limit the terminal
velocity.
Example 4.3.1
An object with mass m moves under constant gravitational force through a medium that exerts a resistance with
magnitude proportional to the speed of the object. (Recall that the speed of an object is |v|, the absolute value of its
velocity v .) Find the velocity of the object as a function of t , and find the terminal velocity. Assume that the initial
velocity is v .
0
Solution
The total force acting on the object is
F = −mg + F1 , (4.3.4)
where −mg is the force due to gravity and F is the resisting force of the medium, which has magnitude k|v|, where k is
1
a positive constant. If the object is moving downward (v ≤ 0 ), the resisting force is upward (Figure 4.3.1a), so
F1 = k|v| = k(−v) = −kv.
On the other hand, if the object is moving upward (v ≥ 0 ), the resisting force is downward (Figure 4.3.1b), so
F1 = −k|v| = −kv.
Thus, Equation 4.3.4 can be written as

F = −mg − kv, (4.3.5)
regardless of the sign of the velocity.

From Newton’s second law of motion,
′
F = ma = m v ,
so Equation 4.3.5 yields

′
m v = −mg − kv,
or
′
k
v + v = −g. (4.3.6)
m
Since e −kt/m
is a solution of the complementary equation, the solutions of Equation 4.3.6 are of the form v = ue
−kt/m
,
where u e′ −kt/m
= −g , so u = −ge
′ kt/m
. Hence,
mg kt/m
u =− e + c,
k
so
mg
−kt/m −kt/m
v = ue =− + ce . (4.3.7)
k
Since v(0) = v , 0
mg
v0 = − + c,
k

Figure 4.3.1 : Resistive forces
so
mg
c = v0 +
k
and Equation 4.3.7 becomes

mg mg
−kt/m
v=− + (v0 + )e .
k k
Letting t → ∞ here shows that the terminal velocity is

mg
lim v(t) = − ,
t→∞ k
which is independent of the initial velocity v (Figure 4.3.2).

0
Example 4.3.2 : terminal velocity

A 960-lb object is given an initial upward velocity of 60 ft/s near the surface of Earth. The atmosphere resists the motion
with a force of 3 lb for each ft/s of speed. Assuming that the only other force acting on the object is constant gravity, find
its velocity v as a function of t , and find its terminal velocity.
Solution
Since mg = 960 and g = 32 , m = 960/32 = 30 . The atmospheric resistance is −3v lb if v is expressed in feet per
second. Therefore
′
30 v = −960 − 3v,
which we rewrite as
1
′
v + v = −32.
10

Figure 4.3.2 : Solutions of m v ′
= −mg − kv
Since e −t/10
is a solution of the complementary equation, the solutions of this equation are of the form v = ue
−t/10
,
where u e′ −t/10
= −32 , so u = −32 e
′
. Hence,
t/10
t/10
u = −320 e + c,
so
−t/10 −t/10
v = ue = −320 + c e . (4.3.8)
The initial velocity is 60 ft/s in the upward (positive) direction; hence, v0 = 60 . Substituting t =0 and v = 60 in
60 = −320 + c,
so c = 380 , and Equation 4.3.8 becomes

−t/10
v = −320 + 380 e ft/s
The terminal velocity is
lim v(t) = −320 ft/s.

t→∞
Example 4.3.3
A 10 kg mass is given an initial velocity v ≤ 0 near Earth’s surface. The only forces acting on it are gravity and
0
atmospheric resistance proportional to the square of the speed. Assuming that the resistance is 8 N if the speed is 2 m/s,
find the velocity of the object as a function of t , and find the terminal velocity.
Solution
Since the object is falling, the resistance is in the upward (positive) direction. Hence,
′ 2
m v = −mg + kv , (4.3.9)
where k is a constant. Since the magnitude of the resistance is 8 N when v = 2 m/s,

2
k(2 ) = 8,
so k = 2 N-s 2 2
/m . Since m = 10 and g = 9.8 , Equation 4.3.9 becomes
′ 2 2
10 v = −98 + 2 v = 2(v − 49). (4.3.10)

If v 0 = −7 , then v ≡ −7 for all t ≥ 0 . If v 0 ≠ −7 , we separate variables to obtain
1 ′
1
v = , (4.3.11)
2
v − 49 5
which is convenient for the required partial fraction expansion

1 1 1 1 1
= = [ − ]. (4.3.12)
2
v − 49 (v − 7)(v + 7) 14 v−7 v+7
Substituting Equation 4.3.12 into Equation 4.3.11 yields

1 1 1 ′
1
[ − ]v = ,
14 v−7 v+7 5
so
1 1 14
′
[ − ]v = .
v−7 v+7 5
ln |v − 7| − ln |v + 7| = 14t/5 + k.
Therefore
∣ v−7 ∣ k 14t/5
∣ ∣ =e e .
∣ v+7 ∣
Since Theorem [thmtype:2.3.1} implies that (v − 7)/(v + 7) cannot change sign (why?), we can rewrite the last equation
as
v−7 14t/5
= ce , (4.3.13)
v+7
which is an implicit solution of Equation 4.3.10. Solving this for v yields

−14t/5
c +e
v = −7 . (4.3.14)
−14t/5
c −e
Since v(0) = v , it Equation 4.3.13 implies that

0
v0 − 7
c = .
v0 + 7
Substituting this into Equation 4.3.14 and simplifying yields

−14t/5 −14t/5
v0 (1 + e − 7(1 − e
v = −7 .
v0 (1 − e−14t/5 − 7(1 + e−14t/5
Since v 0 ≤0 , v is defined and negative for all t > 0 . The terminal velocity is
lim v(t) = −7 m/s,

t→∞
independent of v . More generally, it can be shown (Exercise 4.3.11) that if v is any solution of Equation 4.3.9 such that
0
v ≤ 0 then
0
−−−
mg
lim v(t) = −√
t→∞ k
(Figure 4.3.3).

Figure 4.3.3 : Solutions of m v ′ 2
= −mg + kv , v(0) = v0 ≤ 0
Example 4.3.4 :
A 10-kg mass is launched vertically upward from Earth’s surface with an initial velocity of v m/s. The only forces acting 0
on the mass are gravity and atmospheric resistance proportional to the square of the speed. Assuming that the atmospheric
resistance is 8 N if the speed is 2 m/s, find the time T required for the mass to reach maximum altitude.
Solution
The mass will climb while v > 0 and reach its maximum altitude when v=0 . Therefore v>0 for 0 ≤t <T and
v(T ) = 0 ; therefore, we replace Equation 4.3.10 by
′ 2
10 v = −98 − 2 v . (4.3.15)

5
′
v = −1,
2
v + 49
and integrating this yields

5 v
−1
tan = −t + c.
7 7
(Recall that tan −1

u is the number θ such that −π/2 < θ < π/2 and tan θ = u .) Since v(0) = v , 0
5 −1
v0
c = tan ,
7 7
so v is defined implicitly by
5 v 5 v0
−1 −1
tan = −t + tan , 0 ≤ t ≤ T. (4.3.16)
7 7 7 7

Figure 4.3.4 : Solutions of 4.3.15 for various v 0 > 0
Solving this for v yields

7t −1
v0
v = 7 tan(− + tan ). (4.3.17)
5 7
Using the identity

tan A − tan B
tan(A − B) =
1 + tan A tan B
with A = tan −1
(v0 /7) and B = 7t/5 , and noting that tan(tan −1
θ) = θ , we can simplify Equation 4.3.17 to
v0 − 7 tan(7t/5)
v=7 .
7 + v0 tan(7t/5)
Since v(T ) = 0 and tan −1

(0) = 0 , Equation 4.3.16 implies that
5 v0
−1
−T + tan = 0.
7 7
Therefore
5 −1
v0
T = tan .
7 7
Since tan −1
(v0 /7) < π/2 for all v , the time required for the mass to reach its maximum altitude is less than
0
5π
≈ 1.122 s
14
regardless of the initial velocity. Figure 4.3.4 shows graphs of v over [0, T ] for various values of v . 0

Figure 4.3.5 : Escape velocity
Escape Velocity
Suppose a space vehicle is launched vertically and its fuel is exhausted when the vehicle reaches an altitude h above Earth,
where h is sufficiently large so that resistance due to Earth’s atmosphere can be neglected. Let t = 0 be the time when burnout
occurs. Assuming that the gravitational forces of all other celestial bodies can be neglected, the motion of the vehicle for t > 0
is that of an object with constant mass m under the influence of Earth’s gravitational force, which we now assume to vary
inversely with the square of the distance from Earth’s center; thus, if we take the upward direction to be positive then
gravitational force on the vehicle at an altitude y above Earth is
K
F =− , (4.3.18)
2
(y + R)
where R is Earth’s radius (Figure 4.3.5).

Since F = −mg when y = 0 , setting y = 0 in Equation 4.3.18 yields
K
−mg = − ;
R2
therefore K = mgR and Equation 4.3.18 can be written more specifically as

2
2
mgR
F =− . (4.3.19)
2
(y + R)

2
d y
F =m ,
2
dt
so Equation 4.3.19 implies that

2 2
d y gR
=− . (4.3.20)
dt2 (y + R)2
We’ll show that there’s a number v , called the escape velocity, with these properties:
e
1. If v ≥ v then v(t) > 0 for all t > 0 , and the vehicle continues to climb for all t > 0 ; that is, it “escapes” Earth. (Is it
0 e
really so obvious that lim y(t) = ∞ in this case? For a proof, see Exercise 4.3.20.)
t→∞
2. If v < v then v(t) decreases to zero and becomes negative. Therefore the vehicle attains a maximum altitude y and
0 e m
falls back to Earth.

Since Equation 4.3.20 is second order, we cannot solve it by methods discussed so far. However, we are concerned with v
rather than y , and v is easier to find. Since v = y the chain rule implies that
′

2
d y dv dv dy dv
= = =v .
2
dt dt dy dt dy
Substituting this into Equation 4.3.20 yields the first order separable equation
2
dv gR
v =− . (4.3.21)
2
dy (y + R)
When t = 0 , the velocity is v and the altitude is h . Therefore we can obtain v as a function of y by solving the initial value
0
problem
2
dv gR
v =− , v(h) = v0 .
2
dy (y + R)
Integrating Equation 4.3.21 with respect to y yields

2 2
v gR
= + c. (4.3.22)
2 y +R
Since v(h) = v , 0
2 2
v gR
0
c = − ,
2 h +R

2 2 2 2
v gR v gR
0
= +( − ). (4.3.23)
2 y +R 2 h +R
If
2 1/2
2gR
v0 ≥ ( ) ,
h +R
the parenthetical expression in Equation 4.3.23 is nonnegative, so v(y) > 0 for y >h . This proves that there’s an escape
velocity v . We’ll now prove that
e
1/2
2
2gR
ve = ( )
h +R
by showing that the vehicle falls back to Earth if

2 1/2
2gR
v0 < ( ) . (4.3.24)
h +R
If Equation 4.3.24 holds then the parenthetical expression in Equation 4.3.23 is negative and the vehicle will attain a
maximum altitude y > h that satisfies the equation
m
2 2 2
gR v gR
0
0 = +( − ).
ym + R 2 h +R
The velocity will be zero at the maximum altitude, and the object will then fall to Earth under the influence of gravity.

4.3E: Elementary Mechanics (Exercises)
Except where directed otherwise, assume that the magnitude of the gravitational force on an object with mass m is constant
and equal to mg. In exercises involving vertical motion take the upward direction to be positive.
Q4.3.1
1. A firefighter who weighs 192 lb slides down an infinitely long fire pole that exerts a frictional resistive force with
magnitude proportional to his speed, with k = 2.5 lb-s/ft. Assuming that he starts from rest, find his velocity as a function of
time and find his terminal velocity.
2. A firefighter who weighs 192 lb slides down an infinitely long fire pole that exerts a frictional resistive force with
magnitude proportional to her speed, with constant of proportionality k . Find k , given that her terminal velocity is −16 ft/s,
and then find her velocity v as a function of t . Assume that she starts from rest.
3. A boat weighs 64, 000 lb. Its propellor produces a constant thrust of 50, 000 lb and the water exerts a resistive force with
magnitude proportional to the speed, with k = 2000 lb-s/ft. Assuming that the boat starts from rest, find its velocity as a
function of time, and find its terminal velocity.
4. A constant horizontal force of 10 N pushes a 20 kg-mass through a medium that resists its motion with .5 N for every m/s
of speed. The initial velocity of the mass is 7 m/s in the direction opposite to the direction of the applied force. Find the
velocity of the mass for t > 0 .
5. A stone weighing 1/2 lb is thrown upward from an initial height of 5 ft with an initial speed of 32 ft/s. Air resistance is
proportional to speed, with k = 1/128 lb-s/ft. Find the maximum height attained by the stone.
6. A 3200-lb car is moving at 64 ft/s down a 30-degree grade when it runs out of fuel. Find its velocity after that if friction
exerts a resistive force with magnitude proportional to the square of the speed, with k = 1 lb-s /ft . Also find its terminal
2 2
velocity.
7. A 96 lb weight is dropped from rest in a medium that exerts a resistive force with magnitude proportional to the speed. Find
its velocity as a function of time if its terminal velocity is −128 ft/s.
8. An object with mass m moves vertically through a medium that exerts a resistive force with magnitude proportional to the
speed. Let y = y(t) be the altitude of the object at time t , with y(0) = y . Use the results of Example 4.3.1 to show that
0
m
y(t) = y0 + (v0 − v − gt).
k
9. An object with mass m is launched vertically upward with initial velocity v from Earth’s surface (y = 0 ) in a medium
0 0
that exerts a resistive force with magnitude proportional to the speed. Find the time T when the object attains its maximum
altitude y . Then use the result of Exercise 4.3.8 to find y .
m m
10. An object weighing 256 lb is dropped from rest in a medium that exerts a resistive force with magnitude proportional to
the square of the speed. The magnitude of the resisting force is 1 lb when |v| = 4 ft/s . Find v for t > 0 , and find its terminal
velocity.
11. An object with mass m is given an initial velocity v ≤ 0 in a medium that exerts a resistive force with magnitude
0
proportional to the square of the speed. Find the velocity of the object for t > 0 , and find its terminal velocity.
12. An object with mass m is launched vertically upward with initial velocity v in a medium that exerts a resistive force with
0
magnitude proportional to the square of the speed.

a. Find the time T when the object reaches its maximum altitude.
b. Use the result of Exercise 4.3.11 to find the velocity of the object for t > T .
13. An object with mass m is given an initial velocity v0 ≤ 0 in a medium that exerts a resistive force of the form
a|v|/(1 + |v|) , where a is positive constant.
a. Set up a differential equation for the speed of the object.

b. Use your favorite numerical method to solve the equation you found in (a), to convince yourself that there’s a unique
number a such that lim
0 s(t) = ∞ if a ≤ a
t→∞ and lim0 s(t) exists (finite) if a > a . (We say that a
t→∞ is the 0 0
bifurcation value of a .) Try to find a and lim

0 s(t) in the case where a > a .
t→∞ 0
14. An object of mass m falls in a medium that exerts a resistive force f = f (s) , where s = |v| is the speed of the object.
Assume that f (0) = 0 and f is strictly increasing and differentiable on (0, ∞).
a. Write a differential equation for the speed s = s(t) of the object. Take it as given that all solutions of this equation with
s(0) ≥ 0 are defined for all t > 0 (which makes good sense on physical grounds).
b. Show that if lim f (s) ≤ mg then lim

s→∞ s(t) = ∞ .
t→∞
c. Show that if lim f (s) > mg then lim

s→∞ s(t) = s
t→∞ (terminal speed), where f (s ) = mg ..
T T
15. A 100-g mass with initial velocity v ≤ 0 falls in a medium that exerts a resistive force proportional to the fourth power of
0
the speed. The resistance is .1 N if the speed is 3 m/s.

a. Set up the initial value problem for the velocity v of the mass for t > 0 .
b. Use Exercise 4.3.14 (c) to determine the terminal velocity of the object.
c. To confirm your answer to (b), use one of the numerical methods studied in Chapter 3 to compute approximate solutions on
[0, 1] (seconds) of the initial value problem of (a) , with initial values v = 0 , −2, −4, …, −12. Present your results in
0
graphical form similar to Figure 4.3.3.

16. A 64-lb object with initial velocity v ≤ 0 falls through a dense fluid that exerts a resistive force proportional to the square
0
root of the speed. The resistance is 64 lb if the speed is 16 ft/s.

a. Set up the initial value problem for the velocity v of the mass for t > 0 .
b. Use Exercise 4.3.14 (c) to determine the terminal velocity of the object.
c. To confirm your answer to (b), use one of the numerical methods studied in Chapter 3 to compute approximate solutions on
[0, 4] (seconds) of the initial value problem of (a), with initial values v = 0 , −5, −10, …, −30. Present your results in
0
graphical form similar to Figure 4.3.3.
Q4.3.2
In Exercises 4.3.17-4.3.20, assume that the force due to gravity is given by Newton’s law of gravitation. Take the upward
direction to be positive.
17. A space probe is to be launched from a space station 200 miles above Earth. Determine its escape velocity in miles/s. Take
Earth’s radius to be 3960 miles.
18. A space vehicle is to be launched from the moon, which has a radius of about 1080 miles. The acceleration due to gravity
at the surface of the moon is about 5.31 ft/s . Find the escape velocity in miles/s.
2
19.
a. Show that (Equation 4.3.27) can be rewritten as
h −y
2 2 2
v = ve + v .
0
y +R
b. Show that if v 0 = ρve with 0 ≤ ρ < 1 , then the maximum altitude y attained by the space vehicle is
m
2
h + Rρ
ym = .
2
1 −ρ
c. By requiring that v(y m) =0 , use (Equation 4.3.26) to deduce that if v 0 < ve then
1/2
2
(1 − ρ )(ym − y)
|v| = ve [ ] ,
y +R
where y and ρ are as defined in (b) and y ≥ h .

m
d. Deduce from (c) that if v < v , the vehicle takes equal times to climb from y = h to y = y
e m and to fall back from y = y m
to y = h .

20. In the situation considered in the discussion of escape velocity, show that lim
t→∞ y(t) = ∞ if v(t) > 0 for all t > 0 .

4.4: Autonomous Second Order Equations
A second order differential equation that can be written as
′′ ′
y = F (y, y ) (4.4.1)
where F is independent of t , is said to be autonomous. An autonomous second order equation can be converted into a first
order equation relating v = y and y . If we let v = y , Equation 4.4.1 becomes
′ ′
′
v = F (y, v). (4.4.2)
Since
′
dv dv dy dv
v = = =v , (4.4.3)
dt dy dt dy
Equation 4.4.2 can be rewritten as

dv
v = F (y, v). (4.4.4)
dy
The integral curves of Equation 4.4.4 can be plotted in the (y, v) plane, which is called the Poincaré phase plane of Equation
4.4.1. If y is a solution of Equation 4.4.1 then y = y(t), v = y (t) is a parametric equation for an integral curve of Equation
′
4.4.4. we will call these integral curves trajectories of Equation 4.4.1, and we will call Equation 4.4.4 the phase plane
equivalent of Equation 4.4.1.

In this section we will consider autonomous equations that can be written as
′′ ′ ′
y + q(y, y )y + p(y) = 0. (4.4.5)
Equations of this form often arise in applications of Newton’s second law of motion. For example, suppose y is the
displacement of a moving object with mass m. It’s reasonable to think of two types of time-independent forces acting on the
object. One type - such as gravity - depends only on position. We could write such a force as −mp(y). The second type - such
as atmospheric resistance or friction - may depend on position and velocity. (Forces that depend on velocity are called
damping forces.) We write this force as −mq(y, y )y , where q(y, y ) is usually a positive function and we’ve put the factor y
′ ′ ′ ′
outside to make it explicit that the force is in the direction opposing the motion. In this case Newton’s, second law of motion
leads to Equation 4.4.5.
The phase plane equivalent of Equation 4.4.5 is
dv
v + q(y, v)v + p(y) = 0. (4.4.6)
dy
Some statements that we will be making about the properties of Equation 4.4.5 and Equation 4.4.6 are intuitively reasonable,
but difficult to prove. Therefore our presentation in this section will be informal: we will just say things without proof, all of
which are true if we assume that p = p(y) is continuously differentiable for all y and q = q(y, v) is continuously differentiable
for all (y, v). We begin with the following statements:
Properties
Statement 1: If y and v are arbitrary real numbers then Equation 4.4.5 has a unique solution on (−∞, ∞) such that
0 0
y(0) = y and y (0) = v .

′
0 0
Statement 2: If y = y(t) is a solution of Equation 4.4.5 and τ is any constant then y = y(t − τ ) is also a solution of
1
Equation 4.4.5, and y and y have the same trajectory.

1
Statement 3: If two solutions y and y of Equation 4.4.5 have the same trajectory then y (t) = y(t − τ ) for some
1 1
constant τ .
Statement 4: Distinct trajectories of Equation 4.4.5 cannot intersect; that is, if two trajectories of Equation 4.4.5
intersect, they are identical.

Statement 5: If the trajectory of a solution of Equation 4.4.5 is a closed curve then (y(t), v(t)) traverses the
trajectory in a finite time T , and the solution is periodic with period T ; that is, y(t + T ) = y(t) for all t in (−∞, ∞).
If y is a constant such that p(y) = 0 then y ≡ y is a constant solution of Equation 4.4.5. We say that y is an equilibrium of
¯
¯¯ ¯
¯¯ ¯
¯¯ ¯
¯¯
Equation 4.4.5 and (y, 0) is a critical point of the phase plane equivalent equation Equation 4.4.6. We say that the equilibrium
¯
¯¯
and the critical point are stable if, for any given ϵ > 0 no matter how small, there’s a δ > 0 , sufficiently small, such that if
−−−−−−−−−−−
¯¯ 2 2
√ (y0 − ȳ ) +v <δ
0
then the solution of the initial value problem

′′ ′ ′ ′
y + q(y, y )y + p(y) = 0, y(0) = y0 , y (0) = v0
satisfies the inequality

−−−−−−−−−−−−−−−−
2 2
√ (y(t) − ¯
¯¯
y) + (v(t)) <ϵ
for all t > 0 . Figure 4.4.1 illustrates the geometrical interpretation of this definition in the Poincaré phase plane: if (y , v 0 0) is
in the smaller shaded circle (with radius δ ), then (y(t), v(t)) must be in in the larger circle (with radius ϵ) for all t > 0 .
Figure 4.4.1 : Stability: if (y 0, v0 ) is in the smaller circle then (y(t), v(t))is in the larger circle for all t > 0
If an equilibrium and the associated critical point are not stable, we say they are unstable. To see if you really understand what
stable means, try to give a direct definition of unstable (Exercise 4.4.22). we will illustrate both definitions in the following
examples.
The Undamped Case

We will begin with the case where q ≡ 0 , so Equation 4.4.5 reduces to
′′
y + p(y) = 0. (4.4.7)
We say that this equation - as well as any physical situation that it may model - is undamped. The phase plane equivalent of
Equation 4.4.7 is the separable equation
dv
v + p(y) = 0.
dy

2
v
+ P (y) = c, (4.4.8)
2
where c is a constant of integration and P (y) = ∫ p(y) dy is an antiderivative of p.

If Equation 4.4.7 is the equation of motion of an object of mass m, then mv /2 is the kinetic energy and mP (y) is the
2
potential energy of the object; thus, Equation 4.4.8 says that the total energy of the object remains constant, or is conserved. In
particular, if a trajectory passes through a given point (y , v ) then
0 0
2
v
0
c = + P (y0 ).
2
Example 4.4.1 : The Undamped Spring - Mass System

Consider an object with mass m suspended from a spring and moving vertically. Let y be the displacement of the object
from the position it occupies when suspended at rest from the spring (Figure 4.4.2).
Figure 4.4.2 : (a) y > 0 (b) y = 0 (c) y < 0 .

Assume that if the length of the spring is changed by an amount ΔL (positive or negative), then the spring exerts an
opposing force with magnitude k|ΔL|, where k is a positive constant. In Section 6.1, it will be shown that if the mass of
the spring is negligible compared to m and no other forces act on the object then Newton’s second law of motion implies
that
′′
my = −ky, (4.4.9)
which can be written in the form Equation 4.4.7 with p(y) = ky/m. This equation can be solved easily by a method that
we will study in Section 5.2, but that method isn’t available here. Instead, we will consider the phase plane equivalent of
Equation 4.4.9.
From Equation 4.4.3, we can rewrite Equation 4.4.9 as the separable equation
dv
mv = −ky.
dy

2 2
mv ky
=− + c,
2 2

Figure 4.4.3 : Trajectories of m y ′′
+ ky = 0
which implies that

2 2
mv + ky =ρ (4.4.10)
(ρ = 2c ). This defines an ellipse in the Poincaré phase plane (Figure 4.4.3).

We can identify ρ by setting t = 0 in Equation 4.4.10; thus, ρ = mv + ky , where 2
0 0
2
y0 = y(0) and v0 = v(0) . To
determine the maximum and minimum values of y we set v = 0 in Equation 4.4.10; thus,
−
−
ρ
ymax = R and ymin = −R, withR = √ . (4.4.11)
k
Equation 4.4.9 has exactly one equilibrium, y = 0 , and it is stable. You can see intuitively why this is so: if the object is
¯
¯¯
displaced in either direction from equilibrium, the spring tries to bring it back.
In this case we can find y explicitly as a function of t . (Don’t expect this to happen in more complicated problems!) If
v > 0 on an interval I , Equation 4.4.10 implies that
−−−−−−−
dy ρ − ky 2
=v=√
dt m
on I . This is equivalent to
−
− −
−−
√k dy k
− −−−− − = ω0 , where ω0 =√ . (4.4.12)
2 dt m
√ ρ − ky
Since
−
− −
−
√k dy k
−1
∫ = sin (√ y) + c
− −−−− −
2 ρ
√ ρ − ky
y
−1
= sin ( )+c
R

Figure 4.4.4 : y = R sin(ω 0t + ϕ)
(see Equation 4.4.11), Equation 4.4.12 implies that that there’s a constant ϕ such that
y
−1
sin ( ) = ω0 t + ϕ
R
or
y = R sin(ω0 t + ϕ)
for all t in I . Although we obtained this function by assuming that v > 0 , you can easily verify that y satisfies Equation
4.4.9 for all values of t . Thus, the displacement varies periodically between −R and R , with period T = 2π/ω (Figure 0
4.4.4). (If you’ve taken a course in elementary mechanics you may recognize this as simple harmonic motion.)
Example 4.4.2 : The Undamped Pendulum

Now we consider the motion of a pendulum with mass m, attached to the end of a weightless rod with length L that
rotates on a frictionless axle (Figure 4.4.5). We assume that there’s no air resistance.
Let y be the angle measured from the rest position (vertically downward) of the pendulum, as shown in Figure 4.4.5.
Newton’s second law of motion says that the product of m and the tangential acceleration equals the tangential
component of the gravitational force; therefore, from Figure 4.4.5,
′′
mLy = −mg sin y,
or
′′
g
y =− sin y. (4.4.13)
L
Since sin nπ = 0 if n is any integer, Equation 4.4.13 has infinitely many equilibria ȳ = nπ . If¯
¯
n
n is even, the mass is
directly below the axle (Figure 4.4.6a) and gravity opposes any deviation from the equilibrium.
Figure 4.4.5 : The undamped pendulum

Figure 4.4.6 : (a) Stable equilibrium (b) Unstable equilibrium
However, if n is odd, the mass is directly above the axle (Figure 4.4.6b) and gravity increases any deviation from the
equilibrium. Therefore we conclude on physical grounds that ȳ = 2mπ is stable and ȳ ¯
¯
2m
= (2m + 1)π is unstable.
¯
¯
2m+1
The phase plane equivalent of Equation 4.4.13 is

dv g
v =− sin y,
dy L
where v = y is the angular velocity of the pendulum. Integrating this yields

′
2
v g
= cos y + c. (4.4.14)
2 L
If v = v when y = 0 , then
0
2
v g
0
c = − ,
2 L

2 2
v v g
0
= − (1 − cos y)
2 2 L
2
v 2g y
0 2
= − sin ,
2 L 2
2 2
y
2 2
v =v − vc sin , (4.4.15)
0
2
where
−−
g
vc = 2 √ .
L
The curves defined by Equation 4.4.15 are the trajectories of Equation 4.4.13. They are periodic with period 2π in y ,
which isn’t surprising, since if y = y(t) is a solution of Equation 4.4.13 then so is y = y(t) + 2nπ for any integer n .
n
Figure 4.4.7 shows trajectories over the interval [−π, π]. From Equation 4.4.15, you can see that if |v | > v then v is 0 c
nonzero for all t , which means that the object whirls in the same direction forever, as in Figure 4.4.8. The trajectories
associated with this whirling motion are above the upper dashed curve and below the lower dashed curve in Figure 4.4.7.
You can also see from Equation 4.4.15 that if 0 < |v | < v , then v = 0 when y = ±y
0 c , where max
−1
ymax = 2 sin (| v0 |/ vc ).

In this case the pendulum oscillates periodically between −y and y , as shown in Figure 4.4.9. The trajectories
max max
associated with this kind of motion are the ovals between the dashed curves in Figure 4.4.7. It can be shown (see Exercise
4.4.21 for a partial proof) that the period of the oscillation is
π/2
dθ
T =8∫ . (4.4.16)
−−−−−−−−−−
0 2 2 2
√ vc − v sin θ
0
Although this integral cannot be evaluated in terms of familiar elementary functions, you can see that it is finite if
| v | < v .>
0 c
The dashed curves in Figure 4.4.7 contain four trajectories. The critical points (π, 0) and (−π, 0) are the trajectories of
the unstable equilibrium solutions ȳ = ±π . The upper dashed curve connecting (but not including) them is obtained from
¯
¯
initial conditions of the form y(t ) = 0, v(t ) = v . If y is any solution with this trajectory then
0 0 c
lim y(t) = π and lim y(t) = −π.

t→∞ t→−∞
The lower dashed curve connecting (but not including) them is obtained from initial conditions of the form
y(t ) = 0, v(t ) = −v . If y is any solution with this trajectory then
0 0 c
lim y(t) = −π and lim y(t) = π.

t→∞ t→−∞
Consistent with this, the integral Equation 4.4.16 diverges to ∞ if v 0 = ±vc . (Exercise 4.4.21).
Since the dashed curves separate trajectories of whirling solutions from trajectories of oscillating solutions, each of these
curves is called a separatrix.
In general, if Equation 4.4.7 has both stable and unstable equilibria then the separatrices are the curves given by Equation
4.4.8 that pass through unstable critical points. Thus, if (y , 0) is an unstable critical point, then
¯
¯¯
2
v
¯
¯¯
+ P (y) = P (y ) (4.4.17)
2
defines a separatrix passing through (y, 0).

¯
¯¯
Figure 4.4.7 : Trajectories of the undamped pendulum
Figure 4.4.8 : The whirling undamped pendulum

Figure 4.4.9 : The oscillating undamped pendulum
Stability and Instability Conditions for y'' + p(y) = 0

It can be shown (Exercise 4.4.23) that an equilibrium ȳ of an undamped equation
¯
¯
′′
y + p(y) = 0 (4.4.18)
is stable if there’s an open interval (a, b) containing ȳ such that

¯
¯
¯¯ ¯¯
p(y) < 0 if a < y < ȳ and p(y) > 0 if ȳ <y <b (4.4.19)
If we regard p(y) as a force acting on a unit mass, Equation 4.4.19 means that the force resists all sufficiently small
displacements from ȳ .
¯
¯
We’ve already seen examples illustrating this principle. The equation Equation 4.4.9 for the undamped spring-mass system is
of the form Equation 4.4.18 with p(y) = ky/m, which has only the stable equilibrium ȳ = 0 . In this case Equation 4.4.19
¯
¯
holds with a = −∞ and b = ∞ . The equation Equation 4.4.13 for the undamped pendulum is of the form Equation 4.4.18
with p(y) = (g/L) sin y . We’ve seen that ȳ = 2mπ is a stable equilibrium if m is an integer. In this case
¯
¯
p(y) = sin y < 0 if (2m − 1)π < y < 2mπ
and
p(y) > 0 if 2mπ < y < (2m + 1)π
It can also be shown (Exercise 4.4.24) that ȳ is unstable if there’s a b > ȳ such that
¯
¯ ¯
¯
¯
¯¯
p(y) < 0 if y < y < b (4.4.20)
or an a < ȳ such that

¯
¯
¯¯
p(y) > 0 if a < y < ȳ (4.4.21)
If we regard p(y) as a force acting on a unit mass, Equation 4.4.20 means that the force tends to increase all sufficiently small
positive displacements from ȳ , while Equation 4.4.21 means that the force tends to increase the magnitude of all sufficiently
¯
¯
small negative displacements from ȳ .¯

¯
The undamped pendulum also illustrates this principle. We’ve seen that ȳ = (2m + 1)π is an unstable equilibrium if m is an
¯
¯
integer. In this case
sin y < 0 if (2m + 1)π < y < (2m + 2)π
so Equation 4.4.20 holds with b = (2m + 2)π , and
sin y > 0 if 2mπ < y < (2m + 1)π
so Equation 4.4.21 holds with a = 2mπ .
Example 4.4.3
The equation
′′
y + y(y − 1) = 0 (4.4.22)

is of the form Equation 4.4.18 with p(y) = y(y − 1) . Therefore ȳ = 0 and ¯
¯ ¯¯
ȳ =1 are the equilibria of Equation 4.4.22 .
Since
y(y − 1) > 0 if y < 0 or y > 1
<0 if 0 < y < 1
¯
¯¯
y =0 is unstable and y = 1 is stable.
¯
¯¯
The phase plane equivalent of Equation 4.4.22 is the separable equation

dv
v + y(y − 1) = 0.
dy
Integrating yields
2 3 2
v y y
+ − = C,
2 3 2
which we rewrite as
1
2 2
v + y (2y − 3) = c (4.4.23)
3
after renaming the constant of integration. These are the trajectories of Equation 4.4.22. If y is any solution of Equation
4.4.22, the point (y(t), v(t)) moves along the trajectory of y in the direction of increasing y in the upper half plane (
v = y > 0 ), or in the direction of decreasing y in the lower half plane (v = y < 0 ).

′ ′
Figure 4.4.10 shows typical trajectories. The dashed curve through the critical point (0, 0), obtained by setting c = 0 in
Equation 4.4.23, separates the y -v plane into regions that contain different kinds of trajectories; again, we call this curve a
separatrix. Trajectories in the region bounded by the closed loop (b) are closed curves, so solutions associated with them
are periodic. Solutions associated with other trajectories are not periodic. If y is any such solution with trajectory not on
the separatrix, then
lim y(t) = −∞, lim y(t) = −∞,
t→∞ t→−∞
lim v(t) = −∞, lim v(t) = ∞.

t→∞ t→−∞
Figure 4.4.10 : Trajectories of y ′′

+ y(y − 1) = 0
The separatrix contains four trajectories of Equation 4.4.22. One is the point (0, 0), the trajectory of the equilibrium
ȳ = 0 . Since distinct trajectories cannot intersect, the segments of the separatrix marked (a), b, and (c)– which don’t
¯
¯
include (0, 0) – are distinct trajectories, none of which can be traversed in finite time. Solutions with these trajectories
have the following asymptotic behavior:

lim y(t) = 0, lim y(t) = −∞,
t→∞ t→−∞
lim v(t) = 0, lim v(t) =∞ (on (a))

t→∞ t→−∞
lim y(t) = 0, lim y(t) = 0,

t→∞ t→−∞
.
lim v(t) = 0, lim v(t) =0 (on (b))
t→∞ t→−∞
lim y(t) = −∞, lim y(t) = 0,

t→∞ t→−∞
lim v(t) = −∞, lim v(t) =0 (on (c)).

t→∞ t→−∞
The Damped Case

The phase plane equivalent of the damped autonomous equation
′′ ′ ′
y + q(y, y )y + p(y) = 0 (4.4.24)
is
dv
v + q(y, v)v + p(y) = 0.
dy
This equation isn’t separable, so we cannot solve it for v in terms of y , as we did in the undamped case, and conservation of
energy doesn’t hold. (For example, energy expended in overcoming friction is lost.) However, we can study the qualitative
behavior of its solutions by rewriting it as
dv p(y)
= −q(y, v) − (4.4.25)
dy v
and considering the direction fields for this equation.

In the following examples we will also be showing computer generated trajectories of this equation, obtained by numerical
methods. The exercises call for similar computations. The methods discussed in Chapter 3 are not suitable for this task, since
p(y)/v in Equation 4.4.25 is undefined on the y axis of the Poincaré phase plane. Therefore we are forced to apply numerical
methods briefly discussed in Section 10.1 to the system

′
y =v
′
v = −q(y, v)v − p(y),
which is equivalent to Equation 4.4.24 in the sense defined in Section 10.1. Fortunately, most differential equation software
packages enable you to do this painlessly.
In the text we will confine ourselves to the case where q is constant, so Equation 4.4.24 and Equation 4.4.25 reduce to
′′ ′
y + c y + p(y) = 0 (4.4.26)
and
dv p(y)
= −c − .
dy v
(we will consider more general equations in the exercises.) The constant c is called the damping constant. In situations where
Equation 4.4.26 is the equation of motion of an object, c is positive; however, there are situations where c may be negative.
The Damped Spring-Mass System

Earlier we considered the spring - mass system under the assumption that the only forces acting on the object were gravity and
the spring’s resistance to changes in its length. Now we will assume that some mechanism (for example, friction in the spring
or atmospheric resistance) opposes the motion of the object with a force proportional to its velocity. In Section 6.1 it will be
shown that in this case Newton’s second law of motion implies that
′′ ′
my + c y + ky = 0, (4.4.27)

where c > 0 is the damping constant. Again, this equation can be solved easily by a method that we will study in Section 5.2,
but that method isn’t available here. Instead, we will consider its phase plane equivalent, which can be written in the form
Equation 4.4.25 as
dv c ky
=− − . (4.4.28)
dy m mv
(A minor note: the c in Equation 4.4.26 actually corresponds to c/m in this equation.) Figure 4.4.11 shows a typical direction
field for an equation of this form. Recalling that motion along a trajectory must be in the direction of increasing y in the upper
half plane (v > 0 ) and in the direction of decreasing y in the lower half plane (v < 0 ), you can infer that all trajectories
approach the origin in clockwise fashion. To confirm this, Figure 4.4.12 shows the same direction field with some trajectories
filled in. All the trajectories shown there correspond to solutions of the initial value problem
′′ ′ ′
my + c y + ky = 0, y(0) = y0 , y (0) = v0 ,
where
2 2
mv + ky =ρ (a positive constant);
0 0
thus, if there were no damping (c = 0 ), all the solutions would have the same dashed elliptic trajectory, shown in Figure
4.4.14.
Figure 4.4.11 : A typical direction field for m y ′′ ′

+ cy + ky = 0 with 0 < c < c 1
Figure 4.4.11 with some trajectories added

Solutions corresponding to the trajectories in Figure 4.4.12 cross the y -axis infinitely many times. The corresponding
solutions are said to be oscillatory (Figure 4.4.13) It is shown in Section 6.2 that there’s a number c such that if 0 ≤ c < c 1 1
then all solutions of (4.4.27) are oscillatory, while if c ≥ c , no solutions of (4.4.27) have this property. (In fact, no solution
1
not identically zero can have more than two zeros in this case.) Figure 4.4.14 shows a direction field and some integral curves
for (4.4.28) in this case.

Figure 4.4.13 : An oscillatory solution of m y ′′ ′
+ cy + ky = 0
Example 4.4.4
Now we return to the pendulum. If we assume that some mechanism (for example, friction in the axle or atmospheric
resistance) opposes the motion of the pendulum with a force proportional to its angular velocity, Newton’s second law of
motion implies that
′′ ′
mLy = −c y − mg sin y, (4.4.29)
where c > 0 is the damping constant. (Again, a minor note: the c in Equation 4.4.26 actually corresponds to c/mL in this
equation.) To plot a direction field for Equation 4.4.29 we write its phase plane equivalent as
dv c g
=− − sin y.
dy mL Lv
Figure 4.4.15 shows trajectories of four solutions of Equation 4.4.29, all satisfying y(0) = 0 . For each m = 0 , 1, 2, 3,
imparting the initial velocity v(0) = v causes the pendulum to make m complete revolutions and then settle into
m
decaying oscillation about the stable equilibrium ȳ = 2mπ .

¯
¯
Figure 4.4.14 : A typical direction field for m y ′′ ′

+ cy + ky = 0 with c > c 1
Figure 4.4.15 : Four trajectories of the damped pendulum.

4.4E: Autonomous Second Order Equations (Exercises)
Q4.4.1
In Exercises 4.4.1-4.4.4 find the equations of the trajectories of the given undamped equation. Identify the equilibrium solutions,
determine whether they are stable or unstable, and plot some trajectories.
1. y ′′
+y
3
=0
2. y ′′
+y
2
=0
3. y ′′
+ y|y| = 0
4. y ′′
+ ye
−y
=0
Q4.4.2
In Exercises 4.4.5–4.4.8 find the equations of the trajectories of the given undamped equation. Identify the equilibrium solutions,
determine whether they are stable or unstable, and find the equations of the separatrices (that is, the curves through the unstable
equilibria). Plot the separatrices and some trajectories in each of the regions of Poincaré plane determined by them.
5. y ′′
−y
3
+ 4y = 0
6. y ′′
+y
3
− 4y = 0
7. y ′′
+ y(y
2
− 1)(y
2
− 4) = 0
8. y ′′
+ y(y − 2)(y − 1)(y + 2) = 0
Q4.4.3
In Exercises 4.4.9–4.4.12 plot some trajectories of the given equation for various values (positive, negative, zero) of the parameter a.
Find the equilibria of the equation and classify them as stable or unstable. Explain why the phase plane plots corresponding to
positive and negative values of a differ so markedly. Can you think of a reason why zero deserves to be called the critical value of
a?
9. y ′′
+y
2
−a = 0
10. y ′′
+y
3
− ay = 0
11. y ′′
−y
3
+ ay = 0
12. y ′′
+ y − ay
3
=0
Q4.4.4
In Exercises 4.4.13-4.4.18 plot trajectories of the given equation for c = 0 and small nonzero (positive and negative) values of c to
observe the effects of damping.
13. y ′′
+ cy + y
′ 3
=0
14. y ′′ ′
+ cy − y = 0
15. y ′′
+ cy + y
′ 3
=0
16. y ′′
+ cy + y
′ 2
=0
17. y ′′ ′
+ c y + y|y| = 0
18. y ′′
+ y(y − 1) + cy = 0
Q4.4.5
19. The van der Pol equation
′′ 2 ′
y − μ(1 − y )y + y = 0, (A)
where μ is a positive constant and y is electrical current (Section 6.3), arises in the study of an electrical circuit whose resistive
properties depend upon the current. The damping term −μ(1 − y )y works to reduce |y| if |y| < 1 or to increase |y| if |y| > 1. It
2 ′
can be shown that van der Pol’s equation has exactly one closed trajectory, which is called a limit cycle. Trajectories inside the limit

cycle spiral outward to it, while trajectories outside the limit cycle spiral inward to it (Figure [figure:4.4.16}). Use your favorite
differential equations software to verify this for μ = .5, 1.1.5, 2. Use a grid with −4 < y < 4 and −4 < v < 4 .
Figure 4.4E. 16: Trajectories of van der Pol's euqation

20. Rayleigh’s equation,
′′ ′ 2 ′
y − μ(1 − (y ) /3)y + y = 0 (4.4E.1)
also has a limit cycle. Follow the directions of Exercise 4.4.19 for this equation.
21. In connection with Equation 4.4.16, suppose y(0) = 0 and y ′
(0) = v0 , where 0 < v 0 < vc .
a. Let T be the time required for y to increase from zero to y
1 max = 2 sin
−1
(v0 / vc ) . Show that
dy −−−−−−−−−−−−
2 2 2
= √v − vc sin y/2 , 0 ≤ t < T1 . (A)
0
dt
b. Separate variables in (A) and show that

ym ax
du
T1 = ∫ −−−−−−−−−−−− (B)
0 2 2 2
√v − vc sin u/2
0
c. Substitute sin u/2 = (v 0 / vc ) sin θ in (B) to obtain

π/2
dθ
T1 = 2 ∫ −−−−−−−−−−. (C)
0 2 2 2
√ vc − v sin θ
0
d. Conclude from symmetry that the time required for (y(t), v(t)) to traverse the trajectory
2 2 2 2
v =v − vc sin y/2 (4.4E.2)
0
is T = 4T , and that consequently y(t + T ) = y(t) and v(t + T ) = v(t) ; that is, the oscillation is periodic with period T .
1
e. Show that if v = v , the integral in (C) is improper and diverges to ∞. Conclude from this that y(t) < π for all t and
0 c
limt→∞ y(t) = π .
22. Give a direct definition of an unstable equilibrium of y ′′

+ p(y) = 0 .
23. Let p be continuous for all y and p(0) = 0 . Suppose there’s a positive number ρ such that p(y) > 0 if 0 < y ≤ ρ and p(y) < 0
if −ρ ≤ y < 0 . For 0 < r ≤ ρ let

r 0 r 0
α(r) = min {∫ p(x) dx, ∫ |p(x)| dx} \quad and \quadβ(r) = max {∫ p(x) dx, ∫ |p(x)| dx} . (4.4E.3)
0 −r 0 −r
Let y be the solution of the initial value problem

′′ ′
y + p(y) = 0, y(0) = v0 , y (0) = v0 , (4.4E.4)
y0
and define c(y 0,
2
v0 ) = v
0
+2 ∫
0
p(x) dx .
a. Show that

2
0 < c(y0 , v0 ) < v + 2β(| y0 |)\quad if \quad0 < | y0 | ≤ ρ. (4.4E.5)
0
b. Show that
y
2
v +2 ∫ p(x) dx = c(y0 , v0 ), t > 0. (4.4E.6)
0
c. Conclude from (b) that if c(y , v ) < 2α(r) then |y| < r, t > 0 .
0 0
d. Given ϵ > 0 , let δ > 0 be chosen so that

2 2 –
δ + 2β(δ) < max { ϵ /2, 2α(ϵ/ √2)} . (4.4E.7)
−−−−−−
−− −−−−
Show that if √y 2
0
2
+v
0
<δ then √y 2
+v
2
<ϵ for t > 0 , which implies that y = 0 is a stable equilibrium of y
¯
¯¯ ′′
+ p(y) = 0 .
e. Now let p be continuous for all y and p(ȳ) = 0 , where ȳ is not necessarily zero. Suppose there’s a positive number ρ such that
¯
¯ ¯
¯
p(y) > 0 if ȳ < y ≤ ȳ + ρ and p(y) < 0 if ȳ − ρ ≤ y < ȳ . Show that ȳ is a stable equilibrium of y + p(y) = 0 .
¯
¯ ¯
¯ ¯
¯ ¯
¯ ¯
¯ ′′
24. Let p be continuous for all y .

a. Suppose p(0) = 0 and there’s a positive number ρ such that p(y) < 0 if 0 < y ≤ ρ . Let ϵ be any number such that 0 <ϵ<ρ .
Show that if y is the solution of the initial value problem
′′ ′
y + p(y) = 0, y(0) = y0 , y (0) = 0 (4.4E.8)
with 0 < y < ϵ , then y(t) ≥ ϵ for some t > 0 . Conclude that y = 0 is an unstable equilibrium of y + p(y) = 0 .
0
¯
¯¯ ′′
b. Now let p(y) = 0 , where y isn’t necessarily zero. Suppose there’s a positive number ρ such that p(y) < 0 if y < y ≤ y + ρ .
¯
¯¯ ¯
¯¯ ¯
¯¯ ¯
¯¯
Show that y is an unstable equilibrium of y + p(y) = 0 .

¯
¯¯ ′′
c. Modify your proofs of (a) and (b) to show that if there’s a positive number ρ such that p(y) > 0 if y − ρ ≤ y < y , then y is an ¯
¯¯ ¯
¯¯ ¯
¯¯
unstable equilibrium of y + p(y) = 0 .

′′

4.5: Applications to Curves
We begin with two examples of families of curves generated by varying a parameter over a set of real numbers.
Example 4.5.1
For each value of the parameter c , the equation
2
y − cx =0 (4.5.1)
defines a curve in the xy-plane. If c ≠ 0 , the curve is a parabola through the origin, opening upward if c >0 or
downward if c < 0 . If c = 0 , the curve is the x axis (Figure 4.5.1).
Figure 4.5.1 : A family of curves defined by y − cx2

= 0
Example 4.5.2
For each value of the parameter c the equation
y = x +c (4.5.2)
defines a line with slope 1(Figure 4.5.2).
Theorem 4.5.1
An equation that can be written in the form
H (x, y, c) = 0 (4.5.3)
is said to define a one-parameter family of curves if, for each value of c in in some nonempty set of real numbers, the set
of points (x, y) that satisfy Equation 4.5.3 forms a curve in the xy-plane.
Figure 4.5.2 : A family of lines defined by y = x + c

Figure 4.5.3 : A family of circles defined by x 2
+y
2
−c
2
= 0
Equations Equation 4.5.1 and Equation 4.5.2 define one–parameter families of curves. (Although Equation 4.5.2 isn’t in the
form Equation 4.5.3, it can be written in this form as y − x − c = 0 .)
Example 4.5.3
If c > 0 , the graph of the equation
2 2
x +y −c = 0 (4.5.4)
is a circle with center at (0, 0) and radius √c. If c = 0 , the graph is the single point (0, 0). (We don’t regard a single point
as a curve.) If c < 0 , the equation has no graph. Hence, Equation 4.5.4 defines a one–parameter family of curves for
positive values of c . This family consists of all circles centered at (0, 0) (Figure 4.5.3).
Example 4.5.4
The equation
2 2 2
x +y +c =0
does not define a one-parameter family of curves, since no (x, y) satisfies the equation if c ≠ 0 , and only the single point
(0, 0) satisfies it if c = 0 .
Recall from Section 1.2 that the graph of a solution of a differential equation is called an integral curve of the equation.
Solving a first order differential equation usually produces a one–parameter family of integral curves of the equation. Here we
are interested in the converse problem: given a one–parameter family of curves, is there a first order differential equation for
which every member of the family is an integral curve. This suggests the next definition.
Theorem 4.5.1
4.5.2 If every curve in a one-parameter family defined by the equation
H (x, y, c) = 0 (4.5.5)
is an integral curve of the first order differential equation

′
F (x, y, y ) = 0, (4.5.6)
then Equation 4.5.6 is said to be a differential equation for the family.
To find a differential equation for a one–parameter family we differentiate its defining equation Equation 4.5.5 implicitly with
respect to x, to obtain
′
Hx (x, y, c) + Hy (x, y, c)y = 0. (4.5.7)
If this equation doesn’t, then it is a differential equation for the family. If it does contain c , it may be possible to obtain a
differential equation for the family by eliminating c between Equation 4.5.5 and Equation 4.5.7.

Example 4.5.5
Find a differential equation for the family of curves defined by
2
y = cx . (4.5.8)
Solution
Differentiating Equation 4.5.8 with respect to x yields
′
y = 2cx.
Therefore c = y ′
/2x , and substituting this into Equation 4.5.8 yields
′
xy
y =
2
as a differential equation for the family of curves defined by Equation 4.5.8. The graph of any function of the form
y = cx is an integral curve of this equation.
2
The next example shows that members of a given family of curves may be obtained by joining integral curves for more than
one differential equation.
Example 4.5.6
a. Try to find a differential equation for the family of lines tangent to the parabola y = x . 2
b. Find two tangent lines to the parabola y = x that pass through (2, 3), and find the points of tangency.
2
Solution
a. The equation of the line through a given point (x 0, y0 ) with slope m is
y = y0 + m(x − x0 ). (4.5.9)
If (x , y ) is on the parabola, then

0 0 y0 = x
2
0
and the slope of the tangent line through (x 0,
2
x )
0
is m = 2x0 ; hence,
2
y =x + 2 x0 (x − x0 ),
0
or, equivalently,
2
y = −x + 2 x0 x. (4.5.10)
0
Here x plays the role of the constant

0 c in Definition 4.5.1 ; that is, varying x0 over (−∞, ∞) produces the family of
tangent lines to the parabola y = x . 2
Differentiating Equation 4.5.10 with respect to x yields y ′

= 2 x0 .. We can express x0 in terms of x and y by rewriting
Equation 4.5.10 as
2
x − 2 x0 x + y = 0
0
and using the quadratic formula to obtain

−−−−−
2
x0 = x ± √ x −y. (4.5.11)
We must choose the plus sign in Equation 4.5.11 if x < x and the minus sign if x > x ; thus,
0 0
−−−−−
2
x0 = (x + √ x − y ) if x < x0 .
and
−−−−−
2
x0 = (x − √ x − y ) if x > x0 .

Since y ′
= 2 x0 , this implies that
−−−−−
′ 2
y = 2 (x + √ x −y) , if x < x0 (4.5.12)
and
−−−−−
′ 2
y = 2 (x − √ x −y) , if x > x0 . (4.5.13)
Neither Equation 4.5.12 nor Equation 4.5.13 is a differential equation for the family of tangent lines to the parabola
y = x . However, if each tangent line is regarded as consisting of two tangent half lines joined at the point of tangency,
2
Equation 4.5.12 is a differential equation for the family of tangent half lines on which x is less than the abscissa of the
point of tangency (Figure 4.5.4a), while Equation 4.5.13 is a differential equation for the family of tangent half lines on
which x is greater than this abscissa (Figure 4.5.4(b)). The parabola y = x is also an integral curve of both Equation
2
4.5.12 and Equation 4.5.13.
Figure 4.5.4
b. From Equation 4.5.10 the point (x, y) = (2, 3) is on the tangent line through (x 0,
2
x )
0
if and only if
2
3 = −x + 4 x0 ,
0
2
x − 4 x0 + 3 = (x0 − 3)(x0 − 1) = 0.
0
Letting x 0 =3 in Equation 4.5.10 shows that (2, 3) is on the line
y = −9 + 6x,
which is tangent to the parabola at (x 0,

2
x ) = (3, 9)
0
, as shown in Figure 4.5.5
Letting x 0 =1 in Equation 4.5.10 shows that (2, 3) is on the line
y = −1 + 2x,
which is tangent to the parabola at (x 0,

2
x ) = (1, 1)
0
, as shown in Figure 4.5.5.

Figure 4.5.5
Geometric Problems
We now consider some geometric problems that can be solved by means of differential equations.
Example 4.5.7
Find curves y = y(x) such that every point (x , y(x )) on the curve is the midpoint of the line segment with endpoints
0 0
on the coordinate axes and tangent to the curve at (x , y(x )) (Figure 4.5.6).
0 0
Solution
The equation of the line tangent to the curve at P = (x0 , y(x0 ) is
′
y = y(x0 ) + y (x0 )(x − x0 ).
If we denote the x and y intercepts of the tangent line by x and y (Figure 4.5.6), then
I I
′
0 = y(x0 ) + y (x0 )(xI − x0 ) (4.5.14)
and
′
yI = y(x0 ) − y (x0 )x0 . (4.5.15)
From Figure 4.5.6, P is the midpoint of the line segment connecting (x , 0) and (0, y ) if and only if x = 2 x
I I and
I 0
yI = 2y(x0 ) . Substituting the first of these conditions into Equation 4.5.14 or the second into Equation 4.5.15 yields
′
y(x0 ) + y (x0 )x0 = 0.
Since x is arbitrary we drop the subscript and conclude that y = y(x) satisfies
0
′
y + xy = 0,
which can be rewritten as

′
(xy ) = 0.
Integrating yields xy = c, or
c
y = .
x
If c = 0 this curve is the line y = 0 , which does not satisfy the geometric requirements imposed by the problem; thus,
c ≠ 0 , and the solutions define a family of hyperbolas (Figure 4.5.7).

Figure 4.5.6
Figure 4.5.7
Example 4.5.8
Find curves y = y(x) such that the tangent line to the curve at any point (x , y(x )) intersects the x-axis at
0 0
2
(x , 0)
0
.
Figure 4.5.8 illustrates the situation in the case where the curve is in the first quadrant and 0 < x < 1 .
Solution
The equation of the line tangent to the curve at (x 0, y(x0 )) is
′
y = y(x0 ) + y (x0 )(x − x0 ).
Since (x 2
0
, 0) is on the tangent line,
′ 2
0 = y(x0 ) + y (x0 )(x − x0 ).
0
Since x is arbitrary we drop the subscript and conclude that y = y(x) satisfies
0
′ 2
y + y (x − x) = 0.
Figure 4.5.8

Figure 4.5.9
Therefore
′
y 1 1 1 1
=− =− = − ,
2
y x −x x(x − 1) x x −1
so
∣ x ∣
ln |y| = ln |x| − ln |x − 1| + k = ln∣ ∣ + k,
∣ x −1 ∣
and
cx
y = .
x −1
If c = 0 , the graph of this function is the x-axis. If c ≠ 0 , it is a hyperbola with vertical asymptote x = 1 and horizontal
asymptote y = c . Figure 4.5.9 shows the graphs for c ≠ 0 .
Orthogonal Trajectories
Two curves C and C are said to be orthogonal at a point of intersection (x , y ) if they have perpendicular tangents at
1 2 0 0
(x , y ) . (Figure 4.5.10). A curve is said to be an orthogonal trajectory of a given family of curves if it is orthogonal to every
0 0
curve in the family. For example, every line through the origin is an orthogonal trajectory of the family of circles centered at
the origin. Conversely, any such circle is an orthogonal trajectory of the family of lines through the origin (Figure 4.5.11).
Orthogonal trajectories occur in many physical applications. For example, if u = u(x, y) is the temperature at a point (x, y) ,
the curves defined by
u(x, y) = c (4.5.16)
are called isothermal curves. The orthogonal trajectories of this family are called heat-flow lines, because at any given point
the direction of maximum heat flow is perpendicular to the isothermal through the point. If u represents the potential energy of
an object moving under a force that depends upon (x, y), the curves Equation 4.5.16 are called equipotentials, and the
orthogonal trajectories are called lines of force.
From analytic geometry we know that two nonvertical lines L and L with slopes m and m , respectively, are perpendicular
1 2 1 2
if and only if m = −1/m ; therefore, the integral curves of the differential equation
2 1
′
1
y =−
f (x, y)
Figure 4.5.10

Figure 4.5.11
are orthogonal trajectories of the integral curves of the differential equation
′
y = f (x, y),
because at any point (x 0, y0 ) where curves from the two families intersect the slopes of the respective tangent lines are
1
m1 = f (x0 , y0 ) and m2 = − .
f (x0 , y0 )
This suggests a method for finding orthogonal trajectories of a family of integral curves of a first order equation.
Finding Orthogonal Trajectories

Step 1. Find a differential equation
′
y = f (x, y)
for the given family.

Step 2. Solve the differential equation
′
1
y =−
f (x, y)
to find the orthogonal trajectories.
Example 4.5.9
Find the orthogonal trajectories of the family of circles
2 2 2
x +y =c (c > 0). (4.5.17)
Solution
To find a differential equation for the family of circles we differentiate Equation 4.5.17 implicitly with respect to x to
obtain
′
2x + 2y y = 0,
or
′
x
y =− .
y
Therefore the integral curves of

′
y
y =
x
are orthogonal trajectories of the given family. We leave it to you to verify that the general solution of this equation is
y = kx,

where k is an arbitrary constant. This is the equation of a nonvertical line through (0, 0). The y axis is also an orthogonal
trajectory of the given family. Therefore every line through the origin is an orthogonal trajectory of the given family
Equation 4.5.17 (Figure 4.5.11). This is consistent with the theorem of plane geometry which states that a diameter of a
circle and a tangent line to the circle at the end of the diameter are perpendicular.
Example 4.5.10
Find the orthogonal trajectories of the family of hyperbolas
xy = c (c ≠ 0) (4.5.18)
(Figure 4.5.7).
Solution
Differentiating Equation 4.5.18 implicitly with respect to x yields
′
y + xy = 0,
or
′
y
y =− ;
x
thus, the integral curves of

x
′
y =
y
are orthogonal trajectories of the given family. Separating variables yields

′
y y =x

2 2
y −x = k,
which is the equation of a hyperbola if k ≠ 0 , or of the lines y = x and y = −x if k = 0 (Figure 4.5.12).
Example 4.5.11
Find the orthogonal trajectories of the family of circles defined by
2 2 2
(x − c ) +y =c (c ≠ 0). (4.5.19)
These circles are centered on the x-axis and tangent to the y -axis (Figure 4.5.13a).
Solution
Multiplying out the left side of Equation 4.5.19 yields
2 2
x − 2cx + y = 0, (4.5.20)

Figure 4.5.12
and differentiating this implicitly with respect to x yields
′
2(x − c) + 2y y = 0. (4.5.21)
From Equation 4.5.20,

2 2
x +y
c = ,
2x
so
2 2 2 2
x +y x −y
x −c = x − = .
2x 2x
Substituting this into Equation 4.5.21 and solving for y yields ′
2 2
y −x
′
y = . (4.5.22)
2xy
The curves defined by Equation 4.5.19 are integral curves of Equation 4.5.22, and the integral curves of
′
2xy
y =
2 2
x −y
are orthogonal trajectories of the family Equation 4.5.19. This is a homogeneous nonlinear equation, which we studied in
Section 2.4. Substituting y = ux yields
2x(ux) 2u
′
u x +u = = ,
2 2 2
x − (ux ) 1 −u
so
2
2u u(u + 1)
′
u x = −u = ,
1 − u2 1 − u2

2
1 −u 1
′
u = ,
2
u(u + 1) x
or, equivalently,
1 2u ′
1
[ − ]u = .
2
u u +1 x
Therefore
2
ln |u| − ln(u + 1) = ln |x| + k.

By substituting u = y/x, we see that
2 2 2
ln |y| − ln |x| − ln(x + y ) + ln(x ) = ln |x| + k,
which, since ln(x 2

) = 2 ln |x| , is equivalent to
2 2
ln |y| − ln(x + y ) = k,
or
k 2 2
|y| = e (x + y ).
To see what these curves are we rewrite this equation as

2 2 −k
x + |y | −e |y| = 0
and complete the square to obtain

2 −k 2 −k 2
x + (|y| − e /2 ) = (e /2 ) .
This can be rewritten as

2 2 2
x + (y − h ) =h ,
where
−k
e
⎧ if y ≥ 0,
2
h =⎨
⎩ e
−k
− if y ≤ 0.
2
Thus, the orthogonal trajectories are circles centered on the y axis and tangent to the x axis (Figure 4.5.13 (b)). The
circles for which h > 0 are above the x-axis, while those for which h < 0 are below.
Figure 4.5.13 : (a) The circles (x − c) 2

+y
2
= c
2
(b) The circles x 2
+ (y − b)
2 2
= b

4.5E: Applications to Curves (Exercises)
Q4.5.1
In Exercises 4.5.1-4.5.8 find a first order differential equation for the given family of curves.
1. y(x 2 2
+y ) = c
2. e xy
= cy
3. ln |xy| = c(x 2
+y )
2
4. y = x 1/2
+ cx
5. y = e
2 2
x −x
+ ce
6. y = x 3
+
c
7. y = sin x + ce x
8. y = e x
+ c(1 + x )
2
Q4.5.2
9. Show that the family of circles
2 2
(x − x0 ) +y = 1, −∞ < x0 < ∞,
can be obtained by joining integral curves of two first order differential equations. More specifically, find differential
equations for the families of semicircles
2 2
(x − x0 ) +y = 1, x0 < x < x0 + 1, −∞ < x0 < ∞,
2 2
(x − x0 ) +y = 1, x0 − 1 < x < x0 , −∞ < x0 < ∞.
10. Suppose f and g are differentiable for all x . Find a differential equation for the family of functions y = f + cg (
c =constant).
Q4.5.3
In Exercises 4.5.11-4.5.13 find a first order differential equation for the given family of curves.
11. Lines through a given point (x 0, y0 ) .
12. Circles through (−1, 0) and (1, 0).
13. Circles through (0, 0) and (0, 2).
Q4.5.4
14. Use the method Example 4.5.6 (a) to find the equations of lines through the given points tangent to the parabola y =x
2
.
Also, find the points of tangency.
a. (5, 9)
b. (6, 11)
c. (−6, 20)
d. (−3, 5)
15.
a. Show that the equation of the line tangent to the circle
2 2
x +y =1 (A)
at a point (x 0, y0 ) on the circle is

1 − x0 x
y = if x0 ≠ ±1. (B)
y0
b. Show that if y is the slope of a nonvertical tangent line to the circle (A) and (x, y) is a point on the tangent line then
′
′ 2 2 ′ 2
(y ) (x − 1) − 2xy y + y − 1 = 0. (C)
c. Show that the segment of the tangent line (B) on which (x − x 0 )/ y0 >0 is an integral curve of the differential equation
− −−−−−−−−
2 2
xy − √ x + y − 1
′
y = , (D)
2
x −1
while the segment on which (x − x 0 )/ y0 <0 is an integral curve of the differential equation
− −−−−−−−−
2 2
xy + √ x + y − 1
′
y = . (E)
2
x −1
HINT: Use the quadratic formula to solve (C) for y . Then substitute (B) for y and choose the ± sign in the quadratic
′
formula so that the resulting expression for y reduces to the known slope y = −x /y
′ ′
0 0
d. Show that the upper and lower semicircles of (A) are also integral curves of (D) and (E).
e. Find the equations of two lines through (5,5) tangent to the circle (A), and find the points of tangency.
16.
a. Show that the equation of the line tangent to the parabola
2
x =y (A)
at a point (x 0, y0 ) ≠ (0, 0) on the parabola is

y0 x
y = + . (B)
2 2y0
b. Show that if y is the slope of a nonvertical tangent line to the parabola (A) and (x, y) is a point on the tangent line then
′
2 ′ 2 ′
4 x (y ) − 4xy y + x = 0. (C)
c. Show that the segment of the tangent line defined in (a) on which x > x is an integral curve of the differential equation
0
−−−−−
2
y + √y − x
′
y = , (D)
2x
while the segment on which x < x is an integral curve of the differential equation
0
−−−−−
2
y − √y − x
′
y = , (E)
2x
HINT: Use the quadratic formula to solve (c) for y . Then substitute (B) for y and choose the
′
± sign in the quadratic
formula so that the resulting expression for y reduces to the known slope of y =
′ ′ 1
2y0
− −
d. Show that the upper and lower halves of the parabola (A), given by y = √x and y = −√x for x >0 , are also integral
curves of (D) and (E).
17. Use the results of Exercise 4.5.16 to find the equations of two lines tangent to the parabola x = y and passing through the 2
given point. Also find the points of tangency.

a. (−5, 2)
b. (−4, 0)
c. (7, 4)
d. (5, −3)
18. Find a curve y = y(x) through (1,2) such that the tangent to the curve at any point (x0 , y(x0 )) intersects the x axis at
x = x /2 .
I 0
19. Find all curves y = y(x) such that the tangent to the curve at any point (x 0, y(x0 )) intersects the x axis at x I =x
3
0
.

20. Find all curves y = y(x) such that the tangent to the curve at any point passes through a given point (x 1, y1 ) .
21. Find a curve y = y(x) through (1, −1) such that the tangent to the curve at any point (x 0, y(x0 )) intersects the y axis at
y =x .
3
I 0
22. Find all curves y = y(x) such that the tangent to the curve at any point (x 0, y(x0 )) intersects the y axis at y I = x0 .
23. Find a curve y = y(x) through (0, 2) such that the normal to the curve at any point (x0 , y(x0 )) intersects the x axis at
x = x +1 .
I 0
24. Find a curve y = y(x) through (2, 1) such that the normal to the curve at any point (x0 , y(x0 )) intersects the y axis at
y = 2y(x ) .
I 0
Q4.5.5
In Exercises 4.5.25-2.5.29 find the orthogonal trajectories of the given family of curves.
25. x 2
+ 2y
2 2
=c
26. x 2
+ 4xy + y
2
=c
27. y = ce 2x
28. xy e
2
x
=c
29. y = ce
Q4.5.6
30. Find a curve through (−1, 3) orthogonal to every parabola of the form
2
y = 1 + cx
that it intersects. Which of these parabolas does the desired curve intersect?
31. Show that the orthogonal trajectories of
2 2
x + 2axy + y =c
satisfy
a+1 a−1
|y − x | |y + x | = k.
32. If lines L and L intersect at (x , y ) and α is the smallest angle through which L must be rotated counterclockwise about
1 0 0
(x , y ) to bring it into coincidence with L , we say that α is the angle from L to L ; thus, 0 ≤ α < π . If L and L
0 0 1 are 1 1
tangents to curves C and C , respectively, that intersect at (x , y ), we say that C intersects C at the angle α . Use the
1 0 0 1
identity
tan A + tan B
tan(A + B) =
1 − tan A tan B
to show that if C and C are intersecting integral curves of

1
f (x, y) + tan α π
′ ′
y = f (x, y) and y = (α ≠ ),
1 − f (x, y) tan α 2
respectively, then C intersects C at the angle α .

1
33. Use the result of Exercise 4.5.32 to find a family of curves that intersect every nonvertical line through the origin at the
angle α = π/4.
34. Use the result of Exercise 4.5.32 to find a family of curves that intersect every circle centered at the origin at a given angle
α ≠ π/2.

CHAPTER OVERVIEW
5: LINEAR SECOND ORDER EQUATIONS
In this Chapter, we study a particularly important class of second order equations. Because of their
many applications in science and engineering, second order differential equation have historically
been the most thoroughly studied class of differential equations. Research on the theory of second
order differential equations continues to the present day.
5.1: HOMOGENEOUS LINEAR EQUATIONS

This section is devoted to the theory of homogeneous linear equations.
5.1E: HOMOGENEOUS LINEAR EQUATIONS (EXERCISES)

5.2: CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS
This section deals with homogeneous equations of the special form
′′ ′
ay + by + cy = 0,
where a , b, and c are constant (a ). When you've completed this section you'll know everything there is to know about solving
≠ 0
such equations.
5.2E: CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS (EXERCISES)

5.3: NONHOMOGENEOUS LINEAR EQUATIONS
This section presents the theory of nonhomogeneous linear equations.
5.3E: NONHOMGENEOUS LINEAR EQUATIONS (EXERCISES)

5.4: THE METHOD OF UNDETERMINED COEFFICIENTS I
This section present the method of undetermined coefficients, which can be used to solve nonhomogeneous equations of the form
ay''+by'+cy=F(x) where a, b, and c are constants and F(x) has a special form that is still sufficiently general to occur in many
applications. This sections makes extensive use of the idea of variation of parameters introduced previously.
5.4E: THE METHOD OF UNDETERMINED COEFFICIENTS I (EXERCISES)

5.5: THE METHOD OF UNDETERMINED COEFFICIENTS II
In this section, we use the Method of Undetermined Coefficients to find solutions to the constant coefficient equation
ay''+by'+cy=exp{λx}(P(x) cos ω x + Q(x) sin ω x) where λ and ω are real numbers, ω is not zero, and P and Q are polynomials.
5.5E: THE METHOD OF UNDETERMINED COEFFICIENTS II (EXERCISES)

5.6: REDUCTION OF ORDER
This section deals with reduction of order, a technique based on the idea of variation of parameters, which enables us to find the
general solution of a nonhomogeneous linear second order equation provided that we know one nontrivial (not identically zero)
solution of the associated homogeneous equation.
5.6E: REDUCTION OF ORDER (EXERCISES)

5.7: VARIATION OF PARAMETERS
This section deals with the method traditionally called variation of parameters, which enables us to find the general solution of a
nonhomogeneous linear second order equation provided that we know two nontrivial solutions (with nonconstant ratio) of the
associated homogeneous equation.
5.7E: VARIATION OF PARAMETERS (EXERCISES)
1 9/16/2020
5.1: Homogeneous Linear Equations
A second order differential equation is said to be linear if it can be written as
′′ ′
y + p(x)y + q(x)y = f (x). (5.1.1)
We call the function f on the right a forcing function, since in physical applications it is often related to a force acting on some
system modeled by the differential equation. We say that Equation 5.1.1 is homogeneous if f ≡ 0 or nonhomogeneous if
f ≢ 0 . Since these definitions are like the corresponding definitions in Section 2.1 for the linear first order equation
′
y + p(x)y = f (x), (5.1.2)
it is natural to expect similarities between methods of solving Equation 5.1.1 and Equation 5.1.2. However, solving Equation
5.1.1 is more difficult than solving Equation 5.1.2. For example, while Theorem 5.1.1 gives a formula for the general solution
of Equation 5.1.2 in the case where f ≡ 0 and Theorem 5.1.2 gives a formula for the case where f ≢ 0 , there are no formulas
for the general solution of Equation 5.1.1 in either case. Therefore we must be content to solve linear second order equations
of special forms.
In Section 2.1 we considered the homogeneous equation y + p(x)y = 0 first, and then used a nontrivial solution of this
′
equation to find the general solution of the nonhomogeneous equation y + p(x)y = f (x) . Although the progression from the
′
homogeneous to the nonhomogeneous case isn’t that simple for the linear second order equation, it is still necessary to solve
the homogeneous equation
′′ ′
y + p(x)y + q(x)y = 0 (5.1.3)
in order to solve the nonhomogeneous equation Equation 5.1.1. This section is devoted to Equation 5.1.3.
The next theorem gives sufficient conditions for existence and uniqueness of solutions of initial value problems for Equation
5.1.3. We omit the proof.
Theorem 5.1.1
Suppose p and q are continuous on an open interval (a, b), let x be any point in (a, b), and let k and k be arbitrary real
0 0 1
numbers. Then the initial value problem

′′ ′ ′
y + p(x)y + q(x)y = 0, y(x0 ) = k0 , y (x0 ) = k1
has a unique solution on (a, b).
Since y ≡ 0 is obviously a solution of Equation 5.1.3 we call it the trivial solution. Any other solution is nontrivial. Under the
assumptions of Theorem 5.1.1 , the only solution of the initial value problem
′′ ′ ′
y + p(x)y + q(x)y = 0, y(x0 ) = 0, y (x0 ) = 0
on (a, b) is the trivial solution (Exercise 5.1.24).

The next three examples illustrate concepts that we’ll develop later in this section. You shouldn’t be concerned with how to
find the given solutions of the equations in these examples. This will be explained in later sections.
Example 5.1.1
The coefficients of y and y in
′
′′
y −y = 0 (5.1.4)
are the constant functions p ≡ 0 and q ≡ −1 , which are continuous on (−∞, ∞). Therefore Theorem 5.1.1 implies that
every initial value problem for Equation 5.1.4 has a unique solution on (−∞, ∞).
a. Verify that y = e and y = e are solutions of Equation 5.1.4 on (−∞, ∞).
1
x
2
−x
b. Verify that if c and c are arbitrary constants, y = c e + c e

1 2 is a solution of Equation 5.1.4 on (−∞, ∞).
1
x
2
−x

′′ ′
y − y = 0, y(0) = 1, y (0) = 3. (5.1.5)
Solution:
a. If y1 = e
x
then y
′
1
=e
x
and y
1
′′
=e
x
= y1 , so y
′′
1
− y1 = 0 . If y2 = e
−x
, then y
2
′
= −e
−x
and y
2
′′
=e
−x
= y2 , so
y
2
′′
− y2 = 0 .
b. If
x −x
y = c1 e + c2 e (5.1.6)
then
′ x −x
y = c1 e − c2 e (5.1.7)
and
′′ x −x
y = c1 e + c2 e ,
so
′′ x −x x −x
y −y = (c1 e + c2 e ) − (c1 e + c2 e )
x x −x −x
= c1 (e − e ) + c2 (e −e ) =0
for all x. Therefore y = c 1e

x
+ c2 e
−x
is a solution of Equation 5.1.4 on (−∞, ∞).
c.
We can solve Equation 5.1.5 by choosing c and c in Equation 5.1.6 so that 1 2 y(0) = 1 and ′
y (0) = 3 . Setting x =0 in
Equation 5.1.6 and Equation 5.1.7 shows that this is equivalent to
c1 + c2 = 1
c1 − c2 = 3.
Solving these equations yields c1 = 2 and c2 = −1 . Therefore y = 2e

x
−e
−x
is the unique solution of Equation 5.1.5
on (−∞, ∞).
Example 5.1.2
Let ω be a positive constant. The coefficients of y and y in ′
′′ 2
y +ω y = 0 (5.1.8)
are the constant functions p ≡ 0 and q ≡ ω , which are continuous on (−∞, ∞). Therefore Theorem
2
5.1.1 implies that
every initial value problem for Equation 5.1.8 has a unique solution on (−∞, ∞).
a. Verify that y = cos ωx and y = sin ωx are solutions of Equation 5.1.8 on (−∞, ∞).
1 2
b. Verify that if c and c are arbitrary constants then y = c cos ωx + c sin ωx is a solution of Equation
1 2 1 2 5.1.8 on
(−∞, ∞) .

′′ 2 ′
y + ω y = 0, y(0) = 1, y (0) = 3. (5.1.9)
Solution:
a. If y1 = cos ωx then y
′
1
= −ω sin ωx and y
1
′′
= −ω
2
cos ωx = −ω y1
2
, so y
′′
1
2
+ ω y1 = 0 . If y2 = sin ωx then,
y
2
′
= ω cos ωx and y ′′
2
= −ω
2
sin ωx = −ω y2
2
, so y 2
′′
+ ω y2 = 0
2
.
b. If
y = c1 cos ωx + c2 sin ωx (5.1.10)
then

′
y = ω(−c1 sin ωx + c2 cos ωx) (5.1.11)
and
′′ 2
y = −ω (c1 cos ωx + c2 sin ωx),
so
′′ 2 2 2
y +ω y = −ω (c1 cos ωx + c2 sin ωx) + ω (c1 cos ωx + c2 sin ωx)
2 2
= c1 ω (− cos ωx + cos ωx) + c2 ω (− sin ωx + sin ωx) = 0
for all x. Therefore y = c 1 cos ωx + c2 sin ωx is a solution of Equation 5.1.8 on (−∞, ∞).
c. To solve Equation 5.1.9, we must choosing c and c in Equation 1 2 so that y(0) = 1 and
5.1.10
′
y (0) = 3 . Setting
x = 0 in Equation 5.1.10 and Equation 5.1.11 shows that c = 1 and c 1 2 = 3/ω . Therefore
3
y = cos ωx + sin ωx
ω
is the unique solution of Equation 5.1.9 on (−∞, ∞).

Theorem 5.1.1 implies that if k and k are arbitrary real numbers then the initial value problem
0 1
′′ ′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0, y(x0 ) = k0 , y (x0 ) = k1 (5.1.12)
has a unique solution on an interval (a, b) that contains x , provided that P , P , and 0 0 1 P2 are continuous and P0 has no
zeros on (a, b). To see this, we rewrite the differential equation in Equation 5.1.12 as
P1 (x) P2 (x)
′′ ′
y + y + y =0
P0 (x) P0 (x)
and apply Theorem 5.1.1 with p = P 1 / P0 and q = P 2 / P0 .
Example 5.1.3
The equation
2 ′′ ′
x y + x y − 4y = 0 (5.1.13)
has the form of the differential equation in Equation 5.1.12, with P (x) = x , P (x) = x , and P (x) = −4 , which are 0
2
1 2
are all continuous on (−∞, ∞). However, since P (0) = 0 we must consider solutions of Equation 5.1.13 on (−∞, 0)
and (0, ∞). Since P has no zeros on these intervals, Theorem 5.1.1 implies that the initial value problem
0
2 ′′ ′ ′
x y + x y − 4y = 0, y(x0 ) = k0 , y (x0 ) = k1
has a unique solution on (0, ∞) if x 0 >0 , or on (−∞, 0) if x 0 <0 .

a. Verify that y = x is a solution of Equation 5.1.13 on (−∞, ∞) and y = 1/x is a solution of Equation 5.1.13 on
1
2
2
2
(−∞, 0) and (0, ∞).
b. Verify that if c and c are any constants then y = c x + c /x is a solution of Equation 5.1.13 on (−∞, 0) and
1 2 1
2
2
2
(0, ∞).

2 ′′ ′ ′
x y + x y − 4y = 0, y(1) = 2, y (1) = 0. (5.1.14)
d. Solve the initial value problem

2 ′′ ′ ′
x y + x y − 4y = 0, y(−1) = 2, y (−1) = 0. (5.1.15)
Solution:
a. If y1 =x
2
then y ′
1
= 2x and y ′′
1
=2 , so
2 ′′ ′ 2 2
x y + xy − 4 y1 = x (2) + x(2x) − 4 x =0
1 1

for x in (−∞, ∞). If y 2 = 1/ x
2
, then y ′
2
= −2/ x
3
and y ′′
2
= 6/ x
4
, so
2 2
6 2 4
′′ ′
x y + xy − 4 y2 = x ( ) −x ( )− =0
2 2 4 3 2
x x x
for x in (−∞, 0) or (0, ∞).

b. If
c2
2
y = c1 x + (5.1.16)
2
x
then
2c2
′
y = 2 c1 x − (5.1.17)
3
x
and
6c2
′′
y = 2 c1 + ,
4
x
so
2 ′′ ′ 2
6c2 2c2 2
c2
x y + x y − 4y =x (2 c1 + ) + x (2 c1 x − ) − 4 (c1 x + )
4 3 2
x x x
2 2 2
6 2 4
= c1 (2 x + 2x − 4 x ) + c2 ( − − )
x2 x2 x2
= c1 ⋅ 0 + c2 ⋅ 0 = 0
for x in (−∞, 0) or (0, ∞).

c. To solve Equation 5.1.14, we choose c and c in Equation 5.1.16 so that
1 2 y(1) = 2 and ′
y (1) = 0 . Setting x =1 in
Equation 5.1.16 and Equation 5.1.17 shows that this is equivalent to
c1 + c2 = 2
2 c1 − 2 c2 = 0.
Solving these equations yields c 1 =1 and c 2 =1 . Therefore y = x 2

+ 1/ x
2
is the unique solution of Equation 5.1.14 on
(0, ∞).
d. We can solve Equation 5.1.15 by choosing c and c in Equation 5.1.16 so that

1 2 y(−1) = 2 and y ′
(−1) = 0 . Setting
x = −1 in Equation 5.1.16 and Equation 5.1.17 shows that this is equivalent to
c1 + c2 = 2
−2 c1 + 2 c2 = 0.
Solving these equations yields c 1 =1 and c 2 =1 . Therefore y = x 2

+ 1/ x
2
is the unique solution of Equation 5.1.15 on
(−∞, 0).
Although the formulas for the solutions of Equation 5.1.14 and Equation 5.1.15 are both y = x + 1/x , you should not 2 2
conclude that these two initial value problems have the same solution. Remember that a solution of an initial value problem is
defined on an interval that contains the initial point; therefore, the solution of Equation 5.1.14 is y = x + 1/x on the 2 2
interval (0, ∞), which contains the initial point x = 1 , while the solution of Equation 5.1.15 is y = x + 1/x on the
0
2 2
interval (−∞, 0), which contains the initial point x = −1 . 0
The General Solution of a Homogeneous Linear Second Order Equation

If y and y are defined on an interval (a, b) and c and c are constants, then
1 2 1 2
y = c1 y1 + c2 y2

is a linear combination of y and y . For example, y = 2 cos x + 7 sin x is a linear combination of y
1 2 1 = cos x and y
2 = sin x ,
with c = 2 and c = 7 .
1 2
The next theorem states a fact that we’ve already verified in Examples 5.1.1, 5.1.2, 5.1.3.
Theorem 5.1.2
If y and y are solutions of the homogeneous equation
1 2
′′ ′
y + p(x)y + q(x)y = 0 (5.1.18)
on (a, b), then any linear combination

y = c1 y1 + c2 y2 (5.1.19)
of y and y is also a solution of (5.1.18) on (a, b).

1 2
Proof
If
y = c1 y1 + c2 y2
then
′ ′ ′ ′′ ′′ ′′
y = c1 y + c2 y and y = c1 y + c2 y .
1 2 1 2
Therefore
′′ ′ ′′ ′′ ′ ′
y + p(x)y + q(x)y = (c1 y + c2 y ) + p(x)(c1 y + c2 y ) + q(x)(c1 y1 + c2 y2 )
1 2 1 2
′′ ′ ′′ ′
= c1 (y + p(x)y + q(x)y1 ) + c2 (y + p(x)y + q(x)y2 )
1 1 2 2
= c1 ⋅ 0 + c2 ⋅ 0 = 0,
since y and y are solutions of Equation 5.1.18.

1 2
We say that {y , y } is a fundamental set of solutions of (5.1.18) on (a, b) if every solution of Equation 5.1.18 on
1 2
(a, b) can be written as a linear combination of y and y as in Equation 5.1.19. In this case we say that Equation
1 2
5.1.19is general solution of (5.1.18)on (a, b) .
Linear Independence
We need a way to determine whether a given set {y , y } of solutions of Equation 1 2 5.1.18 is a fundamental set. The next
definition will enable us to state necessary and sufficient conditions for this.
We say that two functions y and y defined on an interval (a, b) are linearly independent on (a, b) if neither is a constant
1 2
multiple of the other on (a, b). (In particular, this means that neither can be the trivial solution of Equation 5.1.18, since, for
example, if y ≡ 0 we could write y = 0y .) We’ll also say that the set {y , y } is linearly independent on (a, b).
1 1 2 1 2
Theorem 5.1.3
Suppose p and q are continuous on (a, b). Then a set {y 1, y2 } of solutions of
′′ ′
y + p(x)y + q(x)y = 0 (5.1.20)
on (a, b) is a fundamental set if and only if {y 1, y2 } is linearly independent on (a, b).
Proof
We’ll present the proof of Theorem 5.1.3 in steps worth regarding as theorems in their own right. However, let’s first
interpret Theorem 5.1.3in terms of Examples 5.1.1, 5.1.2, 5.1.3.

Example 5.1.4
Since e /e = e
x −x 2x
is nonconstant, Theorem 5.1.3 implies that y = c 1e
x
+ c2 e
−x
is the general solution of y ′′
−y = 0
on (−∞, ∞).
Since cos ωx/ sin ωx = cot ωx is nonconstant, Theorem 5.1.3 implies that y = c1 cos ωx + c2 sin ωx is the general
solution of y + ω y = 0 on (−∞, ∞).
′′ 2
Since x /x
2 −2
=x
4
is nonconstant, Theorem 5.1.3 implies that y = c1 x
2
+ c2 / x
2
is the general solution of
2
x y
′′ ′
+ x y − 4y = 0 on (−∞, 0) and (0, ∞).
The Wronskian and Abel's Formula

To motivate a result that we need in order to prove Theorem 5.1.3, let’s see what is required to prove that {y , y } is a 1 2
fundamental set of solutions of Equation 5.1.20 on (a, b). Let x be an arbitrary point in (a, b), and suppose y is an arbitrary0
solution of Equation 5.1.20 on (a, b). Then y is the unique solution of the initial value problem
′′ ′ ′
y + p(x)y + q(x)y = 0, y(x0 ) = k0 , y (x0 ) = k1 ; (5.1.21)
that is, k and k are the numbers obtained by evaluating y and y at x . Moreover, k and k can be any real numbers, since
0 1
′
0 0 1
Theorem 5.1.1 implies that Equation 5.1.21 has a solution no matter how k and k are chosen. Therefore {y , y } is a 0 1 1 2
fundamental set of solutions of Equation 5.1.20on (a, b) if and only if it is possible to write the solution of an arbitrary initial
value problem Equation 5.1.21as y = c y + c y . This is equivalent to requiring that the system
1 1 2 2
c1 y1 (x0 ) + c2 y2 (x0 ) = k0
(5.1.22)
′ ′
c1 y (x0 ) + c2 y (x0 ) = k1
1 2
has a solution (c 1, c2 ) for every choice of (k 0, k1 ) . Let’s try to solve Equation 5.1.22.
Multiplying the first equation in Equation 5.1.22 by y ′
2
(x0 ) and the second by y 2 (x0 ) yields
′ ′ ′
c1 y1 (x0 )y (x0 ) + c2 y2 (x0 )y (x0 ) = y (x0 )k0
2 2 2
′ ′
c1 y (x0 )y2 (x0 ) + c2 y (x0 )y2 (x0 ) = y2 (x0 )k1 ,
1 2
and subtracting the second equation here from the first yields
′ ′ ′
(y1 (x0 )y (x0 ) − y (x0 )y2 (x0 )) c1 = y (x0 )k0 − y2 (x0 )k1 . (5.1.23)
2 1 2
Multiplying the first equation in Equation 5.1.22by y ′

1
(x0 ) and the second by y 1 (x0 ) yields
′ ′ ′
c1 y1 (x0 )y (x0 ) + c2 y2 (x0 )y (x0 ) = y (x0 )k0
1 1 1
′ ′
c1 y (x0 )y1 (x0 ) + c2 y (x0 )y1 (x0 ) = y1 (x0 )k1 ,
1 2
and subtracting the first equation here from the second yields
′ ′ ′
(y1 (x0 )y (x0 ) − y (x0 )y2 (x0 )) c2 = y1 (x0 )k1 − y (x0 )k0 . (5.1.24)
2 1 1
If
′ ′
y1 (x0 )y (x0 ) − y (x0 )y2 (x0 ) = 0,
2 1
it is impossible to satisfy Equation 5.1.23 and Equation 5.1.24 (and therefore Equation 5.1.22 ) unless k0 and k1 happen to
satisfy
′
y1 (x0 )k1 − y (x0 )k0 = 0
1
′
y (x0 )k0 − y2 (x0 )k1 = 0.
2
On the other hand, if

′ ′
y1 (x0 )y (x0 ) − y (x0 )y2 (x0 ) ≠ 0 (5.1.25)
2 1
we can divide Equation 5.1.23and Equation 5.1.24through by the quantity on the left to obtain

′
y ( x0 ) k0 −y2 ( x0 ) k1
2
c1 = ′ ′
y1 ( x0 ) y ( x0 )−y ( x0 ) y2 ( x0 )
2 1
′
(5.1.26)
y ( x0 ) k1 −y ( x0 ) k0
1 1
c2 = ′ ′
,
y1 ( x0 ) y2 ( x0 )−y1 ( x0 ) y2 ( x0 )
no matter how k and k are chosen. This motivates us to consider conditions on y and y that imply Equation 5.1.25.
0 1 1 2
Theorem 5.1.4
Suppose p and q are continuous on (a, b), let y and y be solutions of 1 2
′′ ′
y + p(x)y + q(x)y = 0 (5.1.27)
on (a, b) , and define

′ ′
W = y1 y − y y2 . (5.1.28)
2 1
Let x be any point in (a, b). Then

0
x
−∫ p(t) dt
x
W (x) = W (x0 )e 0
, a <x <b (5.1.29)
Therefore either W has no zeros in (a, b) or W ≡0 on (a, b).
Proof
Differentiating Equation 5.1.28yields
′ ′ ′ ′′ ′ ′ ′′ ′′ ′′
W =y y + y1 y −y y −y y2 = y1 y −y y2 . (5.1.30)
1 2 2 1 2 1 2 1
Since y and y both satisfy Equation 5.1.27,

1 2
′′ ′ ′′ ′
y = −p y − q y1 and y = −p y − q y2 .
1 1 2 2
Substituting these into Equation 5.1.30yields

′ ′ ′
W = −y1 (p y + q y2 ) + y2 (p y + q y1 )
2 1
′ ′
= −p(y1 y − y2 y ) − q(y1 y2 − y2 y1 )
2 1
′ ′
= −p(y1 y − y2 y ) = −pW .
2 1
Therefore W ′
+ p(x)W = 0 ; that is, W is the solution of the initial value problem
′
y + p(x)y = 0, y(x0 ) = W (x0 ).
We leave it to you to verify by separation of variables that this implies Equation 5.1.29. If W (x ) ≠ 0 , Equation 0
5.1.29 implies that W has no zeros in (a, b) , since an exponential is never zero. On the other hand, if W (x ) = 0 , 0
Equation 5.1.29implies that W (x) = 0 for all x in (a, b) .
The function W defined in Equation 5.1.28 is the Wronskian of {y 1, y2 } . Formula Equation 5.1.29 is Abel’s formula.
The Wronskian of {y 1, y2 } is usually written as the determinant
∣ y1 y2 ∣
W =∣ ∣.
′ ′
∣y y ∣
1 2
The expressions in Equation 5.1.26for c and c can be written in terms of determinants as

1 2
1 ∣ k0 y2 (x0 ) ∣ 1 ∣ y1 (x0 ) k0 ∣
c1 = ∣ ∣ and c2 = ∣ ∣.
′ ′
W (x0 ) ∣ k1 y (x0 ) ∣ W (x0 ) ∣ y (x0 ) k1 ∣
2 1
If you’ve taken linear algebra you may recognize this as Cramer’s rule.
Example 5.1.5

Verify Abel’s formula for the following differential equations and the corresponding solutions, from Examples 5.1.1 ,
5.1.2, 5.1.3.
a. y′′
− y = 0;
x
y1 = e , y2 = e
−x
b. y′′
+ ω y = 0;
2
y1 = cos ωx, y2 = sin ωx
c. x 2
y
′′ ′
+ x y − 4y = 0;
2
y1 = x , y2 = 1/ x
2
Solution:
a. Since p ≡ 0 , we can verify Abel’s formula by showing that W is constant, which is true, since
x −x
∣e e ∣
x −x x −x
W (x) = ∣ ∣ = e (−e )−e e = −2
x −x
∣e −e ∣
for all x.
b. Again, since p ≡ 0 , we can verify Abel’s formula by showing that W is constant, which is true, since
∣ cos ωx sin ωx
∣
W (x) = ∣ ∣
∣ −ω sin ωx ω cos ωx ∣
= cos ωx(ω cos ωx) − (−ω sin ωx) sin ωx
2 2
= ω(cos ωx + sin ωx) = ω
for all x.
c. Computing the Wronskian of y 1 =x
2
and y 2 = 1/ x
2
directly yields
∣ x2 1/x
2
∣ 2 1 4
2
W =∣ ∣ =x (− ) − 2x ( ) =− . (5.1.31)
3
∣ 2x −2/x ∣ x3 x
2
x
To verify Abel’s formula we rewrite the differential equation as

1 4
′′ ′
y + y − y =0
2
x x
to see that p(x) = 1/x . If x and x are either both in (−∞, 0) or both in (0, ∞) then
0
x x
dt x
∫ p(t) dt = ∫ = ln( ),
x0 x0
t x0
so Abel’s formula becomes

− ln(x/ x0 )
x0
W (x) = W (x0 )e = W (x0 )
x
4 x0
= −( )( ) from(5.1.31)
x0 x
4
=− ,
x
which is consistent with Equation 5.1.31.
The next theorem will enable us to complete the proof of Theorem 5.1.3.
Theorem 5.1.5
Suppose p and q are continuous on an open interval (a, b), let y and y be solutions of 1 2
′′ ′
y + p(x)y + q(x)y = 0 (5.1.32)
on (a, b), and let W = y1 y

2
′ ′
− y y2 .
1
Then y and y are linearly independent on (a, b) if and only if W has no zeros on
1 2
(a, b).
Proof

We first show that if W (x ) = 0 for some x in (a, b) , then y and y are linearly dependent on (a, b) . Let
0 0 1 2 I be a
subinterval of (a, b) on which y has no zeros. (If there’s no such subinterval, y ≡ 0 on (a, b) , so y and
1 1 1 y2 are
linearly independent, and we are finished with this part of the proof.) Then y /y is defined on I , and 2 1
′ ′ ′
y2 y1 y − y y2 W
2 1
( ) = = . (5.1.33)
2 2
y1 y y
1 1
However, if , Theorem 5.1.4 implies that W ≡ 0 on (a, b) . Therefore Equation 5.1.33 implies that
W (x0 ) = 0
′
(y2 / y1 ) ≡ 0 , so (constant) on I . This shows that y (x) = cy (x) for all x in I . However, we want to
y2 / y1 = c 2 1
show that y = cy (x) for all x in (a, b) . Let Y = y − cy . Then Y is a solution of Equation 5.1.32 on (a, b) such
2 1 2 1
that Y ≡ 0 on I , and therefore Y ≡ 0 on I . Consequently, if x is chosen arbitrarily in I then Y is a solution of the

′
0

′′ ′ ′
y + p(x)y + q(x)y = 0, y(x0 ) = 0, y (x0 ) = 0,
which implies that Y ≡ 0 on (a, b) , by the paragraph following Theorem 5.1.1 . (See also Exercise 5.1.24). Hence,
y − c y ≡ 0 on (a, b) , which implies that y and y are not linearly independent on (a, b) .
2 1 1 2
Now suppose W has no zeros on (a, b). Then y can’t be identically zero on (a, b) (why not?), and therefore there is a
1
subinterval I of (a, b) on which y has no zeros. Since Equation 5.1.33 implies that y /y is nonconstant on I , y isn’t a
1 2 1 2
constant multiple of y on (a, b). A similar argument shows that y isn’t a constant multiple of y on (a, b), since
1 1 2
′ ′ ′
y1 y y2 − y1 y W
1 2
( ) = =−
2 2
y2 y y
2 2
on any subinterval of (a, b) where y has no zeros. 2
We can now complete the proof of Theorem 5.1.3. From Theorem 5.1.5, two solutions y and y of Equation 5.1.32 are 1 2
linearly independent on (a, b) if and only if W has no zeros on (a, b). From Theorem 5.1.4 and the motivating comments
preceding it, {y , y } is a fundamental set of solutions of Equation 5.1.32 if and only if W has no zeros on (a, b). Therefore
1 2
{ y , y } is a fundamental set for Equation 5.1.32 on (a, b) if and only if { y , y } is linearly independent on (a, b) .
1 2 1 2
The next theorem summarizes the relationships among the concepts discussed in this section.
Theorem 5.1.6
Suppose p and q are continuous on an open interval (a, b) and let y and y be solutions of 1 2
′′ ′
y + p(x)y + q(x)y = 0 (5.1.34)
on (a, b). Then the following statements are equivalent; that is, they are either all true or all false.
a. The general solution of (5.1.34) on (a, b) is y = c y + c y . 1 1 2 2
b. {y , y } is a fundamental set of solutions of (5.1.34) on (a, b).

1 2
c. {y , y } is linearly independent on (a, b).

1 2
d. The Wronskian of {y , y } is nonzero at some point in (a, b).

1 2
e. The Wronskian of {y , y } is nonzero at all points in (a, b).

1 2
We can apply this theorem to an equation written as

′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0
on an interval (a, b) where P0 , P1 , and P2 are continuous and P0 has no zeros.dd proof here and it will automatically be
hidden
Theorem 5.1.7

Suppose c is in (a, b) and α and β are real numbers, not both zero. Under the assumptions of Theorem 5.1.7, suppose y 1
and y are solutions of Equation 5.1.34 such that

2
′ ′
α y1 (c) + β y (c) = 0 and α y2 (c) + β y (c) = 0. (5.1.35)
1 2
Then {y 1, y2 } isn’t linearly independent on (a, b).
Proof
Since α and β are not both zero, Equation 5.1.35implies that
′
∣ y1 (c) y (c) ∣ ∣ y1 (c) y2 (c) ∣
1
∣ ∣ = 0, so ∣ ∣ =0
′ ′ ′
∣ y2 (c) y (c) ∣ ∣ y1 (c) y (c) ∣
2
2
and Theorem 5.1.6implies the stated conclusion.

5.1E: Homogeneous Linear Equations (Exercises)
Q5.1.1
1.
a. Verify that y 1 =e
2x
and y 2 =e
5x
are solutions of
′′ ′
y − 7 y + 10y = 0 (A)
on (−∞, ∞).
b. Verify that if c and c are arbitrary constants then y = c
1 2 1e
2x
+ c2 e
5x
is a solution of (A) on (−∞, ∞).
′′ ′ ′
y − 7 y + 10y = 0, y(0) = −1, y (0) = 1.

′′ ′ ′
y − 7 y + 10y = 0, y(0) = k0 , y (0) = k1 .
2.
x
cos x and y 2 =e
x
sin x are solutions of
′′ ′
y − 2 y + 2y = 0 (A)
on (−∞, ∞).
b. Verify that if c and c are arbitrary constants then y = c
1 2 1e
x
cos x + c2 e
x
sin x is a solution of (A) on (−∞, ∞).
′′ ′ ′
y − 2 y + 2y = 0, y(0) = 3, y (0) = −2.

′′ ′ ′
y − 2 y + 2y = 0, y(0) = k0 , y (0) = k1 .
3.
x
and y 2 = xe
x
are solutions of
′′ ′
y − 2y + y = 0 (A)
on (−∞, ∞).
b. Verify that if c and c are arbitrary constants then y = e
1 2
x
(c1 + c2 x) is a solution of (A) on (−∞, ∞).
′′ ′ ′
y − 2 y + y = 0, y(0) = 7, y (0) = 4.

′′ ′ ′
y − 2 y + y = 0, y(0) = k0 , y (0) = k1 .
4.
a. Verify that y 1 = 1/(x − 1) and y 2 = 1/(x + 1) are solutions of
2 ′′ ′
(x − 1)y + 4x y + 2y = 0 (A)
on (−∞, −1), (−1, 1), and (1, ∞). What is the general solution of (A) on each of these intervals?
2 ′′ ′ ′
(x − 1)y + 4x y + 2y = 0, y(0) = −5, y (0) = 1.
What is the interval of validity of the solution?

c. Graph the solution of the initial value problem.
d. Verify Abel’s formula for y and y , with x = 0 . 1 2 0
5. Compute the Wronskians of the given sets of functions.

a. {1, e } x
b. {e , e sin x}
x x
c. {x + 1, x + 2} 2
d. {x , x 1/2
}
−1/3
e. { sin x
,
x
}
cos x
f. {x ln |x|, x ln |x|} 2
g. {e cos √−
x −
x , e sin √x }
x
6. Find the Wronskian of a given set {y 1, y2 } of solutions of

′′ 2 ′
y + 3(x + 1)y − 2y = 0,
given that W (π) = 0.

2 ′′ ′
(1 − x )y − 2x y + α(α + 1)y = 0,
given that W (0) = 1 . (This is Legendre’s equation.)

2 ′′ ′ 2 2
x y + x y + (x −ν )y = 0,
given that W (1) = 1 . (This is Bessel’s equation.)

9. (This exercise shows that if you know one nontrivial solution of y
′′ ′
+ p(x)y + q(x)y = 0 , you can use Abel’s formula to
find another.)
Suppose p and q are continuous and y is a solution of 1
′′ ′
y + p(x)y + q(x)y = 0 (A)
that has no zeros on (a, b). Let P (x) = ∫ p(x) dx be any antiderivative of p on (a, b).
a. Show that if K is an arbitrary nonzero constant and y satisfies 2
′ ′ −P (x)
y1 y − y y2 = K e (B)
2 1
on (a, b), then y also satisfies (A) on (a, b), and {y

2 1, y2 } is a fundamental set of solutions on (A) on (a, b).
−P( x)
b. Conclude from (a) that if y 2 = u y1 where u ′

=K
e
y
2
(x)
, then {y 1, y2 } is a fundamental set of solutions of (A) on (a, b).
1
Q5.1.2
In Exercises 5.1.10-5.1.23 use the method suggested by Exercise 5.1.9 to find a second solution y2 that isn’t a constant
multiple of the solution y . Choose K conveniently to simplify y .
1 2
10. y ′′
− 2 y − 3y = 0
′
;y 1 =e
3x
11. y ′′
− 6 y + 9y = 0
′
;y 1 =e
3x
12. y ′′
− 2ay + a y = 0
′ 2
(a = constant); y 1 =e
ax
13. x 2
y
′′
+ xy − y = 0
′
;y 1 =x
14. x 2
y
′′
− xy + y = 0
′
;y 1 =x
15. x 2
y
′′
− (2a − 1)x y + a y = 0
′ 2
(a = nonzero constant); x >0 ;y 1 =x
a
16. 4x 2
y
′′ ′
− 4x y + (3 − 16 x )y = 0
2
;y 1 =x
1/2
e
2x
17. (x − 1)y ′′
− xy + y = 0
′
;y 1 =e
x

18. x 2
y
′′
− 2x y + (x
′ 2
+ 2)y = 0 ;y 1 = x cos x
19. 4x 2
(sin x)y
′′
− 4x(x cos x + sin x)y + (2x cos x + 3 sin x)y = 0
′
;y 1 =x
1/2
20. (3x − 1)y ′′

− (3x + 2)y − (6x − 8)y = 0
′
;y 1 =e
2x
21. (x 2
− 4)y
′′
+ 4x y + 2y = 0
′
;y 1 =
x−2
1
22. (2x + 1)x y ′′

− 2(2 x
2 ′
− 1)y − 4(x + 1)y = 0 ;y 1 =
1
23. (x 2
− 2x)y
′′ 2
+ (2 − x )y + (2x − 2)y = 0
′
;y 1 =e
x
Q5.1.3
24. Suppose p and q are continuous on an open interval (a, b) and let x be in (a, b). Use Theorem 5.1.1 to show that the only 0

′′ ′ ′
y + p(x)y + q(x)y = 0, y(x0 ) = 0, y (x0 ) = 0
on (a, b) is the trivial solution y ≡ 0 .

25. Suppose P , P , and P are continuous on (a, b) and let x be in (a, b). Show that if either of the following statements is
0 1 2 0
true then P (x) = 0 for some x in (a, b).

0
a. The initial value problem

′′ ′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0, y(x0 ) = k0 , y (x0 ) = k1
has more than one solution on (a, b).

b. The initial value problem
′′ ′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0, y(x0 ) = 0, y (x0 ) = 0
has a nontrivial solution on (a, b).

26. Suppose p and q are continuous on (a, b) and y and y are solutions of 1 2
′′ ′
y + p(x)y + q(x)y = 0 (A)
on (a, b). Let
z1 = α y1 + β y2 and z2 = γ y1 + δy2 ,
where α , β, γ, and δ are constants. Show that if {z 1, z2 } is a fundamental set of solutions of (A) on (a, b) then so is {y 1, y2 } .
27. Suppose p and q are continuous on (a, b) and {y 1, y2 } is a fundamental set of solutions of
′′ ′
y + p(x)y + q(x)y = 0 (A)
on (a, b). Let
z1 = α y1 + β y2 and z2 = γ y1 + δy2 ,
where α, β, γ , and δ are constants. Show that { z1 , z2 } is a fundamental set of solutions of (A) on (a, b) if and only if
αγ − βδ ≠ 0 .
28. Suppose y is differentiable on an interval
1 (a, b) and y2 = ky1 , where k is a constant. Show that the Wronskian of
{ y , y } is identically zero on (a, b) .
1 2
29. Let
3
x , x ≥ 0,
3
y1 = x and y2 = {
3
−x , x < 0.
a. Show that the Wronskian of {y , y } is defined and identically zero on (−∞, ∞).
1 2
b. Suppose a < 0 < b . Show that {y , y } is linearly independent on (a, b). 1 2

c. Use Exercise 5.1.25b to show that these results don’t contradict Theorem 5.1.5, because neither y nor y can be a solution 1 2
of an equation
′′ ′
y + p(x)y + q(x)y = 0
on (a, b) if p and q are continuous on (a, b).

30. Suppose p and q are continuous on (a, b) and {y 1, y2 } is a set of solutions of
′′ ′
y + p(x)y + q(x)y = 0
on (a, b) such that either y 1 (x0 ) = y2 (x0 ) = 0 or y ′

1
(x0 ) = y (x0 ) = 0
2
′
for some x in (a, b). Show that
0 { y1 , y2 } is linearly
dependent on (a, b).
′′ ′
y + p(x)y + q(x)y = 0
on (a, b). Show that if y 1 (x1 ) = y1 (x2 ) = 0 , where a < x 1 < x2 < b , then y 2 (x) =0 for some x in (x 1, x2 ) .
32. Suppose p and q are continuous on (a, b) and every solution of
′′ ′
y + p(x)y + q(x)y = 0 (A)
on (a, b) can be written as a linear combination of the twice differentiable functions {y 1, y2 } . Use Theorem 5.1.1 to show that
y and y are themselves solutions of (A) on (a, b) .
1 2
33. Suppose p , p , q , and q are continuous on (a, b) and the equations

1 2 1 2
′′ ′ ′′ ′
y + p1 (x)y + q1 (x)y = 0 and y + p2 (x)y + q2 (x)y = 0
have the same solutions on (a, b). Show that p 1 = p2 and q 1 = q2 on (a, b).
34. (For this exercise you have to know about 3 ×3 determinants.) Show that if y and 1 y2 are twice continuously
differentiable on (a, b) and the Wronskian W of {y 1, y2 } has no zeros in (a, b) then the equation
∣ y y1 y2 ∣
1 ∣ ∣
′ ′ ′
∣ y y y ∣ =0
1 2
W ∣ ∣
′′ ′′ ′′
∣y y y ∣
1 2
can be written as
′′ ′
y + p(x)y + q(x)y = 0, (A)
where p and q are continuous on (a, b) and {y 1, y2 } is a fundamental set of solutions of (A) on (a, b).
35. Use the method suggested by Exercise 5.1.34 to find a linear homogeneous equation for which the given functions form a
fundamental set of solutions on some interval.
a. e cos 2x, e sin 2x
x x
b. x, e 2x
c. x, x ln x
d. cos(ln x), sin(ln x)
e. cosh x, sinh x
f. x − 1, x + 1
2 2
′′ ′
y + p(x)y + q(x)y = 0 (A)
on (a, b). Show that if y is a solution of (A) on (a, b), there’s exactly one way to choose c and c so that y = c 1 2 1 y1 + c2 y2 on
(a, b) .
37. Suppose p and q are continuous on (a, b) and x is in (a, b). Let y and y be the solutions of
0 1 2

′′ ′
y + p(x)y + q(x)y = 0 (A)
such that
′ ′
y1 (x0 ) = 1, y (x0 ) = 0 and y2 (x0 ) = 0, y (x0 ) = 1.
1 2
(Theorem 5.1.1 implies that each of these initial value problems has a unique solution on (a, b).)
a. Show that {y , y } is linearly independent on (a, b).
1 2
b. Show that an arbitrary solution y of (A) on (a, b) can be written as y = y(x 0 )y1
′
+ y (x0 )y2 .
c. Express the solution of the initial value problem
′′ ′ ′
y + p(x)y + q(x)y = 0, y(x0 ) = k0 , y (x0 ) = k1
as a linear combination of y and y .1 2
38. Find solutions y and y of the equation y

1 2
′′
=0 that satisfy the initial conditions
′ ′
y1 (x0 ) = 1, y (x0 ) = 0 and y2 (x0 ) = 0, y (x0 ) = 1.
1 2
Then use Exercise 5.1.37 (c) to write the solution of the initial value problem
′′ ′
y = 0, y(0) = k0 , y (0) = k1
as a linear combination of y and y . 1 2
39. Let x be an arbitrary real number. Given (Example 5.1.1) that e and e
0
x −x
are solutions of y ′′
−y = 0 , find solutions y1
and y of y − y = 0 such that

2
′′
′ ′
y1 (x0 ) = 1, y (x0 ) = 0 and y2 (x0 ) = 0, y (x0 ) = 1.
1 2
′′ ′
y − y = 0, y(x0 ) = k0 , y (x0 ) = k1
40. Let x be an arbitrary real number. Given (Example 5.1.2) that

0 cos ωx and sin ωx are solutions of y
′′ 2
+ω y = 0 , find
solutions of y + ω y = 0 such that
′′ 2
′ ′
y1 (x0 ) = 1, y (x0 ) = 0 and y2 (x0 ) = 0, y (x0 ) = 1.
1 2
′′ 2 ′
y + ω y = 0, y(x0 ) = k0 , y (x0 ) = k1
as a linear combination of y and y . Use the identities

1 2
cos(A + B) = cos A cos B − sin A sin B
sin(A + B) = sin A cos B + cos A sin B
to simplify your expressions for y , y , and y.1 2
41. Recall from Exercise 5.1.4 that 1/(x − 1) and 1/(x + 1) are solutions of
2 ′′ ′
(x − 1)y + 4x y + 2y = 0 (A)
on (−1, 1). Find solutions of (A) such that

′ ′
y1 (0) = 1, y (0) = 0 and y2 (0) = 0, y (0) = 1.
1 2
Then use Exercise 5.1.37 (c) to write the solution of initial value problem
2 ′′ ′ ′
(x − 1)y + 4x y + 2y = 0, y(0) = k0 , y (0) = k1
42.

a. Verify that y 1 =x
2
and y 2 =x
3
satisfy
2 ′′ ′
x y − 4x y + 6y = 0 (A)
on (−∞, ∞) and that {y , y } is a fundamental set of solutions of (A) on (−∞, 0) and (0, ∞).
1 2
b. Let a , a , b , and b be constants. Show that

1 2 1 2
2 3
a1 x + a2 x , x ≥ 0,
y ={
2 3
b1 x + b2 x , x <0
is a solution of (A) on (−∞, ∞) if and only if a1 = b1 . From this, justify the statement that y is a solution of (A) on
(−∞, ∞) if and only if
2 3
c1 x + c2 x , x ≥ 0,
y ={
2 3
c1 x + c3 x , x < 0,
where c , c , and c are arbitrary constants.

1 2 3
c. For what values of k and k does the initial value problem

0 1
2 ′′ ′ ′
x y − 4x y + 6y = 0, y(0) = k0 , y (0) = k1
have a solution? What are the solutions?

d. Show that if x ≠ 0 and k , k are arbitrary constants, the initial value problem
0 0 1
2 ′′ ′ ′
x y − 4x y + 6y = 0, y(x0 ) = k0 , y (x0 ) = k1 (B)
has infinitely many solutions on (−∞, ∞). On what interval does (B) have a unique solution?
43.
a. Verify that y 1 =x and y 2 =x
2
satisfy
2 ′′ ′
x y − 2x y + 2y = 0 (A)
on (−∞, ∞) and that {y , y } is a fundamental set of solutions of (A) on (−∞, 0) and (0, ∞).
1 2
b. Let a , a , b , and b be constants. Show that

1 2 1 2
2
a1 x + a2 x , x ≥ 0,
y ={
2
b1 x + b2 x , x <0
is a solution of (A) on (−∞, ∞) if and only if a = b and a = b . From this, justify the statement that the general
1 1 2 2
solution of (A) on (−∞, ∞) is y = c x + c x , where c and c are arbitrary constants.

1 2
2
1 2

0 1
2 ′′ ′ ′
x y − 2x y + 2y = 0, y(0) = k0 , y (0) = k1

d. Show that if x ≠ 0 and k , k are arbitrary constants then the initial value problem
0 0 1
2 ′′ ′ ′
x y − 2x y + 2y = 0, y(x0 ) = k0 , y (x0 ) = k1
has a unique solution on (−∞, ∞).

44.
a. Verify that y 1 =x
3
and y 2 =x
4
satisfy
2 ′′ ′
x y − 6x y + 12y = 0 (A)
on (−∞, ∞), and that {y , y } is a fundamental set of solutions of (A) on (−∞, 0) and (0, ∞).
1 2
b. Show that y is a solution of (A) on (−∞, ∞) if and only if

3 4
a1 x + a2 x , x ≥ 0,
y ={
3 4
b1 x + b2 x , x < 0,

where a , a , b , and b are arbitrary constants.
1 2 1 2

0 1
2 ′′ ′ ′
x y − 6x y + 12y = 0, y(0) = k0 , y (0) = k1

d. Show that if x ≠ 0 and k , k are arbitrary constants then the initial value problem
0 0 1
2 ′′ ′ ′
x y − 6x y + 12y = 0, y(x0 ) = k0 , y (x0 ) = k1 (B)
has infinitely many solutions on (−∞, ∞). On what interval does (B) have a unique solution?

5.2: Constant Coefficient Homogeneous Equations
This section deals with homogeneous equations of the special form
′′ ′
ay + b y + cy = 0, (5.2.1)
where a , b , and c are constant (a ≠ 0 ). When you've completed this section you'll know everything there is to know about
solving such equations.
If a, b, and c are real constants and a ≠ 0 , then
′′ ′
ay + b y + cy = F (x)
is said to be a constant coefficient equation. In this section we consider the homogeneous constant coefficient equation
′′ ′
ay + b y + cy = 0. (5.2.2)
As we’ll see, all solutions of Equation 5.2.2 are defined on (−∞, ∞). This being the case, we’ll omit references to the
interval on which solutions are defined, or on which a given set of solutions is a fundamental set, etc., since the interval will
always be (−∞, ∞).
The key to solving Equation 5.2.2 is that if y = e where r is a constant then the left side of Equation 5.2.2 is a multiple of
rx
e ; thus, if y = e then y = re and y = r e , so

rx rx ′ rx ′′ 2 rx
′′ ′ 2 rx rx rx 2 rx
ay + b y + cy = ar e + bre + ce = (ar + br + c)e . (5.2.3)
The quadratic polynomial

2
p(r) = ar + br + c
is the characteristic polynomial of Equation 5.2.2, and p(r) = 0 is the characteristic equation. From Equation 5.2.3 we can
see that y = e is a solution of Equation 5.2.2 if and only if p(r) = 0 .
rx
The roots of the characteristic equation are given by the quadratic formula
− − −−−−−
2
−b ± √ b − 4ac
r = . (5.2.4)
2a
Consider Three Cases:

Case 1: b2
− 4ac > 0 , so the characteristic equation has two distinct real roots.
Case 2: b2
− 4ac = 0 , so the characteristic equation has a repeated real root.
Case 3: b2
− 4ac < 0 , so the characteristic equation has complex roots.
In each case we’ll start with an example.
Case 1: Distinct Real Roots
Example 5.2.1
′′ ′
y + 6 y + 5y = 0. (5.2.5)

′′ ′ ′
y + 6 y + 5y = 0, y(0) = 3, y (0) = −1. (5.2.6)
Solution
a. The characteristic polynomial of Equation 5.2.5 is

2
p(r) = r + 6r + 5 = (r + 1)(r + 5).
Since p(−1) = p(−5) = 0 , y = e 1 and y = e

−x
are solutions of Equation
2
−5x
5.2.5 . Since y2 / y1 = e
−4x
is
nonconstant, Theorem 5.1.6 implies that the general solution of Equation 5.2.5 is
−x −5x
y = c1 e + c2 e . (5.2.7)
b. We must determine c and c in Equation

1 2 5.2.7 so that y satisfies the initial conditions in Equation 5.2.6 .
Differentiating Equation 5.2.7 yields
′ −x −5x
y = −c1 e − 5 c2 e . (5.2.8)
Imposing the initial conditions y(0) = 3, ′

y (0) = −1 in Equation 5.2.7 and Equation 5.2.8 yields
c1 + c2 = 3
−c1 − 5 c2 = −1.
The solution of this system is c 1 = 7/2, c2 = −1/2 . Therefore the solution of Equation 5.2.6 is
7 1
−x −5x
y = e − e .
2 2
Figure 5.2.1 is a graph of this solution.

If the characteristic equation has arbitrary distinct real roots r and r , then y = e and y 1 2 1
r1 x
2 =e are solutions of
r2 x
ay + b y + cy = 0 . Since y / y = e is nonconstant, Theorem 5.1.6 implies that {y , y is a fundamental set of

′′ ′ ( r2 −r1 )x
2 1 1 2}
solutions of ay + by + cy = 0 .
′′ ′
Figure 5.2.1 : y = 7
2
e
−x
−
1
2
−5x
e
Case 2: A Repeated Real Root
Example 5.2.2
′′ ′
y + 6 y + 9y = 0. (5.2.9)

′′ ′ ′
y + 6 y + 9y = 0, y(0) = 3, y (0) = −1. (5.2.10)
Solution
2 2
p(r) = r + 6r + 9 = (r + 3 ) ,

so the characteristic equation has the repeated real root r = −3 . Therefore y = e is a solution of Equation 5.2.9.
1 1
−3x
Since the characteristic equation has no other roots, Equation 5.2.9 has no other solutions of the form e . We look for rx
solutions of the form y = u y = u e , where u is a function that we’ll now determine. (This should remind you of the
1
−3x
method of variation of parameters used in Section 2.1 to solve the nonhomogeneous equation y + p(x)y = f (x) , given a ′
solution y of the complementary equation y + p(x)y = 0 . It’s also a special case of a method called reduction of order
1
′
that we’ll study in Section 5.6. For other ways to obtain a second solution of Equation 5.2.9 that’s not a multiple of e , −3x
see Exercises 5.1.9, 5.1.12, and 5.1.33.

If y = ue −3x
, then
′ ′ −3x −3x ′′ ′′ −3x ′ −3x −3x
y =u e − 3u e and y =u e − 6u e + 9u e ,
so
′′ ′ −3x ′′ ′ ′
y + 6 y + 9y =e [(u − 6 u + 9u) + 6(u − 3u) + 9u]
−3x ′′ ′ ′′ −3x
=e [u − (6 − 6)u + (9 − 18 + 9)u] = u e .
Therefore y = ue is a solution of Equation 5.2.9 if and only if u

−3x ′′
=0 , which is equivalent to u = c 1 + c2 x , where c
1
and c are constants. Therefore any function of the form

2
−3x
y =e (c1 + c2 x) (5.2.11)
is a solution of Equation 5.2.9. Letting c = 1 and c = 0 yields the solution y = e

1 that we already knew. Letting
2 1
−3x
c = 0 and c = 1 yields the second solution y = x e . Since y /y = x is nonconstant, Theorem 5.1.6 implies that
−3x
1 2 2 2 1
{ y , y } is fundamental set of solutions of Equation 5.2.9, and Equation 5.2.11 is the general solution.
1 2
b. Differentiating Equation 5.2.11 yields

′ −3x −3x
y = −3 e (c1 + c2 x) + c2 e . (5.2.12)
Imposing the initial conditions y(0) = 3, y (0) = −1 in Equation 5.2.11 and Equation
′
5.2.12 yields c1 = 3 and
−3 c + c = −1 , so c = 8 . Therefore the solution of Equation 5.2.10 is
1 2 2
−3x
y =e (3 + 8x).
Figure 5.2.2 : y = e −3x

(3 + 8x)
If the characteristic equation of ay ′′ ′

+ b y + cy = 0 has an arbitrary repeated root r , the characteristic polynomial must be1
2 2 2
p(r) = a(r − r1 ) = a(r − 2 r1 r + r ).
1
Therefore

2 2 2
ar + br + c = ar − (2ar1 )r + ar ,
1
which implies that b = −2ar and c = ar . Therefore

1
2
1
ay
′′
+ b y + cy = 0
′
can be written as a(y
′′ ′ 2
− 2 r1 y + r y) = 0
1
.
Since a ≠ 0 this equation has the same solutions as
′′ ′ 2
y − 2 r1 y + r y = 0. (5.2.13)
1
Since p(r ) = 0 , t y = e
1 1is a solution of ay + by + cy = 0 , and therefore of Equation 5.2.13. Proceeding as in Example
r1 x ′′ ′
5.2.1, we look for other solutions of Equation 5.2.13 of the form y = ue ; then r1 x
′ ′ r1 x r1 x ′′ ′′ r1 x ′ r1 x 2 r1 x
y =u e + ru e and y =u e + 2 r1 u e + r ue ,
1
so
′′ ′ 2 rx ′′ ′ 2 ′ 2
y − 2 r1 y + r y =e [(u + 2 r1 u + r u) − 2 r1 (u + r1 u) + r u]
1 1 1
r1 x ′′ ′ 2 2 2 ′′ r1 x
=e [u + (2 r1 − 2 r1 )u + (r − 2r + r )u] = u e .
1 1 1
Therefore y = ue is a solution of Equation 5.2.13 if and only if

r1 x
u
′′
=0 , which is equivalent to u = c1 + c2 x , where c1
and c are constants. Hence, any function of the form

2
r1 x
y =e (c1 + c2 x) (5.2.14)
is a solution of Equation 5.2.13. Letting c = 1 and c = 0 here yields the solution y = e

1 that we already knew. Letting
2 1
r1 x
c = 0 and c = 1 yields the second solution y = x e . Since y /y = x is nonconstant, Theorem 5.1.6 implies that
r1 x
1 2 2 2 1
{ y , y } is a fundamental set of solutions of Equation 5.2.13, and Equation 5.2.14 is the general solution.
1 2
Case 3: Complex Conjugate Roots
Example 5.2.3
′′ ′
y + 4 y + 13y = 0. (5.2.15)

′′ ′ ′
y + 4 y + 13y = 0, y(0) = 2, y (0) = −3. (5.2.16)
Solution:
2 2 2
p(r) = r + 4r + 13 = r + 4r + 4 + 9 = (r + 2 ) + 9.
The roots of the characteristic equation are r = −2 + 3i and r = −2 − 3i . By analogy with Case 1, it is reasonable to
1 2
expect that e and e

(−2+3i)x
are solutions of Equation 5.2.15. This is true (see Exercise 5.2.34); however, there are
(−2−3i)x
difficulties here, since you are probably not familiar with exponential functions with complex arguments, and even if you
are, it is inconvenient to work with them, since they are complex–valued. We’ll take a simpler approach, which we
motivate as follows: the exponential notation suggests that
(−2+3i)x −2x 3ix (−2−3i)x −2x −3ix
e =e e and e =e e ,
so even though we haven’t defined e 3ix

and e , it is reasonable to expect that every linear combination of e
−3ix (−2+3i)x
and e can be written as y = ue

(−2−3i)x −2x
, where u depends upon x . To determine u , we note that if y = ue then −2x
′ ′ −2x −2x ′′ ′′ −2x ′ −2x −2x

y =u e − 2u e and y =u e − 4u e + 4u e ,
so
′′ ′ −2x ′′ ′ ′
y + 4 y + 13y =e [(u − 4 u + 4u) + 4(u − 2u) + 13u]
−2x ′′ ′ −2x ′′
=e [u − (4 − 4)u + (4 − 8 + 13)u] = e (u + 9u).
Therefore y = ue −2x
is a solution of Equation 5.2.15if and only if

′′
u + 9u = 0.
From Example
Example 5.2.1:
Add text here. For the automatic number to work, you need to add the “AutoNum” template (preferably at5.1.2}, the
general solution of this equation is
u = c1 cos 3x + c2 sin 3x.
Therefore any function of the form

−2x
y =e (c1 cos 3x + c2 sin 3x) (5.2.17)
is a solution of Equation 5.2.15. Letting c = 1 and c = 0 yields the solution y = e

1 2 cos 3x . Letting c = 0 and 1
−2x
1
c = 1 yields the second solution y = e sin 3x . Since y / y = tan 3x is nonconstant, Theorem 5.1.6 implies that
−2x
2 2 2 1
1 2
b. Imposing the condition y(0) = 2 in Equation 5.2.17 shows that c 1 =2 . Differentiating Equation 5.2.17 yields
′ −2x −2x
y = −2 e (c1 cos 3x + c2 sin 3x) + 3 e (−c1 sin 3x + c2 cos 3x),
and imposing the initial condition ′

y (0) = −3 here yields −3 = −2 c1 + 3 c2 = −4 + 3 c2 , so c2 = 1/3 . Therefore the
solution of Equation 5.2.16is
1
−2x
y =e (2 cos 3x + sin 3x).
3
Figure 5.2.3 is a graph of this function.

Now suppose the characteristic equation of ay
′′ ′
+ b y + cy = 0 has arbitrary complex roots; thus, b
2
− 4ac < 0 and,
from Equation 5.2.4, the roots are
− −−−−−− − −−−−−−
2 2
−b + i √ 4ac − b −b − i √ 4ac − b
r1 = , r2 = ,
2a 2a
which we rewrite as
r1 = λ + iω, r2 = λ − iω, (5.2.18)
with
− −−−−−−
b √ 4ac − b2
λ =− , ω = .
2a 2a
Don’t memorize these formulas. Just remember that r and r are of the form Equation 5.2.18, where λ is an arbitrary
1 2
real number and ω is positive; λ and ω are the real and imaginary parts, respectively, of r . Similarly, λ and −ω are the 1
real and imaginary parts of r . We say that r and r are complex conjugates,
2 1 2

Figure 5.2.3 : y = e −2x
(2 cos 3x +
1
3
sin 3x)
which means that they have the same real part and their imaginary parts have the same absolute values, but opposite signs.
As in Example 5.2.3, it is reasonable to to expect that the solutions of ay ′′ ′

+ b y + cy = 0 are linear combinations of e (λ+iω)x
and e . Again, the exponential notation suggests that

(λ−iω)x
(λ+iω)x λx iωx (λ−iω)x λx −iωx

e =e e and e =e e ,
so even though we haven’t defined e and e , it is reasonable to expect that every linear combination of e
iωx −iωx
and (λ+iω)x
e
(λ−iω)x
can be written as y = ue , where u depends upon x. To determine u we first observe that since r = λ + iω and
λx
1
r = λ − iω are the roots of the characteristic equation, p must be of the form

2
p(r) = a(r − r1 )(r − r2 )
= a(r − λ − iω)(r − λ + iω)
2 2
= a [(r − λ ) +ω ]
2 2 2
= a(r − 2λr + λ + ω ).
Therefore ay ′′ ′
+ b y + cy = 0 can be written as
′′ ′ 2 2
a [y − 2λ y + (λ + ω )y] = 0.
Since a ≠ 0 this equation has the same solutions as

′′ ′ 2 2
y − 2λ y + (λ + ω )y = 0. (5.2.19)
To determine u we note that if y = ue λx

then
′ ′ λx λx ′′ ′′ λx ′ λx 2 λx
y =u e + λu e and y =u e + 2λ u e + λ ue .
Substituting these expressions into Equation 5.2.19and dropping the common factor e λx
yields
′′ ′ 2 ′ 2 2
(u + 2λ u + λ u) − 2λ(u + λu) + (λ + ω )u = 0,
which simplifies to
′′ 2
u + ω u = 0.
From Example 5.1.2, the general solution of this equation is

u = c1 cos ωx + c2 sin ωx.
Therefore any function of the form

λx
y =e (c1 cos ωx + c2 sin ωx) (5.2.20)

is a solution of Equation 5.2.19. Letting c = 1 and c = 0 here yields the solution y = e cos ωx . Letting c = 0 and
1 2 1
λx
1
c = 1 yields a second solution y = e sin ωx . Since y / y = tan ωx is nonconstant, so Theorem 5.1.6 implies that
λx
2 2 2 1
1 2
Summary
The next theorem summarizes the results of this section.
Theorem 5.2.1
Let p(r) = ar 2
+ br + c be the characteristic polynomial of
′′ ′
ay + b y + cy = 0. (5.2.21)
Then:
a. If p(r) = 0 has distinct real roots r and r , then the general solution of Equation 5.2.21 is
1 2
r1 x r2 x
y = c1 e + c2 e .
b. If p(r) = 0 has a repeated root r , then the general solution of Equation 5.2.21 is
1
r1 x
y =e (c1 + c2 x).
c. If p(r) = 0 has complex conjugate roots r1 = λ + iω and r2 = λ − iω ( where ω > 0), then the general solution of
Equation 5.2.21 is
λx
y =e (c1 cos ωx + c2 sin ωx).

5.2E: Constant Coefficient Homogeneous Equations (Exercises)
Q5.2.1
1. y ′′
+ 5 y − 6y = 0
′
2. y ′′
− 4 y + 5y = 0
′
3. y ′′
+ 8 y + 7y = 0
′
4. y ′′
− 4 y + 4y = 0
′
5. y ′′
+ 2 y + 10y = 0
′
6. y ′′
+ 6 y + 10y = 0
′
7. y ′′
− 8 y + 16y = 0
′
8. y ′′
+y
′
=0
9. y ′′
− 2 y + 3y = 0
′
10. y ′′
+ 6 y + 13y = 0
′
11. 4y ′′
+ 4 y + 10y = 0
′
12. 10y ′′
− 3y − y = 0
′
Q5.2.2
13. y ′′
+ 14 y + 50y = 0,
′
y(0) = 2,
′
y (0) = −17
14. 6y ′′
− y − y = 0,
′
y(0) = 10,
′
y (0) = 0
15. 6y ′′
+ y − y = 0,
′
y(0) = −1,
′
y (0) = 3
16. 4y ′′
− 4 y − 3y = 0,
′
y(0) =
13
12
,
′
y (0) =
23
24
17. 4y ′′
− 12 y + 9y = 0,
′
y(0) = 3,
′
y (0) =
5
Q5.2.3
18. y ′′
+ 7 y + 12y = 0,
′
y(0) = −1,
′
y (0) = 0
19. y ′′
− 6 y + 9y = 0,
′
y(0) = 0,
′
y (0) = 2
20. 36y ′′
− 12 y + y = 0,
′
y(0) = 3,
′
y (0) =
5
21. y ′′
+ 4 y + 10y = 0,
′
y(0) = 3,
′
y (0) = −2
Q5.2.4
22.
a. Suppose y is a solution of the constant coefficient homogeneous equation
′′ ′
ay + b y + cy = 0. (A)
Let z(x) = y(x − x 0) , where x is an arbitrary real number. Show that

0
′′ ′
az + b z + cz = 0.
b. Let z1 (x) = y1 (x − x0 ) and z (x) = y (x − x ) , where

2 2 0 { y1 , y2 } is a fundamental set of solutions of (A). Show that
{ z1 , z2 } is also a fundamental set of solutions of (A).

c. The statement of Theorem 5.2.1 is convenient for solving an initial value problem
′′ ′ ′
ay + b y + cy = 0, y(0) = k0 , y (0) = k1 ,
where the initial conditions are imposed at x 0 =0 . However, if the initial value problem is
′′ ′ ′
ay + b y + cy = 0, y(x0 ) = k0 , y (x0 ) = k1 , (B)
where x 0 ≠0 , then determining the constants in

r1 x r2 x r1 x λx
y = c1 e + c2 e , y =e (c1 + c2 x), or y = e (c1 cos ωx + c2 sin ωx)
(whichever is applicable) is more complicated. Use (b) to restate Theorem 5.2.1 in a form more convenient for solving (B).
Q5.2.5
In Exercises 5.2.23-5.2.28 use a method suggested by Exercise 5.2.22 to solve the initial value problem.
23. y ′′ ′
+ 3 y + 2y = 0, y(1) = −1,
′
y (1) = 4
24. y ′′ ′
− 6 y − 7y = 0, y(2) = −
1
3
,
′
y (2) = −5
25. y ′′ ′
− 14 y + 49y = 0, y(1) = 2,
′
y (1) = 11
26. 9y ′′ ′
+ 6 y + y = 0, y(2) = 2,
′
y (2) = −
14
27. 9y ′′
+ 4y = 0, y(π/4) = 2,
′
y (π/4) = −2
28. y ′′
+ 3y = 0, y(π/3) = 2,
′
y (π/3) = −1
Q5.2.6
29. Prove: If the characteristic equation of
′′ ′
ay + b y + cy = 0 (A)
has a repeated negative root or two roots with negative real parts, then every solution of (A) approaches zero as x → ∞ .
30. Suppose the characteristic polynomial of ay + by ′′ ′
+ cy = 0 has distinct real roots r and r . Use a method suggested by
1 2
Exercise 5.2.22 to find a formula for the solution of

′′ ′ ′
ay + b y + cy = 0, y(x0 ) = k0 , y (x0 ) = k1 .
31 Suppose the characteristic polynomial of ay + by ′′ ′

+ cy = 0 has a repeated real root r1 . Use a method suggested by
Exercise 5.2.22 to find a formula for the solution of
′′ ′ ′
ay + b y + cy = 0, y(x0 ) = k0 , y (x0 ) = k1 .
32. Suppose the characteristic polynomial of ay + by + cy = 0 has complex conjugate roots ′′ ′

λ ± iω . Use a method
suggested by Exercise 5.2.22 to find a formula for the solution of
′′ ′ ′
ay + b y + cy = 0, y(x0 ) = k0 , y (x0 ) = k1 .
33. Suppose the characteristic equation of

′′ ′
ay + b y + cy = 0 (A)
has a repeated real root r . Temporarily, think of e

1
rx
as a function of two real variables x and r.
a. Show that
2
∂ rx
∂ rx rx 2 rx
a (e )+b (e ) + ce = a(r − r1 ) e . (B)
2
∂ x ∂x
b. Differentiate (B) with respect to r to obtain

2
∂ ∂ rx
∂ ∂ rx rx rx
a ( (e )) + b ( (e )) + c(x e ) = [2 + (r − r1 )x]a(r − r1 )e . (C)
∂r ∂2x ∂r ∂x

c. Reverse the orders of the partial differentiations in the first two terms on the left side of (C) to obtain
2
∂ rx
∂ rx rx rx
a (x e )+b (x e ) + c(x e ) = [2 + (r − r1 )x]a(r − r1 )e . (D)
2
∂x ∂x
d. Set r = r in (B) and (D) to see that y

1 1 =e
r1 x
and y 2 = xe
r1 x
are solutions of (A)
34. In calculus you learned that e , cos u, and sin u can be represented by the infinite series
u
∞ n 2 3 n
u
u u u u u
e =∑ =1+ + + +⋯ + +⋯ (A)
n! 1! 2! 3! n!
n=0
∞ 2n 2 4 2n
n
u u u n
u
cos u = ∑(−1 ) =1− + + ⋯ + (−1 ) +⋯ , (B)
(2n)! 2! 4! (2n)!
n=0
and
∞ 2n+1 3 5 2n+1
n
u u u n
u
sin u = ∑(−1 ) =u− + + ⋯ + (−1 ) +⋯ (C)
(2n + 1)! 3! 5! (2n + 1)!
n=0
for all real values of u. Even though you have previously considered (A) only for real values of u, we can set u = iθ , where θ
is real, to obtain
∞ n
(iθ)
iθ
e =∑ . (D)
n!
n=0
Given the proper background in the theory of infinite series with complex terms, it can be shown that the series in (D)
converges for all real θ .
a. Recalling that i2
= −1, write enough terms of the sequence {i n
} to convince yourself that the sequence is repetitive:
1, i, −1, −i, 1, i, −1, −i, 1, i, −1, −i, 1, i, −1, −i, ⋯ .
Use this to group the terms in (D) as

2 4 3 5
θ θ θ θ
iθ
e = (1 − + + ⋯) + i (θ − + + ⋯)
2 4 3! 5!
∞ 2n ∞ 2n+1
n
θ n
θ
= ∑(−1 ) + i ∑(−1 ) .
(2n)! (2n + 1)!
n=0 n=0
By comparing this result with (B) and (C), conclude that

iθ
e = cos θ + i sin θ. (E)
This is Euler’s identity.

b. Starting from
iθ1 iθ2
e e = (cos θ1 + i sin θ1 )(cos θ2 + i sin θ2 ),
collect the real part (the terms not multiplied by i) and the imaginary part (the terms multiplied by i) on the right, and use
the trigonometric identities
cos(θ1 + θ2 ) = cos θ1 cos θ2 − sin θ1 sin θ2
sin(θ1 + θ2 ) = sin θ1 cos θ2 + cos θ1 sin θ2
to verify that
i( θ1 +θ2 ) iθ1 iθ2
e =e e ,
as you would expect from the use of the exponential notation e . iθ
c. If α and β are real numbers, define

α+iβ α iβ α
e =e e =e (cos β + i sin β). (F)
Show that if z 1 = α1 + i β1 and z2 = α2 + i β2 then

z1 +z2 z1 z2
e =e e .
d. Let a , b , and c be real numbers, with a ≠0 . Let z = u + iv where u and v are real-valued functions of x. Then we say
that z is a solution of
′′ ′
ay + b y + cy = 0 (G)
if u and v are both solutions of (G). Use Theorem 5.2.1 (c) to verify that if the characteristic equation of (G) has complex
conjugate roots λ ± iω then z = e 1 and z = e
(λ+iω)x
are both solutions of (G).
2
(λ−iω)x

5.3: Nonhomogeneous Linear Equations
We’ll now consider the nonhomogeneous linear second order equation
′′ ′
y + p(x)y + q(x)y = f (x), (5.3.1)
where the forcing function f isn’t identically zero. The next theorem, an extension of Theorem 5.1.1, gives sufficient
conditions for existence and uniqueness of solutions of initial value problems for Equation 5.3.1. We omit the proof, which is
beyond the scope of this book.
Theorem 5.3.1 : Uniqueness

Suppose p, , q and f are continuous on an open interval (a, b), let x be any point in (a, b), and let k and k be arbitrary 0 0 1
real numbers. Then the initial value problem

′′ ′ ′
y + p(x)y + q(x)y = f (x), y(x0 ) = k0 , y (x0 ) = k1
To find the general solution of Equation 5.3.1 on an interval (a, b) where p, q, and f are continuous, it is necessary to find the
general solution of the associated homogeneous equation
′′ ′
y + p(x)y + q(x)y = 0 (5.3.2)
on (a, b). We call Equation 5.3.2 the complementary equation for Equation 5.3.1.
The next theorem shows how to find the general solution of Equation 5.3.1 if we know one solution y of Equation 5.3.1 and p
a fundamental set of solutions of Equation 5.3.2. We call y a particular solution of Equation 5.3.1 ; it can be any solution
p
that we can find, one way or another.
Theorem 5.3.2
Suppose p, q, and f are continuous on (a, b). Let y be a particular solution of p
′′ ′
y + p(x)y + q(x)y = f (x) (5.3.3)
on (a, b), and let {y 1, y2 } be a fundamental set of solutions of the complementary equation
′′ ′
y + p(x)y + q(x)y = 0 (5.3.4)
on (a, b). Then y is a solution of (5.3.3) on (a, b) if and only if
y = yp + c1 y1 + c2 y2 , (5.3.5)
where c and c are constants.

1 2
Proof
We first show that y in Equation 5.3.5 is a solution of Equation 5.3.3 for any choice of the constants c1 and c2 .
Differentiating Equation 5.3.5twice yields
′ ′ ′ ′ ′′ ′′ ′′ ′′
y = yp + c1 y + c2 y and y = yp + c1 y + c2 y ,
1 2 1 2
so
′′ ′ ′′ ′′ ′′ ′ ′ ′
y + p(x)y + q(x)y = (yp + c1 y + c2 y ) + p(x)(yp + c1 y + c2 y ) + q(x)(yp + c1 y1 + c2 y2 )
1 2 1 2
′′ ′ ′′ ′ ′′ ′
= (yp + p(x)yp + q(x)yp ) + c1 (y + p(x)y + q(x)y1 ) + c2 (y + p(x)y + q(x)y2 )
1 1 2 2
= f + c1 ⋅ 0 + c2 ⋅ 0 = f ,
since y satisfies Equation 5.3.3and y and y satisfy Equation 5.3.4.

p 1 2

Now we’ll show that every solution of Equation 5.3.3 has the form Equation 5.3.5 for some choice of the constants c 1
and c . Suppose y is a solution of Equation 5.3.3. We’ll show that y − y is a solution of Equation 5.3.4, and therefore
2 p
of the form y − y = c y + c y , which implies Equation 5.3.5. To see this, we compute

p 1 1 2 2
′′ ′ ′′ ′′ ′ ′
(y − yp ) + p(x)(y − yp ) + q(x)(y − yp ) = (y − yp ) + p(x)(y − yp ) + q(x)(y − yp )
′′ ′ ′′ ′
= (y + p(x)y + q(x)y) − (yp + p(x)yp + q(x)yp )
= f (x) − f (x) = 0,
since y and y both satisfy Equation 5.3.3.

p
We say that Equation 5.3.5is the general solution of (5.3.3) on (a, b) .
If P , P , and F are continuous and P has no zeros on (a, b), then Theorem 5.3.2 implies that the general solution of
0 1 0
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F (x) (5.3.6)
on (a, b) is y = y p + c1 y1 + c2 y2 , where y is a particular solution of Equation 5.3.6 on (a, b) and {y

p 1, y2 } is a fundamental
set of solutions of
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0
on (a, b). To see this, we rewrite Equation 5.3.6 as

P1 (x) P2 (x) F (x)
′′ ′
y + y + y =
P0 (x) P0 (x) P0 (x)
and apply Theorem 5.3.2 with p = P 1 / P0 ,q=P 2 / P0 , and f = F /P0 .
To avoid awkward wording in examples and exercises, we will not specify the interval (a, b) when we ask for the general
solution of a specific linear second order equation, or for a fundamental set of solutions of a homogeneous linear second
order equation. Let’s agree that this always means that we want the general solution (or a fundamental set of solutions, as
the case may be) on every open interval on which p, q, and f are continuous if the equation is of the form Equation 5.3.3,
or on which P , P , P , and F are continuous and P has no zeros, if the equation is of the form Equation 5.3.6. We
0 1 2 0
leave it to you to identify these intervals in specific examples and exercises.
For completeness, we point out that if P , P , P , and F are all continuous on an open interval (a, b), but P does have a zero
0 1 2 0
in (a, b), then Equation 5.3.6 may fail to have a general solution on (a, b) in the sense just defined. Exercises 5.1.42-5.1.44
illustrate this point for a homogeneous equation.
In this section we to limit ourselves to applications of Theorem 5.3.2 where we can guess at the form of the particular solution.
Example 5.3.1
′′
y + y = 1. (5.3.7)

′′ ′
y + y = 1, y(0) = 2, y (0) = 7. (5.3.8)
Solution
a. We can apply Theorem 5.3.2 with (a, b) = (−∞, ∞) , since the functions p ≡ 0 , q ≡ 1 , and f ≡ 1 in Equation 5.3.7
are continuous on (−∞, ∞). By inspection we see that y ≡ 1 is a particular solution of Equation 5.3.7. Since p
y = cos x and y = sin x form a fundamental set of solutions of the complementary equation y + y = 0 , the general
′′
1 2
solution of Equation 5.3.7 is
y = 1 + c1 cos x + c2 sin x. (5.3.9)

b. Imposing the initial condition y(0) = 2 in Equation 5.3.9 yields 2 = 1 + c , so c 1 1 =1 . Differentiating Equation 5.3.9
yields
′
y = −c1 sin x + c2 cos x.
Imposing the initial condition y ′

(0) = 7 here yields c 2 =7 , so the solution of Equation 5.3.8 is
y = 1 + cos x + 7 sin x.
Figure 5.3.1 is a graph of this function.
Figure 5.3.1 : y = 1 + cos x + 7 sin x
Example 5.3.2
′′ ′ 2
y − 2 y + y = −3 − x + x . (5.3.10)

′′ ′ 2 ′
y − 2 y + y = −3 − x + x , y(0) = −2, y (0) = 1. (5.3.11)
Solution
a. The characteristic polynomial of the complementary equation
′′ ′
y − 2y + y = 0
is r − 2r + 1 = (r − 1) , so y = e and y = x e form a fundamental set of solutions of the complementary equation.

2 2
1
x
2
x
To guess a form for a particular solution of Equation 5.3.10, we note that substituting a second degree polynomial
y = A + Bx + C x
p
2
into the left side of Equation 5.3.10 will produce another second degree polynomial with
coefficients that depend upon A , B , and C . The trick is to choose A , B , and C so the polynomials on the two sides of
Equation 5.3.10 have the same coefficients; thus, if
2 ′ ′′
yp = A + Bx + C x then yp = B + 2C x and yp = 2C ,
so
′′ ′ 2
yp − 2 yp + yp = 2C − 2(B + 2C x) + (A + Bx + C x )
2 2
= (2C − 2B + A) + (−4C + B)x + C x = −3 − x + x .
Equating coefficients of like powers of x on the two sides of the last equality yields

C = 1
B − 4C = −1
A − 2B + 2C = −3,
so C = 1 , B = −1 + 4C = 3 , and A = −3 − 2C + 2B = 1 . Therefore yp = 1 + 3x + x
2
is a particular solution of
Equation 5.3.10 and Theorem 5.3.2 implies that
2 x
y = 1 + 3x + x + e (c1 + c2 x) (5.3.12)

b. Imposing the initial condition y(0) = −2 in Equation 5.3.12 yields −2 = 1 + c1 , so c1 = −3 . Differentiating
′ x x
y = 3 + 2x + e (c1 + c2 x) + c2 e ,
and imposing the initial condition y ′

(0) = 1 here yields 1 = 3 + c 1 + c2 , so c 2 =1 . Therefore the solution of Equation
5.3.11 is
2 x
y = 1 + 3x + x − e (3 − x).
Figure 5.3.2 : y = 1 + 3x + x 2 x
− e (3 − x)
Example 5.3.3
2 ′′ ′ 4
x y + x y − 4y = 2 x (5.3.13)
on (−∞, 0) and (0, ∞).

Solution
In Example 5.1.3, we verified that y1 = x
2
and y2 = 1/ x
2
form a fundamental set of solutions of the complementary
equation
2 ′′ ′
x y + x y − 4y = 0
on (−∞, 0) and (0, ∞). To find a particular solution of Equation 5.3.13, we note that if y = Ax , where A is a constant p
4
then both sides of Equation 5.3.13 will be constant multiples of x and we may be able to choose A so the two sides are
4
equal. This is true in this example, since if y = Ax then

p
4

2 ′′ ′ 2 2 3 4 4 4
x yp + x yp − 4 yp = x (12Ax ) + x(4Ax ) − 4Ax = 12Ax = 2x
if A = 1/6 ; therefore, y = x /6 is a particular solution of Equation

p
4
5.3.13 on (−∞, ∞) . Theorem 5.3.2 implies that
the general solution of Equation 5.3.13 on (−∞, 0) and (0, ∞) is
4
x c2
2
y = + c1 x + .
2
6 x
The Principle of Superposition

The next theorem enables us to break a nonhomogeous equation into simpler parts, find a particular solution for each part, and
then combine their solutions to obtain a particular solution of the original problem.
Theorem 5.3.3 the principle of superposition

Suppose y is a particular solution of
p1
′′ ′
y + p(x)y + q(x)y = f1 (x)
on (a, b) and y is a particular solution of

p2
′′ ′
y + p(x)y + q(x)y = f2 (x)
on (a, b). Then
yp = yp + yp
1 2
′′ ′
y + p(x)y + q(x)y = f1 (x) + f2 (x)
on (a, b).
Proof
If yp = yp
1
+ yp
2
then
′′ ′ ′′ ′
yp + p(x)yp + q(x)yp = (yp + yp ) + p(x)(yp + yp ) + q(x)(yp + yp )
1 2 1 2 1 2
′′ ′ ′′ ′
= (yp + p(x)yp + q(x)yp ) + (yp + p(x)yp + q(x)yp )
1 1 1 2 2 2
= f1 (x) + f2 (x).
It’s easy to generalize Theorem 5.3.3 to the equation

′′ ′
y + p(x)y + q(x)y = f (x) (5.3.14)
where
f = f1 + f2 + ⋯ + fk ;
thus, if y is a particular solution of

pi
′′ ′
y + p(x)y + q(x)y = fi (x)
on (a, b) for i = 1 , 2, …, k , then y + y + ⋯ + y is a particular solution of Equation 5.3.14 on (a, b). Moreover, by a
p1 p2 pk
proof similar to the proof of Theorem 5.3.3 we can formulate the principle of superposition in terms of a linear equation
written in the form
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F (x)
(Exercise 5.3.39); that is, if y is a particular solution of

p
1
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F1 (x)

on (a, b) and y is a particular solution of
p2
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F2 (x)
on (a, b), then y p1 + yp

2
is a solution of
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F1 (x) + F2 (x)
on (a, b).
Example 5.3.4
The function y p1 = x /15
4
2 ′′ ′ 4
x y + 4x y + 2y = 2 x (5.3.15)
on (−∞, ∞) and y p2
2
= x /3 is a particular solution of
2 ′′ ′ 2
x y + 4x y + 2y = 4 x (5.3.16)
on (−∞, ∞) . Use the principle of superposition to find a particular solution of

2 ′′ ′ 4 2
x y + 4x y + 2y = 2 x + 4x (5.3.17)
on (−∞, ∞).
Solution
The right side F (x) = 2x 4
+ 4x
2
in Equation 5.3.17 is the sum of the right sides
4 2
F1 (x) = 2 x and F2 (x) = 4 x .
in Equation 5.3.15and Equation 5.3.16. Therefore the principle of superposition implies that
4 2
x x
yp = yp + yp = +
1 2
15 3
is a particular solution of Equation 5.3.17.

5.3E: Nonhomgeneous Linear Equations (Exercises)
Q5.3.1
In Exercises 5.3.1-5.3.6 find a particular solution by the method used in Example 5.3.2. Then find the general solution and,
where indicated, solve the initial value problem and graph the solution.
1. y ′′ ′
+ 5 y − 6y = 22 + 18x − 18 x
2
2. y ′′ ′
− 4 y + 5y = 1 + 5x
3. y ′′ ′
+ 8 y + 7y = −8 − x + 24 x
2
+ 7x
3
4. y ′′ ′
− 4 y + 4y = 2 + 8x − 4 x
2
5. y ′′ ′
+ 2 y + 10y = 4 + 26x + 6 x
2
+ 10 x ,
3
y(0) = 2,
′
y (0) = 9
6. y ′′ ′
+ 6 y + 10y = 22 + 20x, y(0) = 2, y (0) = −2
′
Q5.3.2
7. Show that the method used in Example 5.3.2 will not yield a particular solution of
′′ ′ 2
y +y = 1 + 2x + x ; (A)
that is, (A) does’nt have a particular solution of the form y p = A + Bx + C x

2
, where A , B , and C are constants.
Q5.3.3
In Exercises 5.3.8-5.3.13 find a particular solution by the method used in Example 5.3.3.
8. x 2
y
′′ ′
+ 7x y + 8y =
6
9. x 2
y
′′ ′
− 7x y + 7y = 13 x
1/2
10. x 2
y
′′ ′
− xy + y = 2x
3
11. x 2
y
′′ ′
+ 5x y + 4y =
1
x
3
12. x 2
y
′′ ′
+ x y + y = 10 x
1/3
13. x 2
y
′′ ′
− 3x y + 13y = 2 x
4
Q5.3.4
14. Show that the method suggested for finding a particular solution in Exercises 5.3.8-5.3.13 will not yield a particular
solution of
2 ′′ ′
1
x y + 3x y − 3y = ; (A)
3
x
that is, (A) doesn’t have a particular solution of the form y p = A/ x

3
.
15. Prove: If a , b , c , α , and M are constants and M ≠0 then
2 ′′ ′ α
ax y + bx y + cy = M x (5.3E.1)
has a particular solution y p = Ax

α
(A = constant) if and only if aα(α − 1) + bα + c ≠ 0 .
Q5.3.5
If a, b, c, and α are constants, then
αx ′′ αx ′ αx 2 αx
α(e ) + b(e ) + ce = (aα + bα + c)e . (5.3E.2)
Use this in Exercises 5.3.16-5.3.21 to find a particular solution. Then find the general solution and, where indicated, solve the
initial value problem and graph the solution.

16. y ′′ ′
+ 5 y − 6y = 6 e
3x
17. y ′′ ′
− 4 y + 5y = e
2x
18. y ′′ ′
+ 8 y + 7y = 10 e
−2x
, y(0) = −2, y (0) = 10
′
19. y ′′ ′
− 4 y + 4y = e ,
x
y(0) = 2,
′
y (0) = 0
20. y ′′ ′
+ 2 y + 10y = e
x/2
21. y ′′ ′
+ 6 y + 10y = e
−3x
Q5.3.6
solution of
′′ ′ 4x
y − 7 y + 12y = 5 e ; (A)
that is, (A) doesn’t have a particular solution of the form y p = Ae

4x
.
23. Prove: If α and M are constants and M ≠0 then constant coefficient equation
′′ ′ αx
ay + b y + cy = M e (5.3E.3)
has a particular solution y p = Ae

αx
(A = constant) if and only if e αx
isn’t a solution of the complementary equation.
Q5.3.7
If ω is a constant, differentiating a linear combination of cos ωx and sin ωx with respect to x yields another linear combination
of cos ωx and sin ωx. In Exercises 5.3.24-5.3.29 use this to find a particular solution of the equation. Then find the general
solution and, where indicated, solve the initial value problem and graph the solution.
24. y ′′ ′
− 8 y + 16y = 23 cos x − 7 sin x
25. y ′′
+y
′
= −8 cos 2x + 6 sin 2x
26. y ′′ ′
− 2 y + 3y = −6 cos 3x + 6 sin 3x
27. y ′′ ′
+ 6 y + 13y = 18 cos x + 6 sin x
28. y ′′ ′
+ 7 y + 12y = −2 cos 2x + 36 sin 2x, y(0) = −3,
′
y (0) = 3
29. y ′′ ′
− 6 y + 9y = 18 cos 3x + 18 sin 3x, y(0) = 2,
′
y (0) = 2
Q5.3.8
30. Find the general solution of
′′ 2
y + ω y = M cos ωx + N sin ωx, (5.3E.4)
0
where M and N are constants and ω and ω are distinct positive numbers.
0
solution of
′′
y + y = cos x + sin x; (A)
that is, (A) does not have a particular solution of the form y p = A cos x + B sin x .
32. Prove: If M , N are constants (not both zero) and ω > 0 , the constant coefficient equation
′′ ′
ay + b y + cy = M cos ωx + N sin ωx (A)
has a particular solution that’s a linear combination of cos ωx and sin ωx if and only if the left side of (A) is not of the form
a(y + ω y) , so that cos ωx and sin ωx are solutions of the complementary equation.
′′ 2
Q5.3.9

In Exercises 5.3.33-5.3.38 refer to the cited exercises and use the principal of superposition to find a particular solution. Then
find the general solution.
33. y ′′ ′
+ 5 y − 6y = 22 + 18x − 18 x
2
+ 6e
3x
(See Exercises 5.3.1 and 5.3.16.)
34. y ′′ ′
− 4 y + 5y = 1 + 5x + e
2x
35. y ′′ ′
+ 8 y + 7y = −8 − x + 24 x
2
+ 7x
3
+ 10 e
−2x
36. y ′′ ′
− 4 y + 4y = 2 + 8x − 4 x
2
+e
x
37. y ′′ ′
+ 2 y + 10y = 4 + 26x + 6 x
2
+ 10 x
3
+e
x/2
38. y ′′ ′
+ 6 y + 10y = 22 + 20x + e
−3x
Q5.3.10
39. Prove: If y is a particular solution of
p
1
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F1 (x) (5.3E.5)
on (a, b) and y p2 is a particular solution of

′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F2 (x) (5.3E.6)
on (a, b) , then y p = yp
1
+ yp
2
is a solution of
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F1 (x) + F2 (x) (5.3E.7)
on (a, b) .
40. Suppose p , q, and f are continuous on (a, b) . Let y1 , y2 , and yp be twice differentiable on (a, b) , such that
y = c1 y1 + c2 y2 + yp is a solution of
′′ ′
y + p(x)y + q(x)y = f (5.3E.8)
on (a, b) for every choice of the constants c 1, c2 . Show that y and y are solutions of the complementary equation on (a, b) .
1 2

5.4: The Method of Undetermined Coefficients I
In this section we consider the constant coefficient equation
′′ ′ αx
ay + b y + cy = e G(x), (5.4.1)
where α is a constant and G is a polynomial.

From Theorem 5.3.2, the general solution of Equation 5.4.1 is y = y + c y + c y , where p 1 1 2 2 yp is a particular solution of
Equation 5.4.1 and {y , y } is a fundamental set of solutions of the complementary equation
1 2
′′ ′
ay + b y + cy = 0.
In Section 5.2 we showed how to find {y , y }. In this section we’ll show how to find y . The procedure that we’ll use is
1 2 p
called the method of undetermined coefficients. Our first example is similar to Exercises 5.3.16-5.3.21.
Example 5.4.1 :
Find a particular solution of
′′ ′ 2x
y − 7 y + 12y = 4 e . (5.4.2)
Then find the general solution.

Solution
Substituting y = Ae for y in Equation 5.4.2 will produce a constant multiple of Ae on the left side of Equation
p
2x 2x
5.4.2, so it may be possible to choose A so that y is a solution of Equation 5.4.2. Let’s try it; if y = Ae then 2x
p p
′′ ′ 2x 2x 2x 2x 2x
yp − 7 yp + 12 yp = 4Ae − 14Ae + 12Ae = 2Ae = 4e
if A = 2 . Therefore y = 2e is a particular solution of Equation

p
2x
5.4.2 . To find the general solution, we note that the
characteristic polynomial of the complementary equation
′′ ′
y − 7 y + 12y = 0 (5.4.3)
is p(r) = r − 7r + 12 = (r − 3)(r − 4) , so
2
{e
3x
,e
4x
} is a fundamental set of solutions of Equation 5.4.3. Therefore
the general solution of Equation 5.4.2 is
2x 3x 4x
y = 2e + c1 e + c2 e .
Example 5.4.2
′′ ′ 4x
y − 7 y + 12y = 5 e . (5.4.4)

Solution
Fresh from our success in finding a particular solution of Equation 5.4.2 — where we chose y = Ae because the right p
2x
side of Equation 5.4.2 is a constant multiple of e — it may seem reasonable to try y = Ae as a particular solution of
2x
p
4x
Equation 5.4.4. However, this will not work, since we saw in Example 5.4.1 that e is a solution of the complementary 4x
equation Equation 5.4.3, so substituting y = Ae into the left side of Equation 5.4.4) produces zero on the left, no
p
4x
matter how we chooseA . To discover a suitable form for y , we use the same approach that we used in Section 5.2 to find
p
a second solution of
′′ ′
ay + b y + cy = 0

in the case where the characteristic equation has a repeated real root: we look for solutions of Equation 5.4.4 in the form
y = ue , where u is a function to be determined. Substituting
4x
4x ′ ′ 4x 4x ′′ ′′ 4x ′ 4x 4x
y = ue , y =u e + 4u e , and y =u e + 8u e + 16u e (5.4.5)
into Equation 5.4.4and canceling the common factor e 4x

yields
′′ ′ ′
(u + 8 u + 16u) − 7(u + 4u) + 12u = 5,
or
′′ ′
u + u = 5.
By inspection we see that up = 5x is a particular solution of this equation, so yp = 5x e

4x
Equation 5.4.4. Therefore
4x 3x 4x
y = 5x e + c1 e + c2 e
is the general solution.
Example 5.4.3
′′ ′ 4x
y − 8 y + 16y = 2 e . (5.4.6)
Solution
Since the characteristic polynomial of the complementary equation
′′ ′
y − 8 y + 16y = 0 (5.4.7)
is p(r) = r − 8r + 16 = (r − 4) , both y = e and y = x e are solutions of Equation 5.4.7. Therefore Equation

2 2
1
4x
2
4x
5.4.6) does not have a solution of the form y = Ae or y = Ax e . As in Example 5.4.2, we look for solutions of
4x 4x
p p
Equation 5.4.6 in the form y = ue , where u is a function to be determined. Substituting from Equation 5.4.5 into
4x
Equation 5.4.6 and canceling the common factor e yields 4x
′′ ′ ′
(u + 8 u + 16u) − 8(u + 4u) + 16u = 2,
or
′′
u = 2.
Integrating twice and taking the constants of integration to be zero shows that up = x
2
is a particular solution of this
equation, so y = x e is a particular solution of Equation 5.4.4. Therefore
p
2 4x
4x 2
y =e (x + c1 + c2 x)
is the general solution.
The preceding examples illustrate the following facts concerning the form of a particular solution yp of a constant coefficent
equation
′′ ′ αx
ay + b y + cy = ke ,
where k is a nonzero constant:

a. If e αx
isn’t a solution of the complementary equation
′′ ′
ay + b y + cy = 0, (5.4.8)
then y = Ae , where A is a constant. (See Example 5.4.1).

p
αx
b. If e is a solution of Equation 5.4.8 but xe is not, then y = Ax e , where A is a constant. (See Example 5.4.2.)
αx αx
p
αx
c. If both e and xe are solutions of Equation 5.4.8, then y = Ax e , where A is a constant. (See Example 5.4.3.)
αx αx
p
2 αx

See Exercise 5.4.30 for the proofs of these facts.
In all three cases you can just substitute the appropriate form for y and its derivatives directly into p
′′ ′ αx
ayp + b yp + c yp = ke ,
and solve for the constant A , as we did in Example 5.4.1. (See Exercises 5.4.31-5.4.33.) However, if the equation is
′′ ′ αx
ay + b y + cy = ke G(x),
where G is a polynomial of degree greater than zero, we recommend that you use the substitution y = ue
αx
as we did in
Examples 5.4.2 and 5.4.3. The equation for u will turn out to be
′′ ′ ′
au + p (α)u + p(α)u = G(x), (5.4.9)
where p(r) = ar + br + c is the characteristic polynomial of the complementary equation and p (r) = 2ar + b (Exercise
2 ′
5.4.30); however, you shouldn’t memorize this since it is easy to derive the equation for u in any particular case. Note,
however, that if e is a solution of the complementary equation then p(α) = 0 , so Equation 5.4.9 reduces to
αx
′′ ′ ′
au + p (α)u = G(x),
while if both and xe are solutions of the complementary equation then

e
αx αx
p(r) = a(r − α)
2
and ′
p (r) = 2a(r − α) , so
p(α) = p (α) = 0 and Equation 5.4.9) reduces to
′
′′
au = G(x).
Example 5.4.4
′′ ′ 3x 2
y − 3 y + 2y = e (−1 + 2x + x ). (5.4.10)
Solution
Substituting
3x ′ ′ 3x 3x ′′ ′′ 3x ′ 3x 3x
y = ue , y =u e + 3u e , andy =u e + 6u e + 9u e
into Equation 5.4.10) and canceling e 3x

yields
′′ ′ ′ 2
(u + 6 u + 9u) − 3(u + 3u) + 2u = −1 + 2x + x ,
or
′′ ′ 2
u + 3 u + 2u = −1 + 2x + x . (5.4.11)
As in Example 5.3.2, in order to guess a form for a particular solution of Equation 5.4.11), we note that substituting a
second degree polynomial u = A + Bx + C x for u in the left side of Equation 5.4.11) produces another second
p
2
degree polynomial with coefficients that depend upon A , B , and C ; thus,

2 ′ ′′
if up = A + Bx + C x then up = B + 2C x and up = 2C .
If u is to satisfy Equation 5.4.11), we must have

p
′′ ′ 2
up + 3 up + 2 up = 2C + 3(B + 2C x) + 2(A + Bx + C x )
2 2
= (2C + 3B + 2A) + (6C + 2B)x + 2C x = −1 + 2x + x .
Equating coefficients of like powers of x on the two sides of the last equality yields
2C =1
2B + 6C =2
2A + 3B + 2C = −1.
Solving these equations for C , B, and A (in that order) yields C = 1/2, B = −1/2, A = −1/4 . Therefore

1 2
up = − (1 + 2x − 2 x )
4
is a particular solution of Equation 5.4.11, and

3x
e
3x 2
yp = up e =− (1 + 2x − 2 x )
4
Example 5.4.5
′′ ′ 3x 2
y − 4 y + 3y = e (6 + 8x + 12 x ). (5.4.12)
Solution
Substituting
3x ′ ′ 3x 3x ′′ ′′ 3x ′ 3x 3x
y = ue , y =u e + 3u e , and y =u e + 6u e + 9u e
into Equation 5.4.12) and canceling e 3x

yields
′′ ′ ′ 2
(u + 6 u + 9u) − 4(u + 3u) + 3u = 6 + 8x + 12 x ,
or
′′ ′ 2
u + 2 u = 6 + 8x + 12 x . (5.4.13)
There’s no u term in this equation, since e is a solution of the complementary equation for Equation 5.4.12). (See
3x
Exercise 5.4.30.) Therefore Equation 5.4.13) does not have a particular solution of the form u = A + Bx + C x that p
2
we used successfully in Example 5.4.4, since with this choice of u , p
′′ ′
up + 2 up = 2C + (B + 2C x)
can’t contain the last term (12x ) on the right side of Equation
2
5.4.13 ). Instead, let’s try u p = Ax + Bx
2
+ Cx
3
on the
grounds that
′ 2 ′′
up = A + 2Bx + 3C x and up = 2B + 6C x
together contain all the powers of x that appear on the right side of Equation 5.4.13).
Substituting these expressions in place of u and u in Equation 5.4.13) yields
′ ′′
2 2 2
(2B + 6C x) + 2(A + 2Bx + 3C x ) = (2B + 2A) + (6C + 4B)x + 6C x = 6 + 8x + 12 x .
Comparing coefficients of like powers of x on the two sides of the last equality shows that u satisfies Equation 5.4.13) if p
6C = 12
4B + 6C =8
2A + 2B = 6.
Solving these equations successively yields C =2 , B = −1 , and A = 4 . Therefore

2
up = x(4 − x + 2 x )
is a particular solution of Equation 5.4.13), and

3x 3x 2
yp = up e = xe (4 − x + 2 x )
is a particular solution of Equation 5.4.12).
Example 5.4.6

′′ ′ −x/2 2
4y + 4y + y = e (−8 + 48x + 144 x ). (5.4.14)
Solution
Substituting
−x/2 ′ ′ −x/2
1 −x/2 ′′ ′′ −x/2 ′ −x/2
1 −x/2
y = ue , y =u e − ue , and y =u e −u e + ue
2 4
into Equation 5.4.14) and canceling e −x/2

yields
u u
′′ ′ ′ ′′ 2
4 (u −u + ) + 4 (u − ) + u = 4u = −8 + 48x + 144 x ,
4 2
or
′′ 2
u = −2 + 12x + 36 x , (5.4.15)
which does not contain u or u because e ′

and xe are both solutions of the complementary equation. (See
−x/2 −x/2
Exercise 5.4.30.) To obtain a particular solution of Equation 5.4.15) we integrate twice, taking the constants of integration
to be zero; thus,
′ 2 3 2 3 4 2 2
up = −2x + 6 x + 12 x and up = −x + 2x + 3x = x (−1 + 2x + 3 x ).
Therefore
−x/2 2 −x/2 2
yp = up e =x e (−1 + 2x + 3 x )
Summary
The preceding examples illustrate the following facts concerning particular solutions of a constant coefficent equation of the
form
′′ ′ αx
ay + b y + cy = e G(x),
where G is a polynomial (see Exercise 5.4.30):

a. If e αx
′′ ′
ay + b y + cy = 0, (5.4.16)
then y = e Q(x), where Q is a polynomial of the same degree as G. (See Example 5.4.4).
p
αx
b. If e is a solution of Equation 5.4.16 but xe is not, then y = x e Q(x), where Q is a polynomial of the same degree
αx αx
p
αx
as G. (See Example 5.4.5.)

c. If both e and xe are solutions of Equation 5.4.16, then y = x e Q(x), where Q is a polynomial of the same degree
αx αx
p
2 αx
as G. (See Example 5.4.6.)

In all three cases, you can just substitute the appropriate form for y and its derivatives directly into p
′′ ′ αx
ayp + b yp + c yp = e G(x),
and solve for the coefficients of the polynomial Q. However, if you try this you will see that the computations are more
tedious than those that you encounter by making the substitution y = ue and finding a particular solution of the resulting αx
equation for u. (See Exercises 5.4.34-5.4.36.) In Case (a) the equation for u will be of the form
′′ ′ ′
au + p (α)u + p(α)u = G(x),
with a particular solution of the form u = Q(x) , a polynomial of the same degree as G, whose coefficients can be found by
p
the method used in Example 5.4.4. In Case (b) the equation for u will be of the form
′′ ′ ′
au + p (α)u = G(x)

(no u term on the left), with a particular solution of the form u = xQ(x), where Q is a polynomial of the same degree as G
p
whose coefficents can be found by the method used in Example 5.4.5. In Case (c), the equation for u will be of the form
′′
au = G(x)
with a particular solution of the form u = x Q(x) that can be obtained by integrating G(x)/a twice and taking the constants
p
2
of integration to be zero, as in Example 5.4.6.
Using the Principle of Superposition

The next example shows how to combine the method of undetermined coefficients and Theorem 5.3.3, the principle of
superposition.
Example 5.4.7
′′ ′ 2x 4x
y − 7 y + 12y = 4 e + 5e . (5.4.17)
Solution
In Example 5.4.1 we found that y p1 = 2e
2x
′′ ′ 2x
y − 7 y + 12y = 4 e ,
and in Example 5.4.2we found that y p

2
= 5x e
4x
′′ ′ 4x
y − 7 y + 12y = 5 e .
Therefore the principle of superposition implies that y p = 2e

2x
+ 5x e
4x

5.4E: The Method of Undetermined Coefficients I (Exercises)
Q5.4.1
In Exercises 5.4.1-5.4.14 find a particular solution.
1. y ′′
− 3 y + 2y = e
′ 3x
(1 + x)
2. y ′′
− 6 y + 5y = e
′ −3x
(35 − 8x)
3. y ′′ ′
− 2 y − 3y = e (−8 + 3x)
x
4. y ′′
+ 2y + y = e
′ 2x
(−7 − 15x + 9 x )
2
5. y ′′
+ 4y = e
−x
(7 − 4x + 5 x )
2
6. y ′′ ′
− y − 2y = e (9 + 2x − 4 x )
x 2
7. y ′′ ′
− 4 y − 5y = −6x e
−x
8. y ′′ ′
− 3 y + 2y = e (3 − 4x)
x
9. y ′′ ′
+ y − 12y = e
3x
(−6 + 7x)
10. 2y ′′ ′
− 3 y − 2y = e
2x
(−6 + 10x)
11. y ′′
+ 2y + y = e
′ −x
(2 + 3x)
12. y ′′ ′
− 2 y + y = e (1 − 6x)
x
13. y ′′
− 4 y + 4y = e
′ 2x
(1 − 3x + 6 x )
2
14. 9y ′′
+ 6y + y = e
′ −x/3
(2 − 4x + 4 x )
2
Q5.4.2
15. y ′′
− 3 y + 2y = e
′ 3x
(1 + x)
16. y ′′ ′
− 6 y + 8y = e (11 − 6x)
x
17. y ′′
+ 6 y + 9y = e
′ 2x
(3 − 5x)
18. y ′′ ′
+ 2 y − 3y = −16x e
x
19. y ′′ ′
− 2 y + y = e (2 − 12x)
x
Q5.4.3
In Exercises 5.4.20-5.4.23 solve the initial value problem and plot the solution.
20. y ′′ ′
− 4 y − 5y = 9 e
2x
(1 + x), y(0) = 0,
′
y (0) = −10
21. y ′′
+ 3 y − 4y = e
′ 2x
(7 + 6x), y(0) = 2,
′
y (0) = 8
22. y ′′ ′
+ 4 y + 3y = −e
−x
(2 + 8x), y(0) = 1,
′
y (0) = 2
23. y ′′ ′
− 3 y − 10y = 7 e
−2x
, y(0) = 1,
′
y (0) = −17
Q5.4.4
In Exercises 5.4.24-5.4.29 use the principle of superposition to find a particular solution.
24. y ′′ ′
+ y + y = xe
x
+e
−x
(1 + 2x)
25. y ′′ ′
− 7 y + 12y = −e (17 − 42x) − e
x 3x
26. y ′′ ′
− 8 y + 16y = 6x e
4x
+ 2 + 16x + 16 x
2
27. y ′′ ′
− 3 y + 2y = −e
2x
(3 + 4x) − e
x

28. y ′′ ′ x
− 2 y + 2y = e (1 + x) + e
−x
(2 − 8x + 5 x )
2
29. y ′′
+y = e
−x
(2 − 4x + 2 x ) + e
2 3x
(8 − 12x − 10 x )
2
Q5.4.5
30.
a. Prove that y is a solution of the constant coefficient equation
′′ ′ αx
ay + b y + cy = e G(x) (A)
if and only if y = ue αx
, where u satisfies
′′ ′ ′
au + p (α)u + p(α)u = G(x) (B)
and p(r) = ar 2
+ br + c is the characteristic polynomial of the complementary equation
′′ ′
ay + b y + cy = 0.
For the rest of this exercise, let G be a polynomial. Give the requested proofs for the case where
2 3
G(x) = g0 + g1 x + g2 x + g3 x .
b. Prove that if e isn’t a solution of the complementary equation then (B) has a particular solution of the form u = A(x) ,
αx
p
where A is a polynomial of the same degree as G, as in Example 5.4.4. Conclude that (A) has a particular solution of the
form y = e A(x) .
p
αx
c. Show that if e is a solution of the complementary equation and xe isn’t , then (B) has a particular solution of the form
αx αx
u = xA(x) , where A is a polynomial of the same degree as G, as in Example 5.4.5. Conclude that (A) has a particular
p
solution of the form y = x e A(x) . p

αx
d. Show that if e and xe are both solutions of the complementary equation then (B) has a particular solution of the form
αx αx
u = x A(x) , where A is a polynomial of the same degree as G, and x A(x) can be obtained by integrating G/a twice,
2 2
p
taking the constants of integration to be zero, as in Example 5.4.6. Conclude that (A) has a particular solution of the form
A(x) .
2 αx
y =x e
p
Q5.4.6
Exercises 5.4.31–5.4.36 treat the equations considered in Examples 5.4.1–5.4.6. Substitute the suggested form of y into the p
equation and equate the resulting coefficients of like functions on the two sides of the resulting equation to derive a set of
simultaneous equations for the coefficients in y . Then solve for the coefficients to obtain y . Compare the work you’ve done
p p
with the work required to obtain the same results in Examples 5.4.1–5.4.6.
31. Compare with Example 5.4.1:
′′ ′ 2x 2x
y − 7 y + 12y = 4 e ; yp = Ae

′′ ′ 4x 4x
y − 7 y + 12y = 5 e ; yp = Ax e

′′ ′ 4x 2 4x
y − 8 y + 16y = 2 e ; yp = Ax e

′′ ′ 3x 2 3x 2
y − 3 y + 2y = e (−1 + 2x + x ), yp = e (A + Bx + C x )

′′ ′ 3x 2 3x 2 3
y − 4 y + 3y = e (6 + 8x + 12 x ), yp = e (Ax + Bx + Cx )

′′ ′ −x/2 2 −x/2 2 3 4
4y + 4y + y = e (−8 + 48x + 144 x ), yp = e (Ax + Bx + Cx )

Q5.4.7
37. Write y = ue αx
to find the general solution.
−x
a. y
′′
+ 2y + y =
′ e
√x
b. y ′′ ′
+ 6 y + 9y = e
−3x
ln x
2x
c. y − 4y + 4y =
′′ ′ e
1+x
d. 4y + 4y + y = 4e
′′ ′ −x/2
(
1
x
+ x)
38. Suppose α ≠ 0 and k is a positive integer. In most calculus books integrals like ∫ x e
k αx
dx are evaluated by integrating
by parts k times. This exercise presents another method. Let
αx
y =∫ e P (x) dx
with
k
P (x) = p0 + p1 x + ⋯ + pk x
(where p k ≠0 ).
a. Show that y = e αx
u , where
′
u + αu = P (x). (A)
b. Show that (A) has a particular solution of the form

k
up = A0 + A1 x + ⋯ + Ak x ,
where A , k Ak−1 , …, A0 can be computed successively by equating coefficients of x ,x

k k−1
,…,1 on both sides of the
equation
′
up + α up = P (x).
c. Conclude that
αx k αx
∫ e P (x) dx = (A0 + A1 x + ⋯ + Ak x ) e + c,
where c is a constant of integration.

39. Use the method of Exercise 5.4.38 to evaluate the integral.
a. ∫ e x
(4 + x)dx
b. ∫ e −x
(−1 + x )dx
2
c. ∫ x 3
e
−2x
dx
d. ∫ e x
(1 + x ) dx
2
e. ∫ e 3x
(−14 + 30x + 27 x )dx
2
f. ∫ e −x
(1 + 6 x
2
− 14 x
3
+ 3 x )dx
4
40. Use the method suggested in Exercise 5.4.38 to evaluate ∫ x k

e
αx
dx , where k is an arbitrary positive integer and α ≠ 0 .

5.5: The Method of Undetermined Coefficients II
′′ ′ λx
ay + b y + cy = e (P (x) cos ωx + Q(x) sin ωx) (5.5.1)
where λ and ω are real numbers, ω ≠ 0 , and P and Q are polynomials. We want to find a particular solution of Equation 5.5.1. As in Section 5.4,
the procedure that we will use is called the method of undetermined coefficients.
Forcing Functions Without Exponential Factors

We begin with the case where λ = 0 in Equation 5.5.1 ; thus, we we want to find a particular solution of
′′ ′
ay + b y + cy = P (x) cos ωx + Q(x) sin ωx, (5.5.2)
where P and Q are polynomials.

Differentiating xr
cos ωx and x
r
sin ωx yields
d
r r r−1
x cos ωx = −ωx sin ωx + rx cos ωx
dx
and
d
r r r−1
x sin ωx = ωx cos ωx + rx sin ωx.
dx
This implies that if
yp = A(x) cos ωx + B(x) sin ωx
where A and B are polynomials, then

′′ ′
ayp + b yp + c yp = F (x) cos ωx + G(x) sin ωx,
where F and G are polynomials with coefficients that can be expressed in terms of the coefficients of A and B . This suggests that we try to choose
A and B so that F = P and G = Q , respectively. Then y will be a particular solution of Equation 5.5.2. The next theorem tells us how to choose
p
the proper form for y . For the proof see Exercise 5.5.37.
p
Theorem 5.5.1
Suppose ω is a positive number and P and Q are polynomials. Let k be the larger of the degrees of P and Q. Then the equation
′′ ′
ay + b y + cy = P (x) cos ωx + Q(x) sin ωx
has a particular solution
yp = A(x) cos ωx + B(x) sin ωx, (5.5.3)
where
k k
A(x) = A0 + A1 x + ⋯ + Ak x and B(x) = B0 + B1 x + ⋯ + Bk x ,
provided that cos ωx and sin ωx are not solutions of the complementary equation. The solutions of
′′ 2
a(y + ω y) = P (x) cos ωx + Q(x) sin ωx
for which cos ωx and sin ωx are solutions of the complementary equation are of the form of Equation 5.5.3, where
2 k+1 2 k+1
A(x) = A0 x + A1 x + ⋯ + Ak x and B(x) = B0 x + B1 x + ⋯ + Bk x .
For an analog of this theorem that’s applicable to Equation 5.5.1, see Exercise 5.5.38.
Example 5.5.1
′′ ′
y − 2 y + y = 5 cos 2x + 10 sin 2x. (5.5.4)
Solution
In Equation 5.5.4 the coefficients of cos 2x and sin 2x are both zero degree polynomials (constants). Therefore Theorem 5.5.1 implies that
Equation 5.5.4 has a particular solution

yp = A cos 2x + B sin 2x.
Since
′ ′′
yp = −2A sin 2x + 2B cos 2x and yp = −4(A cos 2x + B sin 2x),
replacing y by y in Equation 5.5.4 yields

p
′′ ′
yp − 2 yp + yp = −4(A cos 2x + B sin 2x) − 4(−A sin 2x + B cos 2x)
+(A cos 2x + B sin 2x)
= (−3A − 4B) cos 2x + (4A − 3B) sin 2x.
Equating the coefficients of cos 2x and sin 2x here with the corresponding coefficients on the right side of Equation 5.5.4 shows that yp is a
solution of Equation 5.5.4 if
−3A − 4B = 5
4A − 3B = 10.
Solving these equations yields A = 1 , B = −2 . Therefore
yp = cos 2x − 2 sin 2x
Example 5.5.2
′′
y + 4y = 8 cos 2x + 12 sin 2x. (5.5.5)
Solution
The procedure used in Example 5.5.1 doesn’t work here; substituting y p = A cos 2x + B sin 2x for y in Equation 5.5.5 yields
′′
yp + 4 yp = −4(A cos 2x + B sin 2x) + 4(A cos 2x + B sin 2x) = 0
for any choice of A and B , since cos 2x and sin 2x are both solutions of the complementary equation for Equation 5.5.5 . We’re dealing with
the second case mentioned in Theorem 5.5.1, and should therefore try a particular solution of the form
yp = x(A cos 2x + B sin 2x). (5.5.6)
Then
′
yp = A cos 2x + B sin 2x + 2x(−A sin 2x + B cos 2x)
′′
andyp = −4A sin 2x + 4B cos 2x − 4x(A cos 2x + B sin 2x)
= −4A sin 2x + 4B cos 2x − 4 yp (see (5.5.6)),
so
′′
yp + 4 yp = −4A sin 2x + 4B cos 2x.
Therefore y is a solution of Equation 5.5.5if

p
−4A sin 2x + 4B cos 2x = 8 cos 2x + 12 sin 2x,
which holds if A = −3 and B = 2 . Therefore

yp = −x(3 cos 2x − 2 sin 2x)
Example 5.5.3
′′ ′
y + 3 y + 2y = (16 + 20x) cos x + 10 sin x. (5.5.7)
Solution
The coefficients of cos x and sin x in Equation 5.5.7 are polynomials of degree one and zero, respectively. Therefore Theorem 5.5.1 tells us to
look for a particular solution of Equation 5.5.7 of the form
yp = (A0 + A1 x) cos x + (B0 + B1 x) sin x. (5.5.8)
Then

′
yp = (A1 + B0 + B1 x) cos x + (B1 − A0 − A1 x) sin x (5.5.9)
and
′′
yp = (2 B1 − A0 − A1 x) cos x − (2 A1 + B0 + B1 x) sin x, (5.5.10)
so
′′ ′
yp + 3 yp + 2 yp = [ A0 + 3 A1 + 3 B0 + 2 B1 + (A1 + 3 B1 )x] cos x + [ B0 + 3 B1 − 3 A0 − 2 A1 + (B1 − 3 A1 )x] sin x. (5.5.11)
Comparing the coefficients of x cos x , x sin x , cos x , and sin x here with the corresponding coefficients in Equation 5.5.7 shows that yp is a
solution of Equation 5.5.7if
A1 + 3 B1 = 20
−3 A1 + B1 =0
A0 + 3 B0 + 3 A1 + 2 B1 = 16
−3 A0 + B0 − 2 A1 + 3 B1 = 10.
Solving the first two equations yields A 1 =2 ,B1 =6 . Substituting these into the last two equations yields
A0 + 3 B0 = 16 − 3 A1 − 2 B1 = −2
−3 A0 + B0 = 10 + 2 A1 − 3 B1 = −4.
Solving these equations yields A 0 =1 ,B 0 = −1 . Substituting A 0 =1 ,A 1 =2 ,B0 = −1 ,B 1 =6 into Equation 5.5.8shows that
yp = (1 + 2x) cos x − (1 − 6x) sin x
A Useful Observation
In Equations 5.5.9, 5.5.10, and 5.5.11 the polynomials multiplying sin x can be obtained by replacing A , A , B , and B by B , B , −A , and 0 1 0 1 0 1 0
−A , respectively, in the polynomials mutiplying cos x. An analogous result applies in general, as follows (Exercise 5.5.36).
1
Theorem 5.5.2
If
yp = A(x) cos ωx + B(x) sin ωx,
where A(x) and B(x) are polynomials with coefficients A …, A and B , …, B , then the polynomials multiplying sin ωx in
0 k 0 k
′ ′′ ′′ ′ ′′ 2
yp , yp , ayp + b yp + c yp and yp + ω yp
can be obtained by replacing A0 , …, Ak by B0 , …, Bk and B0 , …, Bk by −A0 , …, −Ak in the corresponding polynomials multiplying
cos ωx.
We will not use this theorem in our examples, but we recommend that you use it to check your manipulations when you work the exercises.
Example 5.5.4
′′
y + y = (8 − 4x) cos x − (8 + 8x) sin x. (5.5.12)
Solution
According to Theorem 5.5.1, we should look for a particular solution of the form
2 2
yp = (A0 x + A1 x ) cos x + (B0 x + B1 x ) sin x, (5.5.13)
since cos x and sin x are solutions of the complementary equation. However, let’s try
yp = (A0 + A1 x) cos x + (B0 + B1 x) sin x (5.5.14)
first, so you can see why it doesn’t work. From Equation 5.5.10,
′′
yp = (2 B1 − A0 − A1 x) cos x − (2 A1 + B0 + B1 x) sin x,
which together with Equation 5.5.14implies that

′′
yp + yp = 2 B1 cos x − 2 A1 sin x.
Since the right side of this equation does not contain x cos x or x sin x , Equation 5.5.14 can’t satisfy Equation 5.5.12 no matter how we
choose A , A , B , and B .
0 1 0 1

Now let y be as in Equation 5.5.13. Then
p
′ 2
yp = [ A0 + (2 A1 + B0 )x + B1 x ] cos x
2
+ [ B0 + (2 B1 − A0 )x − A1 x ] sin x
and
′′ 2
yp = [2 A1 + 2 B0 − (A0 − 4 B1 )x − A1 x ] cos x
2
+ [2 B1 − 2 A0 − (B0 + 4 A1 )x − B1 x ] sin x,
so
′′
yp + yp = (2 A1 + 2 B0 + 4 B1 x) cos x + (2 B1 − 2 A0 − 4 A1 x) sin x.
Comparing the coefficients of cos x and sin x here with the corresponding coefficients in Equation 5.5.12 shows that yp is a solution of
Equation 5.5.12if
4B1 = −4
−4A1 = −8
2 B0 + 2 A1 =8
−2 A0 + 2 B1 = −8.
The solution of this system is A 1 =2 ,B 1 = −1 ,A 0 =3 ,B 0 =2 . Therefore

yp = x [(3 + 2x) cos x + (2 − x) sin x]
Forcing Functions with Exponential Factors

To find a particular solution of
′′ ′ λx
ay + b y + cy = e (P (x) cos ωx + Q(x) sin ωx) (5.5.15)
when λ ≠ 0 , we recall from Section 5.4 that substituting y = ue into Equation 5.5.15 will produce a constant coefficient equation for u with the
λx
forcing function P (x) cos ωx + Q(x) sin ωx. We can find a particular solution u of this equation by the procedure that we used in Examples p
5.5.1-5.5.4. Then y = u e is a particular solution of Equation 5.5.15.

λx
p p
Example 5.5.5
′′ ′ −2x
y − 3 y + 2y = e [2 cos 3x − (34 − 150x) sin 3x] . (5.5.16)
Let y = ue −2x
. Then
′′ ′ −2x ′′ ′ ′
y − 3 y + 2y =e [(u − 4 u + 4u) − 3(u − 2u) + 2u]
−2x ′′ ′
=e (u − 7 u + 12u)
−2x
=e [2 cos 3x − (34 − 150x) sin 3x]
if
′′ ′
u − 7 u + 12u = 2 cos 3x − (34 − 150x) sin 3x. (5.5.17)
Since cos 3x and sin 3x aren’t solutions of the complementary equation

′′ ′
u − 7 u + 12u = 0,
Theorem 5.5.1tells us to look for a particular solution of Equation 5.5.17of the form
up = (A0 + A1 x) cos 3x + (B0 + B1 x) sin 3x. (5.5.18)
Then
′
up = (A1 + 3 B0 + 3 B1 x) cos 3x + (B1 − 3 A0 − 3 A1 x) sin 3x
′′
and up = (−9 A0 + 6 B1 − 9 A1 x) cos 3x − (9 B0 + 6 A1 + 9 B1 x) sin 3x,
so
′′ ′
up − 7 up + 12 up = [3 A0 − 21 B0 − 7 A1 + 6 B1 + (3 A1 − 21 B1 )x] cos 3x
+ [21 A0 + 3 B0 − 6 A1 − 7 B1 + (21 A1 + 3 B1 )x] sin 3x.
Comparing the coefficients of x cos 3x, x sin 3x, cos 3x , and sin 3x here with the corresponding coefficients on the right side of Equation
5.5.17shows that u is a solution of Equation 5.5.17if
p

3 A1 − 21 B1 =0
21 A1 + 3 B1 = 150
(5.5.19)
3 A0 − 21 B0 − 7 A1 + 6 B1 = 2
21 A0 + 3 B0 − 6 A1 − 7 B1 = −34.
Solving the first two equations yields A 1 =7 ,B 1 =1 . Substituting these values into the last two equations of Equation 5.5.19yields
3 A0 − 21 B0 = 2 + 7 A1 − 6 B1 = 45
21 A0 + 3 B0 = −34 + 6 A1 + 7 B1 = 15.
Solving this system yields A 0 =1 ,B 0 = −2 . Substituting A 0 =1 ,A 1 =7 ,B0 = −2 , and B 1 =1 into Equation 5.5.18shows that
up = (1 + 7x) cos 3x − (2 − x) sin 3x
is a particular solution of Equation 5.5.17. Therefore

−2x
yp = e [(1 + 7x) cos 3x − (2 − x) sin 3x]
Example 5.5.6
′′ ′ −x
y + 2 y + 5y = e [(6 − 16x) cos 2x − (8 + 8x) sin 2x] . (5.5.20)
Solution
Let y = ue −x
. Then
′′ ′ −x ′′ ′ ′
y + 2 y + 5y =e [(u − 2 u + u) + 2(u − u) + 5u]
−x ′′
=e (u + 4u)
−x
=e [(6 − 16x) cos 2x − (8 + 8x) sin 2x]
if
′′
u + 4u = (6 − 16x) cos 2x − (8 + 8x) sin 2x. (5.5.21)
Since cos 2x and sin 2x are solutions of the complementary equation

′′
u + 4u = 0,
Theorem 5.5.1tells us to look for a particular solution of Equation 5.5.21of the form
2 2
up = (A0 x + A1 x ) cos 2x + (B0 x + B1 x ) sin 2x.
Then
′ 2
up = [ A0 + (2 A1 + 2 B0 )x + 2 B1 x ] cos 2x
2
+ [ B0 + (2 B1 − 2 A0 )x − 2 A1 x ] sin 2x
and
′′ 2
up = [2 A1 + 4 B0 − (4 A0 − 8 B1 )x − 4 A1 x ] cos 2x
2
+ [2 B1 − 4 A0 − (4 B0 + 8 A1 )x − 4 B1 x ] sin 2x,
so
′′
up + 4 up = (2 A1 + 4 B0 + 8 B1 x) cos 2x + (2 B1 − 4 A0 − 8 A1 x) sin 2x.
Equating the coefficients of x cos 2x, x sin 2x, cos 2x , and sin 2x here with the corresponding coefficients on the right side of Equation 5.5.21
shows that u is a solution of Equation 5.5.21if
p
8B1 = −16
−8A1 =− 8
(5.5.22)
4 B0 + 2 A1 =6
−4 A0 + 2 B1 = −8.
The solution of this system is A 1 =1 ,B 1 = −2 ,B 0 =1 ,A 0 =1 . Therefore

up = x[(1 + x) cos 2x + (1 − 2x) sin 2x]
is a particular solution of Equation 5.5.21, and

−x
yp = x e [(1 + x) cos 2x + (1 − 2x) sin 2x]

You can also find a particular solution of Equation 5.5.20 by substituting
−x
yp = x e [(A0 + A1 x) cos 2x + (B0 + B1 x) sin 2x]
for y in Equation 5.5.20and equating the coefficients of x e −x

cos 2x x e , −x
,
sin 2x e
−x
, and e
cos 2x
−x
sin 2x in the resulting expression for
′′ ′
yp + 2 yp + 5 yp
with the corresponding coefficients on the right side of Equation 5.5.20. (See Exercise 5.5.38). This leads to the same system Equation 5.5.22
of equations for A , A , B , and B that we obtained in Example 5.5.6. However, if you try this approach you’ll see that deriving Equation
0 1 0 1
5.5.22 this way is much more tedious than the way we did it in Example 5.5.6.

5.5E: The Method of Undetermined Coefficients II (Exercises)
Q5.5.1
1. y ′′ ′
+ 3 y + 2y = 7 cos x − sin x
2. y ′′ ′
+ 3 y + y = (2 − 6x) cos x − 9 sin x
3. y ′′ ′ x
+ 2 y + y = e (6 cos x + 17 sin x)
4. y ′′ ′
+ 3 y − 2y = −e
2x
(5 cos 2x + 9 sin 2x)
5. y ′′ ′ x
− y + y = e (2 + x) sin x
6. y ′′ ′
+ 3 y − 2y = e
−2x
[(4 + 20x) cos 3x + (26 − 32x) sin 3x]
7. y ′′
+ 4y = −12 cos 2x − 4 sin 2x
8. y ′′
+ y = (−4 + 8x) cos x + (8 − 4x) sin x
9. 4y ′′
+ y = −4 cos x/2 − 8x sin x/2
10. y ′′ ′
+ 2 y + 2y = e
−x
(8 cos x − 6 sin x)
11. y ′′ ′
− 2 y + 5y = e
x
[(6 + 8x) cos 2x + (6 − 8x) sin 2x]
12. y ′′ ′
+ 2y + y = 8x
2
cos x − 4x sin x
13. y ′′ ′
+ 3 y + 2y = (12 + 20x + 10 x ) cos x + 8x sin x
2
14. y ′′ ′ 2
+ 3 y + 2y = (1 − x − 4 x ) cos 2x − (1 + 7x + 2 x ) sin 2x
2
15. y ′′ ′
− 5 y + 6y = −e
x 2
[(4 + 6x − x ) cos x − (2 − 4x + 3 x ) sin x]
2
16. y ′′ ′
− 2 y + y = −e
x 2
[(3 + 4x − x ) cos x + (3 − 4x − x ) sin x]
2
17. y ′′ ′
− 2 y + 2y = e
x 2
[(2 − 2x − 6 x ) cos x + (2 − 10x + 6 x ) sin x]
2
Q5.5.2
In Exercises 5.5.18-5.5.21 find a particular solution and graph it.
18. y ′′ ′
+ 2y + y = e
−x
[(5 − 2x) cos x − (3 + 3x) sin x]
19. y ′′
+ 9y = −6 cos 3x − 12 sin 3x
20. y ′′ ′ 2
+ 3 y + 2y = (1 − x − 4 x ) cos 2x − (1 + 7x + 2 x ) sin 2x
2
21. y ′′ ′
+ 4 y + 3y = e
−x 2
[(2 + x + x ) cos x + (5 + 4x + 2 x ) sin x]
2
Q5.5.3
22. y ′′ ′ x
− 7 y + 6y = −e (17 cos x − 7 sin x),
′
y(0) = 4, y (0) = 2
23. y ′′ ′ x
− 2 y + 2y = −e (6 cos x + 4 sin x),
′
y(0) = 1, y (0) = 4
24. y ′′ ′
+ 6 y + 10y = −40 e
x
sin x, y(0) = 2,
′
y (0) = −3
25. y ′′ ′
− 6 y + 10y = −e
3x
(6 cos x + 4 sin x), y(0) = 2,
′
y (0) = 7
26. y ′′ ′
− 3 y + 2y = e
3x
[21 cos x − (11 + 10x) sin x] , y(0) = 0,
′
y (0) = 6
Q5.5.4
In Exercises 5.5.27-5.5.32 use the principle of superposition to find a particular solution. Where indicated, solve the initial
value problem.

27. y ′′ ′
− 2 y − 3y = 4 e
3x x
+ e (cos x − 2 sin x)
28. y ′′
+ y = 4 cos x − 2 sin x + x e
x
+e
−x
29. y ′′ ′
− 3 y + 2y = x e
x
+ 2e
2x
+ sin x
30. y ′′ ′
− 2 y + 2y = 4x e
x
cos x + x e
−x
+1 +x
2
31. y ′′ ′
− 4 y + 4y = e
2x
(1 + x) + e
2x
(cos x − sin x) + 3 e
3x
+1 +x
32. y ′′ ′
− 4 y + 4y = 6 e
2x
+ 25 sin x, y(0) = 5, y (0) = 3
′
Q5.5.5
33. y ′′
+ 4y = −e
−2x
[(4 − 7x) cos x + (2 − 4x) sin x] , y(0) = 3,
′
y (0) = 1
34. y ′′ ′
+ 4 y + 4y = 2 cos 2x + 3 sin 2x + e
−x
, y(0) = −1, y (0) = 2
′
35. y ′′ x
+ 4y = e (11 + 15x) + 8 cos 2x − 12 sin 2x, y(0) = 3, y (0) = 5
′
Q5.5.6
36.
a. Verify that if
yp = A(x) cos ωx + B(x) sin ωx (5.5E.1)
where A and B are twice differentiable, then

′ ′ ′
yp = (A + ωB) cos ωx + (B − ωA) sin ωx and
′′ ′′ ′ 2 ′′ ′ 2
yp = (A + 2ωB − ω A) cos ωx + (B − 2ωA − ω B) sin ωx.
b. Use the results of (a) to verify that

′′ ′ 2 ′ ′ ′′
ayp + b yp + c yp = [(c − aω )A + bωB + 2aωB + b A + aA ] cos ωx+
2 ′ ′ ′′
[−bωA + (c − aω )B − 2aωA + b B + aB ] sin ωx.
c. Use the results of (a) to verify that

′′ 2 ′′ ′ ′′ ′
yp + ω yp = (A + 2ωB ) cos ωx + (B − 2ωA ) sin ωx. (5.5E.2)
d. Prove Theorem 5.5.2.

37. Let a , b , c , and ω be constants, with a ≠ 0 and ω > 0 , and let
k k
P (x) = p0 + p1 x + ⋯ + pk x and Q(x) = q0 + q1 x + ⋯ + qk x , (5.5E.3)
where at least one of the coefficients p , q is nonzero, so k is the larger of the degrees of P and Q.
k k
a. Show that if cos ωx and sin ωx are not solutions of the complementary equation
′′ ′
ay + b y + cy = 0, (5.5E.4)
then there are polynomials

k k
A(x) = A0 + A1 x + ⋯ + Ak x and B(x) = B0 + B1 x + ⋯ + Bk x (A)
such that
2 ′ ′ ′′
(c − aω )A + bωB + 2aωB + b A + aA =P
(5.5E.5)
2 ′ ′ ′′
−bωA + (c − aω )B − 2aωA + b B + aB = Q,
where (A k, ,
Bk ) (Ak−1 , Bk−1 ) , …,(A 0, B0 ) can be computed successively by solving the systems

2
(c − aω )Ak + bωBk = pk
(5.5E.6)
2
−bωAk + (c − aω )Bk = qk ,
and, if 1 ≤ r ≤ k ,
2
(c − aω )Ak−r + bωBk−r = pk−r + ⋯
(5.5E.7)
2
−bωAk−r + (c − aω )Bk−r = qk−r + ⋯ ,
where the terms indicated by “⋯” depend upon the previously computed coefficients with subscripts greater than k−r .
Conclude from this and Exercise 5.5.36b that
yp = A(x) cos ωx + B(x) sin ωx (B)
′′ ′
ay + b y + cy = P (x) cos ωx + Q(x) sin ωx. (5.5E.8)
b. Conclude from Exercise 5.5.36c that the equation

′′ 2
a(y + ω y) = P (x) cos ωx + Q(x) sin ωx (C)
does not have a solution of the form (B) with A and B as in (A). Then show that there are polynomials
2 k+1 2 k+1
A(x) = A0 x + A1 x + ⋯ + Ak x and B(x) = B0 x + B1 x + ⋯ + Bk x (5.5E.9)
such that
′′ ′
a(A + 2ωB ) =P
(5.5E.10)
′′ ′
a(B − 2ωA ) = Q,
where the pairs (A k, ,

Bk ) (Ak−1 , Bk−1 ) , …, (A 0, B0 ) can be computed successively as follows:
qk
Ak = −
2aω(k + 1)
pk
Bk = ,
2aω(k + 1)
and, if k ≥ 1 ,
1 qk−j
Ak−j =− [ − (k − j + 2)Bk−j+1 ]
2ω a(k − j + 1)
1 pk−j
Bk−j = [ − (k − j + 2)Ak−j+1 ]
2ω a(k − j + 1)
for 1 ≤ j ≤ k . Conclude that (B) with this choice of the polynomials A and B is a particular solution of (C).
38. Show that Theorem 5.5.1 implies the next theorem:
Theorem 5.5E . 1
Suppose ω is a positive number and P and Q are polynomials. Let k be the larger of the degrees of P and . Then the
Q
equation
′′ ′ λx
ay + b y + cy = e (P (x) cos ωx + Q(x) sin ωx) (5.5E.11)
has a particular solution

λx
yp = e (A(x) cos ωx + B(x) sin ωx) , (A)
where
k k
A(x) = A0 + A1 x + ⋯ + Ak x and B(x) = B0 + B1 x + ⋯ + Bk x , (5.5E.12)
provided that e λx
cos ωx and e λx
sin ωx are not solutions of the complementary equation. The equation

′′ ′ 2 2 λx
a [y − 2λ y + (λ + ω )y] = e (P (x) cos ωx + Q(x) sin ωx) (5.5E.13)
(for which e λx
cos ωx and e λx
sin ωx are solutions of the complementary equation) has a particular solution of the form
(A), where
2 k+1 2 k+1
A(x) = A0 x + A1 x + ⋯ + Ak x and B(x) = B0 x + B1 x + ⋯ + Bk x . (5.5E.14)
39. This exercise presents a method for evaluating the integral
λx
y =∫ e (P (x) cos ωx + Q(x) sin ωx) dx (5.5E.15)
where ω ≠ 0 and
k k
P (x) = p0 + p1 x + ⋯ + pk x , Q(x) = q0 + q1 x + ⋯ + qk x . (5.5E.16)
a. Show that y = e λx
u , where
′
u + λu = P (x) cos ωx + Q(x) sin ωx. (A)
b. Show that (A) has a particular solution of the form

up = A(x) cos ωx + B(x) sin ωx, (5.5E.17)
where
k k
A(x) = A0 + A1 x + ⋯ + Ak x , B(x) = B0 + B1 x + ⋯ + Bk x , (5.5E.18)
and the pairs of coefficients (A , B ), (A , B ), …,(A , B ) can be computed successively as the solutions of pairs
k k k−1 k−1 0 0
of equations obtained by equating the coefficients of x cos ωx and x sin ωx for r = k , k − 1 , …, 0.

r r
c. Conclude that
λx λx
∫ e (P (x) cos ωx + Q(x) sin ωx) dx = e (A(x) cos ωx + B(x) sin ωx) + c, (5.5E.19)
where c is a constant of integration.

40. Use the method of Exercise 5.5.39 to evaluate the integral.
a. ∫ x cos xdx
2
b. ∫ x e cos xdx
2 x
c. ∫ x e sin 2xdx
−x
d. ∫ x e sin xdx
2 −x
e. ∫ x e sin xdx
3 x
f. ∫ e [x cos x − (1 + 3x) sin x]dx

x
g. ∫ e [(1 + x ) cos x + (1 ) sin x]dx

−x 2 2
x

5.6: Reduction of Order
In this section we give a method for finding the general solution of
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F (x) (5.6.1)
if we know a nontrivial solution y of the complementary equation

1
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0. (5.6.2)
The method is called reduction of order because it reduces the task of solving Equation 5.6.1 to solving a first order equation.
Unlike the method of undetermined coefficients, it does not require P , P , and P to be constants, or F to be of any special 0 1 2
form.
By now you shoudn’t be surprised that we look for solutions of Equation 5.6.1 in the form
y = uy1 (5.6.3)
where u is to be determined so that y satisfies Equation 5.6.1. Substituting Equation 5.6.3 and
′ ′ ′
y = u y1 + u y
1
′′ ′′ ′ ′ ′′
y = u y1 + 2 u y + uy
1 1
into Equation 5.6.1 yields

′′ ′ ′ ′′ ′ ′
P0 (x)(u y1 + 2 u y + uy ) + P1 (x)(u y1 + u y ) + P2 (x)u y1 = F (x).
1 1 1
Collecting the coefficients of u, u , and u yields′ ′′
′′ ′ ′ ′′ ′
(P0 y1 )u + (2 P0 y + P1 y1 )u + (P0 y + P1 y + P2 y1 )u = F . (5.6.4)
1 1 1
However, the coefficient of u is zero, since y satisfies Equation 5.6.2. Therefore Equation 5.6.4 reduces to
1
′′ ′
Q0 (x)u + Q1 (x)u = F , (5.6.5)
with
′
Q0 = P0 y1 and Q1 = 2 P0 y + P1 y1 .
1
(It isn’t worthwhile to memorize the formulas for Q and Q !) Since Equation 5.6.5 is a linear first order equation in u , we
0 1
′
can solve it for u by variation of parameters as in Section 1.2, integrate the solution to obtain u, and then obtain y from
′
Equation 5.6.3.
Example 5.6.1
′′ ′ 2
xy − (2x + 1)y + (x + 1)y = x , (5.6.6)
given that y 1 =e
x
is a solution of the complementary equation
′′ ′
xy − (2x + 1)y + (x + 1)y = 0. (5.6.7)
b. As a byproduct of (a), find a fundamental set of solutions of Equation 5.6.7.

Solution
a. If y = ue , then y
x ′ ′
=u e
x
+ ue
x
and y ′′
=u e
′′ x
+ 2u e
′ x
+ ue
x
, so
′′ ′ ′′ x ′ x x ′ x x x
xy − (2x + 1)y + (x + 1)y = x(u e + 2u e + u e ) − (2x + 1)(u e + u e ) + (x + 1)u e
′′ ′ x
= (x u − u )e .
Therefore y = ue is a solution of Equation 5.6.6 if and only if

x

′′ ′ x 2
(x u − u )e =x ,
which is a first order equation in u . We rewrite it as

′
′
u
′′ −x
u − = xe . (5.6.8)
x
To focus on how we apply variation of parameters to this equation, we temporarily write z =u

′
, so that Equation 5.6.8
becomes
′
z −x
z − = xe . (5.6.9)
x
We leave it to you to show (by separation of variables) that z 1 =x is a solution of the complementary equation
′
z
z − =0
x
for Equation 5.6.9. By applying variation of parameters as in Section 1.2, we can now see that every solution of Equation
5.6.9 is of the form
′ −x ′ −x −x
z = vx where v x = xe , so v =e and v = −e + C1 .
Since u ′
= z = vx , u is a solution of Equation 5.6.8 if and only if
′ −x
u = vx = −x e + C1 x.
−x
C1 2
u = (x + 1)e + x + C2 .
2
Therefore the general solution of Equation 5.6.6 is
x
C1 2 x x
y = ue = x +1 + x e + C2 e . (5.6.10)
2
b. By letting C = C = 0 in Equation 5.6.10, we see that y = x + 1 is a solution of Equation 5.6.6. By letting

1 2 p
1
C = 2 and C = 0 , we see that y is also a solution of Equation 5.6.6. Since the difference of two
2 x
1 2 = x +1 +x e p
2
solutions of Equation 5.6.6 is a solution of Equation 5.6.7, y = y − y = x e is a solution of Equation 5.6.7. Since
2 p
1
p
2
2 x
y /y
2 is nonconstant and we already know that y = e is a solution of Equation 5.6.6, Theorem 5.1.6 implies that
1 1
x
{ e , x e } is a fundamental set of solutions of Equation 5.6.7.

x 2 x
Although Equation 5.6.10 is a correct form for the general solution of Equation 5.6.6, it is silly to leave the arbitrary
coefficient of x e as C /2 where C is an arbitrary constant. Moreover, it is sensible to make the subscripts of the
2 x
1 1
coefficients of y = e and y = x e consistent with the subscripts of the functions themselves. Therefore we rewrite
1
x
2
2 x
Equation 5.6.10 as
x 2 x
y = x + 1 + c1 e + c2 x e
by simply renaming the arbitrary constants. We’ll also do this in the next two examples, and in the answers to the
exercises.
Example 5.6.2
2 ′′ ′ 2
x y + xy − y = x + 1,
given that y 1 =x is a solution of the complementary equation

2 ′′ ′
x y + x y − y = 0. (5.6.11)
As a byproduct of this result, find a fundamental set of solutions of Equation 5.6.11.

2 ′′ ′ 2 ′
x y + xy − y = x + 1, y(1) = 2, y (1) = −3. (5.6.12)
Solution
a. If y = ux, then y ′ ′
= u x +u and y ′′ ′′
= u x + 2u
′
, so
2 ′′ ′ 2 ′′ ′ ′
x y + xy − y = x (u x + 2 u ) + x(u x + u) − ux
3 ′′ 2 ′
=x u + 3x u .
Therefore y = ux is a solution of Equation 5.6.12 if and only if

3 ′′ 2 ′ 2
x u + 3x u = x + 1,
which is a first order equation in u . We rewrite it as

′
3 1 1
′′ ′
u + u = + . (5.6.13)
3
x x x
To focus on how we apply variation of parameters to this equation, we temporarily write z = u , so that Equation 5.6.13 ′
becomes
3 1 1
′
z + z = + . (5.6.14)
3
x x x
We leave it to you to show by separation of variables that z 1 = 1/ x

3
is a solution of the complementary equation
′
3
z + z =0
x
for Equation 5.6.14. By variation of parameters, every solution of Equation 5.6.14 is of the form
′ 3
v v 1 1 ′ 2
x
z = where = + , so v =x +1 and v= + x + C1 .
x3 x3 x x3 3
Since u ′ 3
= z = v/ x , u is a solution of Equation 5.6.14 if and only if
v 1 1 C1
′
u = = + + .
3 2 3
x 3 x x

x 1 C1
u = − − + C2 .
2
3 x 2x

2
x C1
y = ux = −1 − + C2 x. (5.6.15)
3 2x
Reasoning as in the solution of Example 5.6.1a , we conclude that y1 = x and y2 = 1/x form a fundamental set of
solutions for Equation 5.6.11.
As we explained above, we rename the constants in Equation 5.6.15 and rewrite it as
2
x c2
y = − 1 + c1 x + . (5.6.16)
3 x
b. Differentiating Equation 5.6.16 yields

2x c2
′
y = + c1 − . (5.6.17)
2
3 x
Setting x =1 in Equation 5.6.16 and Equation 5.6.17 and imposing the initial conditions y(1) = 2 and ′
y (1) = −3
yields

8
c1 + c2 =
3
11
c1 − c2 = − .
3
Solving these equations yields c 1 = −1/2 ,c 2 = 19/6 . Therefore the solution of Equation 5.6.12is
2
x x 19
y = −1 − + .
3 2 6x
Using reduction of order to find the general solution of a homogeneous linear second order equation leads to a homogeneous
linear first order equation in u that can be solved by separation of variables. The next example illustrates this.
′
Example 5.6.3
Find the general solution and a fundamental set of solutions of
2 ′′ ′
x y − 3x y + 3y = 0, (5.6.18)
given that y 1 =x is a solution.

Solution
If y = ux then y ′ ′
= u x +u and y ′′ ′′
= u x + 2u
′
, so
2 ′′ ′ 2 ′′ ′ ′
x y − 3x y + 3y = x (u x + 2 u ) − 3x(u x + u) + 3ux
3 ′′ 2 ′
=x u −x u .
Therefore y = ux is a solution of Equation 5.6.18if and only if

3 ′′ 2 ′
x u − x u = 0.
Separating the variables u and x yields

′
′′
u 1
= ,
′
u x
so
′ ′
ln | u | = ln |x| + k, or equivalently u = C1 x.
Therefore
C1
2
u = x + C2 ,
2
so the general solution of Equation 5.6.18is

C1
3
y = ux = x + C2 x,
2
which we rewrite as
3
y = c1 x + c2 x .
Therefore {x, x 3
} is a fundamental set of solutions of Equation 5.6.18.

5.6E: Reduction of Order (Exercises)
Q5.6.1
In Exercises 5.6.1-5.6.17 find the general solution, given that y satisfies the complementary equation. As a byproduct, find a 1
fundamental set of solutions of the complementary equation.

1. (2x + 1)y ′′ ′
− 2 y − (2x + 3)y = (2x + 1 ) ;
2
y1 = e
−x
2. x 2
y
′′
+ xy − y =
′ 4
x2
; y1 = x
3. x 2
y
′′
− x y + y = x;
′
y1 = x
4. y ′′
− 3 y + 2y =
′
1+e
1
−x
; y1 = e
2x
5. y ′′
− 2y + y = 7x
′ 3/2
e ;
x
y1 = e
x
6. 4x 2
y
′′
+ (4x − 8 x )y + (4 x
2 ′ 2
− 4x − 1)y = 4 x
1/2 x
e (1 + 4x); y1 = x
1/2
e
x
7. y ′′
− 2 y + 2y = e
′ x
sec x; y1 = e
x
cos x
8. y ′′
+ 4x y + (4 x
′ 2
+ 2)y = 8 e
−x(x+2)
; y1 = e
−x
9. x 2
y
′′ ′
+ x y − 4y = −6x − 4; y1 = x
2
10. x 2
y
′′
+ 2x(x − 1)y + (x
′ 2
− 2x + 2)y = x e
3 2x
; y1 = x e
−x
11. x 2
y
′′
− x(2x − 1)y + (x
′ 2
− x − 1)y = x e ;
2 x
y1 = x e
x
12. (1 − 2x)y ′′ ′
+ 2 y + (2x − 3)y = (1 − 4x + 4 x )e ;
2 x
y1 = e
x
13. x 2
y
′′
− 3x y + 4y = 4 x ;
′ 4
y1 = x
2
14. 2x y ′′
+ (4x + 1)y + (2x + 1)y = 3 x
′ 1/2
e
−x
; y1 = e
−x
15. x y ′′
− (2x + 1)y + (x + 1)y = −e ;
′ x
y1 = e
x
16. 4x 2
y
′′
− 4x(x + 1)y + (2x + 3)y = 4 x
′ 5/2
e
2x
; y1 = x
1/2
17. x 2
y
′′
− 5x y + 8y = 4 x ;
′ 2
y1 = x
2
Q5.6.2
In Exercises 5.6.18-5.6.30 find a fundamental set of solutions, given that y is a solution. 1
18. x y ′′
+ (2 − 2x)y + (x − 2)y = 0;
′
y1 = e
x
19. x 2
y
′′
− 4x y + 6y = 0;
′
y1 = x
2
20. x 2
(ln |x| ) y
2 ′′
− (2x ln |x|)y + (2 + ln |x|)y = 0;
′
y1 = ln |x|
−
21. 4x y ′′
+ 2 y + y = 0;
′
y1 = sin √x
22. x y ′′
− (2x + 2)y + (x + 2)y = 0;
′
y1 = e
x
23. x 2
y
′′
− (2a − 1)x y + a y = 0;
′ 2
y1 = x
a
24. x 2
y
′′
− 2x y + (x
′ 2
+ 2)y = 0; y1 = x sin x
25. x y ′′
− (4x + 1)y + (4x + 2)y = 0;
′
y1 = e
2x
26. 4x 2
(sin x)y
′′
− 4x(x cos x + sin x)y + (2x cos x + 3 sin x)y = 0;
′
y1 = x
1/2
27. 4x 2
y
′′ ′
− 4x y + (3 − 16 x )y = 0;
2
y1 = x
1/2
e
2x
28. (2x + 1)x y ′′

− 2(2 x
2 ′
− 1)y − 4(x + 1)y = 0; y1 = 1/x
29. (x 2
− 2x)y
′′
+ (2 − x )y + (2x − 2)y = 0;
2 ′
y1 = e
x
30. x y ′′
− (4x + 1)y + (4x + 2)y = 0;
′
y1 = e
2x

Q5.6.3
In Exercises 5.6.31-5.6.33 solve the initial value problem, given that y satisfies the complementary equation. 1
31. x2
y
′′
− 3x y + 4y = 4 x ,
′ 4
y(−1) = 7,
′
y (−1) = −8; y1 = x
2
32. (3x − 1)y ′′ ′

− (3x + 2)y − (6x − 8)y = 0, y(0) = 2, y (0) = 3;
′
y1 = e
2x
33. (x + 1) 2
y
′′
− 2(x + 1)y − (x
′ 2
+ 2x − 1)y = (x + 1 ) e ,
3 x
y(0) = 1,
′
y (0) = − 1; y1 = (x + 1)e
x
Q5.6.4
In Exercises 5.6.34 and 5.6.35 solve the initial value problem and graph the solution, given that y satisfies the complementary 1
equation.
34. x2
y
′′
+ 2x y − 2y = x ,
′ 2
y(1) =
5
4
′
, y (1) =
3
2
; y1 = x
35. (x 2
− 4)y
′′ ′
+ 4x y + 2y = x + 2, y(0) = −
1
3
,
′
y (0) = −1; y1 =
1
x−2
Q5.6.5
36. Suppose p and p are continuous on (a, b). Let y be a solution of
1 2 1
′′ ′
y + p1 (x)y + p2 (x)y = 0 (A)
that has no zeros on (a, b), and let x be in (a, b). Use reduction of order to show that y and
0 1
x t
1
y2 (x) = y1 (x) ∫ exp(− ∫ p1 (s) ds) dt (5.6E.1)
2
x0 y (t) x0
1
form a fundamental set of solutions of (A) on (a, b).

37. The nonlinear first order equation
′ 2
y +y + p(x)y + q(x) = 0 (A)
is a Riccati equation. (See Exercise 2.4.55.) Assume that p and q are continuous.
a. Show that y is a solution of (A) if and only if y = z ′
/z , where
′′ ′
z + p(x)z + q(x)z = 0. (B)
b. Show that the general solution of (A) is

′ ′
c1 z + c2 z
1 2
y = , (C)
c1 z1 + c2 z2
where {z , z } is a fundamental set of solutions of (B) and c and c are arbitrary constants.
1 2 1 2
c. Does the formula (C) imply that the first order equation (A) has a two–parameter family of solutions? Explain your answer.
38. Use a method suggested by Exercise 5.6.37 to find all solutions. of the equation.
a. y + y + k = 0
′ 2 2
b. y + y − 3y + 2 = 0
′ 2
c. y + y + 5y − 6 = 0
′ 2
d. y + y + 8y + 7 = 0
′ 2
e. y + y + 14y + 50 = 0
′ 2
f. 6y + 6y − y − 1 = 0
′ 2
g. 36y + 36y − 12y + 1 = 0

′ 2
39. Use a method suggested by Exercise 5.6.37 and reduction of order to find all solutions of the equation, given that y1 is a
solution.
a. x (y + y ) − x(x + 2)y + x + 2 = 0; y = 1/x
2 ′ 2
1
b. y + y + 4xy + 4x + 2 = 0; y = −2x
′ 2 2
1
c. (2x + 1)(y + y ) − 2y − (2x + 3) = 0; y = −1

′ 2
1

d. (3x − 1)(y + y ) − (3x + 2)y − 6x + 8 = 0; y = 2
′ 2
1
e. x (y + y ) + xy + x − = 0; y = − tan x −
2 ′ 2 2 1
4
1
1
2x
f. x (y + y ) − 7xy + 7 = 0; y = 1/x
2 ′ 2
1
40. The nonlinear first order equation

′ 2
y + r(x)y + p(x)y + q(x) = 0 (A)
is the generalized Riccati equation. (See Exercise 2.4.55.) Assume that p and q are continuous and r is differentiable.
a. Show that y is a solution of (A) if and only if y = z ′
/rz , where
′
r (x)
′′ ′
z + [p(x) − ] z + r(x)q(x)z = 0. (B)
r(x)
b. Show that the general solution of (A) is

′ ′
c1 z + c2 z
1 2
y = , (5.6E.2)
r(c1 z1 + c2 z2 )
where {z 1, z2 } is a fundamental set of solutions of (B) and c and c are arbitrary constants.
1 2

5.7: Variation of Parameters
In this section we give a method called variation of parameters for finding a particular solution of
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F (x) (5.7.1)
if we know a fundamental set {y 1, y2 } of solutions of the complementary equation

′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0. (5.7.2)
Having found a particular solution y by this method, we can write the general solution of Equation 5.7.1 as
p
y = yp + c1 y1 + c2 y2 .
Since we need only one nontrivial solution of Equation 5.7.2 to find the general solution of Equation 5.7.1 by reduction of
order, it is natural to ask why we are interested in variation of parameters, which requires two linearly independent solutions of
Equation 5.7.2 to achieve the same goal. Here’s the answer:
If we already know two linearly independent solutions of Equation 5.7.2 then variation of parameters will probably be
simpler than reduction of order.
Variation of parameters generalizes naturally to a method for finding particular solutions of higher order linear equations
(Section 9.4) and linear systems of equations (Section 10.7), while reduction of order doesn’t.
Variation of parameters is a powerful theoretical tool used by researchers in differential equations. Although a detailed
discussion of this is beyond the scope of this book, you can get an idea of what it means from Exercises 5.7.37–5.7.39.
We’ll now derive the method. As usual, we consider solutions of Equation 5.7.1 and Equation 5.7.2 on an interval (a, b)
where P , P , P , and F are continuous and P has no zeros. Suppose that {y , y } is a fundamental set of solutions of the
0 1 2 0 1 2
complementary equation Equation 5.7.2. We look for a particular solution of Equation 5.7.1 in the form
yp = u1 y1 + u2 y2 (5.7.3)
where u and u are functions to be determined so that y satisfies Equation 5.7.1. You may not think this is a good idea,
1 2 p
since there are now two unknown functions to be determined, rather than one. However, since u and u have to satisfy only 1 2
one condition (that y is a solution of Equation 5.7.1), we can impose a second condition that produces a convenient
p
simplification, as follows.
′ ′ ′ ′ ′
yp = u1 y + u2 y + u y1 + u y2 . (5.7.4)
1 2 1 2
As our second condition on u and u we require that

1 2
′ ′
u y1 + u y2 = 0. (5.7.5)
1 2
Then Equation 5.7.4 becomes

′ ′ ′
yp = u1 y + u2 y ; (5.7.6)
1 2
that is, Equation 5.7.5 permits us to differentiate y (once!) as if u and u are constants. Differentiating Equation 5.7.4 yields
p 1 2
′′ ′′ ′′ ′ ′ ′ ′
yp = u1 y + u2 y +u y +u y . (5.7.7)
1 2 1 1 2 2
(There are no terms involving u and u here, as there would be if we hadn’t required Equation 5.7.5.) Substituting Equation
′′
1
′′
2
5.7.3, Equation 5.7.6, and Equation 5.7.7 into Equation 5.7.1 and collecting the coefficients of u and u yields 1 2
′′ ′ ′′ ′ ′ ′ ′ ′
u1 (P0 y + P1 y + P2 y1 ) + u2 (P0 y + P1 y + P2 y2 ) + P0 (u y +u y ) = F.
1 1 2 2 1 1 2 2
As in the derivation of the method of reduction of order, the coefficients of u and u here are both zero because 1 2 y1 and y2
satisfy the complementary equation. Hence, we can rewrite the last equation as
′ ′ ′ ′
P0 (u y +u y ) = F. (5.7.8)
1 1 2 2

Therefore y in Equation 5.7.3 satisfies Equation 5.7.1 if
p
′ ′
u y1 + u y2 =0
1 2
(5.7.9)
′ ′ ′ ′ F
u y +u y = ,
1 1 2 2 P0
where the first equation is the same as Equation 5.7.5 and the second is from Equation 5.7.8.
We’ll now show that you can always solve Equation 5.7.9 for u and u . (The method that we use here will always work, but
′
1
′
2
simpler methods usually work when you’re dealing with specific equations.) To obtain u , multiply the first equation in ′
1
Equation 5.7.9 by y and the second equation by y . This yields

′
2 2
′ ′ ′ ′
u y1 y + u y2 y =0
1 2 2 2
F y2
′ ′ ′ ′
u y y2 + u y y2 = .
1 1 2 2
P0
Subtracting the second equation from the first yields

F y2
′ ′ ′
u (y1 y − y y2 ) = − . (5.7.10)
1 2 1
P0
Since { y1 , y2 } is a fundamental set of solutions of Equation 5.7.2 on (a, b), Theorem 5.1.6 implies that the Wronskian
y1 y
2
′ ′
− y y2
1
has no zeros on (a, b). Therefore we can solve Equation 5.7.10 for u , to obtain ′
1
F y2
′
u =− . (5.7.11)
1 ′ ′
P0 (y1 y − y y2 )
2 1
We leave it to you to start from Equation 5.7.9 and show by a similar argument that
F y1
′
u = . (5.7.12)
2 ′ ′
P0 (y1 y − y y2 )
2 1
We can now obtain u and u by integrating u and u . The constants of integration can be taken to be zero, since any choice
1 2
′
1
′
2
of u and u in Equation 5.7.3 will suffice.

1 2
You should not memorize Equation 5.7.11 and Equation 5.7.12. On the other hand, you don’t want to rederive the whole
procedure for every specific problem. We recommend the a compromise:
a. Write
yp = u1 y1 + u2 y2 (5.7.13)
to remind yourself of what you’re doing.

b. Write the system
′ ′
u y1 + u y2 =0
1 2
F
(5.7.14)
′ ′ ′ ′
u y +u y =
1 1 2 2 P0
for the specific problem you’re trying to solve.

c. Solve Equation 5.7.14 for u and u by any convenient method.
′
1
′
2
d. Obtain u and u by integrating u and u , taking the constants of integration to be zero.

1 2
′
1
′
2
e. Substitute u and u into Equation 5.7.13 to obtain y .

1 2 p
Example 5.7.1
Find a particular solution y of p
2 ′′ ′ 9/2
x y − 2x y + 2y = x , (5.7.15)
given that y 1 =x and y 2 =x

2
are solutions of the complementary equation
2 ′′ ′
x y − 2x y + 2y = 0.

Then find the general solution of Equation 5.7.15.
Solution
We set
2
yp = u1 x + u2 x ,
where
′ ′ 2
u x+ u x =0
1 2
9/2
x 5/2
′ ′
u + 2u x = =x .
1 2
x2
From the first equation, u = −u x . Substituting this into the second equation yields u x = x
′
1
′
2
′
2
5/2
, so ′
u
2
3/2
=x and
therefore u = −u x = −x . Integrating and taking the constants of integration to be zero yields
′
1
′
2
5/2
2 7/2
2 5/2
u1 = − x and u2 = x .
7 5
Therefore
2
2 7/2
2 5/2 2
4 9/2
yp = u1 x + u2 x =− x x+ x x = x ,
7 5 35
and the general solution of Equation 5.7.15 is

4 9/2 2
y = x + c1 x + c2 x .
35
Example 5.7.2
Find a particular solution y of p
′′ ′ 2
(x − 1)y − x y + y = (x − 1 ) , (5.7.16)
given that y 1 =x and y2 =e

x
′′ ′
(x − 1)y − x y + y = 0.
Then find the general solution of Equation 5.7.16.

Solution
We set
x
yp = u1 x + u2 e ,
where
′ ′ x
u x +u e =0
1 2
2
(x − 1)
′ ′ x
u +u e = = x − 1.
1 2
x −1
Subtracting the first equation from the second yields −u (x − 1) = x − 1 , so u = −1 . From this and the first equation,
′
1
′
1
′
u = −x e
2
−x
u = xe
′
1
. Integrating and taking the constants of integration to be zero yields
−x
−x
u1 = −x and u2 = −(x + 1)e .
Therefore
x −x x 2
yp = u1 x + u2 e = (−x)x + (−(x + 1)e )e = −x − x − 1,
so the general solution of Equation 5.7.16 is

x 2 x 2 x
y = yp + c1 x + c2 e = −x − x − 1 + c1 x + c2 e = −x − 1 + (c1 − 1)x + c2 e . (5.7.17)
However, since c is an arbitrary constant, so is c

1 1 −1 ; therefore, we improve the appearance of this result by renaming
the constant and writing the general solution as
2 x
y = −x − 1 + c1 x + c2 e . (5.7.18)
There’s nothing wrong with leaving the general solution of Equation 5.7.16 in the form Equation 5.7.17; however, we think
you’ll agree that Equation 5.7.18 is preferable. We can also view the transition from Equation 5.7.17 to Equation 5.7.18
differently. In this example the particular solution y = −x − x − 1 contained the term −x, which satisfies the
p
2
complementary equation. We can drop this term and redefine y = −x − 1 , since −x − x − 1 is a solution of Equation p
2 2
5.7.16 and x is a solution of the complementary equation; hence, −x − 1 = (−x − x − 1) + x is also a solution of 2 2
Equation 5.7.16. In general, it is always legitimate to drop linear combinations of {y , y } from particular solutions obtained 1 2
by variation of parameters. (See Exercise 5.7.36 for a general discussion of this question.) We’ll do this in the following
examples and in the answers to exercises that ask for a particular solution. Therefore, don’t be concerned if your answer to
such an exercise differs from ours only by a solution of the complementary equation.
Example 5.7.3
′′ ′
1
y + 3 y + 2y = . (5.7.19)
x
1 +e

Solution
The characteristic polynomial of the complementary equation
′′ ′
y + 3 y + 2y = 0 (5.7.20)
is p(r) = r
2
+ 3r + 2 = (r + 1)(r + 2), so y = e and y = e
1 form a fundamental set of solutions of Equation
−x
2
−2x
5.7.20. We look for a particular solution of Equation 5.7.19 in the form
−x −2x
yp = u1 e + u2 e ,
where
′ −x ′ −2x
u e + u e =0
1 2
−x −2x
1
′ ′
−u e − 2u e = .
1 2 x
1 +e
Adding these two equations yields

2x
1 e
′ −2x ′
−u e = , so u =− .
2 x 2 x
1 +e 1 +e
From the first equation,

x
e
′ ′ −x
u = −u e = .
1 2 x
1 +e
Integrating by means of the substitution v = e and taking the constants of integration to be zero yields
x
x
e dv x
u1 = ∫ dx = ∫ = ln(1 + v) = ln(1 + e )
x
1 +e 1 +v
and

2x
e v 1
u2 = −∫ dx = − ∫ dv = ∫ [ − 1] dv
x
1 +e 1 +v 1 +v
x x
= ln(1 + v) − v = ln(1 + e ) − e .
Therefore
−x −2x
yp = u1 e + u2 e
x −x x x −2x
= [ln(1 + e )] e + [ln(1 + e ) − e ] e ,
so
−x −2x x −x
yp = (e +e ) ln(1 + e ) − e .
Since the last term on the right satisfies the complementary equation, we drop it and redefine
−x −2x x
yp = (e +e ) ln(1 + e ).
The general solution of Equation 5.7.19 is

−x −2x −x −2x x −x −2x
y = yp + c1 e + c2 e = (e +e ) ln(1 + e ) + c1 e + c2 e .
Example 5.7.4
2
2 ′′ ′ ′
(x − 1)y + 4x y + 2y = , y(0) = −1, y (0) = −5, (5.7.21)
x +1
given that
1 1
y1 = and y2 =
x −1 x +1

2 ′′ ′
(x − 1)y + 4x y + 2y = 0.
Solution
We first use variation of parameters to find a particular solution of
2 ′′ ′
2
(x − 1)y + 4x y + 2y =
x +1
on (−1, 1) in the form

u1 u2
yp = + ,
x −1 x +1
where
′ ′
u u
1 2
+ =0 (5.7.22)
x −1 x +1
′ ′
u u 2
1 2
− − =
2 2 2
(x − 1) (x + 1) (x + 1)(x − 1)
Multiplying the first equation by 1/(x − 1) and adding the result to the second equation yields
1 1 2
′
[ − ]u = . (5.7.23)
2 2 2 2
x −1 (x + 1) (x + 1)(x − 1)
Since

1 1 (x + 1) − (x − 1) 2
[ − ] = = ,
2 2 2 2
x −1 (x + 1) (x + 1)(x − 1) (x + 1)(x − 1)
Equation 5.7.23 implies that u ′

2
=1 . From Equation 5.7.22,
x −1 x −1
′ ′
u =− u =− .
1 2
x +1 x +1
Integrating and taking the constants of integration to be zero yields

x −1 x +1 −2
u1 = − ∫ dx = − ∫ dx
x +1 x +1
2
=∫ [ − 1] dx = 2 ln(x + 1) − x
x +1
and
u2 = ∫ dx = x.
Therefore
u1 u2 1 1
yp = + = [2 ln(x + 1) − x] +x
x −1 x +1 x −1 x +1
2 ln(x + 1) 1 1 2 ln(x + 1) 2x
= +x [ − ] = − .
x −1 x +1 x −1 x −1 (x + 1)(x − 1)
However, since
2x 1 1
=[ + ]
(x + 1)(x − 1) x +1 x −1
is a solution of the complementary equation, we redefine

2 ln(x + 1)
yp = .
x −1

2 ln(x + 1) c1 c2
y = + + . (5.7.24)
x −1 x −1 x +1
Differentiating this yields
2 2 ln(x + 1) c1 c2
′
y = − − − .
x2 − 1 (x − 1)2 (x − 1)2 (x + 1)2
Setting x = 0 in the last two equations and imposing the initial conditions y(0) = −1 and y ′
(0) = −5 yields the system
−c1 + c2 = −1
−2 − c1 − c2 = −5.
The solution of this system is c 1 = 2, c2 = 1 . Substituting these into Equation 5.7.24 yields
2 ln(x + 1) 2 1
y = + +
x −1 x −1 x +1
2 ln(x + 1) 3x + 1
= +
x −1 x2 − 1
as the solution of Equation 5.7.21. Figure 5.7.1 is a graph of the solution.

2 ln(x+1)
Figure 5.7.1 : y =
3x+1
+
2
x−1 x −1
We’ve now considered three methods for solving nonhomogeneous linear equations: undetermined coefficients, reduction of
order, and variation of parameters. It’s natural to ask which method is best for a given problem. The method of undetermined
coefficients should be used for constant coefficient equations with forcing functions that are linear combinations of
polynomials multiplied by functions of the form e , e cos ωx, or e sin ωx. Although the other two methods can be used
αx λx λx
to solve such problems, they will be more difficult except in the most trivial cases, because of the integrations involved.
If the equation isn’t a constant coefficient equation or the forcing function isn’t of the form just specified, the method of
undetermined coefficients does not apply and the choice is necessarily between the other two methods. The case could be
made that reduction of order is better because it requires only one solution of the complementary equation while variation of
parameters requires two. However, variation of parameters will probably be easier if you already know a fundamental set of
solutions of the complementary equation.

5.7E: Variation of Parameters (Exercises)
Q5.7.1
In Exercises 5.7.1–5.7.6 use variation of parameters to find a particular solution.
1. y ′′
+ 9y = tan 3x
2. y ′′
+ 4y = sin 2x sec
2
2x
3. y ′′
− 3 y + 2y =
′
1+e
4
−x
4. y ′′
− 2 y + 2y = 3 e
′ x
sec x
5. y ′′
− 2 y + y = 14 x
′ 3/2
e
x
−x
6. y ′′
−y =
4e
1−e
−2x
Q5.7.2
In Exercises 5.7.7-5.7.29 use variation of parameters to find a particular solution, given the solutions y1 , y2 of the
complementary equation.
7. x 2
y
′′
+ xy − y = 2x
′ 2
+ 2; y1 = x, y2 =
1
8. x y ′′
+ (2 − 2x)y + (x − 2)y = e
′ 2x
; y1 = e ,
x
y2 =
e
9. 4x 2
y
′′
+ (4x − 8 x )y + (4 x
2 ′ 2
− 4x − 1)y = 4 x
1/2
e ,
x
x >0 ;y 1 =x
1/2 x
e , y2 = x
−1/2
e
x
2 2
10. y ′′
+ 4x y + (4 x
′ 2
+ 2)y = 4 e
−x(x+2)
; y1 = e
−x
, y2 = x e
−x
11. x 2
y
′′
− 4x y + 6y = x
′ 5/2
, x > 0; y1 = x , y2 = x
2 3
12. x 2
y
′′
− 3x y + 3y = 2 x
′ 4
sin x; y1 = x, y2 = x
3
13. (2x + 1)y ′′ ′

− 2 y − (2x + 3)y = (2x + 1 ) e
2 −x
; y1 = e
−x
, y2 = x e
x
− − −
14. 4x y ′′ ′
+ 2 y + y = sin √x ; y1 = cos √x , y2 = sin √x
15. x y ′′
− (2x + 2)y + (x + 2)y = 6 x e ;
′ 3 x
y1 = e ,
x
y2 = x e
3 x
16. x 2
y
′′
− (2a − 1)x y + a y = x
′ 2 a+1
; y1 = x ,
a
y2 = x
a
ln x
17. x 2
y
′′
− 2x y + (x
′ 2
+ 2)y = x
3
cos x; y1 = x cos x, y2 = x sin x
2 2
18. x y ′′ ′
− y − 4x y = 8x ;
3 5
y1 = e
x
, y2 = e
−x
19. (sin x)y ′′

+ (2 sin x − cos x)y + (sin x − cos x)y = e
′ −x
; y1 = e
−x
, y2 = e
−x
cos x
− 2x − −2x
20. 4x 2
y
′′ ′
− 4x y + (3 − 16 x )y = 8 x
2 5/2
; y1 = √x e , y2 = √x e
− −
21. 4x 2
y
′′
− 4x y + (4 x
′ 2
+ 3)y = x
7/2
; y1 = √x sin x, y2 = √x cos x
22. x 2
y
′′
− 2x y − (x
′ 2
− 2)y = 3 x ;
4
y1 = x e , y2 = x e
x −x
23. x 2
y
′′
− 2x(x + 1)y + (x
′ 2
+ 2x + 2)y = x e ;
3 x
y1 = x e ,
x
y2 = x e
2 x
24. x 2
y
′′
− x y − 3y = x
′ 3/2
; y1 = 1/x, y2 = x
3
25. x 2
y
′′
− x(x + 4)y + 2(x + 3)y = x e ;
′ 4 x
y1 = x ,
2
y2 = x e
2 x
26. x 2
y
′′
− 2x(x + 2)y + (x
′ 2
+ 4x + 6)y = 2x e ;
x
y1 = x e ,
2 x
y2 = x e
3 x
27. x 2
y
′′
− 4x y + (x
′ 2
+ 6)y = x ;
4
y1 = x
2
cos x, y2 = x
2
sin x
28. (x − 1)y ′′ ′
− x y + y = 2(x − 1 ) e ;
2 x
y1 = x, y2 = e
x
− − x
29. 4x 2
y
′′
− 4x(x + 1)y + (2x + 3)y = x
′ 5/2 x
e ; y1 = √x , y2 = √x e
Q5.7.3
In Exercises 5.7.30–5.7.32 use variation of parameters to solve the initial value problem, given y1 , y2 are solutions of the
complementary equation.
30. (3x − 1)y ′′ ′
− (3x + 2)y − (6x − 8)y = (3x − 1 ) e
2 2x
, y(0) = 1, y (0) = 2
′
;y 1 =e
2x
, y2 = x e
−x
31. (x − 1) 2
y
′′ ′
− 2(x − 1)y + 2y = (x − 1 ) ,
2
y(0) = 3,
′
y (0) = −6 ;
y1 = x − 1 ,y 2 =x
2
−1
32. (x − 1) 2
y
′′
− (x
2 ′
− 1)y + (x + 1)y = (x − 1 ) e ,
3 x
y(0) = 4,
′
y (0) = −6 ;
x
y1 = (x − 1)e , y2 = x − 1
Q5.7.4
In Exercises 5.7.33-5.7.35 use variation of parameters to solve the initial value problem and graph the solution, given that y , 1
y are solutions of the complementary equation.

2
33. (x 2
− 1)y
′′ ′
+ 4x y + 2y = 2x,
′
y(0) = 0, y (0) = −2; y1 =
1
x−1
, y2 =
x+1
1
34. x2
y
′′ ′
+ 2x y − 2y = −2 x ,
2
y(1) = 1, y (1) = −1;
′
y1 = x, y2 =
1
x
2
35. (x + 1)(2x + 3)y ′′ ′

+ 2(x + 2)y − 2y = (2x + 3 ) ,
2
y(0) = 0,
′
y (0) = 0 ;y 1 = x + 2, y2 =
1
x+1
Q5.7.5
36. Suppose
¯
¯¯
yp = y + a1 y1 + a2 y2 (5.7E.1)
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = F (x), (A)
where y and y are solutions of the complementary equation

1 2
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0. (5.7E.2)
Show that ȳ is also a solution of (A).

¯
¯
37. Suppose p, q, and f are continuous on (a, b) and let x be in (a, b). Let y and y be the solutions of 0 1 2
′′ ′
y + p(x)y + q(x)y = 0 (5.7E.3)
such that
′ ′
y1 (x0 ) = 1, y (x0 ) = 0, y2 (x0 ) = 0, y (x0 ) = 1. (5.7E.4)
1 2
Use variation of parameters to show that the solution of the initial value problem
′′ ′ ′
y + p(x)y + q(x)y = f (x), y(x0 ) = k0 , y (x0 ) = k1 , (5.7E.5)
is
y(x) = k0 y1 (x) + k1 y2 (x)
x t (5.7E.6)
+∫ (y1 (t)y2 (x) − y1 (x)y2 (t)) f (t) exp(∫ p(s) ds) dt.
x0 x0
HINT: Use Abel's formula for the Wronskian of {y 1, y2 } , and integrate u and u from x to x. ′
1
′
2 0
Show also that

′ ′ ′
y (x) = k0 y (x) + k1 y (x)
1 2
x t
(5.7E.7)
′ ′
+∫ (y1 (t)y (x) − y (x)y2 (t)) f (t) exp(∫ p(s) ds) dt.
x0 2 1 x0
38. Suppose f is continuous on an open interval that contains x0 = 0 . Use variation of parameters to find a formula for the

′′ ′
y − y = f (x), y(0) = k0 , y (0) = k1 . (5.7E.8)
39. Suppose f is continuous on (a, ∞), where a < 0 , so x 0 =0 is in (a, ∞).

a. Use variation of parameters to find a formula for the solution of the initial value problem
′′ ′
y + y = f (x), y(0) = k0 , y (0) = k1 . (5.7E.9)
HINT: You will need the addition formulas for the sine and cosine.
sin(A + B) = sin A cos B + cos A sin B
∞
For the rest of this exercise assume that the improper integral ∫ 0
f (t) dt is absolutely convergent.
b. Show that if y is a solution of
′′
y + y = f (x) (A)
on (a, ∞), then
lim (y(x) − A0 cos x − A1 sin x) = 0 (B)

x→∞
and
′
lim (y (x) + A0 sin x − A1 cos x) = 0, (C)
x→∞
where
∞ ∞
A0 = k0 − ∫ f (t) sin t dt and A1 = k1 + ∫ f (t) cos t dt. (5.7E.10)

0 0
∞ ∞
HINT: Recall from calculus that if ∫ f (t)dt converges absolutely, then lim
0
∫ x→∞
x
|f (t)|dt = 0 .
c. Show that if A and A are arbitrary constants, then there’s a unique solution of y
0 1
′′
+ y = f (x) on (a, ∞) that satisfies
(B) and (C).

CHAPTER OVERVIEW
6: APPLICATIONS OF LINEAR SECOND ORDER EQUATIONS
In this Chapter, we study applications of linear second order equations.
6.1: SPRING PROBLEMS I

about spring--mass systems.
6.1E: SPRING PROBLEMS I (EXERCISES)

6.2: SPRING PROBLEMS II
about spring--mass systems.
6.2E: SPRING PROBLEMS II (EXERCISES)
1 9/16/2020
6.1: Spring Problems I
We consider the motion of an object of mass m, suspended from a spring of negligible mass. We say that the spring–mass
system is in equilibrium when the object is at rest and the forces acting on it sum to zero. The position of the object in this case
is the equilibrium position. We define y to be the displacement of the object from its equilibrium position (Figure 6.1.1),
measured positive upward.
Figure 6.1.1 : (a) y > 0 (b) y = 0 , (c) y < 0

Our model accounts for the following kinds of forces acting on the object:
The force −mg, due to gravity.
A force F exerted by the spring resisting change in its length. The natural length of the spring is its length with no mass
s
attached. We assume that the spring obeys Hooke’s law: If the length of the spring is changed by an amount ΔL from its
natural length, then the spring exerts a force F = kΔL , where k is a positive number called the spring constant. If the
s
spring is stretched then ΔL > 0 and F > 0 , so the spring force is upward, while if the spring is compressed then ΔL < 0
s
and F < 0 , so the spring force is downward.

s
A damping force F = −cy that resists the motion with a force proportional to the velocity of the object. It may be due to
d
′
air resistance or friction in the spring. However, a convenient way to visualize a damping force is to assume that the object
is rigidly attached to a piston with negligible mass immersed in a cylinder (called a dashpot) filled with a viscous liquid
(Figure 6.1.1). As the piston moves, the liquid exerts a damping force. We say that the motion is undamped if c = 0 , or
damped if c > 0 .
An external force F , other than the force due to gravity, that may vary with t , but is independent of displacement and
velocity. We say that the motion is free if F ≡ 0 , or forced if F ≢ 0 .
Figure 6.1.2 : A spring – mass system with damping

′′ ′
my = −mg + Fd + Fs + F = −mg − c y + Fs + F . (6.1.1)
We must now relate F to y . In the absence of external forces the object stretches the spring by an amount Δl to assume its
s
equilibrium position (Figure 6.1.3). Since the sum of the forces acting on the object is then zero, Hooke’s Law implies that
mg = kΔl . If the object is displaced y units from its equilibrium position, the total change in the length of the spring is
ΔL = Δl − y , so Hooke’s law implies that
Fs = kΔL = kΔl − ky.

′′ ′
my = −mg − c y + kΔL − ky + F .
Since mg = kΔl this can be written as

′′ ′
my + c y + ky = F . (6.1.2)
We call this the equation of motion.
Figure 6.1.3 :(a) Natural length of spring (b) Spring stretched by mass
Simple Harmonic Motion

Throughout the rest of this section we’ll consider spring–mass systems without damping; that is, c = 0 . We’ll consider
systems with damping in the next section. We first consider the case where the motion is also free; that is, F =0. We begin with
an example.
Example 6.1.1
An object stretches a spring 6 inches in equilibrium.
a. Set up the equation of motion and find its general solution.
b. Find the displacement of the object for t > 0 if it is initially displaced 18 inches above equilibrium and given a
downward velocity of 3 ft/s.
Solution a:
Setting c = 0 and F =0 in Equation 6.1.2 yields the equation of motion
′′
my + ky = 0,
which we rewrite as
k
′′
y + y = 0. (6.1.3)
m
Although we would need the weight of the object to obtain k from the equation mg = kΔl we can obtain k/m from Δl
alone; thus, k/m = g/Δl . Consistent with the units used in the problem statement, we take g = 32 ft/s . Although Δl is
2

stated in inches, we must convert it to feet to be consistent with this choice of g ; that is, Δl = 1/2 ft. Therefore
k 32
= = 64
m 1/2

′′
y + 64y = 0. (6.1.4)
The characteristic equation of Equation 6.1.4 is

2
r + 64 = 0,
which has the zeros r = ±8i . Therefore the general solution of Equation 6.1.4 is
y = c1 cos 8t + c2 sin 8t. (6.1.5)
Solution b:
The initial upward displacement of 18 inches is positive and must be expressed in feet. The initial downward velocity is
negative; thus,
3 ′
y(0) = and y (0) = −3.
2

′
y = −8 c1 sin 8t + 8 c2 cos 8t. (6.1.6)
Setting t =0 in Equation 6.1.5 and Equation 6.1.6 and imposing the initial conditions shows that c1 = 3/2 and
c2 = −3/8 . Therefore
3 3
y = cos 8t − sin 8t,
2 8
where y is in feet (Figure 6.1.4).

We’ll now consider the equation
′′
my + ky = 0
−−−−
where m and k are arbitrary positive numbers. Dividing through by m and defining ω 0 = √k/m yields
′′ 2
y + ω y = 0.
0
The general solution of this equation is

y = c1 cos ω0 t + c2 sin ω0 t. (6.1.7)
We can rewrite this in a more useful form by defining

−−−−−−
2 2
R = √c +c , (6.1.8)
1 2

Figure 6.1.4 : y = 3
2
cos 8t −
3
8
sin 8t
and
c1 = R cos ϕ and c2 = R sin ϕ. (6.1.9)
Substituting from Equation 6.1.9 into Equation 6.1.7 and applying the identity
cos ω0 t cos ϕ + sin ω0 t sin ϕ = cos(ω0 t − ϕ)
yields
y = R cos(ω0 t − ϕ). (6.1.10)
From Equation 6.1.8 and Equation 6.1.9 we see that the R and ϕ can be interpreted as polar coordinates of the point with
rectangular coordinates (c , c ) (Figure 6.1.5). Given c and c , we can compute R from Equation 6.1.8. From Equation
1 2 1 2
6.1.8 and Equation 6.1.9, we see that ϕ is related to c and c by

1 2
c1 c2
cos ϕ = −−−−−− and sin ϕ = −−−−−−.
2 2 2 2
√c +c √c +c
1 2 1 2
There are infinitely many angles ϕ , differing by integer multiples of 2π , that satisfy these equations. We will always
choose ϕ so that −π ≤ ϕ < π .
The motion described by Equation 6.1.7 or Equation 6.1.10 is simple harmonic motion. We see from either of these
equations that the motion is periodic, with period
T = 2π/ ω0 .
This is the time required for the object to complete one full cycle of oscillation (for example, to move from its highest
position to its lowest position and back to its highest position). Since the highest and lowest positions of the object are
y = R and y = −R , we say that R is the amplitude of the oscillation. The angle ϕ in Equation 6.1.10 is the phase angle.
It’s measured in radians. Equation Equation 6.1.10 is the amplitude–phase form of the displacement. If t is in seconds
then ω is in radians per second (rad/s); it is the frequency of the motion. It is also called the natural frequency of the
0
spring–mass system without damping.

−−−−−−
Figure 6.1.5 : R = √c 2
1
+c
2
2
;c 1 = R cos ϕ ;c 2 = R sin ϕ
Example 6.1.2
We found the displacement of the object in Example 6.1.1 to be
3 3
y = cos 8t − sin 8t.
2 8
Find the frequency, period, amplitude, and phase angle of the motion.
Solution:
The frequency is ω 0 =8 rad/s, and the period is T = 2π/ ω0 = π/4 s. Since c 1 = 3/2 and c 2 = −3/8 , the amplitude is
−−−−−−−−−−−−
2 2
−−−−−− 3 3 3
2 2
−−
R = √c + c = √( ) +( ) = √17.
1 2
2 8 8
The phase angle is determined by

3
2 4
cos ϕ = = −− (6.1.11)
3 −−
√17 √17
8
and
3
− 1
8
sin ϕ = =− −−. (6.1.12)
3 −−
√17 √17
8
Using a calculator, we see from Equation 6.1.11that

ϕ ≈ ±.245 rad.
Since sin ϕ < 0 (see Equation 6.1.12), the minus sign applies here; that is,
ϕ ≈ −.245 rad.
Example 6.1.3
The natural length of a spring is 1 m. An object is attached to it and the length of the spring increases to 102 cm when the
object is in equilibrium. Then the object is initially displaced downward 1 cm and given an upward velocity of 14 cm/s.
Find the displacement for t > 0 . Also, find the natural frequency, period, amplitude, and phase angle of the resulting
motion. Express the answers in cgs units.
Solution:
In cgs units g = 980 cm/s . Since Δl = 2 cm, ω
2 2
0
= g/Δl = 490 . Therefore

′′ ′
y + 490y = 0, y(0) = −1, y (0) = 14.
The general solution of the differential equation is

−− −−
y = c1 cos 7 √10t + c2 sin 7 √10t,
so
′ −− −− −−
y = 7 √10 (−c1 sin 7 √10t + c2 cos 7 √10t) .
−−
Substituting the initial conditions into the last two equations yields c 1 = −1 and c2 = 2/ √10 . Hence,
−− 2 −−
y = − cos 7 √10t + sin 7 √10t.
−−
√10
−− −−
The frequency is 7√10 rad/s, and the period is T = 2π/(7 √10) s. The amplitude is
−−−−−−−−−−−−−−
2 −
−
−−−−−− 2 7
2 2 2
R = √c +c = √ (−1 ) +( ) =√ cm
1 2 −−
√10 5
The phase angle is determined by

−
− −
−
c1 5 c2 2
cos ϕ = = −√ and sin ϕ = =√
R 7 R 7
Therefore ϕ is in the second quadrant and

−
−
5
−1
ϕ = cos (−√ ) ≈ 2.58rad
7
Undamped Forced Oscillation

In many mechanical problems a device is subjected to periodic external forces. For example, soldiers marching in cadence on
a bridge cause periodic disturbances in the bridge, and the engines of a propeller driven aircraft cause periodic disturbances in
its wings. In the absence of sufficient damping forces, such disturbances – even if small in magnitude – can cause structural
breakdown if they are at certain critical frequencies. To illustrate, this we’ll consider the motion of an object in a spring–mass
system without damping, subject to an external force
F (t) = F0 cos ωt
where F is a constant. In this case the equation of motion Equation 6.1.2 is

0
′′
my + ky = F0 cos ωt,
which we rewrite as
′′
F0
2
y +ω y = cos ωt (6.1.13)
0
m
−−−−
with ω = √k/m . We’ll see from the next two examples that the solutions of Equation
0 6.1.13 with ω ≠ ω0 behave very
differently from the solutions with ω = ω . 0
Example 6.1.4
′′
F0 ′
2
y +ω y = cos ωt, y(0) = 0, y (0) = 0, (6.1.14)
0
m
given that ω ≠ ω . 0
Solution:

We first obtain a particular solution of Equation 6.1.13 by the method of undetermined coefficients. Since ω ≠ ω , cos ωt 0

′′ 2
y + ω y = 0.
0
Therefore Equation 6.1.13has a particular solution of the form

yp = A cos ωt + B sin ωt.
Since
′′ 2
yp = −ω (A cos ωt + B sin ωt),
F0
′′ 2
yp + ω yp = cos ωt
0
m
if and only if
2
F0
2
(ω − ω ) (A cos ωt + B sin ωt) = cos ωt.
0
m
This holds if and only if

F0
A = and B = 0,
2 2
m(ω −ω )
0
so
F0
yp = cos ωt.
2 2
m(ω −ω )
0
The general solution of Equation 6.1.13is

F0
y = cos ωt + c1 cos ω0 t + c2 sin ω0 t, (6.1.15)
2
m(ω − ω2 )
0
so
′
−ωF0
y = sin ωt + ω0 (−c1 sin ω0 t + c2 cos ω0 t).
2 2
m(ω −ω )
0
The initial conditions y(0) = 0 and y ′

(0) = 0 in Equation 6.1.14imply that
F0
c1 = − and c2 = 0.
2 2
m(ω −ω )
0
Substituting these into Equation 6.1.15yields

F0
y = (cos ωt − cos ω0 t). (6.1.16)
2 2
m(ω −ω )
0
It is revealing to write this in a different form. We start with the trigonometric identities
cos(α − β) = cos α cos β + sin α sin β
cos(α + β) = cos α cos β − sin α sin β.
Subtracting the second identity from the first yields

cos(α − β) − cos(α + β) = 2 sin α sin β (6.1.17)
Now let
α − β = ωt and α + β = ω0 t, (6.1.18)
so that

(ω0 + ω)t (ω0 − ω)t
α = and β = . (6.1.19)
2 2
Substituting Equation 6.1.18and Equation 6.1.19into Equation 6.1.17yields

(ω0 − ω)t (ω0 + ω)t
cos ωt − cos ω0 t = 2 sin sin ,
2 2
and substituting this into Equation 6.1.16yields

(ω0 + ω)t
y = R(t) sin , (6.1.20)
2
where
2F0 (ω0 − ω)t
R(t) = sin . (6.1.21)
2 2
m(ω −ω ) 2
0
From Equation 6.1.20 we can regard y as a sinusoidal variation with frequency (ω + ω)/2 and variable amplitude 0
|R(t)|. In Figure 6.1.6 the dashed curve above the t axis is y = |R(t)| , the dashed curve below the t axis is y = −|R(t)| ,
and the displacement y appears as an oscillation bounded by them. The oscillation of y for t on an interval between
successive zeros of R(t) is called a beat.
Figure 6.1.6 : Undamped oscillation with beats

You can see from Equation 6.1.20 and Equation 6.1.21 that
2| F0 |
|y(t)| ≤ ;
2 2
m| ω −ω |
0
moreover, if ω + ω is sufficiently large compared with ω − ω , then |y| assumes values close to (perhaps equal to) this
0 0
upper bound during each beat. However, the oscillation remains bounded for all t . (This assumes that the spring can
withstand deflections of this size and continue to obey Hooke’s law.) The next example shows that this isn’t so if ω = ω . 0
Example 6.1.5 :
′′
F0
2
y +ω y = cos ω0 t. (6.1.22)
0
m
Solution:
We first obtain a particular solution y of Equation 6.1.22. Since cos ω
p 0t is a solution of the complementary equation, the
form for y is
p
yp = t(A cos ω0 t + B sin ω0 t). (6.1.23)
Then
′
yp = A cos ω0 t + B sin ω0 t + ω0 t(−A sin ω0 t + B cos ω0 t)
and
′′ 2
yp = 2 ω0 (−A sin ω0 t + B cos ω0 t) − ω t(A cos ω0 t + B sin ω0 t). (6.1.24)
0

From Equation 6.1.23and Equation 6.1.24, we see that y satisfies Equation 6.1.22if
p
F0
−2Aω0 sin ω0 t + 2Bω0 cos ω0 t = cos ω0 t;
m
that is, if
F0
A =0 and B = .
2mω0
Therefore
F0 t
yp = sin ω0 t
2mω0
is a particular solution of Equation 6.1.22. The general solution of Equation 6.1.22is

F0 t
y = sin ω0 t + c1 cos ω0 t + c2 sin ω0 t.
2mω0
Figure 6.1.7 : Undamped displacement due to resonance.

The graph of y is shown in Figure 6.1.7, where it can be seen that y oscillates between the dashed lines
p p
F0 t F0 t
y = and y =−
2mω0 2mω0
with increasing amplitude that approaches ∞ as t → ∞ . Of course, this means that the spring must eventually fail to
obey Hooke’s law or break.
This phenomenon of unbounded displacements of a spring–mass system in response to a periodic forcing function at its
natural frequency is called resonance. More complicated mechanical structures can also exhibit resonance–like
phenomena. For example, rhythmic oscillations of a suspension bridge by wind forces or of an airplane wing by periodic
vibrations of reciprocating engines can cause damage or even failure if the frequencies of the disturbances are close to
critical frequencies determined by the parameters of the mechanical system in question.

6.1E: Spring Problems I (Exercises)
In the following exercises assume that there’s no damping.
Q6.1.1
1. An object stretches a spring 4 inches in equilibrium. Find and graph its displacement for t > 0 if it is initially displaced 36
inches above equilibrium and given a downward velocity of 2 ft/s.
2. An object stretches a string 1.2 inches in equilibrium. Find its displacement for t >0 if it is initially displaced 3 inches
below equilibrium and given a downward velocity of 2 ft/s.
3. A spring with natural length .5 m has length 50.5 cm with a mass of 2 gm suspended from it. The mass is initially displaced
1.5 cm below equilibrium and released with zero velocity. Find its displacement for t > 0 .
4. An object stretches a spring 6 inches in equilibrium. Find its displacement for t > 0 if it is initially displaced 3 inches above
equilibrium and given a downward velocity of 6 inches/s. Find the frequency, period, amplitude and phase angle of the motion.
5. An object stretches a spring 5 cm in equilibrium. It is initially displaced 10 cm above equilibrium and given an upward
velocity of .25 m/s. Find and graph its displacement for t > 0 . Find the frequency, period, amplitude, and phase angle of the
motion.
6. A 10 kg mass stretches a spring 70 cm in equilibrium. Suppose a 2 kg mass is attached to the spring, initially displaced 25
cm below equilibrium, and given an upward velocity of 2 m/s. Find its displacement for t > 0 . Find the frequency, period,
amplitude, and phase angle of the motion.
7. A weight stretches a spring 1.5 inches in equilibrium. The weight is initially displaced 8 inches above equilibrium and given
a downward velocity of 4 ft/s. Find its displacement for t > 0 .
8. A weight stretches a spring 6 inches in equilibrium. The weight is initially displaced 6 inches above equilibrium and given a
downward velocity of 3 ft/s. Find its displacement for t > 0 .
−−
9. A spring–mass system has natural frequency 7√10 rad/s. The natural length of the spring is .7 m. What is the length of the
spring when the mass is in equilibrium?
10. A 64 lb weight is attached to a spring with constant k = 8 lb/ft and subjected to an external force F (t) = 2 sin t . The
weight is initially displaced 6 inches above equilibrium and given an upward velocity of 2 ft/s. Find its displacement for
t > 0.
11. A unit mass hangs in equilibrium from a spring with constant k = 1/16 . Starting at t =0 , a force F (t) = 3 sin t is
applied to the mass. Find its displacement for t > 0 .
12. A 4 lb weight stretches a spring 1 ft in equilibrium. An external force F (t) = .25 sin 8t lb is applied to the weight, which
is initially displaced 4 inches above equilibrium and given a downward velocity of 1 ft/s. Find and graph its displacement for
t > 0.
13. A 2 lb weight stretches a spring 6 inches in equilibrium. An external force F (t) = sin 8t lb is applied to the weight, which
is released from rest 2 inches below equilibrium. Find its displacement for t > 0 .
14. A 10 gm mass suspended on a spring moves in simple harmonic motion with period 4 s. Find the period of the simple
harmonic motion of a 20 gm mass suspended from the same spring.
15. A 6 lb weight stretches a spring 6 inches in equilibrium. Suppose an external force F (t) = sin ωt + cos ωt lb is
3
16
3
applied to the weight. For what value of ω will the displacement be unbounded? Find the displacement if ω has this value.
Assume that the motion starts from equilibrium with zero initial velocity.
16. A 6 lb weight stretches a spring 4 inches in equilibrium. Suppose an external force F (t) = 4 sin ωt − 6 cos ωt lb is
applied to the weight. For what value of ω will the displacement be unbounded? Find and graph the displacement if ω has this
value. Assume that the motion starts from equilibrium with zero initial velocity.
17. A mass of one kg is attached to a spring with constant k = 4 N/m. An external force F (t) = − cos ωt − 2 sin ωt n is
applied to the mass. Find the displacement y for t > 0 if ω equals the natural frequency of the spring–mass system. Assume

that the mass is initially displaced 3 m above equilibrium and given an upward velocity of 450 cm/s.
18. An object is in simple harmonic motion with frequency ω , with y(0) = y and y (0) = v . Find its displacement for
0 0
′
0
t > 0 . Also, find the amplitude of the oscillation and give formulas for the sine and cosine of the initial phase angle.
19. Two objects suspended from identical springs are set into motion. The period of one object is twice the period of the other.
How are the weights of the two objects related?
20. Two objects suspended from identical springs are set into motion. The weight of one object is twice the weight of the other.
How are the periods of the resulting motions related?
21. Two identical objects suspended from different springs are set into motion. The period of one motion is 3 times the period
of the other. How are the two spring constants related?

6.2: Spring Problems II
Free Vibrations with Damping
In this section we consider the motion of an object in a spring–mass system with damping. We start with unforced motion, so
the equation of motion is
′′ ′
my + c y + ky = 0. (6.2.1)
Now suppose the object is displaced from equilibrium and given an initial velocity. Intuition suggests that if the damping force
is sufficiently weak the resulting motion will be oscillatory, as in the undamped case considered in the previous section, while
if it is sufficiently strong the object may just move slowly toward the equilibrium position without ever reaching it. We’ll now
confirm these intuitive ideas mathematically. The characteristic equation of Equation 6.2.1 is
2
mr + cr + k = 0.
The roots of this equation are

− −−−− −−− − −−−− −−−
−c − √ c2 − 4mk −c + √ c2 − 4mk
r1 = and r2 = . (6.2.2)
2m 2m
We saw in Section 5.2 that the form of the solution of Equation 6.2.1 depends upon whether c 2
− 4mk is positive, negative, or
zero. We’ll now consider these three cases.
Underdamped Motion
−−−−
We say the motion is underdamped if c < √4mk . In this case r and r in Equation 6.2.2 are complex conjugates, which we
1 2
write as
c c
r1 = − − i ω1 and r2 = − + i ω1 ,
2m 2m
where
− −−−−−−−
√ 4mk − c2
ω1 = .
2m
The general solution of Equation 6.2.1 in this case is

−ct/2m
y =e (c1 cos ω1 t + c2 sin ω1 t).
By the method used in Section 6.1 to derive the amplitude–phase form of the displacement of an object in simple harmonic
motion, we can rewrite this equation as
−ct/2m
y = Re cos(ω1 t − ϕ), (6.2.3)

Figure 6.2.1 : Underdamped motion
where
−−−−−−
2 2
R = √c +c , R cos ϕ = c1 , and R sin ϕ = c2 .
1 2
The factor Re in Equation 6.2.3 is called the time–varying amplitude of the motion, the quantity ω is called the
−ct/2m
1
frequency, and T = 2π/ω (which is the period of the cosine function in Equation 6.2.3 is called the quasi–period. A typical
1
graph of Equation 6.2.3 is shown in Figure 6.2.1. As illustrated in that figure, the graph of y oscillates between the dashed
exponential curves y = ±Re . −ct/2m
Overdamped Motion
−−−−
We say the motion is overdamped if c > √4mk . In this case the zeros r and r of the characteristic polynomial are real, with
1 2
r < r < 0 (see 6.2.2), and the general solution of 6.2.1 is

1 2
r1 t r2 t
y = c1 e + c2 e .
Again lim t→∞y(t) = 0 as in the underdamped case, but the motion isn’t oscillatory, since y can’t equal zero for more than
one value of t unless c = c = 0 . (Exercise 6.2.23.)
1 2
Critically Damped Motion

−−−−
We say the motion is critically damped if c = √4mk . In this case r1 = r2 = −c/2m and the general solution of Equation
6.2.1 is
−ct/2m
y =e (c1 + c2 t).
Again limt→∞ y(t) = 0 and the motion is nonoscillatory, since y can’t equal zero for more than one value of t unless
c1 = c2 = 0 . (Exercise 6.2.22).
Example 6.2.1
Suppose a 64 lb weight stretches a spring 6 inches in equilibrium and a dashpot provides a damping force of c lb for each
ft/sec of velocity.
a. Write the equation of motion of the object and determine the value of c for which the motion is critically damped.
b. Find the displacement y for t > 0 if the motion is critically damped and the initial conditions are y(0) = 1 and
y (0) = 20 .
′
c. Find the displacement y for t > 0 if the motion is critically damped and the initial conditions are y(0) = 1 and
y (0) = −20 .
′
Solution a

Here m = 2 slugs and k = 64/.5 = 128 lb/ft. Therefore the equation of motion Equation 6.2.1 is
′′ ′
2y + c y + 128y = 0. (6.2.4)
The characteristic equation is

2
2r + cr + 128 = 0,
which has roots

− −−−−−−− −
−c ± √ c2 − 8 ⋅ 128
r = .
4
Therefore the damping is critical if

− −−− −
c = √ 8 ⋅ 128 = 32 lb--sec/ft.
Solution b
Setting c = 32 in Equation 6.2.4 and cancelling the common factor 2 yields
′′
y + 16y + 64y = 0.
The characteristic equation is

2 2
r + 16r + 64y = (r + 8 ) = 0.
Hence, the general solution is

−8t
y =e (c1 + c2 t). (6.2.5)
Figure 6.2.2 : (a) y = e

−8t
(1 + 28t) and (b) y = e −8t
(1 − 12t)

′ −8t
y = −8y + c2 e . (6.2.6)
Imposing the initial conditions y(0) = 1 and y (0) = 20 in the last two equations shows that
6.2.2
′
1 = c1 and
20 = −8 + c . Hence, the solution of the initial value problem is
2
−8t
y =e (1 + 28t).
Therefore the object approaches equilibrium from above as t → ∞ . There’s no oscillation.

Solution c
Imposing the initial conditions y(0) = 1 and y (0) = −20 in Equation
′
6.2.5 and Equation 6.2.6 yields 1 = c1 and
−20 = −8 + c . Hence, the solution of this initial value problem is
2

−8t
y =e (1 − 12t).
Therefore the object moves downward through equilibrium just once, and then approaches equilibrium from below as
t → ∞ . Again, there’s no oscillation. The solutions of these two initial value problems are graphed in Figure 6.2.2.
Example 6.2.2
Find the displacement of the object in Example 6.2.1 if the damping constant is c = 4 lb–sec/ft and the initial conditions
are y(0) = 1.5 ft and y (0) = −3 ft/sec.
′
Solution
With c = 4 , the equation of motion Equation 6.2.4 becomes
′′ ′
y + 2 y + 64y = 0 (6.2.7)
after cancelling the common factor 2. The characteristic equation

2
r + 2r + 64 = 0
has complex conjugate roots

−−−−−− −
−2 ± √ 4 − 4 ⋅ 64 –
r = = −1 ± 3 √7i.
2
Therefore the motion is underdamped and the general solution of Equation 6.2.7is
−t – –
y =e (c1 cos 3 √7t + c2 sin 3 √7t).

′ – −t – –
y = −y + 3 √7e (−c1 sin 3 √7t + c2 cos 3 √7t).
Imposing the initial conditions y(0) = 1.5 and y (0) = −3 in the last two equations yields
′
1.5 = c1 and
–
−3 = −1.5 + 3 √7c . Hence, the solution of the initial value problem is
2
−t
3 – 1 –
y =e ( cos 3 √7t − – sin 3 √7t) . (6.2.8)
2 2 √7
The amplitude of the function in parentheses is

−−−−−−−−−−−−−−−
2 2 −−−−−−− − −−−−
3 1 9 1 64 4
R = √( ) +( –) =√ + =√ = –
2 2 √7 4 4⋅7 4⋅7 √7
Therefore we can rewrite Equation 6.2.8as

4 −t –
y = –e cos(3 √7t − ϕ),
√7
where
–
3 3 √7 1 1
cos ϕ = = and sin ϕ = − – =− .
2R 8 2 √7R 8
Therefore ϕ ≅−0.125 radians.
Example 6.2.3
Let the damping constant in Example 6.2.1 be c = 40 lb–sec/ft. Find the displacement y for t >0 if y(0) = 1 and
y (0) = 1 .
′
Solution
With c = 40 , the equation of motion Equation 6.2.4 reduces to

′′ ′
y + 20 y + 64y = 0 (6.2.9)
after cancelling the common factor 2. The characteristic equation

2
r + 20r + 64 = (r + 16)(r + 4) = 0
has the roots r 1 = −4 and r

2 = −16 . Therefore the general solution of Equation 6.2.9 is
−4t −16t
y = c1 e + c2 e . (6.2.10)

′ −4t −16t
y = −4 e − 16 c2 e .
Figure 6.2.3 : y = 17
12
e
−4t
−
5
12
−16t
e
The last two equations and the initial conditions y(0) = 1 and y ′
(0) = 1 imply that
c1 + c2 =1
−4c1 − 16c2 = 1.
The solution of this system is c 1 ,

= 17/12 c2 = −5/12 . Substituting these into Equation 6.2.10 yields
17 5
−4t −16t
y = e − e
12 12
as the solution of the given initial value problem (Figure 6.2.3).
Forced Vibrations with Damping

Now we consider the motion of an object in a spring-mass system with damping, under the influence of a periodic forcing
function F (t) = F cos ωt , so that the equation of motion is
0
′′ ′
my + c y + ky = F0 cos ωt. (6.2.11)
In Section 6.1 we considered this equation with c = 0 and found that the resulting displacement y assumed arbitrarily large
−−−−
values in the case of resonance (that is, when ω = ω = √k/m ). Here we’ll see that in the presence of damping the
0
displacement remains bounded for all t , and the initial conditions have little effect on the motion as t → ∞ . In fact, we’ll see
that for large t the displacement is closely approximated by a function of the form
y = R cos(ωt − ϕ), (6.2.12)
where the amplitude R depends upon m, c , k , F , and ω. We’re interested in the following question:
0
QUESTION

Assuming that m, c , k , and F are held constant, what value of ! produces the largest amplitude
0 R in Equation ,
6.2.12
and what is this largest amplitude?
To answer this question, we must solve Equation 6.2.11 and determine R in terms of F , ω 0 0, , and c . We can obtain a
ω
particular solution of Equation 6.2.11 by the method of undetermined coefficients. Since cos ωt does not satisfy the
complementary equation
′′ ′
my + c y + ky = 0,
we can obtain a particular solution of Equation 6.2.11 in the form

yp = A cos ωt + B sin ωt. (6.2.13)

′
yp = ω(−A sin ωt + B cos ωt)
and
′′ 2
yp = −ω (A cos ωt + B sin ωt).
From the last three equations,

′′ ′ 2 2
m yp + c yp + kyp = (−m ω A + cωB + kA) cos ωt + (−m ω B − cωA + kB) sin ωt,
so y satisfies Equation 6.2.11 if

p
2
(k − m ω )A + cωB = F0
2
−cωA + (k − m ω )B = 0.
Solving for A and B and substituting the results into Equation 6.2.13 yields
F0 2
yp = [(k − m ω ) cos ωt + cω sin ωt] ,
2 2 2 2
(k − m ω ) +c ω
which can be written in amplitude–phase form as
(6.2.14)
where
2
k − mω cω
cos ϕ = − −−−−−−−−−−−−− − and sin ϕ = − −−−−−−−−−−−−− −. (6.2.15)
√ (k − m ω2 )2 + c2 ω2 √ (k − m ω2 )2 + c2 ω2
To compare this with the undamped forced vibration that we considered in Section 6.1 it is useful to write
k
2 2 2 2
k − mω =m( − ω ) = m(ω − ω ), (6.2.16)
0
m
−−−−
where ω = √k/m is the natural angular frequency of the undamped simple harmonic motion of an object with mass m on a
0
spring with constant k . Substituting Equation 6.2.16 into Equation 6.2.14 yields
F0
yp = cos(ωt − ϕ). (6.2.17)
−−−−−−−−−−−−−−−−
2 2 2 2 2 2
√ m (ω −ω ) +c ω
0
The solution of an initial value problem

′′ ′ ′
my + c y + ky = F0 cos ωt, y(0) = y0 , y (0) = v0 ,
is of the form y = y c + yp , where y has one of the three forms

c

−ct/2m
yc = e (c1 cos ω1 t + c2 sin ω1 t),
−ct/2m
yc = e (c1 + c2 t),
r1 t r2 t
yc = c1 e + c2 e (r1 , r2 < 0).
In all three cases lim t→∞y (t) = 0 for any choice of c and c . For this reason we say that y is the transient component of
c 1 2 c
the solution y . The behavior of y for large t is determined by y , which we call the steady state component of y . Thus, for
p
large t the motion is like simple harmonic motion at the frequency of the external force.
The amplitude of y in Equation 6.2.17 is
p
F0
R = −−−−−−−−−−−−−−−−, (6.2.18)
2 2 2 2 2 2
√ m (ω −ω ) +c ω
0
which is finite for all ω; that is, the presence of damping precludes the phenomenon of resonance that we encountered in
studying undamped vibrations under a periodic forcing function. We’ll now find the value ω of ω for which R is max
maximized. This is the value of ω for which the function

2 2 2 2 2 2
ρ(ω) = m (ω −ω ) +c ω
0
in the denominator of Equation 6.2.18 attains its minimum value. By rewriting this as
2 4 4 2 2 2 2
ρ(ω) = m (ω + ω ) + (c − 2 m ω )ω , (6.2.19)
0 0
you can see that ρ is a strictly increasing function of ω if 2
−−−−−−
2 2
−−−−
c ≥ √2m ω = √2mk.
0
(Recall that ω = k/m ). Therefore ω

2
0 max =0 if this inequality holds. From Equation 6.2.15, you can see that ϕ = 0 if ω = 0 .
In this case, Equation 6.2.14 reduces to
F0 F0
yp = −−− − =
− ,
2 4 k
√m ω
0
which is consistent with Hooke’s law: if the mass is subjected to a constant force F , its displacement should approach a 0
−−−−
constant y such that ky = F . Now suppose c < √2mk . Then, from Equation 6.2.19,
p p 0
′ 2 2 2 2 2
ρ (ω) = 2ω(2 m ω +c − 2 m ω ),
0
and ω max is the value of ω for which the expression in parentheses equals zero; that is,
−−−−−−−− −−−−−−−−−−−−−
2 2
c k c
2
ωmax = √ ω − =√ (1 − ).
0 2
2m m 2km
(To see that ρ(ω max ) is the minimum value of ρ(ω), note that ρ (ω) < 0 if ω < ω and ρ (ω) > 0 if′
max
′
ω > ωmax .)
Substituting ω = ω max in Equation 6.2.18 and simplifying shows that the maximum amplitude R is max
2mF0 −−−−
Rmax = if c < √2mk.
− −−−−−−−
c √ 4mk − c2
We summarize our results as follows.
Theorem 6.2.1
Suppose we consider the amplitude R of the steady state component of the solution of
′′ ′
my + c y + ky = F0 cos ωt
as a function of ω.
−−−−
a. If c ≥ √2mk , the maximum amplitude is R max = F0 /k and it is attained when ω = ω max =0 .

−−−−
b. If c < √2mk , the maximum amplitude is
2mF0
Rmax = − −−−−−−−, (6.2.20)
2
c √ 4mk − c
and it is attained when

−−−−−−−−−−−−−
2
k c
ω = ωmax = √ (1 − ). (6.2.21)
m 2km
Note that Rmax and ω

max are continuous functions of c , for c ≥0 , since Equation 6.2.20 and Equation 6.2.21 reduce to
−−−−
Rmax = F0 /k and ωmax =0 if c = √2km .

6.2E: Spring Problems II (Exercises)
Q6.2.1
1. A 64 lb object stretches a spring 4 ft in equilibrium. It is attached to a dashpot with damping constant c = 8 lb-sec/ft. The
object is initially displaced 18 inches above equilibrium and given a downward velocity of 4 ft/sec. Find its displacement and
time–varying amplitude for t > 0 .
2. A 16 lb weight is attached to a spring with natural length 5 ft. With the weight attached, the spring measures 8.2 ft. The
weight is initially displaced 3 ft below equilibrium and given an upward velocity of 2 ft/sec. Find and graph its displacement
for t > 0 if the medium resists the motion with a force of one lb for each ft/sec of velocity. Also, find its time–varying
amplitude.
3. An 8 lb weight stretches a spring 1.5 inches. It is attached to a dashpot with damping constant c = 8 lb-sec/ft. The weight is
initially displaced 3 inches above equilibrium and given an upward velocity of 6 ft/sec. Find and graph its displacement for
t > 0.
4. A 96 lb weight stretches a spring 3.2 ft in equilibrium. It is attached to a dashpot with damping constant c =18 lb-sec/ft. The
weight is initially displaced 15 inches below equilibrium and given a downward velocity of 12 ft/sec. Find its displacement for
t > 0.
5. A 16 lb weight stretches a spring 6 inches in equilibrium. It is attached to a damping mechanism with constant c . Find all
values of c such that the free vibration of the weight has infinitely many oscillations.
6. An 8 lb weight stretches a spring .32 ft. The weight is initially displaced 6 inches above equilibrium and given an upward
velocity of 4 ft/sec. Find its displacement for t > 0 if the medium exerts a damping force of 1.5 lb for each ft/sec of velocity.
7. A 32 lb weight stretches a spring 2 ft in equilibrium. It is attached to a dashpot with constant c = 8 lb-sec/ft. The weight is
initially displaced 8 inches below equilibrium and released from rest. Find its displacement for t > 0 .
8. A mass of 20 gm stretches a spring 5 cm. The spring is attached to a dashpot with damping constant 400 dyne sec/cm.
Determine the displacement for t > 0 if the mass is initially displaced 9 cm above equilibrium and released from rest.
9. A 64 lb weight is suspended from a spring with constant k = 25 lb/ft. It is initially displaced 18 inches above equilibrium
and released from rest. Find its displacement for t > 0 if the medium resists the motion with 6 lb of force for each ft/sec of
velocity.
10. A 32 lb weight stretches a spring 1 ft in equilibrium. The weight is initially displaced 6 inches above equilibrium and
given a downward velocity of 3 ft/sec. Find its displacement for t > 0 if the medium resists the motion with a force equal to 3
times the speed in ft/sec.
11. An 8 lb weight stretches a spring 2 inches. It is attached to a dashpot with damping constant c = 4 lb-sec/ft. The weight is
initially displaced 3 inches above equilibrium and given a downward velocity of 4 ft/sec. Find its displacement for t > 0 .
12. A 2 lb weight stretches a spring .32 ft. The weight is initially displaced 4 inches below equilibrium and given an upward
velocity of 5 ft/sec. The medium provides damping with constant c = 1/8 lb-sec/ft. Find and graph the displacement for
t > 0.
13. An 8 lb weight stretches a spring 8 inches in equilibrium. It is attached to a dashpot with damping constant c = .5 lb-sec/ft
and subjected to an external force F (t) = 4 cos 2t lb. Determine the steady state component of the displacement for t > 0 .
14. A 32 lb weight stretches a spring 1 ft in equilibrium. It is attached to a dashpot with constant c = 12 lb-sec/ft. The weight
is initially displaced 8 inches above equilibrium and released from rest. Find its displacement for t > 0 .
15. A mass of one kg stretches a spring 49 cm in equilibrium. A dashpot attached to the spring supplies a damping force of 4
N for each m/sec of speed. The mass is initially displaced 10 cm above equilibrium and given a downward velocity of 1 m/sec.
Find its displacement for t > 0 .
16. A mass of 100 grams stretches a spring 98 cm in equilibrium. A dashpot attached to the spring supplies a damping force of
600 dynes for each cm/sec of speed. The mass is initially displaced 10 cm above equilibrium and given a downward velocity

of 1 m/sec. Find its displacement for t > 0 .
17. A 192 lb weight is suspended from a spring with constant k = 6 lb/ft and subjected to an external force F (t) = 8 cos 3t
lb. Find the steady state component of the displacement for t > 0 if the medium resists the motion with a force equal to 8
times the speed in ft/sec.
18. A 2 gm mass is attached to a spring with constant 20 dyne/cm. Find the steady state component of the displacement if the
mass is subjected to an external force F (t) = 3 cos 4t − 5 sin 4t dynes and a dashpot supplies 4 dynes of damping for each
cm/sec of velocity.
19. A 96 lb weight is attached to a spring with constant 12 lb/ft. Find and graph the steady state component of the
displacement if the mass is subjected to an external force F (t) = 18 cos t − 9 sin t lb and a dashpot supplies 24 lb of damping
for each ft/sec of velocity.
20. A mass of one kg stretches a spring 49 cm in equilibrium. It is attached to a dashpot that supplies a damping force of 4 N
for each m/sec of speed. Find the steady state component of its displacement if it is subjected to an external force
F (t) = 8 sin 2t − 6 cos 2t N.
21. A mass m is suspended from a spring with constant k and subjected to an external force F (t) = α cos ω t + β sin ω t , 0 0
where ω is the natural frequency of the spring–mass system without damping. Find the steady state component of the
0
displacement if a dashpot with constant c supplies damping.

22. Show that if c and c are not both zero then
1 2
r1 t
y =e (c1 + c2 t) (6.2E.1)
can’t equal zero for more than one value of t .

23. Show that if c and c are not both zero then
1 2
r1 t r2 t
y = c1 e + c2 e (6.2E.2)
can’t equal zero for more than one value of t .

24. Find the solution of the initial value problem
′′ ′ ′
my + c y + ky = 0, y(0) = y0 , y (0) = v0 , (6.2E.3)
given that the motion is underdamped, so the general solution of the equation is
−ct/2m
y =e (c1 cos ω1 t + c2 sin ω1 t). (6.2E.4)

′′ ′ ′
my + c y + ky = 0, y(0) = y0 , y (0) = v0 , (6.2E.5)
given that the motion is overdamped, so the general solution of the equation is
r1 t r2 t
y = c1 e + c2 e (r1 , r2 < 0). (6.2E.6)

′′ ′ ′
my + c y + ky = 0, y(0) = y0 , y (0) = v0 , (6.2E.7)
given that the motion is critically damped, so that the general solution of the equation is of the form
r1 t
y =e (c1 + c2 t) (r1 < 0). (6.2E.8)

CHAPTER OVERVIEW
7: SERIES SOLUTIONS OF LINEAR SECOND ORDER EQUATIONS
In this Chapter, we study a class of second order differential equations that occur in many
applications, but cannot be solved in closed form in terms of elementary functions.
7.1: PRELUDE TO SERIES SOLUTIONS OF LINEAR SECOND ORDER EQUATIONS

Second order differential equations occur in many applications, but cannot be solved in closed
form in terms of elementary functions including Bessel's, Airy's and Langendre's Equations that
can be written in the form P0(x)y′′+P1(x)y′+P2(x)y=0. These equations do not, in general, have
closed form solutions, we seek series representations for solutions.
7.2: REVIEW OF POWER SERIES

Many applications give rise to differential equations with solutions that can’t be expressed in terms
of elementary functions such as polynomials, rational functions, exponential and logarithmic functions, and trigonometric functions.
The solutions of some of the most important of these equations can be expressed in terms of power series. We’ll study such equations
in this chapter. In this section we review relevant properties of power series.
7.2E: REVIEW OF POWER SERIES (EXERCISES)

7.3: SERIES SOLUTIONS NEAR AN ORDINARY POINT I
This section is devoted to finding power series solutions of (A) in the case where $P_0(0)~\ne~0$. The situation is more complicated
if P_0(0)=0; however, if P_1 and P_2 satisfy assumptions that apply to most equations of interest, then we're able to use a modified
series method to obtain solutions of (A).
7.3E: SERIES SOLUTIONS NEAR AN ORDINARY POINT I (EXERCISES)

7.4: SERIES SOLUTIONS NEAR AN ORDINARY POINT II
In this section we continue to find series solutions of initial value problems . For the equations considered here it is difficult or
impossible to obtain an explicit formula for an in terms of n . Nevertheless, we can calculate as many coefficients as we wish. The
next three examples illustrate this.
7.4E: SERIES SOLUTIONS NEAR AN ORDINARY POINT II (EXERCISES)
1 9/16/2020
7.1: Prelude to Series Solutions of Linear Second Order Equations
In this Chapter, we study a class of second order differential equations that occur in many applications, but cannot be solved in
closed form in terms of elementary functions. Here are some examples:
1. Bessel’s equation which occurs in problems displaying cylindrical symmetry, such as diffraction of light through a circular
aperture, propagation of electromagnetic radiation through a coaxial cable, and vibrations of a circular drum head.
2 ′′ ′ 2 2
x y + x y + (x −ν )y = 0,
2. Airy’s equation which occurs in astronomy and quantum physics.

′′
y − xy = 0,
3. Legendre’s equation which occurs in problems displaying spherical symmetry, particularly in electromagnetism.
2 ′′ ′
(1 − x )y − 2x y + α(α + 1)y = 0,
These equations and others considered in this chapter can be written in the form
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0, (7.1.1)
where P , P , and P are polynomials with no common factor. For most equations that occur in applications, these
0 1 2
polynomials are of degree two or less. We’ll impose this restriction, although the methods that we’ll develop can be extended
to the case where the coefficient functions are polynomials of arbitrary degree, or even power series that converge in some
circle around the origin in the complex plane.
Since Equation 7.1.1 does not in general have closed form solutions, we seek series representations for solutions. We’ll see
that if P (0) ≠ 0 then solutions of (A) can be written as power series
0
∞
n
y = ∑ an x
n=0
that converge in an open interval centered at x = 0 .

7.2: Review of Power Series
Many applications give rise to differential equations with solutions that can’t be expressed in terms of elementary functions
such as polynomials, rational functions, exponential and logarithmic functions, and trigonometric functions. The solutions of
some of the most important of these equations can be expressed in terms of power series. We’ll study such equations in this
chapter. In this section we review relevant properties of power series. We’ll omit proofs, which can be found in any standard
calculus text.
Definition 7.2.1
An infinite series of the form
∞
n
∑ an (x − x0 ) , (7.2.1)
n=0
where x and a , a , …, a , …are constants, is called a power series in x − x

0 0 1 n 0. We say that the power series Equation
7.2.1 converges for a given x if the limit
N
n
lim ∑ an (x − x0 )
N →∞
n=0
exists; otherwise, we say that the power series diverges for the given x.
A power series in x − x must converge if x = x , since the positive powers of x − x are all zero in this case. This may be
0 0 0
the only value of x for which the power series converges. However, the next theorem shows that if the power series converges
for some x ≠ x then the set of all values of x for which it converges forms an interval.
0
Theorem 7.2.2
For any power series
∞
n
∑ an (x − x0 ) ,
n=0
exactly one of the these statements is true:

I. The power series converges only for x = x . 0
II. The power series converges for all values of x.

III. There’s a positive number R such that the power series converges if |x − x 0| <R and diverges if |x − x 0| >R .
In case (iii) we say that R is the radius of convergence of the power series. For convenience, we include the other two cases in
this definition by defining R = 0 in case (i) and R = ∞ in case (ii). We define the open interval of convergence of
∑
∞
n=0
a (x − x )
n 0 to be
n
(x0 − R, x0 + R) if 0 < R < ∞, or (−∞, ∞) if R = ∞.
If R is finite, no general statement can be made concerning convergence at the endpoints x = x 0 ±R of the open interval of
convergence; the series may converge at one or both points, or diverge at both.
Recall from calculus that a series of constants ∑ α is said to converge absolutely if the series of absolute values
∞
n=0 n
| α | converges. It can be shown that a power series ∑ with a positive radius of convergence R
∞ ∞ n
∑ n a (x − x ) n 0
n=0 n=0
converges absolutely in its open interval of convergence; that is, the series
∞
n
∑ | an ||x − x0 |
n=0

of absolute values converges if |x − x 0| <R . However, if R < ∞ , the series may fail to converge absolutely at an endpoint
x ± R , even if it converges here.
0
The next theorem provides a useful method for determining the radius of convergence of a power series. It’s derived in
calculus by applying the ratio test to the corresponding series of absolute values. For related theorems see Exercises 7.2.2 and
7.2.4.
Theorem 7.2.3
Suppose there’s an integer N such that a n ≠0 if n ≥ N and
∣ an+1 ∣
lim ∣ ∣ = L,
n→∞ ∣ an ∣
where 0 ≤ L ≤ ∞. Then the radius of convergence of ∑

∞
n=0
an (x − x0 )
n
is R = 1/L, which should be interpreted to
mean that R = 0 if L = ∞, or R = ∞ if L = 0 .
Example 7.2.1
Find the radius of convergence of the series:
a. ∞
n
∑ n! x
n=0
b. ∞
n
x
n
∑ (−1 )
n!
n=10
c. ∞
n 2 n
∑ 2 n (x − 1 ) .
n=0
Solution a
Here a n = n! , so
∣ an+1 ∣ (n + 1)!
lim ∣ ∣ = lim = lim (n + 1) = ∞.
n→∞ ∣ an ∣ n→∞ n! n→∞
Hence, R = 0 .
Solution b
Here a n
n
= (1 ) /n! for n ≥ N = 10 , so
∣ an+1 ∣ n! 1
lim ∣ ∣ = lim = lim = 0.
n→∞ ∣ an ∣ n→∞ (n + 1)! n→∞ n+1
Hence, R = ∞ .
Solution c
Here a n
n
=2 n
2
, so
n+1 2 2
∣ an+1 ∣ 2 (n + 1 ) 1
lim ∣ ∣ = lim = 2 lim (1 + ) = 2.
n
n→∞ ∣ an ∣ n→∞ 2 n2 n→∞ n
Hence, R = 1/2 .
Taylor Series
If a function f has derivatives of all orders at a point x = x , then the Taylor series of f about x is defined by
0 0

∞ (n)
f (x0 )
n
∑ (x − x0 ) .
n!
n=0
In the special case where x

0 =0 , this series is also called the Maclaurin series of f .
Taylor series for most of the common elementary functions converge to the functions on their open intervals of convergence.
For example, you are probably familiar with the following Maclaurin series:
∞ n
x
x
e =∑ , −∞ < x < ∞ (7.2.2)
n!
n=0
∞ 2n+1
n
x
sin x = ∑(−1 ) , −∞ < x < ∞ (7.2.3)
(2n + 1)!
n=0
∞ 2n
n
x
cos x = ∑(−1 ) −∞ < x < ∞ (7.2.4)
(2n)!
n=0
∞
1 n
= ∑x −1 < x < 1 (7.2.5)
1 −x
n=0
Differentiation of Power Series

A power series with a positive radius of convergence defines a function
∞
n
f (x) = ∑ an (x − x0 )
n=0
on its open interval of convergence. We say that the series represents f on the open interval of convergence. A function f
represented by a power series may be a familiar elementary function as in Equations 7.2.2 - 7.2.5; however, it often happens
that f isn’t a familiar function, so the series actually defines f .
The next theorem shows that a function represented by a power series has derivatives of all orders on the open interval of
convergence of the power series, and provides power series representations of the derivatives.
Theorem 7.2.4 : A power series

A power series
∞
n
f (x) = ∑ an (x − x0 )
n=0
with positive radius of convergence R has derivatives of all orders in its open interval of convergence, and successive
derivatives can be obtained by repeatedly differentiating term by term; that is,
∞
′ n−1
f (x) = ∑ nan (x − x0 ) , (7.2.6)
n=1
∞
′′ n−2
f (x) = ∑ n(n − 1)an (x − x0 ) , (7.2.7)
n=2
⋮
∞
(k) n−k
f (x) = ∑ n(n − 1) ⋯ (n − k + 1)an (x − x0 ) . (7.2.8)
n=k
Moreover, all of these series have the same radius of convergence R.
Example 7.2.2

Let f (x) = sin x . From Equation 7.2.3,
∞ 2n+1
x
n
f (x) = ∑(−1 ) .
(2n + 1)!
n=0

∞ 2n+1 ∞ 2n
′ n
d x n
x
f (x) = ∑(−1 ) [ ] = ∑(−1 ) ,
dx (2n + 1)! (2n)!
n=0 n=0
which is the series Equation 7.2.4 for cos x.
Uniqueness of Power Series

The next theorem shows that if f is defined by a power series in x − x with a positive radius of convergence, then the power
0
series is the Taylor series of f about x . 0
Theorem 7.2.5
If the power series
∞
n
f (x) = ∑ an (x − x0 )
n=0
has a positive radius of convergence, then

(n)
f (x0 )
an = ; (7.2.9)
n!
that is, ∑ ∞
n=0
n
an (x − x0 ) is the Taylor series of f about x . 0
This result can be obtained by setting x = x in Equation 7.2.8, which yields

0
(k)
f (x0 ) = k(k − 1) ⋯ 1 ⋅ ak = k! ak .
This implies that

(k)
f (x0 )
ak = .
k!
Except for notation, this is the same as Equation 7.2.9.

The next theorem lists two important properties of power series that follow from Theorem 7.2.4.
Theorem 7.2.6
a. If
∞ ∞
n n
∑ an (x − x0 ) = ∑ bn (x − x0 )
n=0 n=0
for all x in an open interval that contains x , then a

0 n = bn for n = 0 , 1, 2, ….
b. If
∞
n
∑ an (x − x0 ) =0
n=0
for all x in an open interval that contains x , then a

0 n =0 for n = 0 , 1, 2, ….

To obtain (a) we observe that the two series represent the same function f on the open interval; hence, Theorem 7.2.4 implies
that
(n)
f (x0 )
an = bn = , n = 0, 1, 2, … .
n!
(b) can be obtained from (a) by taking b n =0 for n = 0 , 1, 2, ….
Taylor Polynomials
If f has N derivatives at a point x , we say that
0
N (n)
f (x0 )
n
TN (x) = ∑ (x − x0 )
n!
n=0
is the N -th Taylor polynomial of f about x . This definition and Theorem 7.2.4 imply that if
0
n
f (x) = ∑ an (x − x0 ) ,
n=0
where the power series has a positive radius of convergence, then the Taylor polynomials of f about x are given by 0
n
TN (x) = ∑ an (x − x0 ) .
n=0
In numerical applications, we use the Taylor polynomials to approximate f on subintervals of the open interval of convergence
of the power series. For example, Equation 7.2.2 implies that the Taylor polynomial T of f (x) = e is N
x
N n
x
TN (x) = ∑ .
n!
n=0
The solid curve in Figure 7.2.1 is the graph of y = e on the interval [0, 5]. The dotted curves in Figure 7.2.1 are the graphs of
x
the Taylor polynomials T , …, T of y = e about x = 0 . From this figure, we conclude that the accuracy of the
1 6
x
0
approximation of y = e by its Taylor polynomial T improves as N increases.

x
N
Figure 7.2.1 : Approximation of y = e by Taylor polynomials about x = 0

x
Shifting the Summation Index

In Definition 7.2.1 of a power series in x − x , the n -th term is a constant multiple of (x − x ) . This isn’t true in Equation
0 0
n
7.2.6, Equation 7.2.7, and Equation 7.2.8, where the general terms are constant multiples of (x − x ) , (x − x ) , and n−1 n−2
0 0
(x − x )
0 , respectively. However, these series can all be rewritten so that their n -th terms are constant multiples of
n−k
(x − x ) . For example, letting n = k + 1 in the series in Equation 7.2.6 yields

n
0

∞
′ k
f (x) = ∑(k + 1)ak+1 (x − x0 ) , (7.2.10)
k=0
where we start the new summation index k from zero so that the first term in Equation 7.2.10 (obtained by setting k = 0 ) is
the same as the first term in Equation 7.2.6 (obtained by setting n = 1 ). However, the sum of a series is independent of the
symbol used to denote the summation index, just as the value of a definite integral is independent of the symbol used to denote
the variable of integration. Therefore we can replace k by n in Equation 7.2.10 to obtain
∞
′ n
f (x) = ∑(n + 1)an+1 (x − x0 ) , (7.2.11)
n=0
where the general term is a constant multiple of (x − x 0)

n
.
It isn’t really necessary to introduce the intermediate summation index k . We can obtain Equation 7.2.11 directly from
Equation 7.2.6 by replacing n by n + 1 in the general term of Equation 7.2.6 and subtracting 1 from the lower limit of
Equation 7.2.6. More generally, we use the following procedure for shifting indices.
Shifting the Summation Index in a Power Series

For any integer k , the power series
∞
n−k
∑ bn (x − x0 )
n=n0
can be rewritten as
∞
n
∑ bn+k (x − x0 )
n=n0 −k
that is, replacing n by n+k in the general term and subtracting k from the lower limit of summation leaves the series
unchanged.
Example 7.2.3
Rewrite the following power series from Equation 7.2.7 and Equation 7.2.8 so that the general term in each is a constant
multiple of (x − x ) :0
n
∞ ∞
n−2 n−k
(a) ∑ n(n − 1)an (x − x0 ) (b) ∑ n(n − 1) ⋯ (n − k + 1)an (x − x0 ) .
n=2 n=k
Solution a
Replacing n by n + 2 in the general term and subtracting 2 from the lower limit of summation yields
∞ ∞
n−2 n
∑ n(n − 1)an (x − x0 ) = ∑(n + 2)(n + 1)an+2 (x − x0 ) .
n=2 n=0
Solution b
Replacing n by n + k in the general term and subtracting k from the lower limit of summation yields
∞ ∞
n−k n
∑ n(n − 1) ⋯ (n − k + 1)an (x − x0 ) = ∑(n + k)(n + k − 1) ⋯ (n + 1)an+k (x − x0 ) .
n=k n=0
Example 7.2.4
Given that

∞
n
f (x) = ∑ an x ,
n=0
write the function xf as a power series in which the general term is a constant multiple of x .
′′ n
Solution
From Theorem 7.2.4 with x 0 =0 ,
∞
′′ n−2
f (x) = ∑ n(n − 1)an x .
n=2
Therefore
∞
′′ n−1
xf (x) = ∑ n(n − 1)an x .
n=2
Replacing n by n + 1 in the general term and subtracting 1 from the lower limit of summation yields
∞
′′ n
xf (x) = ∑(n + 1)nan+1 x .
n=1
We can also write this as

∞
′′ n
xf (x) = ∑(n + 1)nan+1 x ,
n=0
since the first term in this last series is zero. (We’ll see later that sometimes it is useful to include zero terms at the
beginning of a series.)
Linear Combinations of Power Series

If a power series is multiplied by a constant, then the constant can be placed inside the summation; that is,
∞ ∞
n n
c ∑ an (x − x0 ) = ∑ c an (x − x0 ) .
n=0 n=0
Two power series

∞ ∞
n n
f (x) = ∑ an (x − x0 ) and g(x) = ∑ bn (x − x0 )
n=0 n=0
with positive radii of convergence can be added term by term at points common to their open intervals of convergence; thus, if
the first series converges for |x − x | < R and the second converges for |x − x | < R , then
0 1 0 2
n
f (x) + g(x) = ∑(an + bn )(x − x0 )
n=0
for |x − x | < R , where R is the smaller of R and R . More generally, linear combinations of power series can be formed
0 1 2
term by term; for example,

∞
n
c1 f (x) + c2 f (x) = ∑(c1 an + c2 bn )(x − x0 ) .
n=0
Example 7.2.5
Find the Maclaurin series for cosh x as a linear combination of the Maclaurin series for e and e x −x
.
Solution

By definition,
1 x
1 −x
cosh x = e + e .
2 2
Since
∞ n ∞ n
x
x −x n
x
e =∑ and e = ∑(−1 ) ,
n! n!
n=0 n=0
it follows that
∞ n
1 x
n
cosh x = ∑ [1 + (−1 ) ] . (7.2.12)
2 n!
n=0
Since
1 1 if n = 2m, an even integer,
n
[1 + (−1 ) ] = {
2 0 if n = 2m + 1, an odd integer,
we can rewrite Equation 7.2.12 more simply as

∞ 2m
x
cosh x = ∑ .
(2m)!
m=0
This result is valid on (−∞, ∞), since this is the open interval of convergence of the Maclaurin series for e and e x −x
.
Example 7.2.6
Suppose
∞
n
y = ∑ an x
n=0
on an open interval I that contains the origin.

a. Express
′′
(2 − x)y + 2y
as a power series in x on I .
b. Use the result of (a) to find necessary and sufficient conditions on the coefficients { an } for y to be a solution of the
homogeneous equation
′′
(2 − x)y + 2y = 0 (7.2.13)
on I .
Solution a
From Equation 7.2.7 with x 0 =0 ,
∞
′′ n−2
y = ∑ n(n − 1)an x .
n=2
Therefore
′′ ′′ ′
(2 − x)y + 2y = 2y − x y + 2y
∞ ∞ ∞
(7.2.14)
n−2 n−1 n
= ∑n=2 2n(n − 1)an x − ∑n=2 n(n − 1)an x + ∑n=0 2 an x .
To combine the three series we shift indices in the first two to make their general terms constant multiples of x ; thus, n

∞ ∞
n−2 n
∑ 2n(n − 1)an x = ∑ 2(n + 2)(n + 1)an+2 x (7.2.15)
n=2 n=0
and
∞ ∞ ∞
n−1 n n
∑ n(n − 1)an x = ∑(n + 1)nan+1 x = ∑(n + 1)nan+1 x , (7.2.16)
n=2 n=1 n=0
where we added a zero term in the last series so that when we substitute from Equation 7.2.15 and Equation 7.2.16 into
Equation 7.2.14 all three series will start with n = 0 ; thus,
∞
′′ n
(2 − x)y + 2y = ∑[2(n + 2)(n + 1)an+2 − (n + 1)nan+1 + 2 an ] x . (7.2.17)
n=0
Solution b
From Equation 7.2.17 we see that y satisfies Equation 7.2.13 on I if
2(n + 2)(n + 1)an+2 − (n + 1)nan+1 + 2 an = 0, n = 0, 1, 2, … . (7.2.18)
Conversely, Theorem 7.2.5b implies that if y = ∑ ∞
n=0
an x
n
satisfies Equation 7.2.13 on I , then Equation 7.2.18 holds.
Example 7.2.7
Suppose
∞
n
y = ∑ an (x − 1 )
n=0
on an open interval I that contains x 0 =1 . Express the function

′′ 2 ′
(1 + x)y + 2(x − 1 ) y + 3y (7.2.19)
as a power series in x − 1 on I .
Solution
Since we want a power series in x −1 , we rewrite the coefficient of y
′′
in Equation 7.2.19 as 1 + x = 2 + (x − 1) , so
′′ ′′ 2 ′
2y + (x − 1)y + 2(x − 1 ) y + 3y.
From Equation 7.2.6 and Equation 7.2.7 with x 0 =1 ,

∞ ∞
′ n−1 ′′ n−2
y = ∑ nan (x − 1 ) and y = ∑ n(n − 1)an (x − 1 ) .
n=1 n=2
Therefore
∞
′′ n−2
2y = ∑ 2n(n − 1)an (x − 1 ) ,
n=2
′′ n−1
(x − 1)y = ∑ n(n − 1)an (x − 1 ) ,
n=2
2 ′ n+1
2(x − 1) y = ∑ 2nan (x − 1 ) ,
n=1
n
3y = ∑ 3 an (x − 1 ) .
n=0

Before adding these four series we shift indices in the first three so that their general terms become constant multiples of
(x − 1) . This yields
n
′′ n
2y = ∑ 2(n + 2)(n + 1)an+2 (x − 1 ) , (7.2.20)
n=0
′′ n
(x − 1)y = ∑(n + 1)nan+1 (x − 1 ) , (7.2.21)
n=0
2 ′ n
2(x − 1) y = ∑ 2(n − 1)an−1 (x − 1 ) , (7.2.22)
n=1
n
3y = ∑ 3 an (x − 1 ) , (7.2.23)
n=0
where we added initial zero terms to the series in Equation 7.2.21 and Equation 7.2.22 . Adding Equation 7.2.20 –
′′ 2 ′ ′′ ′′ 2 ′
(1 + x)y + 2(x − 1 ) y + 3y = 2y + (x − 1)y + 2(x − 1 ) y + 3y
∞
n
= ∑ bn (x − 1 ) ,
n=0
where
b0 = 4 a2 + 3 a0 , (7.2.24)
bn = 2(n + 2)(n + 1)an+2 + (n + 1)nan+1 + 2(n − 1)an−1 + 3 an , n ≥ 1. (7.2.25)
The formula Equation 7.2.24 for b can’t be obtained by setting n = 0 in Equation 7.2.25, since the summation in
0
Equation 7.2.22 begins with n = 1 , while those in Equation 7.2.20, Equation 7.2.21, and Equation 7.2.23 begin with
n = 0.

7.2E: Review of Power Series (Exercises)
Q7.1.1
1.For each power series use Theorem 7.1.3 to find the radius of convergence R . If R >0 , find the open interval of
convergence.
∞ n
(−1)
a. ∑
n
(x − 1 )
n
2 n
n=0
∞
b. ∑ 2 n
n(x − 2 )
n
n=0
∞
n!
c. ∑
n
x
n
9
n=0
∞
n(n + 1)
d. ∑ n
(x − 2 )
n
16
n=0
∞ n
7
e. ∑(−1 )
n
x
n
n!
n=0
∞ n
3
f. ∑ n+1
(x + 7 )
n
2
n=0
4 (n + 1 )
2. Suppose there’s an integer M such that b m ≠0 for m ≥ M , and

∣ bm+1 ∣
lim ∣ ∣ = L,
m→∞ ∣ bm ∣
where 0 ≤ L ≤ ∞ . Show that the radius of convergence of

∞
2m
∑ bm (x − x0 )
m=0
−
−
is R = 1/ √L , which is interpreted to mean that R = 0 if L = ∞ or R = ∞ if L = 0 .
3. For each power series, use the result of Exercise 7.1.2 to find the radius of convergence R . If R > 0 , find the open interval
of convergence.
∞
a. ∑ (−1 )
m
(3m + 1)(x − 1 )
2m+1
m=0
∞
m(2m + 1)
b. ∑ (−1) m
m
(x + 2 )
2m
2
m=0
∞
m!
c. ∑ (x − 1 )
2m
(2m)!
m=0
∞
m!
d. ∑ (−1) m
m
(x + 8 )
2m
9
m=0
∞
(2m − 1)
e. ∑ (−1 )
m
m
x
2m+1
3
m=0
∞
f. ∑ (x − 1) 2m
m=0
4. Suppose there’s an integer M such that b m ≠0 for m ≥ M , and

∣ bm+1 ∣
lim ∣ ∣ = L,
m→∞ ∣ bm ∣
where 0 ≤ L ≤ ∞ . Let k be a positive integer. Show that the radius of convergence of

∞
km
∑ bm (x − x0 )
m=0
−
−
is R = 1/ √L
k
, which is interpreted to mean that R = 0 if L = ∞ or R = ∞ if L = 0 .
5. For each power series use the result of Exercise 7.1.4 to find the radius of convergence R . If R > 0 , find the open interval
of convergence.
∞ m
(−1)
a. ∑ (x − 3 )
3m+2
(27)m
m=0
∞ 7m+6
x
b. ∑
m
m=0
∞ m
9 (m + 1)
c. ∑ (x − 3 )
4m+2
(m + 2)
m=0
∞ m
2
d. ∑ (−1) m
x
4m+3
m!
m=0
∞
m!
e. ∑ (x + 1 )
4m+3
(26)m
m=0
∞ m
(−1)
f. ∑ m
(x − 1 )
3m+1
8 m(m + 1)
m=0
6. Graph y = sin x and the Taylor polynomial

M n 2n+1
(−1) x
T2M+1 (x) = ∑
(2n + 1)!
n=0
on the interval (−2π, 2π) for M =1 , 2, 3, …, until you find a value of M for which there’s no perceptible difference between
the two graphs.
7. Graph y = cos x and the Taylor polynomial
M n 2n
(−1) x
T2M (x) = ∑
(2n)!
n=0
on the interval (−2π, 2π) for M =1 , 2, 3, …, until you find a value of M for which there’s no perceptible difference between
the two graphs.
8. Graph y = 1/(1 − x) and the Taylor polynomial
N
n
TN (x) = ∑ x
n=0
on the interval [0, .95] for N = 1 , 2, 3, …, until you find a value of N for which there’s no perceptible difference between the
two graphs. Choose the scale on the y -axis so that 0 ≤ y ≤ 20 .
9. Graph y = cosh x and the Taylor polynomial
M 2n
x
T2M (x) = ∑
(2n)!
n=0
on the interval (−5, 5) for M = 1 , 2, 3, …, until you find a value of M for which there’s no perceptible difference between
the two graphs. Choose the scale on the y -axis so that 0 ≤ y ≤ 75 .
10. Graph y = sinh x and the Taylor polynomial
M 2n+1
x
T2M+1 (x) = ∑
(2n + 1)!
n=0

on the interval (−5, 5) for M = 0 , 1, 2, …, until you find a value of M for which there’s no perceptible difference between
the two graphs. Choose the scale on the y -axis so that −75 ≤ y ≤ 75 .
Q7.1.2
In Exercises 7.1.11-7.1.15 find a power series solution \(y(x)=\sum_{n=0}^{\infty} a_{n}x^{n}\].
11. (2 + x)y ′′
+ x y + 3y
′
12. (1 + 3x 2
)y
′′ 2
+ 3 x y − 2y
′
13. (1 + 2x 2
)y
′′
+ (2 − 3x)y + 4y
′
14. (1 + x 2
)y
′′
+ (2 − x)y + 3y
′
15. (1 + 3x 2
)y
′′ ′
− 2x y + 4y
Q7.1.3
∞
16. Suppose y(x) = ∑ a n (x + 1)

n
on an open interval that contains x 0 = −1 . Find a power series in x + 1 for
n=0
′′ ′
xy + (4 + 2x)y + (2 + x)y.
17. Suppose y(x) = ∑ a n (x − 2)

n
on an open interval that contains x 0 = 2 . Find a power series in x − 2 for
n=0
2 ′′ ′
x y + 2x y − 3xy.
18. Do the following experiment for various choices of real numbers a and a . 0 1
a. Use differential equations software to solve the initial value problem

′′ ′
(2 − x)y + 2y = 0, y(0) = a0 , y (0) = a1 ,
numerically on (−1.95, 1.95). Choose the most accurate method your software package provides. (See Section 10.1 for a
brief discussion of one such method.)
b. For N = 2 , 3, 4, …, compute a , …, a from Equation 7.1.18 and graph
2 N
n
TN (x) = ∑ an x
n=0
and the solution obtained in (a) on the same axes. Continue increasing N until it is obvious that there’s no point in
continuing. (This sounds vague, but you’ll know when to stop.)
19. Follow the directions of Exercise 7.1.18 for the initial value problem
′′ 2 ′ ′
(1 + x)y + 2(x − 1 ) y + 3y = 0, y(1) = a0 , y (1) = a1 ,
on the interval (0, 2). Use Equations 7.1.24 and 7.1.25 to compute {a n} .
∞
20. Suppose the series ∑ a nx

n
converges on an open interval (−R, R) , let r be an arbitrary real number, and define
n=0
∞ ∞
r n n+r
y(x) = x ∑ an x = ∑ an x
n=0 n=0
on (0, R). Use Theorem 7.1.4 and the rule for differentiating the product of two functions to show that

∞
′ n+r−1
y (x) = ∑(n + r)an x ,
n=0
′′ n+r−2
y (x) = ∑(n + r)(n + r − 1)an x ,
n=0
⋮
∞
(k) n+r−k
y (x) = ∑(n + r)(n + r − 1) ⋯ (n + r − k)an x
n=0
on (0, R)
Q7.1.4
21. x2
(1 − x)y
′′ ′
+ x(4 + x)y + (2 − x)y
22. x2
(1 + x)y
′′ ′
+ x(1 + 2x)y − (4 + 6x)y
23. x2
(1 + x)y
′′ 2 ′
− x(1 − 6x − x )y + (1 + 6x + x )y
2
24. x2
(1 + 3x)y
′′ 2
+ x(2 + 12x + x )y + 2x(3 + x)y
′
25. x2 2
(1 + 2 x )y
′′ 2 ′
+ x(4 + 2 x )y + 2(1 − x )y
2
26. x2 2
(2 + x )y
′′ 2 ′
+ 2x(5 + x )y + 2(3 − x )y
2

7.3: Series Solutions Near an Ordinary Point I
Many physical applications give rise to second order homogeneous linear differential equations of the form
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0, (7.3.1)
where P , P , and P are polynomials. Usually the solutions of these equations can’t be expressed in terms of familiar
0 1 2
elementary functions. Therefore we’ll consider the problem of representing solutions of Equation 7.3.1 with series.
We assume throughout that P , P and P have no common factors. Then we say that
0 1 2 x0 is an ordinary point of Equation
7.3.1 if P (x ) ≠ 0 , or a singular point if P (x ) = 0 . For Legendre’s equation,
0 0 0 0
2 ′′ ′
(1 − x )y − 2x y + α(α + 1)y = 0, (7.3.2)
x0 = 1 and x 0 = −1 are singular points and all other points are ordinary points. For Bessel’s equation,
2 ′′ ′ 2 2
x y + x y + (x −ν )y = 0,
x0 = 0 is a singular point and all other points are ordinary points. If P is a nonzero constant as in Airy’s equation, 0
′′
y − xy = 0, (7.3.3)
then every point is an ordinary point.

Since polynomials are continuous everywhere, P /P and P /P are continuous at any point x that isn’t a zero of P .
1 0 2 0 0 0
Therefore, if x is an ordinary point of Equation 7.3.1 and a and a are arbitrary real numbers, then the initial value problem
0 0 1
′′ ′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0, y(x0 ) = a0 , y (x0 ) = a1 (7.3.4)
has a unique solution on the largest open interval that contains x and does not contain any zeros of P . To see this, we rewrite
0 0
the differential equation in Equation 7.3.4 as

P1 (x) P2 (x)
′′ ′
y + y + y =0
P0 (x) P0 (x)
and apply Theorem 5.1.1 with p = P /P and q = P /P . In this section and the next we consider the problem of
1 0 2 0
representing solutions of Equation 7.3.1 by power series that converge for values of x near an ordinary point x . 0
We state the next theorem without proof.
Theorem 7.3.1
Suppose P , P , and P are polynomials with no common factor and P isn’t identically zero. Let x be a point such that
0 1 2 0 0
P (x ) ≠ 0, and let ρ be the distance from x to the nearest zero of P

0 0 in the complex plane. ( If P is constant, then
0 0 0
ρ = ∞ .) Then every solution of
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0 (7.3.5)
can be represented by a power series

∞
n
y = ∑ an (x − x0 ) (7.3.6)
n=0
that converges at least on the open interval (x − ρ, x + ρ) . ( If P is nonconstant, so that ρ is necessarily finite, then
0 0 0
the open interval of convergence of (7.3.6) may be larger than (x − ρ, x + ρ). If P is constant then ρ = ∞ and 0 0 0
(x − ρ, x + ρ) = (−∞, ∞) .
0 0
We call Equation 7.3.6 a power series solution in x − x of Equation 7.3.5. We’ll now develop a method for finding power
0
series solutions of Equation 7.3.5. For this purpose we write Equation 7.3.5 as Ly = 0 , where
′′ ′
Ly = P0 y + P1 y + P2 y. (7.3.7)

Theorem 7.3.1 implies that every solution of Ly = 0 on (x 0 − ρ, x0 + ρ) can be written as
∞
n
y = ∑ an (x − x0 ) .
n=0
Setting x = x in this series and in the series

0
′ n−1
y = ∑ nan (x − x0 )
n=1
shows that y(x ) = a and y (x ) = a . Since every initial value problem Equation 7.3.4 has a unique solution, this means
0 0
′
0 1
that a and a can be chosen arbitrarily, and a , a , …are uniquely determined by them.
0 1 2 3
To find a , a , …, we write P , P , and P in powers of x − x , substitute

2 3 0 1 2 0
n
y = ∑ an (x − x0 ) ,
n=0
′ n−1
y = ∑ nan (x − x0 ) ,
n=1
′′ n−2
y = ∑ n(n − 1)an (x − x0 )
n=2
into Equation 7.3.7, and collect the coefficients of like powers of x − x . This yields 0
n
Ly = ∑ bn (x − x0 ) , (7.3.8)
n=0
where {b , b , … , b , …} are expressed in terms of {a , a , … , a , …} and the coefficients of P , P , and P , written in

0 1 n 0 1 n 0 1 2
powers of x − x . Since Equation 7.3.8 and (a) of Theorem 7.1.6 imply that Ly = 0 if and only if b = 0 for n ≥ 0 , all
0 n
power series solutions in x − x of Ly = 0 can be obtained by choosing a and a arbitrarily and computing a , a , …,
0 0 1 2 3
successively so that b = 0 for n ≥ 0 . For simplicity, we call the power series obtained this way the power series in x − x
n 0
for the general solution of Ly = 0 , without explicitly identifying the open interval of convergence of the series.
Example 7.3.1
Let x be an arbitrary real number. Find the power series in x − x for the general solution of
0 0
′′
y + y = 0. (7.3.9)
Solution
Here
′′
Ly = y + y.
If
∞
n
y = ∑ an (x − x0 ) ,
n=0
then
∞
′′ n−2
y = ∑ n(n − 1)an (x − x0 ) ,
n=2
so

∞ ∞
n−2 n
Ly = ∑ n(n − 1)an (x − x0 ) + ∑ an (x − x0 ) .
n=2 n=0
To collect coefficients of like powers of x − x , we shift the summation index in the first sum. This yields
0
∞ ∞ ∞
n n n
Ly = ∑(n + 2)(n + 1)an+2 (x − x0 ) + ∑ an (x − x0 ) = ∑ bn (x − x0 ) ,
n=0 n=0 n=0
with
bn = (n + 2)(n + 1)an+2 + an .
Therefore Ly = 0 if and only if

−an
an+2 = , n ≥ 0, (7.3.10)
(n + 2)(n + 1)
where a and a are arbitrary. Since the indices on the left and right sides of Equation
0 1 7.3.10 differ by two, we write
Equation 7.3.10 separately for n even (n = 2m) and n odd (n = 2m + 1) . This yields
−a2m
a2m+2 = , m ≥ 0, (7.3.11)
(2m + 2)(2m + 1)
and
−a2m+1
a2m+3 = , m ≥ 0. (7.3.12)
(2m + 3)(2m + 2)
Computing the coefficients of the even powers of x − x from Equation 7.3.11 yields
0
a0
a2 = −
2⋅1
a2 1 a0 a0
a4 = − =− (− ) = ,
4⋅3 4⋅3 2⋅1 4⋅3⋅2⋅1
a4 1 a0 a0
a6 = − =− ( ) =− ,
6⋅5 6⋅5 4⋅3⋅2⋅1 6⋅5⋅4⋅3⋅2⋅1
and, in general,
m
a0
a2m = (−1 ) , m ≥ 0. (7.3.13)
(2m)!
Computing the coefficients of the odd powers of x − x from Equation 7.3.12 yields
0
a1
a3 = −
3⋅2
a3 1 a1 a1
a5 = − =− (− ) = ,
5⋅4 5⋅4 3⋅2 5⋅4⋅3⋅2
a5 1 a1 a1
a7 = − =− ( ) =− ,
7⋅6 7⋅6 5⋅4⋅3⋅2 7⋅6⋅5⋅4⋅3⋅2
and, in general,
m
(−1) a1
a2m+1 = m ≥ 0. (7.3.14)
(2m + 1)!
Thus, the general solution of Equation 7.3.9 can be written as

∞ ∞
2m 2m+1
y = ∑ a2m (x − x0 ) + ∑ a2m+1 (x − x0 ) ,
m=0 m=0
or, from Equation 7.3.13 and Equation 7.3.14, as

∞ 2m ∞ 2m+1
(x − x0 ) (x − x0 )
m m
y = a0 ∑ (−1 ) + a1 ∑ (−1 ) . (7.3.15)
(2m)! (2m + 1)!
m=0 m=0
If we recall from calculus that

∞ 2m ∞ 2m+1
(x − x0 ) (x − x0 )
m m
∑ (−1 ) = cos(x − x0 ) and ∑ (−1 ) = sin(x − x0 ),
(2m)! (2m + 1)!
m=0 m=0
then Equation 7.3.15 becomes
y = a0 cos(x − x0 ) + a1 sin(x − x0 ),
which should look familiar.
Equations like Equations 7.3.10, 7.3.11, and 7.3.12, which define a given coefficient in the sequence {a } in terms of one or n
more coefficients with lesser indices are called recurrence relations. When we use a recurrence relation to compute terms of a
sequence we are computing recursively.
In the remainder of this section we consider the problem of finding power series solutions in x − x for equations of the form 0
2 ′′ ′
(1 + α(x − x0 ) ) y + β(x − x0 )y + γy = 0. (7.3.16)
Many important equations that arise in applications are of this form with x0 = 0 , including Legendre’s equation Equation
7.3.2, Airy’s equation Equation 7.3.3, Chebyshev’s equation,
2 ′′ ′ 2
(1 − x )y − x y + α y = 0,
and Hermite’s equation,

′′ ′
y − 2x y + 2αy = 0.
Since
2
P0 (x) = 1 + α(x − x0 )
in Equation 7.3.16, the point x is an ordinary point of Equation 7.3.16, and Theorem 7.3.1 implies that the solutions of
0
Equation 7.3.16 can be written as power series in x − x that converge on the interval (x − 1/√|α|, x + 1/√|α|) if
0 0 0
α ≠ 0 , or on (−∞, ∞) if α = 0 . We’ll see that the coefficients in these power series can be obtained by methods similar to
the one used in Example 7.3.1.

To simplify finding the coefficients, we introduce some notation for products:
s
∏ bj = br br+1 ⋯ bs if s ≥ r.
j=r
Thus,
7
∏ bj = b2 b3 b4 b5 b6 b7 ,
j=2
∏(2j + 1) = (1)(3)(5)(7)(9) = 945,
j=0
and
2
2 2
∏j =2 = 4.
j=2
We define

s
∏ bj = 1 if s < r,
j=r
no matter what the form of b .

j
Example 7.3.2
Find the power series in x for the general solution of
2 ′′ ′
(1 + 2 x )y + 6x y + 2y = 0. (7.3.17)
Solution
Here
2 ′′ ′
Ly = (1 + 2 x )y + 6x y + 2y.
If
∞
n
y = ∑ an x
n=0
then
∞ ∞
′ n−1 ′′ n−2
y = ∑ nan x and y = ∑ n(n − 1)an x ,
n=1 n=2
so
∞ ∞ ∞
2 n−2 n−1 n
Ly = (1 + 2 x ) ∑ n(n − 1)an x + 6x ∑ nan x + 2 ∑ an x
n=2 n=1 n=0
∞ ∞
n−2 n
= ∑ n(n − 1)an x + ∑ [2n(n − 1) + 6n + 2] an x
n=2 n=0
∞ ∞
n−2 2 n
= ∑ n(n − 1)an x + 2 ∑(n + 1 ) an x .
n=2 n=0
To collect coefficients of x , we shift the summation index in the first sum. This yields
n
∞ ∞ ∞
n 2 n n
Ly = ∑(n + 2)(n + 1)an+2 x + 2 ∑(n + 1 ) an x = ∑ bn x ,
n=0 n=0 n=0
with
2
bn = (n + 2)(n + 1)an+2 + 2(n + 1 ) an , n ≥ 0.
To obtain solutions of Equation 7.3.17, we set b n =0 for n ≥ 0 . This is equivalent to the recurrence relation
n+1
an+2 = −2 an , n ≥ 0. (7.3.18)
n+2
Since the indices on the left and right differ by two, we write Equation 7.3.18separately for n = 2m and n = 2m + 1 , as
in Example 7.3.1. This yields
2m + 1 2m + 1
a2m+2 = −2 a2m = − a2m , m ≥ 0, (7.3.19)
2m + 2 m +1
and
2m + 2 m +1
a2m+3 = −2 a2m+1 = −4 a2m+1 , m ≥ 0. (7.3.20)
2m + 3 2m + 3
Computing the coefficients of even powers of x from Equation 7.3.19yields

1
a2 = − a0 ,
1
3 3 1 1⋅3
a4 = − a2 = (− ) (− ) a0 = a0 ,
2 2 1 1⋅2
5 5 1⋅3 1⋅3⋅5
a6 = − a4 = − ( ) a0 = − a0 ,
3 3 1⋅2 1⋅2⋅3
7 7 1⋅3⋅5 1⋅3⋅5⋅7
a8 = − a6 = − (− ) a0 = a0 .
4 4 1⋅2⋅3 1⋅2⋅3⋅4
In general,
m
∏ (2j − 1)
j=1
m
a2m = (−1 ) a0 , m ≥ 0. (7.3.21)
m!
0
(Note that Equation 7.3.21is correct for m = 0 because we defined ∏ j=1
bj = 1 for any b .) j
Computing the coefficients of odd powers of x from Equation 7.3.20 yields

1
a3 = −4 a1 ,
3
2 2 1 1⋅2
2
a5 = −4 a3 = −4 (−4 ) a1 = 4 a1 ,
5 5 3 3⋅5
3 3 1⋅2 1⋅2⋅3
2 3
a7 = −4 a5 = −4 (4 ) a1 = −4 a1 ,
7 7 3⋅5 3⋅5⋅7
4 4 3
1⋅2⋅3 4
1⋅2⋅3⋅4
a9 = −4 a7 = −4 (4 ) a1 = 4 a1 .
9 9 3⋅5⋅7 3⋅5⋅7⋅9
In general,
m m
(−1 ) 4 m!
a2m+1 = a1 , m ≥ 0. (7.3.22)
m
∏ (2j + 1)
j=1
From Equation 7.3.21and Equation 7.3.22,

m
∞ ∞ m
∏ (2j − 1) 4 m!
j=1
m 2m m 2m+1
y = a0 ∑ (−1 ) x + a1 ∑ (−1 ) x .
m
m! ∏ (2j + 1)
m=0 m=0 j=1
is the power series in x for the general solution of Equation 7.3.17. Since P (x) = 1 + 2x has no real zeros, Theorem 0
2
–
5.1.1 implies that every solution of Equation 7.3.17 is defined on (−∞, ∞). However, since P (±i/√2) = 0 , Theorem 0
– –
7.3.1 implies only that the power series converges in (−1/ √2, 1/ √2) for any choice of a and a . 0 1
The results in Examples 7.3.1 and 7.3.2 are consequences of the following general theorem.
Theorem 7.3.2
The coefficients {a n} in any solution y = ∑ ∞
n=0
an (x − x0 )
n
of
2 ′′ ′
(1 + α(x − x0 ) ) y + β(x − x0 )y + γy = 0 (7.3.23)
satisfy the recurrence relation

p(n)
an+2 = − an , n ≥ 0, (7.3.24)
(n + 2)(n + 1)
where
p(n) = αn(n − 1) + βn + γ. (7.3.25)

Moreover, the coefficients of the even and odd powers of x − x can be computed separately as
0
p(2m)
a2m+2 = − a2m , m ≥0 (7.3.26)
(2m + 2)(2m + 1)
p(2m + 1)
a2m+3 = − a2m+1 , m ≥ 0, (7.3.27)
(2m + 3)(2m + 2)
where a and a are arbitrary.

0 1
Proof
Here
2 ′′ ′
Ly = (1 + α(x − x0 ) )y + β(x − x0 )y + γy.
If
∞
n
y = ∑ an (x − x0 ) ,
n=0
then
∞ ∞
′ n−1 ′′ n−2
y = ∑ nan (x − x0 ) and y = ∑ n(n − 1)an (x − x0 ) .
n=1 n=2
Hence,
∞ n−2 ∞ n
Ly =∑ n(n − 1)an (x − x0 ) +∑ [αn(n − 1) + βn + γ] an (x − x0 )
n=2 n=0
∞ n−2 ∞ n
=∑ n(n − 1)an (x − x0 ) +∑ p(n)an (x − x0 ) ,
n=2 n=0
from Equation 7.3.25. To collect coefficients of powers of x − x , we shift the summation index in the first sum. This
0
yields
∞
n
Ly = ∑ [(n + 2)(n + 1)an+2 + p(n)an ] (x − x0 ) .
n=0
Thus, Ly = 0 if and only if
(n + 2)(n + 1)an+2 + p(n)an = 0, n ≥ 0,
which is equivalent to Equation 7.3.24. Writing Equation 7.3.24 separately for the cases where n = 2m and
n = 2m + 1 yields Equation 7.3.26and Equation 7.3.27.
Example 7.3.3
Find the power series in x − 1 for the general solution of
2 ′′ ′
(2 + 4x − 2 x )y − 12(x − 1)y − 12y = 0. (7.3.28)
Solution
We must first write the coefficient P (x) = 2 + 4x − x in powers of
0
2
x −1 . To do this, we write x = (x − 1) + 1 in
P (x) and then expand the terms, collecting powers of x − 1 ; thus,
0
2 2
2 + 4x − 2x = 2 + 4[(x − 1) + 1] − 2[(x − 1) + 1]
2
= 4 − 2(x − 1 ) .
Therefore we can rewrite Equation 7.3.28as

2 ′′ ′
(4 − 2(x − 1 ) ) y − 12(x − 1)y − 12y = 0,

or, equivalently,
1 2 ′′ ′
(1 − (x − 1 ) ) y − 3(x − 1)y − 3y = 0.
2
This is of the form Equation 7.3.23with α = −1/2 , β = −3 , and γ = −3 . Therefore, from Equation 7.3.25
n(n − 1) (n + 2)(n + 3)
p(n) = − − 3n − 3 = − .
2 2
Hence, Theorem7.3.2implies that

p(2m)
a2m+2 = − a2m
(2m + 2)(2m + 1)
(2m + 2)(2m + 3) 2m + 3
= a2m = a2m , m ≥0
2(2m + 2)(2m + 1) 2(2m + 1)
and
p(2m + 1)
a2m+3 = − a2m+1
(2m + 3)(2m + 2)
(2m + 3)(2m + 4) m +2
= a2m+1 = a2m+1 , m ≥ 0.
2(2m + 3)(2m + 2) 2(m + 1)
We leave it to you to show that

2m + 1 m +1
a2m = m
a0 and a2m+1 = m
a1 , m ≥ 0,
2 2
which implies that the power series in x − 1 for the general solution of Equation 7.3.28is
∞ ∞
2m + 1 2m
m +1 2m+1
y = a0 ∑ (x − 1 ) + a1 ∑ (x − 1 ) .
m m
2 2
m=0 m=0
In the examples considered so far we were able to obtain closed formulas for coefficients in the power series solutions. In
some cases this is impossible, and we must settle for computing a finite number of terms in the series. The next example
illustrates this with an initial value problem.
Example 7.3.4
∞
Compute a , a , …, a in the series solution y = ∑
0 1 7 n=0
an x
n
2 ′′ ′ ′
(1 + 2 x )y + 10x y + 8y = 0, y(0) = 2, y (0) = −3. (7.3.29)
Solution
Since α = 2 , β = 10, and γ = 8 in Equation 7.3.29,
2
p(n) = 2n(n − 1) + 10n + 8 = 2(n + 2 ) .
Therefore
2
(n + 2) n+2
an+2 = −2 an = −2 an , n ≥ 0.
(n + 2)(n + 1) n+1
Writing this equation separately for n = 2m and n = 2m + 1 yields

(2m + 2) m +1
a2m+2 = −2 a2m = −4 a2m , m ≥0 (7.3.30)
2m + 1 2m + 1
2m + 3 2m + 3
a2m+3 = −2 a2m+1 = − a2m+1 , m ≥ 0. (7.3.31)
2m + 2 m +1

Starting with a 0 = y(0) = 2 , we compute a 2, a4 , and a from Equation 7.3.30:
6
1
a2 = −4 2 = −8,
1
2 64
a4 = −4 (−8) = ,
3 3
3 64 256
a6 = −4 ( ) =− .
5 3 5
Starting with a 1
′
= y (0) = −3 , we compute a 3, a5 and a from Equation 7.3.31:
7
3
a3 = − (−3) = 9,
1
5 45
a5 = − 9 =− ,
2 2
7 45 105
a7 = − (− ) = .
3 2 2
Therefore the solution of Equation 7.3.29is
2 3
64 4
45 5
256 6
105 7
y = 2 − 3x − 8 x + 9x + x − x − x + x +⋯ .
3 2 5 2
Using Technology
Computing coefficients recursively as in Example 7.3.4 is tedious. We recommend that you do this kind of computation by
writing a short program to implement the appropriate recurrence relation on a calculator or computer. You may wish to do this
in verifying examples and doing exercises (identified by the symbol C ) in this chapter that call for numerical computation of
the coefficients in series solutions. We obtained the answers to these exercises by using software that can produce answers in
the form of rational numbers. However, it is perfectly acceptable - and more practical - to get your answers in decimal form.
You can always check them by converting our fractions to decimals.
If you’re interested in actually using series to compute numerical approximations to solutions of a differential equation, then
whether or not there’s a simple closed form for the coefficients is essentially irrelevant. For computational purposes it is
usually more efficient to start with the given coefficients a = y(x ) and a = y (x ) , compute a , …, a recursively, and
0 0 1
′
0 2 N
then compute approximate values of the solution from the Taylor polynomial
N
n
TN (x) = ∑ an (x − x0 ) .
n=0
The trick is to decide how to choose N so the approximation y(x) ≈ T (x) is sufficiently accurate on the subinterval of the
N
interval of convergence that you’re interested in. In the computational exercises in this and the next two sections, you will
often be asked to obtain the solution of a given problem by numerical integration with software of your choice (see Section
10.1 for a brief discussion of one such method), and to compare the solution obtained in this way with the approximations
obtained with T for various values of N . This is a typical textbook kind of exercise, designed to give you insight into how
N
the accuracy of the approximation y(x) ≈ T (x) behaves as a function of N and the interval that you’re working on. In real
N
life, you would choose one or the other of the two methods (numerical integration or series solution). If you choose the method
of series solution, then a practical procedure for determining a suitable value of N is to continue increasing N until the
maximum of |T − T N | on the interval of interest is within the margin of error that you’re willing to accept.
N −1
In doing computational problems that call for numerical solution of differential equations you should choose the most accurate
numerical integration procedure your software supports, and experiment with the step size until you’re confident that the
numerical results are sufficiently accurate for the problem at hand.

7.3E: Series Solutions Near an Ordinary Point I (Exercises)
Q7.2.1
In Exercises 7.2.1-7.2.8 find the power series in x for the general solution.
1. (1 + x 2
)y
′′ ′
+ 6x y + 6y = 0
2. (1 + x 2
)y
′′ ′
+ 2x y − 2y = 0
3. (1 + x 2
)y
′′ ′
− 8x y + 20y = 0
4. (1 − x 2
)y
′′ ′
− 8x y − 12y = 0
5. (1 + 2x 2
)y
′′ ′
+ 7x y + 2y = 0
6. (1 + x 2
)y
′′
+ 2x y +
′ 1
4
y =0
7. (1 − x 2
)y
′′ ′
− 5x y − 4y = 0
8. (1 + x 2
)y
′′
− 10x y + 28y = 0
′
Q7.2.2
9.
a. Find the power series in x for the general solution of y + x y + 2y = 0 . ′′ ′
b. For several choices of a and a , use differential equations software to solve the initial value problem
0 1
′′ ′ ′
y + x y + 2y = 0, y(0) = a0 , y (0) = a1 , (A)
numerically on (−5, 5).

c. For fixed r in {1, 2, 3, 4, 5}graph
N
n
TN (x) = ∑ an x
n=0
and the solution obtained in (a) on (−r, r). Continue increasing N until there’s no perceptible difference between the two
graphs.
10. Follow the directions of Exercise [exer:7.2.9} for the differential equation
′′ ′
y + 2x y + 3y = 0.
Q7.2.3
In Exercises 7.2.11-7.2.13 find a0 , . . . , aN for N at least 7 in the power series solution y =∑
∞
n=0
n
an x of the initial value
problem.
11. (1 + x 2
)y
′′ ′
+ x y + y = 0, y(0) = 2,
′
y (0) = −1
12. (1 + 2x 2
)y
′′ ′
− 9x y − 6y = 0, y(0) = 1,
′
y (0) = −1
13. (1 + 8x 2
)y
′′
+ 2y = 0, y(0) = 2,
′
y (0) = −1
Q7.2.4
–
14. Do the next experiment for various choices of real numbers a , a , and r, with 0 < r < 1/√2 . 0 1

2 ′′ ′ ′
(1 − 2 x )y − x y + 3y = 0, y(0) = a0 , y (0) = a1 , (A)
numerically on (−r, r).

b. For N = 2 , 3, 4, …, compute a , …, a 2 N in the power series solution y = ∑ ∞
n=0
n
an x of (A), and graph

N
n
TN (x) = ∑ an x
n=0
graphs.
15. Do (a) and (b) for several values of r in (0, 1):
2 ′′ ′ ′
(1 + x )y + 10x y + 14y = 0, y(0) = 5, y (0) = 1, (A)

b. For N = 2 , 3, 4, …, compute a , …, a 2 N in the power series solution y = ∑ ∞
n=0
an x
n
of (A), and graph
N
n
TN (x) = ∑ an x
n=0
graphs. What happens to the required N as r → 1 ?
c. Try (a) and (b) with r = 1.2. Explain your results.
Q7.2.5
In Exercises 7.2.16-7.2.20 find the power series in x − x for the general solution. 0
16. y ′′
− y = 0; x0 = 3
17. y ′′ ′
− (x − 3)y − y = 0; x0 = 3
18. (1 − 4x + 2x 2
)y
′′
+ 10(x − 1)y + 6y = 0;
′
x0 = 1
19. (11 − 8x + 2x 2
)y
′′
− 16(x − 2)y + 36y = 0;
′
x0 = 2
20. (5 + 6x + 3x 2
)y
′′
+ 9(x + 1)y + 3y = 0;
′
x0 = −1
Q7.2.6
In Exercises 7.2.21-7.2.26 find a , . . . a for N at least 7 in the power series y = ∑
0 N
∞
n=0
n
an (x − x0 ) for the solution of the
initial value problem. Take x to be the point where the initial conditions are imposed.
0
21. (x 2
− 4)y
′′ ′
− x y − 3y = 0, y(0) = −1,
′
y (0) = 2
22. y ′′ ′
+ (x − 3)y + 3y = 0, y(3) = −2,
′
y (3) = 3
23. (5 − 6x + 3x 2
)y
′′ ′
+ (x − 1)y + 12y = 0, y(1) = −1,
′
y (1) = 1
24. (4x 2
− 24x + 37)y
′′
+ y = 0, y(3) = 4,
′
y (3) = −6
25. (x 2
− 8x + 14)y
′′
− 8(x − 4)y + 20y = 0,
′
y(4) = 3,
′
y (4) = −4
26. (2x 2
+ 4x + 5)y
′′
− 20(x + 1)y + 60y = 0,
′
y(−1) = 3,
′
y (−1) = −3
Q7.2.7
27.
a. Find a power series in x for the general solution of
2 ′′ ′
(1 + x )y + 4x y + 2y = 0. (A)
b. Use (a) and the formula

1 2 n
= 1 +r+r +⋯ +r +⋯ (−1 < r < 1)
1 −r
for the sum of a geometric series to find a closed form expression for the general solution of (A) on (−1, 1).

c. Show that the expression obtained in (b) is actually the general solution of of (A) on (−∞, ∞).
28. Use Theorem 7.2.2 to show that the power series in x for the general solution of
2 ′′ ′
(1 + α x )y + βx y + γy = 0
is
∞ m−1 2m ∞ m−1 2m+1
m
x m
x
y = a0 ∑ (−1 ) [ ∏ p(2j)] + a1 ∑ (−1 ) [ ∏ p(2j + 1)] .
(2m)! (2m + 1)!
m=0 j=0 m=0 j=0
29. Use Exercise 7.2.28 to show that all solutions of

2 ′′ ′
(1 + α x )y + βx y + γy = 0
are polynomials if and only if
αn(n − 1) + βn + γ = α(n − 2r)(n − 2s − 1),
where r and s are nonnegative integers.

30.
a. Use Exercise [exer:7.2.28} to show that the power series in x for the general solution of
2 ′′ ′
(1 − x )y − 2bx y + α(α + 2b − 1)y = 0
is y = a0 y1 + a1 y2 , where
∞ m−1 2m
x
y1 = ∑ [ ∏ (2j − α)(2j + α + 2b − 1)]
(2m)!
m=0 j=0
and
∞ m−1 2m+1
x
y2 = ∑ [ ∏ (2j + 1 − α)(2j + α + 2b)]
(2m + 1)!
m=0 j=0
b. Suppose 2b isn’t a negative odd integer and k is a nonnegative integer. Show that y is a polynomial of degree 2k such that 1
y (−x) = y (x) if α = 2k , while y

1 1 is a polynomial of degree 2k + 1 such that y (−x) = −y (−x) if α = 2k + 1 .
2 2 2
Conclude that if n is a nonnegative integer, then there’s a polynomial P of degree n such that P (−x) = (−1) P (x) n n
n
n
and
2 ′′ ′
(1 − x )Pn − 2bx Pn + n(n + 2b − 1)Pn = 0. (A)
c. Show that (A) implies that

2 b ′ ′ 2 b−1
[(1 − x ) Pn ] = −n(n + 2b − 1)(1 − x ) Pn ,
and use this to show that if m and n are nonnegative integers, then
2 b ′ 2 b ′ 2 b−1
[(1 − x ) Pn ] Pm − [(1 − x ) Pm ] Pn = [m(m + 2b − 1) − n(n + 2b − 1)](1 − x ) Pm Pn (B)
d. Now suppose b > 0 . Use (B) and integration by parts to show that if m ≠ n , then
1
2 b−1
∫ (1 − x ) Pm (x)Pn (x) dx = 0.
−1
(We say that P and P are orthogonal on (−1, 1) with respect to the weighting function(1 − x
m n
2 b−1
) .)
31.
a. Use Exercise 7.2.28 to show that the power series in x for the general solution of Hermite’s equation
′′ ′
y − 2x y + 2αy = 0

is y = a 0 y1 + a1 y1 , where
∞ m−1 m 2m
2 x
y1 = ∑ [ ∏ (2j − α)]
(2m)!
m=0 j=0
and
∞ m−1 m 2m+1
2 x
y2 = ∑ [ ∏ (2j + 1 − α)]
(2m + 1)!
m=0 j=0
b. Suppose k is a nonnegative integer. Show that y is a polynomial of degree 2k such that y (−x) = y (x) if α = 2k , while
1 1 1
y2 is a polynomial of degree 2k + 1 such that y (−x) = −y (−x) if α = 2k + 1 . Conclude that if n is a nonnegative

2 2
integer then there’s a polynomial P of degree n such that P (−x) = (−1) P (x) and
n n
n
n
′′ ′
Pn − 2x Pn + 2nPn = 0. (A)
c. Show that (A) implies that

2 2
−x ′ ′ −x
[e Pn ] = −2ne Pn ,
and use this to show that if m and n are nonnegative integers, then
2 2 2
−x ′ ′ −x ′ ′ −x
[e Pn ] Pm − [ e Pm ] Pn = 2(m − n)e Pm Pn . (B)
d. Use (B) and integration by parts to show that if m ≠ n , then

∞
2
−x
∫ e Pm (x)Pn (x) dx = 0.
−∞
(We say that P and P are orthogonal on (−∞, ∞) with respect to the weighting function e
m n
−x
.)
32. Consider the equation
3 ′′ 2 ′
(1 + α x ) y + β x y + γxy = 0, (A)
and let p(n) = αn(n − 1) + βn + γ . (The special case y ′′

− xy = 0 of (A) is Airy’s equation.)
a. Modify the argument used to prove Theorem [thmtype:7.2.2} to show that
∞
n
y = ∑ an x
n=0
is a solution of (A) if and only if a 2 =0 and

p(n)
an+3 = − an , n ≥ 0.
(n + 3)(n + 2)
b. Show from (a) that a n =0 unless n = 3m or n = 3m + 1 for some nonnegative integer m, and that
p(3m)
a3m+3 = − a3m , m ≥ 0,
(3m + 3)(3m + 2)
and
p(3m + 1)
a3m+4 = − a3m+1 , m ≥ 0,
(3m + 4)(3m + 3)
where a and a may be specified arbitrarily.

0 1
c. Conclude from (b) that the power series in x for the general solution of (A) is
3m
∞ m−1 p(3j) x
m
y = a0 ∑ (−1 ) [∏ ] m
m=0 j=0 3j+2 3 m!
∞ m−1 p(3j+1) x
3m+1
m
+a1 ∑ (−1 ) [∏ ] m
.
m=0 j=0 3j+4 3 m!

Q7.2.8
In Exercises 7.2.33-7.2.37 use the method of Exercise 7.2.32 to find the power series in x for the general solution.
33. y ′′
− xy = 0
34. (1 − 2x 3
)y
′′ 2
− 10 x y − 8xy = 0
′
35. (1 + x 3
)y
′′ 2
+ 7 x y + 9xy = 0
′
36. (1 − 2x 3
)y
′′ 2
+ 6 x y + 24xy = 0
′
37. (1 − x 3
)y
′′ 2
+ 15 x y − 63xy = 0
′
Q7.2.9
38. Consider the equation
k+2 ′′ k+1 ′ k
(1 + α x )y + βx y + γ x y = 0, (A)
where k is a positive integer, and let p(n) = αn(n − 1) + βn + γ .

a. Modify the argument used to prove Theorem 7.2.2 to show that
∞
n
y = ∑ an x
n=0
is a solution of (A) if and only if a n =0 for 2 ≤ n ≤ k + 1 and

p(n)
an+k+2 = − an , n ≥ 0.
(n + k + 2)(n + k + 1)
b. Show from (a) that a n =0 unless n = (k + 2)m or n = (k + 2)m + 1 for some nonnegative integer m, and that
p ((k + 2)m)
a(k+2)(m+1) = − a(k+2)m , m ≥ 0,
(k + 2)(m + 1)[(k + 2)(m + 1) − 1]
and
p ((k + 2)m + 1)
a(k+2)(m+1)+1 = − a(k+2)m+1 , m ≥ 0,
[(k + 2)(m + 1) + 1](k + 2)(m + 1)
where a and a may be specified arbitrarily.

0 1
c. Conclude from (b) that the power series in x for the general solution of (A) is
∞ m−1 p((k+2)j) ( k+2) m
m x
y = a0 ∑ (−1 ) [∏ ] m
m=0 j=0 (k+2)(j+1)−1 (k+2 ) m!
( k+2) m+1
∞ m−1 p((k+2)j+1) x
m
+a1 ∑ (−1 ) [∏ ] m
.
m=0 j=0 (k+2)(j+1)+1 (k+2 ) m!
Q7.2.10
In Exercises 7.2.39-7.2.44 use the method of Exercise 7.2.38 to find the power series in x for the general solution.
39. (1 + 2x 5
)y
′′ 4
+ 14 x y + 10 x y = 0
′ 3
40. y ′′ 2
+x y = 0
41. y ′′ 6 ′
+ x y + 7x y = 0
5
42. (1 + x 8
)y
′′ 7
− 16 x y + 72 x y = 0
′ 6
43. (1 − x 6
)y
′′ 5
− 12 x y − 30 x y = 0
′ 4
44. y ′′ 5 ′
+ x y + 6x y = 0
4

7.4: Series Solutions Near an Ordinary Point II
In this section we continue to find series solutions
∞
n
y = ∑ an (x − x0 )
n=0
of initial value problems

′′ ′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0, y(x0 ) = a0 , y (x0 ) = a1 , (7.4.1)
where P , P , and P are polynomials and P (x ) ≠ 0 , so x is an ordinary point of Equation

0 1 2 0 0 0 . However, here we
7.4.1
consider cases where the differential equation in Equation 7.4.1 is not of the form
2 ′′ ′
(1 + α(x − x0 ) ) y + β(x − x0 )y + γy = 0,
so Theorem 7.2.2 does not apply, and the computation of the coefficients {a } is more complicated. For the equations n
considered here it is difficult or impossible to obtain an explicit formula for a in terms of n . Nevertheless, we can calculate as n
many coefficients as we wish. The next three examples illustrate this.
Example 7.4.1
Find the coefficients a , …, a in the series solution y = ∑
0 7
∞
n=0
an x
n
2 ′′ ′ ′
(1 + x + 2 x )y + (1 + 7x)y + 2y = 0, y(0) = −1, y (0) = −2. (7.4.2)
Solution
Here
2 ′′ ′
Ly = (1 + x + 2 x )y + (1 + 7x)y + 2y.
– –
The zeros (−1 ± i √7)/4 of P (x) = 1 + x + 2x have absolute value 1/√2, so Theorem 7.2.2 implies that the series
0
2
– –
solution converges to the solution of Equation 7.4.2 on (−1/√2, 1/√2). Since
∞ ∞ ∞
n ′ n−1 ′′ n−2
y = ∑ an x , y = ∑ nan x and y = ∑ n(n − 1)an x ,
n=0 n=1 n=2
∞ ∞ ∞
n−2 n−1 n
Ly = ∑ n(n − 1)an x + ∑ n(n − 1)an x + 2 ∑ n(n − 1)an x
n=2 n=2 n=2
∞ ∞ ∞
n−1 n n
+ ∑ nan x + 7 ∑ nan x + 2 ∑ an x .
n=1 n=1 n=0
Shifting indices so the general term in each series is a constant multiple of x yields n
∞ ∞ ∞
n n n
Ly = ∑(n + 2)(n + 1)an+2 x + ∑(n + 1)nan+1 x + 2 ∑ n(n − 1)an x
n=0 n=0 n=0
∞ ∞ ∞ ∞
n n n n
+ ∑(n + 1)an+1 x + 7 ∑ nan x + 2 ∑ an x = ∑ bn x ,
n=0 n=0 n=0 n=0
where
2
bn = (n + 2)(n + 1)an+2 + (n + 1 ) an+1 + (n + 2)(2n + 1)an .
Therefore y = ∑ ∞
n=0
an x
n
is a solution of Ly = 0 if and only if
n+1 2n + 1
an+2 = − an+1 − an , n ≥ 0. (7.4.3)
n+2 n+1

From the initial conditions in Equation ,
7.4.2 a0 = y(0) = −1 and a1 = y (0) = −2
′
. Setting n =0 in Equation 7.4.3
yields
1 1
a2 = − a1 − a0 = − (−2) − (−1) = 2.
2 2
Setting n = 1 in Equation 7.4.3 yields

2 3 2 3 5
a3 = − a2 − a1 = − (2) − (−2) = .
3 2 3 2 3
We leave it to you to compute a 4, a5 , a6 , a7 from Equation 7.4.3 and show that
2
5 3
55 4
3 5
61 6
443 7
y = −1 − 2x + 2 x + x − x + x + x − x +⋯ .
3 12 4 8 56
We also leave it to you (Exercise [exer:7.3.13}) to verify numerically that the Taylor polynomials T N (x) = ∑
N
n=0
n
an x
– –
converge to the solution of Equation 7.4.2 on (−1/√2, 1/√2).
Example 7.4.2
Find the coefficients a , …, a in the series solution
0 5
n
y = ∑ an (x + 1 )
n=0

′′ ′ ′
(3 + x)y + (1 + 2x)y − (2 − x)y = 0, y(−1) = 2, y (−1) = −3. (7.4.4)
Since the desired series is in powers of x + 1 we rewrite the differential equation in Equation 7.4.4 as Ly = 0 , with
′′ ′
Ly = (2 + (x + 1)) y − (1 − 2(x + 1)) y − (3 − (x + 1)) y.
Since
∞ ∞ ∞
n ′ n−1 ′′ n−2
y = ∑ an (x + 1 ) , y = ∑ nan (x + 1 ) and y = ∑ n(n − 1)an (x + 1 ) ,
n=0 n=1 n=2
∞ ∞
n−2 n−1
Ly = 2 ∑ n(n − 1)an (x + 1 ) + ∑ n(n − 1)an (x + 1 )
n=2 n=2
∞ ∞
n−1 n
− ∑ nan (x + 1 ) + 2 ∑ nan (x + 1 )
n=1 n=1
∞ ∞
n n+1
− 3 ∑ an (x + 1 ) + ∑ an (x + 1 ) .
n=0 n=0
Shifting indices so that the general term in each series is a constant multiple of (x + 1) yields n
∞ ∞
n n
Ly = 2 ∑(n + 2)(n + 1)an+2 (x + 1 ) + ∑(n + 1)nan+1 (x + 1 )
n=0 n=0
∞ ∞ ∞
n n n
− ∑(n + 1)an+1 (x + 1 ) + ∑(2n − 3)an (x + 1 ) + ∑ an−1 (x + 1 )
n=0 n=0 n=1
∞
n
= ∑ bn (x + 1 ) ,
n=0
where

b0 = 4 a2 − a1 − 3 a0
and
2
bn = 2(n + 2)(n + 1)an+2 + (n − 1)an+1 + (2n − 3)an + an−1 , n ≥ 1.
n=0
an (x + 1 )
n
1
a2 = (a1 + 3 a0 ) (7.4.5)
4
and
1 2
an+2 = − [(n − 1)an+1 + (2n − 3)an + an−1 ] , n ≥ 1. (7.4.6)
2(n + 2)(n + 1)
From the initial conditions in Equation 7.4.4, a = y(−1) = 2 and a = y (−1) = −3 . We leave it to you to compute
0 1
′
a , …, a with Equation 7.4.5 and Equation 7.4.6 and show that the solution of Equation 7.4.4 is
2 5
3 2
5 3
7 4
1 5
y = −2 − 3(x + 1) + (x + 1 ) − (x + 1 ) + (x + 1 ) − (x + 1 ) +⋯ .
4 12 48 60
We also leave it to you (Exercise [exer:7.3.14}) to verify numerically that the Taylor polynomials T N
N
(x) = ∑
n=0
n
an x
converge to the solution of Equation 7.4.4 on the interval of convergence of the power series solution.
Example 7.4.3
Find the coefficients a , …, a in the series solution y = ∑
0 5
∞
n=0
an x
n
′′ ′ 2 ′
y + 3x y + (4 + 2 x )y = 0, y(0) = 2, y (0) = −3. (7.4.7)
Solution
Here
′′ ′ 2
Ly = y + 3x y + (4 + 2 x )y.
Since
∞ ∞ ∞
n ′ n−1 ′′ n−2
y = ∑ an x , y = ∑ nan x , and y = ∑ n(n − 1)an x ,
n=0 n=1 n=2
∞ ∞ ∞ ∞
n−2 n n n+2
Ly = ∑ n(n − 1)an x + 3 ∑ nan x + 4 ∑ an x + 2 ∑ an x .
n=2 n=1 n=0 n=0
Shifting indices so that the general term in each series is a constant multiple of x yields n
∞ ∞ ∞ ∞
n n n n
Ly = ∑(n + 2)(n + 1)an+2 x + ∑(3n + 4)an x + 2 ∑ an−2 x = ∑ bn x
n=0 n=0 n=2 n=0
where
b0 = 2 a2 + 4 a0 , b1 = 6 a3 + 7 a1 ,
and
bn = (n + 2)(n + 1)an+2 + (3n + 4)an + 2 an−2 , n ≥ 2.
n=0
an x
n
7
a2 = −2 a0 , a3 = − a1 , (7.4.8)
6
and

1
an+2 = − [(3n + 4)an + 2 an−2 ] , n ≥ 2. (7.4.9)
(n + 2)(n + 1)
From the initial conditions in Equation 7.4.7, a = y(0) = 2 and a = y (0) = −3 . We leave it to you to compute
0 1
′
a2 ,
…, a with Equation 7.4.8 and Equation 7.4.9 and show that the solution of Equation 7.4.7 is
5
2
7 3 4
79 5
y = 2 − 3x − 4 x + x + 3x − x +⋯ .
2 40
We also leave it to you (Exercise [exer:7.3.15}) to verify numerically that the Taylor polynomials T N (x) = ∑
N
n=0
n
an x
converge to the solution of Equation 7.4.9 on the interval of convergence of the power series solution.

7.4E: Series Solutions Near an Ordinary Point II (Exercises)
Q7.3.1
In Exercises 7.3.1-7.3.12 find the coefficients a0 ,…, a N for N at least 7 in the series solution y =∑
∞
n=0
n
an x of the initial
value problem.
1. (1 + 3x)y ′′ ′
+ x y + 2y = 0, y(0) = 2,
′
y (0) = −3
2. (1 + x + 2x 2
)y
′′
+ (2 + 8x)y + 4y = 0,
′
y(0) = −1,
′
y (0) = 2
3. (1 − 2x 2
)y
′′
+ (2 − 6x)y − 2y = 0,
′
y(0) = 1,
′
y (0) = 0
4. (1 + x + 3x 2
)y
′′
+ (2 + 15x)y + 12y = 0,
′
y(0) = 0,
′
y (0) = 1
5. (2 + x)y ′′
+ (1 + x)y + 3y = 0,
′
y(0) = 4,
′
y (0) = 3
6. (3 + 3x + x 2
)y
′′
+ (6 + 4x)y + 2y = 0,
′
y(0) = 7,
′
y (0) = 3
7. (4 + x)y ′′
+ (2 + x)y + 2y = 0,
′
y(0) = 2,
′
y (0) = 5
8. (2 − 3x + 2x 2
)y
′′
− (4 − 6x)y + 2y = 0,
′
y(1) = 1,
′
y (1) = −1
9. (3x + 2x 2
)y
′′
+ 10(1 + x)y + 8y = 0,
′
y(−1) = 1,
′
y (−1) = −1
10. (1 − x + x 2
)y
′′
− (1 − 4x)y + 2y = 0,
′
y(1) = 2,
′
y (1) = −1
11. (2 + x)y ′′
+ (2 + x)y + y = 0,
′
y(−1) = −2,
′
y (−1) = 3
12. x2
y
′′
− (6 − 7x)y + 8y = 0,
′
y(1) = 1,
′
y (1) = −2
Q7.3.2
–
13. Do the following experiment for various choices of real numbers a , a , and r, with 0 < r < 1/√2 . 0 1

2 ′′ ′ ′
(1 + x + 2 x )y + (1 + 7x)y + 2y = 0, y(0) = a0 , y (0) = a1 , (A)
numerically on (−r, r). (See Example 7.3.1.)

b. For N = 2 , 3, 4, …, compute a , …, a in the power series solution y = ∑ 2 N
∞
n=0
n
an x of (A), and graph
N
n
TN (x) = ∑ an x
n=0
graphs.
14. Do the following experiment for various choices of real numbers a , a , and r, with 0 < r < 2 . 0 1

′′ ′ ′
(3 + x)y + (1 + 2x)y − (2 − x)y = 0, y(−1) = a0 , y (−1) = a1 , (A)
numerically on (−1 − r, −1 + r) . (See Example 7.3.2). Why this interval?)

b. For N = 2 , 3, 4, …, compute a , … , a in the power series solution 2 N
n
y = ∑ an (x + 1 )
n=0
of (A), and graph

N
n
TN (x) = ∑ an (x + 1 )
n=0

and the solution obtained in (a) on (−1 − r, −1 + r) . Continue increasing N until there’s no perceptible difference
between the two graphs.
15. Do the following experiment for several choices of a , a , and r, with r > 0 . 0 1

′′ ′ 2 ′
y + 3x y + (4 + 2 x )y = 0, y(0) = a0 , y (0) = a1 , (A)
numerically on (−r, r). (See Example 7.3.3.)

∞
b. Find the coefficients a , a , …, a in the power series solution y = ∑
0 1 N n=0
an x
n
of (A), and graph
N
n
TN (x) = ∑ an x
n=0
graphs.
16. Do the following experiment for several choices of a and a . 0 1

′′ ′ ′
(1 − x)y − (2 − x)y + y = 0, y(0) = a0 , y (0) = a1 , (A)

b. Find the coefficients a , a , …, a 0 1 N in the power series solution y = ∑ N
n=0
an x
n
of (A), and graph
N
n
TN (x) = ∑ an x
n=0
graphs. What happens as you let r → 1 ?
′′ ′ ′
(1 + x)y + 3 y + 32y = 0, y(0) = a0 , y (0) = a1 .
2 ′′ ′ ′
(1 + x )y + y + 2y = 0, y(0) = a0 , y (0) = a1 .
Q7.3.3
In Exercises 7.3.19-7.3.28 find the coefficients a 0, . . . aN for N at least 7 in the series solution
∞
n
y = ∑ an (x − x0 )
n=0
of the initial value problem. Take x to be the point where the initial conditions are imposed.
0
19. (2 + 4x)y ′′ ′
− 4 y − (6 + 4x)y = 0, y(0) = 2,
′
y (0) = −7
20. (1 + 2x)y ′′ ′
− (1 − 2x)y − (3 − 2x)y = 0, y(1) = 1,
′
y (1) = −2
21. (5 + 2x)y ′′ ′
− y + (5 + x)y = 0, y(−2) = 2,
′
y (−2) = −1
22. (4 + x)y ′′ ′
− (4 + 2x)y + (6 + x)y = 0, y(−3) = 2,
′
y (−3) = −2
23. (2 + 3x)y ′′ ′
− x y + 2xy = 0, y(0) = −1,
′
y (0) = 2
24. (3 + 2x)y ′′ ′
+ 3 y − xy = 0, y(−1) = 2,
′
y (−1) = −3
25. (3 + 2x)y ′′ ′
− 3 y − (2 + x)y = 0, y(−2) = −2,
′
y (−2) = 3
26. (10 − 2x)y ′′

+ (1 + x)y = 0, y(2) = 2,
′
y (2) = −4
27. (7 + x)y ′′ ′
+ (8 + 2x)y + (5 + x)y = 0, y(−4) = 1,
′
y (−4) = 2

28. (6 + 4x)y ′′
+ (1 + 2x)y = 0, y(−1) = −1,
′
y (−1) = 2
Q7.3.4
29. Show that the coefficients in the power series in x for the general solution of
2 ′′ ′
(1 + αx + β x )y + (γ + δx)y + ϵy = 0
satisfy the recurrrence relation

γ + αn βn(n − 1) + δn + ϵ
an+2 = − an+1 − an .
n+2 (n + 2)(n + 1)
30.
a. Let α and β be constants, with β ≠ 0 . Show that y = ∑ ∞
n=0
an x
n
is a solution of
2 ′′ ′
(1 + αx + β x )y + (2α + 4βx)y + 2βy = 0 (A)
if and only if
an+2 + α an+1 + β an = 0, n ≥ 0. (B)
An equation of this form is called a second order homogeneous linear difference equation. The polynomial
p(r) = r + αr + β is called the characteristic polynomial of (B). If r and r are the zeros of p , then 1/r and 1/r are
2
1 2 1 2
the zeros of
2
P0 (x) = 1 + αx + β x
b. Suppose p(r) = (r − r1 )(r − r2 ) where r and r are real and distinct, and let
1 2 ρ be the smaller of the two numbers
{1/| r1 |, 1/| r2 |} . Show that if c and c are constants then the sequence
1 2
n n
an = c1 r + c2 r , n ≥0
1 2
satisfies (B). Conclude from this that any function of the form
∞
n n n
y = ∑(c1 r + c2 r )x
1 2
n=0
is a solution of (A) on (−ρ, ρ).

c. Use (b) and the formula for the sum of a geometric series to show that the functions
1 1
y1 = and y2 =
1 − r1 x 1 − r2 x
form a fundamental set of solutions of (A) on (−ρ, ρ).

d. Show that {y , y } is a fundamental set of solutions of (A) on any interval that does’nt contain either 1/r or 1/r .
1 2 1 2
e. Suppose p(r) = (r − r ) , and let ρ = 1/|r |. Show that if c and c are constants then the sequence
1
2
1 1 2
n
an = (c1 + c2 n)r , n ≥0
1
satisfies (B). Conclude from this that any function of the form
∞
n n
y = ∑(c1 + c2 n)r x
1
n=0
is a solution of (A) on (−ρ, ρ).

f. Use (e) and the formula for the sum of a geometric series to show that the functions
1 x
y1 = and y2 =
2
1 − r1 x (1 − r1 x )
form a fundamental set of solutions of (A) on (−ρ, ρ).

g. Show that {y , y } is a fundamental set of solutions of (A) on any interval that does not contain 1/r .
1 2 1

31. Use the results of Exercise 7.3.30 to find the general solution of the given equation on any interval on which polynomial
multiplying y has no zeros. ′′
a. (1 + 3x + 2x )y + (6 + 8x)y + 4y = 0 2 ′′ ′
b. (1 − 5x + 6x )y − (10 − 24x)y + 12y = 02 ′′ ′
c. (1 − 4x + 4x )y − (8 − 16x)y + 8y = 0 2 ′′ ′
d. (4 + 4x + x )y + (8 + 4x)y + 2y = 0
2 ′′ ′
e. (4 + 8x + 3x )y + (16 + 12x)y + 6y = 0 2 ′′ ′
Q7.3.5
∞
In Exercises 7.3.32-7.3.38 find the coefficients a 0, . . . , aN for N at least 7 in the series solution y = ∑ n=0
n
an x of the initial
value problem.
32. y ′′ ′
+ 2x y + (3 + 2 x )y = 0,
2
y(0) = 1,
′
y (0) = −2
33. y ′′ ′
− 3x y + (5 + 2 x )y = 0,
2
y(0) = 1,
′
y (0) = −2
34. y ′′ ′
+ 5x y − (3 − x )y = 0,
2
y(0) = 6,
′
y (0) = −2
35. y ′′ ′
− 2x y − (2 + 3 x )y = 0,
2
y(0) = 2,
′
y (0) = −5
36. y ′′ ′
− 3x y + (2 + 4 x )y = 0,
2
y(0) = 3,
′
y (0) = 6
37. 2y ′′ ′
+ 5x y + (4 + 2 x )y = 0,
2
y(0) = 3,
′
y (0) = −2
38. 3y ′′ ′
+ 2x y + (4 − x )y = 0,
2
y(0) = −2,
′
y (0) = 3
Q7.3.6
39. Find power series in x for the solutions y and y of 1 2
′′ ′ 2
y + 4x y + (2 + 4 x )y = 0
such that y 1 (0) =1 ,y ′

1
(0) = 0 ,y 2 (0) =0 ,y ′
2
(0) = 1 , and identify y and y in terms of familiar elementary functions.
1 2
Q7.3.7
In Exercises 7.3.40-7.3.49 find the coefficients a 0, . . . , aN for N at least 7 in the series solution
∞
n
y = ∑ an (x − x0 )
n=0
of the initial value problem. Take x to be the point where the initial conditions are imposed. 0
40. (1 + x)y ′′ 2
+ x y + (1 + 2x)y = 0,
′
y(0) − 2,
′
y (0) = 3
41. y ′′
+ (1 + 2x + x )y + 2y = 0,
2 ′
y(0) = 2,
′
y (0) = 3
42. (1 + x 2
)y
′′
+ (2 + x )y + xy = 0,
2 ′
y(0) = −3,
′
y (0) = 5
43. (1 + x)y ′′
+ (1 − 3x + 2 x )y − (x − 4)y = 0,
2 ′
y(1) = −2,
′
y (1) = 3
44. y ′′
+ (13 + 12x + 3 x )y + (5 + 2x),
2 ′
y(−2) = 2,
′
y (−2) = −3
45. (1 + 2x + 3x 2
)y
′′
+ (2 − x )y + (1 + x)y = 0,
2 ′
y(0) = 1,
′
y (0) = −2
46. (3 + 4x + x 2
)y
′′
− (5 + 4x − x )y − (2 + x)y = 0,
2 ′
y(−2) = 2,
′
y (−2) = −1
47. (1 + 2x + x 2
)y
′′
+ (1 − x)y = 0, y(0) = 2,
′
y (0) = −1
48. (x − 2x 2
)y
′′
+ (1 + 3x − x )y + (2 + x)y = 0,
2 ′
y(1) = 1,
′
y (1) = 0
49. (16 − 11x + 2x 2

)y
′′
+ (10 − 6x + x )y − (2 − x)y,
2 ′
y(3) = 1,
′
y (3) = −2

CHAPTER OVERVIEW
8: LAPLACE TRANSFORMS
IN THIS CHAPTER we study the method of Laplace transforms, which illustrates one of the basic
problem solving techniques in mathematics: transform a difcult problem into an easier one, solve
the latter, and then use its solution to obtain a solution of the original problem. The method discussed
here transforms an initial value problem for a constant coefcient equation into an algebraic equation
whose solution can then be used to solve the initial value problem. In some cases this method is
merely an alternative procedure for solvingproblems that can be solved equally well by methods that
we considered previously; however, in other cases the method of Laplace transforms is more
efcient than the methods previously discussed. This is especially true in physical problems dealing
with discontinuous forcing functions.
8.1: INTRODUCTION TO THE LAPLACE TRANSFORM

This section denes the Laplace transform and develops its properties.
8.1E: INTRODUCTION TO THE LAPLACE TRANSFORM (EXERCISES)

8.2: THE INVERSE LAPLACE TRANSFORM
This section deals with the problem of nding a function that has a given Laplace transform.
8.2E: THE INVERSE LAPLACE TRANSFORM (EXERCISES)

8.3: SOLUTION OF INITIAL VALUE PROBLEMS
This section applies the Laplace transform to solve initial value problems for constant coefcient second order differential equations
on (0,∞).
8.3E: SOLUTION OF INITIAL VALUE PROBLEMS (EXERCISES)

8.4: A BRIEF TABLE OF LAPLACE TRANSFORMS
This section is a brief table of Laplace transforms.
1 9/16/2020
8.1: Introduction to the Laplace Transform
Definition of the Laplace Transform
To define the Laplace transform, we first recall the definition of an improper integral. If g is integrable over the interval [a, T ]
for every T > a , then the improper integral of g over [a, ∞) is defined as
∞ T
∫ g(t) dt = lim ∫ g(t) dt. (8.1.1)

T →∞
a a
We say that the improper integral converges if the limit in Equation 8.1.1 exists; otherwise, we say that the improper integral
diverges or does not exist. Here’s the definition of the Laplace transform of a function f .
Defintion 8.1.1 : Laplace transform

Let f be defined for t ≥ 0 and let s be a real number. Then the Laplace transform of f is the function F defined by
∞
−st
F (s) = ∫ e f (t) dt, (8.1.2)
0
for those values of s for which the improper integral converges.
It is important to keep in mind that the variable of integration in Equation 8.1.2 is t , while s is a parameter independent of t .
We use t as the independent variable for f because in applications the Laplace transform is usually applied to functions of
time.
The Laplace transform can be viewed as an operator L that transforms the function f = f (t) into the function F = F (s) .
Thus, Equation 8.1.2 can be expressed as
F = L(f ).
The functions f and F form a transform pair, which we’ll sometimes denote by
f (t) ↔ F (s).
It can be shown that if F (s) is defined for s = s then it is defined for all s > s (Exercise 8.1.14b).
0 0
Computation of Some Simple Laplace Transforms
Example 8.1.1
Find the Laplace transform of f (t) = 1 .
Solution
From Equation 8.1.2 with f (t) = 1 ,
∞ T
−st −st
F (s) = ∫ e dt = lim ∫ e dt.
T →∞
0 0
If s ≠ 0 then
T −sT
1 T 1 −e
−st −st ∣
∫ e dt = − e = . (8.1.3)
∣
0
s 0 s
Therefore
T 1
, s > 0,
−st s
lim ∫ e dt = { (8.1.4)
T →∞
0 ∞, s < 0.

If s = 0 the integrand reduces to the constant 1, and
T T
lim ∫ 1 dt = lim ∫ 1 dt = lim T = ∞.

T →∞ T →∞ T →∞
0 0
Therefore F (0) is undefined, and

∞
−st
1
F (s) = ∫ e dt = , s > 0.
0
s
This result can be written in operator notation as

1
L(1) = , s > 0,
s
or as the transform pair

1
1 ↔ , s > 0.
s
Note
It is convenient to combine the steps of integrating from 0 to T and letting T → ∞ . Therefore, instead of writing
Equation 8.1.3 and 8.1.4 as separate steps we write
∞ ∞ 1
1 ∣ , s >0
−st −st s
∫ e dt = − e ∣ ={
s ∣
0 0 ∞, s <0
We'll follow this practice throughout this chapter.
Example 8.1.2
Find the Laplace transform of f (t) = t .
From Equation 8.1.2 with f (t) = t ,
∞
−st
F (s) = ∫ e t dt. (8.1.5)
0
If s ≠ 0 , integrating by parts yields

∞ −st ∞ ∞ ∞
−st
te ∣ 1 −st
t 1 −st ∣
∫ e t dt =− ∣ + ∫ e dt = − [ + ]e ∣
s ∣ s s 2
s ∣
0 0 0 0
1
2
, s > 0,
s
={
∞, s < 0.
If s = 0 , the integral in Equation 8.1.5 becomes

∞ 2 ∞
t ∣
∫ t dt = ∣ = ∞.
0
2 ∣0
Therefore F (0) is undefined and

1
F (s) = , s > 0.
2
s
This result can also be written as

1
L(t) = , s > 0,
2
s
or as the transform pair

1
t ↔ , s > 0.
s2
Example 8.1.3 :
Find the Laplace transform of f (t) = e , where a is a constant.
at
From Equation 8.1.2 with f (t) = e ,at
∞
−st at
F (s) = ∫ e e dt.
0
Combining the exponentials yields

∞
−(s−a)t
F (s) = ∫ e dt.
0
However, we know from Example 8.1.1 that

∞
−st
1
∫ e dt = , s > 0.
0 s
Replacing s by s − a here shows that

1
F (s) = , s > a.
s−a
This can also be written as

1 1
at at
L(e ) = , s > a, or e ↔ , s > a.
s−a s−a
Example 8.1.4
[Find the Laplace transforms of f (t) = sin ωt and g(t) = cos ωt , where ω is a constant.
Define
∞
−st
F (s) = ∫ e sin ωt dt (8.1.6)
0
and
∞
−st
G(s) = ∫ e cos ωt dt. (8.1.7)
0
If s > 0 , integrating Equation 8.1.6 by parts yields

−st ∞
e ∞ ω
∣ −st
F (s) = − sin ωt + ∫ e cos ωt dt,
∣
s 0 s 0
so
ω
F (s) = G(s). (8.1.8)
s
If s > 0 , integrating Equation 8.1.7 by parts yields

−st ∞
e cos ωt ∞ ω
∣ −st
G(s) = − − ∫ e sin ωt dt,
∣0
s s 0
so

1 ω
G(s) = − F (s).
s s
Now substitute from Equation 8.1.8 into this to obtain

2
1 ω
G(s) = − G(s).
2
s s
Solving this for G(s) yields

s
G(s) = , s > 0.
2 2
s +ω
This and Equation 8.1.8 imply that

ω
F (s) = , s > 0.
s2 + ω2
Tables of Laplace Transforms

Extensive tables of Laplace transforms have been compiled and are commonly used in applications. The brief table of Laplace
transforms in the Appendix will be adequate for our purposes.
Example 8.1.5
Use the table of Laplace transforms to find L(t 3
e
4t
) .
The table includes the transform pair
n at
n!
t e ↔ .
n+1
(s − a)
Setting n = 3 and a = 4 here yields

3! 6
3 4t
L(t e ) = = .
4 4
(s − 4) (s − 4)
We’ll sometimes write Laplace transforms of specific functions without explicitly stating how they are obtained. In such cases
you should refer to the table of Laplace transforms.
Linearity of the Laplace Transform

The next theorem presents an important property of the Laplace transform.
Theorem 8.1.2 Linearity Property

Suppose L(f ) is defined for s > s
i i, 1 ≤ i ≤ n). Let s be the largest of the numbers s , s , …,s
0 1 2 n, and let c ,
1 c2 ,…, cn
be constants. Then
L(c1 f1 + c2 f2 + ⋯ + cn fn ) = c1 L(f1 ) + c2 L(f2 ) + ⋯ + cn L(fn ) for s > s0 .
Proof
We give the proof for the case where n = 2 . If s > s then 0
∞
−st
L(c1 f1 + c2 f2 ) = ∫ e (c1 f1 (t) + c2 f2 (t))) dt
0
∞ ∞
−st −st
= c1 ∫ e f1 (t) dt + c2 ∫ e f2 (t) dt
0 0
= c1 L(f1 ) + c2 L(f2 ).

Example 8.1.6
Use Theorem 8.1.2 and the known Laplace transform
at
1
L(e ) =
s−a
to find L(cosh bt) (b ≠ 0) .

Solution
By definition,
bt −bt
e +e
cosh bt = .
2
Therefore
1 bt 1 −bt
L(cosh bt) = L( e + e )
2 2
1 bt 1 −bt
= L(e )+ L(e ) (linearity property) (8.1.9)
2 2
1 1 1 1
= + ,
2 s−b 2 s+b
where the first transform on the right is defined for s > b and the second for s > −b ; hence, both are defined for s > |b| .
Simplifying the last expression in Equation 8.1.9 yields
s
L(cosh bt) = , s > |b|.
s2 − b2
The next theorem enables us to start with known transform pairs and derive others. (For other results of this kind, see
Exercises 8.1.6 and 8.1.13.)
Theorem 8.1.3 First Shifting Theorem

If
∞
−st
F (s) = ∫ e f (t) dt (8.1.10)
0
is the Laplace transform of f (t) for s > s , then F (s − a) is the Laplace transform of e
0
at
f (t) for s > s
0 +a .
Proof
Replacing s by s − a in Equation 8.1.10yields
∞
−(s−a)t
F (s − a) = ∫ e f (t) dt (8.1.11)
0
if s − a > s ; that is, if s > s

0 0 +a . However, Equation 8.1.11can be rewritten as
∞
−st at
F (s − a) = ∫ e (e f (t)) dt,
0
which implies the conclusion.
Example 8.1.7
Use Theorem 8.1.3 and the known Laplace transforms of 1, t , cos ωt, and sin ωt to find
at at λt λt
L(e ), L(te ), L(e sin ωt), and L(e cos ωt).
Solution

In the following table the known transform pairs are listed on the left and the required transform pairs listed on the right
are obtained by applying Theorem 8.1.3.
Table 8.1.1
at
f (t) ↔ F (s) e f (t) ↔ F (s − a)
1 at 1
1 ↔ , s > 0 e ↔ , s > a
s (s−a)
1 at 1
t ↔ , s > 0 te ↔ , s > a
2 2
s (s−a)
ω λt ω
sin ωt ↔ , s > 0 e sin ωt ↔ , s > λ
2 2 2
s +ω (s−λ ) +ω2
s λt s−λ
cosωt ↔ , s > 0 e sin ωt ↔ , s > λ
2 2 2
s +ω (s−λ ) +ω2
Existence of Laplace Transforms

Not every function has a Laplace transform. For example, it can be shown (Exercise 8.1.3) that
∞
2
−st t
∫ e e dt = ∞
0
for every real number s . Hence, the function f (t) = e does not have a Laplace transform.
2
t
Our next objective is to establish conditions that ensure the existence of the Laplace transform of a function. We first review
some relevant definitions from calculus.
Recall that a limit
lim f (t)
t→t0
exists if and only if the one-sided limits
lim f (t) and lim f (t)

t→t0 − t→t0 +
both exist and are equal; in this case,
lim f (t) = lim f (t) = lim f (t).

t→t0 t→t0 − t→t0 +
Recall also that f is continuous at a point t in an open interval (a, b) if and only if
0
lim f (t) = f (t0 ),

t→t0
lim f (t) = lim f (t) = f (t0 ). (8.1.12)

t→t0 + t→t0 −
For simplicity, we define
f (t0 +) = lim f (t) and f (t0 −) = lim f (t),

t→t0 + t→t0 −
so Equation 8.1.12 can be expressed as
f (t0 +) = f (t0 −) = f (t0 ).
If f (t
0 +) and f (t 0 −) have finite but distinct values, we say that f has a jump discontinuity at t , and 0
f (t0 +) − f (t0 −)
is called the jump in f at t (Figure 8.1.1).

0

Figure 8.1.1 : A jump discontinuity.
If f (t
0 +) and f (t
0 −) are finite and equal, but either f isn’t defined at t or it is defined but
0
f (t0 ) ≠ f (t0 +) = f (t0 −),
we say that f has a removable discontinuity at t (Figure 8.1.2). This terminolgy is appropriate since a function
0 f with a
removable discontinuity at t can be made continuous at t by defining (or redefining)
0 0
f (t0 ) = f (t0 +) = f (t0 −).
Note
We know from calculus that a definite integral isn’t affected by changing the values of its integrand at isolated points.
Therefore, redefining a function f to make it continuous at removable discontinuities does not change L(f ).
Theorem 8.1.4 Piecewise Continuous

A function f is said to be piecewise continuous on a finite closed interval [0, T ] if f (0+) and f (T −) are finite and f
is continuous on the open interval (0, T ) except possibly at finitely many points, where f may have jump
discontinuities or removable discontinuities.
A function f is said to be piecewise continuous on the infinite interval [0, ∞) if it is piecewise continuous on [0, T ]
for every T > 0 .
Figure 8.1.3 shows the graph of a typical piecewise continuous function.

It is shown in calculus that if a function is piecewise continuous on a finite closed interval then it is integrable on that interval.
But if f is piecewise continuous on [0, ∞), then so is e f (t), and therefore
−st
T
−st
∫ e f (t) dt
0
Figure 8.1.2

Figure 8.1.3
exists for every T >0 . However, piecewise continuity alone does not guarantee that the improper integral
∞ T
−st −st
∫ e f (t) dt = lim ∫ e f (t) dt (8.1.13)
T →∞
0 0
converges for s in some interval (s , ∞). For example, we noted earlier that Equation 8.1.13 diverges for all s if f (t) = e .
0
t
Stated informally, this occurs because e increases too rapidly as t → ∞ . The next definition provides a constraint on the
2
t
growth of a function that guarantees convergence of its Laplace transform for s in some interval (s , ∞) . 0
Theorem 8.1.5 : of exponential order

A function f is said to be of exponential order s if there are constants M and t such that
0 0
s0 t
|f (t)| ≤ M e , t ≥ t0 . (8.1.14)
In situations where the specific value of s is irrelevant we say simply that f is of exponential order.
0
The next theorem gives useful sufficient conditions for a function f to have a Laplace transform. The proof is sketched in
Exercise 8.1.10.
Theorem 8.1.6
If f is piecewise continuous on [0, ∞) and of exponential order s , then L(f ) is defined for s > s .
0 0
Note
We emphasize that the conditions of Theorem 8.1.6 are sufficient, but not necessary, for f to have a Laplace transform.
For example, Exercise 8.1.14(c) shows that f may have a Laplace transform even though f isn’t of exponential order
Example 8.1.8
If f is bounded on some interval [t 0, ∞) , say
|f (t)| ≤ M , t ≥ t0 ,
then Equation 8.1.14 holds with s = 0 , so f is of exponential order zero. Thus, for example, sin ωt and cos ωt are of
0
exponential order zero, and Theorem 8.1.6 implies that L(sin ωt) and L(cos ωt) exist for s > 0 . This is consistent with
the conclusion of Example 8.1.4.
Example 8.1.9
It can be shown that if lim e
t→∞ f (t) exists and is finite then f is of exponential order s (Exercise 8.1.9). If α is any
−s0 t
0
real number and s > 0 then f (t) = t is of exponential order s , since

0
α
0
−s0 t α
lim e t = 0,
t→∞

by L’Hôpital’s rule. If α ≥ 0 , f is also continuous on [0, ∞). Therefore Exercise 8.1.9 and Theorem 8.1.6 imply that
L(t ) exists for s ≥ s . However, since s is an arbitrary positive number, this really implies that L(t ) exists for all
α α
0 0
s > 0 . This is consistent with the results of Example 8.1.2 and Exercises 8.1.6 and 8.1.8.
Example 8.1.10
Find the Laplace transform of the piecewise continuous function
1, 0 ≤ t < 1,
f (t) = {
−t
−3 e , t ≥ 1.
Solution
Since f is defined by different formulas on [0, 1) and [1, ∞), we write
∞ 1 ∞
−st −st −st −t
F (s) = ∫ e f (t) dt = ∫ e (1) dt + ∫ e (−3 e ) dt.
0 0 1
Since
−s
1 1−e
−st s ≠0
s
∫ e dt = {
0 1 s =0
and
∞ ∞ −(s+1)
−st −t −(s+1)t
3e
∫ e (−3 e ) dt = −3 ∫ e dt = − , s > −1,
1 1
s+1
it follows that
−s −( s+1)
1−e e
−3 s > −1, s ≠ 0
s s+1
F (s) = {
3
1− s =0
e
This is consistent with Theorem 8.1.6, since

−t
|f (t)| ≤ 3 e , t ≥ 1,
and therefore f is of exponential order s 0 = −1 .
Note
In Section 8.4 we’ll develop a more efficient method for finding Laplace transforms of piecewise continuous functions.
Example 8.1.11
We stated earlier that
∞
2
−st t
∫ e e dt = ∞
0
for all s , so Theorem 8.1.6 implies that f (t) = e is not of exponential order, since
2
t
2
t
e 1 2
t −s0 t
lim = lim e = ∞,
s0 t
t→∞ Me t→∞ M
so
2
t s0 t
e > Me
for sufficiently large values of t , for any choice of M and s (Exercise 8.1.3). 0

8.1E: Introduction to the Laplace Transform (Exercises)
Q8.1.1
∞
1. Find the Laplace transforms of the following functions by evaluating the integral F (s) = ∫ 0
e
−st
f (t) dt .
a. t
b. te −t
c. sinh bt
d. e − 3e
2t t
e. t 2
2. Use the table of Laplace transforms to find the Laplace transforms of the following functions.
a. cosh t sin t
b. sin t 2
c. cos 2t 2
d. cosh t 2
e. t sinh 2t
f. sin t cos t
g. sin(t + ) π
h. cos 2t − cos 3t
i. sin 2t + cos 4t
3. Show that
∞
2
−st t
∫ e e dt = ∞
0
for every real number s .

4. Graph the following piecewise continuous functions and evaluate f (t+), f (t−), and f (t) at each point of discontinuity.
⎧ −t, 0 ≤ t < 2,
⎪
a. f (t) = ⎨ t − 4, 2 ≤ t < 3,
⎩
⎪
1, t ≥ 3.
2
⎧t
⎪ + 2, 0 ≤ t < 1,
b. f (t) = ⎨ 4, t = 1,
⎩
⎪
t, t > 1.
⎧ sin t, 0 ≤ t < π/2,

⎪
c. f (t) = ⎨ 2 sin t, π/2 ≤ t < π,

⎩
⎪
cos t, t ≥ π.
⎧ t, 0 ≤ t < 1,
⎪
⎪
⎪
⎪
⎪ 2, t = 1,
⎪
d. f (t) = ⎨ 2 − t, 1 ≤ t < 2,
⎪
⎪
⎪
⎪ 3, t = 2,
⎪
⎩
⎪
6, t > 2.
5. Find the Laplace transform:

−t
e , 0 ≤ t < 1,
a. f (t) = {
−2t
e , t ≥ 1.
1, 0 ≤ t < 4,
b. f (t) = {
t, t ≥ 4.
t, 0 ≤ t < 1,
c. f (t) = {
1, t ≥ 1.

t
te , 0 ≤ t < 1,
d. f (t) = { t
e , t ≥ 1.
6. Prove that if f (t) ↔ F (s) then t f (t) ↔ (−1) F (s) . HINT: Assume that it's permissible to differentiate the integral
k k (k)
∞
∫
0
e
−st
f (t)dt with respect to s under the integral sign.
7. Use the known Laplace transforms
ω s−λ
λt λt
L(e sin ωt) = and L(e cos ωt) =
2 2 2 2
(s − λ ) +ω (s − λ ) +ω
and the result of Exercise 8.1.6 to find L(te λt

cos ωt) and L(te λt
sin ωt) .
8. Use the known Laplace transform L(1) = 1/s and the result of Exercise 8.1.6 to show that
n
n!
L(t ) = , n = integer.
n+1
s
9.
a. Show that if lim e f (t) exists and is finite then f is of exponential order s .
t→∞
−s0 t
0
b. Show that if f is of exponential order s then lim e f (t) = 0 for all s > s .
0 t→∞
−st
0
c. Show that if f is of exponential order s and g(t) = f (t + τ ) where τ > 0 , then g is also of exponential order s .
0 0
10. Recall the next theorem from calculus.
Theorem 8.1E . 1
Let g be integrable on [0, T ] for every T > 0. Suppose there’s a function w defined on some interval [τ , ∞) (with τ ≥0 )
∞ ∞
such that |g(t)| ≤ w(t) for t ≥ τ and ∫ τ
converges. Then ∫ g(t) dt converges.
w(t) dt
0
Use Theorem 8.1E. 1 to show that if f is piecewise continuous on [0, ∞) and of exponential order s0 , then f has a Laplace
transform F (s) defined for s > s . 0
11. Prove: If f is piecewise continuous and of exponential order then lim s→∞ F (s) = 0 .
12. Prove: If f is continuous on [0, ∞) and of exponential order s 0 >0 , then
t
1
L (∫ f (τ ) dτ ) = L(f ), s > s0 .
0
s
HINT: Use integration by parts to evaluate the transform on the left.

13. Suppose f is piecewise continuous and of exponential order, and that lim t→0+ f (t)/t exists. Show that
∞
f (t)
L( ) =∫ F (r) dr.
t s
HINT: Use the results of Exercises 8.1.6 and 8.1.11.

14. Suppose f is piecewise continuous on [0, ∞).
t
a. Prove: If the integral g(t) = ∫ e f (τ ) dτ satisfies the inequality |g(t)| ≤ M
0
−s0 τ
(t ≥ 0) , then f has a Laplace transform
F (s) defined for s > s . HINT: Use integration by parts to show that
0
T T
−st −(s−s0 )T −(s−s0 )t
∫ e f (t)dt = e g(T ) + (s − s0 ) ∫ e g(t)dt
0 0
b. Show that if L(f ) exists for s = s then it exists for s > s . Show that the function
0 0
2 2
t t
f (t) = te cos(e )
has a Laplace transform defined for s > 0 , even though f isn’t of exponential order.
c. Show that the function

2 2
t t
f (t) = te cos(e )
has a Laplace transform defined for s > 0 , even though f isn’t of exponential order.
15. Use the table of Laplace transforms and the result of Exercise 8.1.13 to find the Laplace transforms of the following
functions.
a. sin ωt
t
(ω > 0)
b. cos ωt−1
t
(ω > 0)
at bt
c. e −e
d. cosh t−1
t
2
e. sinh
t
t
16. The gamma function is defined by

∞
α−1 −x
Γ(α) = ∫ x e dx,
0
which can be shown to converge if α > 0 .

a. Use integration by parts to show that
Γ(α + 1) = αΓ(α), α > 0.
b. Show that Γ(n + 1) = n! if n = 1 , 2, 3,….

c. From (b) and the table of Laplace transforms,
Γ(α + 1)
α
L(t ) = , s > 0,
α+1
s
if α is a nonnegative integer. Show that this formula is valid for any α > −1 . HINT: Change the variable of integration in
the integral for Γ(α + 1) .
17. Suppose f is continuous on [0, T ] and f (t + T ) = f (t) for all t ≥0 . (We say in this case that f is periodic with period
T .)
a. Conclude from Theorem 8.1.6 that the Laplace transform of f is defined for s > 0 .
b. Show that
T
1 −st
F (s) = ∫ e f (t) dt, s > 0.
−sT
1 −e 0
HINT: Write
∞ (n+1)T
−st
F (s) = ∑ ∫ e f (t)dt
n=0 nT
Then show that

(n+1)T T
−st −nsT −st
∫ e f (t)dt = e ∫ e f (t)dt
nT 0
and recall the formula for the sum of a geometric series.

18. Use the formula given in Exercise 8.1.17b to find the Laplace transforms of the given periodic functions:
t, 0 ≤ t < 1,
a. f (t) = { f (t + 2) = f (t), t ≥0
2 − t, 1 ≤ t < 2,
1
1, 0 ≤t < ,
b. f (t) = { 1
2
f (t + 1) = f (t), t ≥0
−1, ≤ t < 1,
2
c. f (t) = | sin t|

sin t, 0 ≤ t < π,
d. f (t) = { f (t + 2π) = f (t)
0, π ≤ t < 2π,

8.2: The Inverse Laplace Transform
Definition of the Inverse Laplace Transform
In Section 8.1 we defined the Laplace transform of f by
∞
−st
F (s) = L(f ) = ∫ e f (t) dt.
0
We’ll also say that f is an inverse Laplace Transform of F , and write

−1
f =L (F ).
To solve differential equations with the Laplace transform, we must be able to obtain f from its transform F . There’s a
formula for doing this, but we can’t use it because it requires the theory of functions of a complex variable. Fortunately, we
can use the table of Laplace transforms to find inverse transforms that we’ll need.
Example 8.2.1 :
Use the table of Laplace transforms to find
a. −1
1
L ( )
2
s −1
b. −1
s
L ( ).
2
s +9
Solution a
Setting b = 1 in the transform pair
b
sinh bt ↔
2 2
s −b
shows that
−1
1
L ( ) = sinh t.
s2 − 1
Solution b
Setting ω = 3 in the transform pair
s
cos ωt ↔
2 2
s +ω
shows that
s
−1
L ( ) = cos 3t.
2
s +9
The next theorem enables us to find inverse transforms of linear combinations of transforms in the table. We omit the proof.
Theorem 8.2.1 : Linearity Property

If F 1, F2 , …, F are Laplace transforms and c
n 1, c2 , …, c are constants, then
n
−1 −1 −1 −1
L (c1 F1 + c2 F2 + ⋯ + cn Fn ) = c1 L (F1 ) + c2 L (F2 ) + ⋯ + cn L Fn .

Example 8.2.2
Find
−1
8 7
L ( + ).
2
s+5 s +3
Solution
From the table of Laplace transforms in Section 8.8,,
1 ω
at
e ↔ and sin ωt ↔ .
2 2
s−a s +ω
–
Theorem 8.2.1 with a = −5 and ω = √3 yields
−1
8 7 −1
1 −1
1
L ( + ) = 8L ( ) + 7L ( )
2 2
s+5 s +3 s+5 s +3
–
1 7 √3
−1 −1
= 8L ( )+ –L ( )
s+5 √3 s2 + 3
−5t
7 –
= 8e + – sin √3t.
√3
Example 8.2.3
Find
−1
3s + 8
L ( ).
2
s + 2s + 5
Solution
Completing the square in the denominator yields
3s + 8 3s + 8
= .
2 2
s + 2s + 5 (s + 1 ) +4
Because of the form of the denominator, we consider the transform pairs

s+1 2
−t −t
e cos 2t ↔ and e sin 2t ↔ ,
2 2
(s + 1 ) +4 (s + 1 ) +4
and write
3s + 8 3s + 3 5
−1 −1 −1
L ( ) =L ( ) +L ( )
2 2 2
(s + 1 ) +4 (s + 1 ) +4 (s + 1 ) +4
−1
s+1 5 −1
2
= 3L ( )+ L ( )
2
(s + 1 ) +4 2 (s + 1 )2 + 4
−t
5
=e (3 cos 2t + sin 2t).
2
Note
We’ll often write inverse Laplace transforms of specific functions without explicitly stating how they are obtained. In
such cases you should refer to the table of Laplace transforms in Section 8.8.
Inverse Laplace Transforms of Rational Functions

Using the Laplace transform to solve differential equations often requires finding the inverse transform of a rational function

P (s)
F (s) = ,
Q(s)
where P and Q are polynomials in s with no common factors. Since it can be shown that lim F (s) = 0 if F is a Laplace
s→∞
transform, we need only consider the case where degree(P ) < degree(Q). To obtain L (F ), we find the partial fraction
−1
expansion of F , obtain inverse transforms of the individual terms in the expansion from the table of Laplace transforms, and
use the linearity property of the inverse transform. The next two examples illustrate this.
Example 8.2.4
Find the inverse Laplace transform of
3s + 2
F (s) = . (8.2.1)
2
s − 3s + 2
Solution
(Method 1)
Factoring the denominator in Equation 8.2.1 yields
3s + 2
F (s) = . (8.2.2)
(s − 1)(s − 2)
The form for the partial fraction expansion is

3s + 2 A B
= + . (8.2.3)
(s − 1)(s − 2) s−1 s−2
Multiplying this by (s − 1)(s − 2) yields
3s + 2 = (s − 2)A + (s − 1)B.
Setting s = 2 yields B = 8 and setting s = 1 yields A = −5 . Therefore

5 8
F (s) = − +
s−1 s−2
and
−1 −1
1 −1
1
t 2t
L (F ) = −5 L ( ) + 8L ( ) = −5 e + 8 e .
s−1 s−2
(Method 2) We don’t really have to multiply Equation 8.2.3 by (s − 1)(s − 2) to compute A and B . We can obtain A by
simply ignoring the factor s − 1 in the denominator of Equation 8.2.2 and setting s = 1 elsewhere; thus,
3s + 2 ∣ 3 ⋅ 1 +2
A = ∣ = = −5. (8.2.4)
s−2 ∣ s=1
1 −2
Similarly, we can obtain B by ignoring the factor s−2 in the denominator of Equation 8.2.2 and setting s =2
elsewhere; thus,
3s + 2 ∣ 3 ⋅ 2 +2
B = ∣ = = 8. (8.2.5)
s − 1 ∣s=2 2 −1
To justify this, we observe that multiplying Equation 8.2.3 by s − 1 yields

3s + 2 B
= A + (s − 1) ,
s−2 s−2
and setting s = 1 leads to Equation 8.2.4. Similarly, multiplying Equation 8.2.3 by s − 2 yields
3s + 2 A
= (s − 2) +B
s−1 s−2

and setting s = 2 leads to Equation 8.2.5. (It isn’t necesary to write the last two equations. We wrote them only to justify
the shortcut procedure indicated in Equation 8.2.4 and Equation 8.2.5.)
The shortcut employed in the second solution of Example 8.2.4 is Heaviside’s method. The next theorem states this method
formally. For a proof and an extension of this theorem, see Exercise 8.2.10.
Theorem 8.2.2
Suppose
P (s)
F (s) = , (8.2.6)
(s − s1 )(s − s2 ) ⋯ (s − sn )
where s , s , …, s are distinct and P is a polynomial of degree less than n. Then

1 2 n
A1 A2 An
F (s) = + +⋯ + ,
s − s1 s − s2 s − sn
where A can be computed from Equation 8.2.6 by ignoring the factor s − s and setting s = s elsewhere.
i i i
Example 8.2.5
2
6 + (s + 1)(s − 5s + 11)
F (s) = . (8.2.7)
s(s − 1)(s − 2)(s + 1)
Solution
The partial fraction expansion of Equation 8.2.7 is of the form
A B C D
F (s) = + + + . (8.2.8)
s s−1 s−2 s+1
To find A , we ignore the factor s in the denominator of Equation 8.2.7 and set s = 0 elsewhere. This yields
6 + (1)(11) 17
A = = .
(−1)(−2)(1) 2
Similarly, the other coefficients are given by

6 + (2)(7)
B = = −10,
(1)(−1)(2)
6 + 3(5) 7
C = = ,
2(1)(3) 2
and
6
D = = −1.
(−1)(−2)(−3)
Therefore
17 1 10 7 1 1
F (s) = − + −
2 s s−1 2 s−2 s+1
and

−1
17 −1
1 −1
1 7 −1
1 −1
1
L (F ) = L ( ) − 10 L ( )+ L ( ) −L ( )
2 s s−1 2 s−2 s+1
17 t
7 2t −t
= − 10 e + e −e .
2 2
Note
We didn’t “multiply out” the numerator in Equation 8.2.7 before computing the coefficients in Equation , since it
8.2.8
wouldn’t simplify the computations.
Example 8.2.6
8 − (s + 2)(4s + 10)
F (s) = . (8.2.9)
2
(s + 1)(s + 2)
Solution
A B C
F (s) = + + . (8.2.10)
2
s+1 s+2 (s + 2)
Because of the repeated factor (s + 2) in Equation 8.2.9, Heaviside’s method doesn’t work. Instead, we find a common
2
denominator in Equation 8.2.10. This yields

2
A(s + 2 ) + B(s + 1)(s + 2) + C (s + 1)
F (s) = . (8.2.11)
2
(s + 1)(s + 2)
If Equation 8.2.9 and Equation 8.2.11 are to be equivalent, then

2
A(s + 2 ) + B(s + 1)(s + 2) + C (s + 1) = 8 − (s + 2)(4s + 10). (8.2.12)
The two sides of this equation are polynomials of degree two. From a theorem of algebra, they will be equal for all s if
they are equal for any three distinct values of s . We may determine A , B and C by choosing convenient values of s .
The left side of Equation 8.2.12 suggests that we take s = −2 to obtain C = −8 , and s = −1 to obtain A = 2 . We can
now choose any third value of s to determine B . Taking s = 0 yields 4A + 2B + C = −12 . Since A = 2 and C = −8
this implies that B = −6 . Therefore
2 6 8
F (s) = − −
2
s+1 s+2 (s + 2)
and
1 1 1
−1 −1 −1 −1
L (F ) = 2 L ( ) − 6L ( ) − 8L ( )
2
s+1 s+2 (s + 2)
−t −2t −2t
= 2e − 6e − 8te .
Example 8.2.7
2
s − 5s + 7
F (s) = .
(s + 2)3
Solution

A B C
F (s) = + + .
2 3
s+2 (s + 2) (s + 2)
The easiest way to obtain A , B , and C is to expand the numerator in powers of s + 2 . This yields
2 2 2
s − 5s + 7 = [(s + 2) − 2 ] − 5[(s + 2) − 2] + 7 = (s + 2 ) − 9(s + 2) + 21.
Therefore
2
(s + 2 ) − 9(s + 2) + 21
F (s) =
3
(s + 2)
1 9 21
= − +
2 3
s+2 (s + 2) (s + 2)
and
−1 −1
1 −1
1 21 −1
2
L (F ) = L ( ) − 9L ( )+ L ( )
2
s+2 (s + 2) 2 (s + 2)3
21
−2t 2
=e (1 − 9t + t ).
2
Example 8.2.8
1 − s(5 + 3s)
F (s) = . (8.2.13)
2
s [(s + 1 ) + 1]
Solution
One form for the partial fraction expansion of F is
A Bs + C
F (s) = + . (8.2.14)
2
s (s + 1 ) +1
However, we see from the table of Laplace transforms that the inverse transform of the second fraction on the right of
Equation 8.2.14 will be a linear combination of the inverse transforms
−t −t
e cos t and e sin t
of
s+1 1
and
(s + 1 )2 + 1 (s + 1 )2 + 1
respectively. Therefore, instead of Equation 8.2.14 we write
A B(s + 1) + C
F (s) = + . (8.2.15)
2
s (s + 1 ) +1
Finding a common denominator yields

2
A [(s + 1 ) + 1] + B(s + 1)s + C s
F (s) = . (8.2.16)
2
s [(s + 1 ) + 1]
If Equation 8.2.13 and Equation 8.2.16 are to be equivalent, then

2
A [(s + 1 ) + 1] + B(s + 1)s + C s = 1 − s(5 + 3s).
This is true for all s if it is true for three distinct values of s . Choosing s = 0 , −1, and 1 yields the system

2A = 1
A−C = 3
5A + 2B + C = −7.
Solving this system yields

1 7 5
A = , B =− , C =− .
2 2 2
Hence, from Equation 8.2.15,

1 7 s+1 5 1
F (s) = − − .
2 2
2s 2 (s + 1 ) +1 2 (s + 1 ) +1
Therefore
−1
1 −1
1 7 −1
s+1 5 −1
1
L (F ) = L ( )− L ( )− L ( )
2 2
2 s 2 (s + 1 ) +1 2 (s + 1 ) +1
1 7 5
−t −t
= − e cos t − e sin t.
2 2 2
Example 8.2.9 :
8 + 3s
F (s) = . (8.2.17)
2 2
(s + 1)(s + 4)
Solution
A + Bs C + Ds
F (s) = + .
2 2
s +1 s +4
The coefficients A , B , C and D can be obtained by finding a common denominator and equating the resulting numerator
to the numerator in Equation 8.2.17. However, since there’s no first power of s in the denominator of Equation 8.2.17,
there’s an easier way: the expansion of
1
F1 (s) =
2 2
(s + 1)(s + 4)
can be obtained quickly by using Heaviside’s method to expand

1 1 1 1
= ( − )
(x + 1)(x + 4) 3 x +1 x +4
and then setting x = s to obtain

2
1 1 1 1
= ( − ).
2 2 2 2
(s + 1)(s + 4) 3 s +1 s +4
Multiplying this by 8 + 3s yields

8 + 3s 1 8 + 3s 8 + 3s
F (s) = = ( − ).
2 2 2 2
(s + 1)(s + 4) 3 s +1 s +4
Therefore
−1
8 4
L (F ) = sin t + cos t − sin 2t − cos 2t.
3 3

Using Technology
Some software packages that do symbolic algebra can find partial fraction expansions very easily. We recommend that you use
such a package if one is available to you, but only after you’ve done enough partial fraction expansions on your own to master
the technique.

8.2E: The Inverse Laplace Transform (Exercises)
Q8.2.1
1. Use the table of Laplace transforms to find the inverse Laplace transform.
a. 3
4
(s−7)
b. 2s−4
s2 −4s+13
c. s2 +4s+20
1
d. 2
s +9
2
2
s −1
e. 2
2
( s +1 )
f. 1
2
(s−2 ) −4
g. 2
12s−24
2
( s −4s+85 )
h. 2
2
(s−3 ) −9
2
i. s −4s+3
2
2
( s −4s+5 )
2. Use Theorem 8.2.1 and the table of Laplace transforms to find the inverse Laplace transform.
a. 2s+3
4
(s−7)
2
s −1
b. 6
(s−2)
s+5
c. 2
s +6s+18
d. 2s+1
s2 +9
e. 2
s +2s+1
s
s+1
f. 2
s −9
3 2
g. s +2 s −s−3
4
(s+1)
h. 2s+3
2
(s−1 ) +4
i. 1
s
−
s2 +1
s
j. 3s+4
s2 −1
k. s−1
3
+
4s+1
s +9
2
l. 3
2
−
2s+6
s +4
2
(s+2)
3. Use Heaviside’s method to find the inverse Laplace transform.

3−(s+1)(s−2)
a. (s+1)(s+2)(s−2)
7+(s+4)(18−3s)
b. (s−3)(s−1)(s+4)
2+(s−2)(3−2s)
c. (s−2)(s+2)(s−3)
3−(s−1)(s+1)
d. (s+4)(s−2)(s−1)
2
3+(s−2)(10−2s−s )
e. (s−2)(s+2)(s−1)(s+3)
2
3+(s−3)(2 s +s−21)
f. (s−3)(s−1)(s+4)(s−2)
4. Find the inverse Laplace transform.

2+3s
a. 2
( s +1)(s+2)(s+1)
2
3 s +2s+1
b. 2
( s +1)( s +2s+2)
2
c. (s−2)( s +2s+5)
3s+2
2
9/11/2020 8.2E.1 https://math.libretexts.org/@go/page/44283

2
d. 3 s +2s+1
2
(s−1 ) (s+2)(s+3)
2
2 s +s+3
e. 2 2
(s−1 ) (s+2 )
3s+2
f. 2
( s2 +1)(s−1 )
5. Use the method of Example 8.2.9 to find the inverse Laplace transform.
a. 3s+2
( s2 +4)( s2 +9)
b. 2
−4s+1
( s +1)( s +16)
2
c. 2
5s+3
( s +1)( s +4)
2
−s+1
d. 2
(4 s +1)( s +1)
2
e. 2
17s−34
( s +16)(16 s +1)
2
f. 2s−1
(4 s2 +1)(9 s2 +1)

a. 17s−15
( s2 −2s+5)( s2 +2s+10)
b. 2
8s+56
( s −6s+13)( s +2s+5)
2
c. 2
( s +4s+5)( s −4s+13)
s+9
2
d. ( s2 −4s+5)( s2 −6s+13)
3s−2
3s−1
e. 2
( s −2s+2)( s +2s+5)
2
f. 2
20s+40
(4 s −4s+5)(4 s +4s+5)
2

a. 1
s( s2 +1)
b. (s−1)( s −2s+17)
1
2
c. 3s+2
(s−2)( s2 +2s+10)
34−17s
d. (2s−1)( s −2s+5)
2
e. s+2
(s−3)( s +2s+5)
2
f. 2s−2
(s−2)( s2 +2s+10)

a. ( s2 +1)(s−1)(s−3)
2s+1
b. 2
( s +2s+2)( s −1)
s+2
2
c. 2
2s−1
( s −2s+2)(s+1)(s−2)
d. s−6
( s2 −1)( s2 +4)
2s−3
e. s(s−2)( s −2s+5)
2
f. 2
5s−15
( s −4s+13)(s−2)(s−1)
9. Given that f (t) ↔ F (s) , find the inverse Laplace transform of F (as − b) , where a > 0 .
10.
a. If s , s , …, s are distinct and P is a polynomial of degree less than n , then
1 2 n
P (s) A1 A2 An
= + +⋯ + .
(s − s1 )(s − s2 ) ⋯ (s − sn ) s − s1 s − s2 s − sn
Multiply through by s − si to show that Ai can be obtained by ignoring the factor s − si on the left and setting s = si
elsewhere.

b. Suppose P and Q are polynomials such that
1 degree(P ) ≤ degree(Q1 ) and Q1 (s1 ) ≠ 0 . Show that the coefficient of
1/(s − s1 ) in the partial fraction expansion of
P (s)
F (s) =
(s − s1 )Q1 (s)
is P (s )/Q (s ).
1 1 1
c. Explain how the results of (a) and (b) are related.

8.3: Solution of Initial Value Problems
Laplace Transforms of Derivatives
In the rest of this chapter we’ll use the Laplace transform to solve initial value problems for constant coefficient second order
equations. To do this, we must know how the Laplace transform of f is related to the Laplace transform of f . The next ′
theorem answers this question.
Theorem 8.3.1
Suppose f is continuous on [0, ∞) and of exponential order s , and f is piecewise continuous on [0, ∞). Then f and f 0
′ ′
have Laplace transforms for s > s , and 0
′
L(f ) = sL(f ) − f (0). (8.3.1)
Proof
We know from Theorem 8.1.6 that L(f ) is defined for s > s . We first consider the case where 0 f
′
is continuous on
[0, ∞). Integration by parts yields
T T
T
−st ′ −st ∣ −st
∫ e f (t) dt = e f (t) +s∫ e f (t) dt
∣
0
0 0
T
−sT −st
=e f (T ) − f (0) + s ∫ e f (t) dt (8.3.2)
0
for any T >0 . Since f is of exponential order s0 , lim e

−sT
f (T ) = 0 and the integral in on the right side of
T →∞
Equation 8.3.2converges as T → ∞ if s > s . Therefore

0
∞ ∞
−st ′ −st
∫ e f (t) dt = −f (0) + s ∫ e f (t) dt
0 0
= −f (0) + sL(f ),
which proves Equation 8.3.1.

Suppose T > 0 and f is only piecewise continuous on [0, T ], with discontinuities at
′
t1 < t2 < ⋯ < tn−1 . For
convenience, let t = 0 and t = T . Integrating by parts yields
0 n
ti ti
ti
−st ′ −st ∣ −st
∫ e f (t) dt = e f (t) +s∫ e f (t) dt
∣
ti− 1
ti− 1 ti− 1
ti
−sti −sti− 1 −st
=e f (ti ) − e f (ti−1 ) + s ∫ e f (t) dt.
ti− 1
Summing both sides of this equation from i = 1 to n and noting that

−st1 −st0 −st2 −st1 −stN −stN − 1
(e f (t1 ) − e f (t0 )) + (e f (t2 ) − e f (t1 )) + ⋯ + (e f (tN ) − e f (tN −1 ))
−stN −st0 −sT

=e f (tN ) − e f (t0 ) = e f (T ) − f (0)
yields Equation 8.3.2, so Equation 8.3.1follows as before.
Example 8.3.1 :
In Example 8.1.4 we saw that
s
L(cos ωt) = .
2 2
s +ω
Solution

Applying Equation 8.3.1 with f (t) = cos ωt shows that
2
s ω
L(−ω sin ωt) = s −1 = − .
2 2 2 2
s +ω s +ω
Therefore
ω
L(sin ωt) = ,
2 2
s +ω
which agrees with the corresponding result obtained in 8.1.4.
In Section 2.1 we showed that the solution of the initial value problem
′
y = ay, y(0) = y0 , (8.3.3)
is y = y 0e
at
. We’ll now obtain this result by using the Laplace transform.
Let Y (s) = L(y) be the Laplace transform of the unknown solution of Equation 8.3.3 . Taking Laplace transforms of both
sides of Equation 8.3.3 yields
′
L(y ) = L(ay),
which, by Theorem 8.3.1, can be rewritten as
sL(y) − y(0) = aL(y),
or
sY (s) − y0 = aY (s).
Solving for Y (s) yields

y0
Y (s) = ,
s−a
so
−1 −1
y0 −1
1 at
y =L (Y (s)) = L ( ) = y0 L ( ) = y0 e ,
s−a s−a
which agrees with the known result.

We need the next theorem to solve second order differential equations using the Laplace transform.
Theorem 8.3.2
Suppose f and f are continuous on [0, ∞) and of exponential order
′
s0 , and that f
′′
is piecewise continuous on [0, ∞).
Then f , f , and f have Laplace transforms for s > s ,

′ ′′
0
′
L(f ) = sL(f ) − f (0), (8.3.4)
and
′′ 2 ′
L(f ) = s L(f ) − f (0) − sf (0). (8.3.5)
Proof
Theorem 8.3.1 implies that L(f ) exists and satisfies Equation 8.3.4 for s > s . To prove that L(f ) exists and
′
0
′′
satisfies Equation 8.3.5 for s > s , we first apply Theorem 8.3.1 to g = f . Since g satisfies the hypotheses of
0
′
Theorem 8.3.1, we conclude that L(g ) is defined and satisfies

′
′
L(g ) = sL(g) − g(0)
for s > s . However, since g

0
′
=f
′′
, this can be rewritten as

′′ ′ ′
L(f ) = sL(f ) − f (0).
Substituting Equation 8.3.4into this yields Equation 8.3.5.
Solving Second Order Equations with the Laplace Transform

We’ll now use the Laplace transform to solve initial value problems for second order equations.
Example 8.3.2
Use the Laplace transform to solve the initial value problem
′′ ′ 2t ′
y − 6 y + 5y = 3 e , y(0) = 2, y (0) = 3. (8.3.6)
Solution
Taking Laplace transforms of both sides of the differential equation in Equation 8.3.6 yields
′′ ′ 2t
3
L(y − 6 y + 5y) = L (3 e ) = ,
s−2
which we rewrite as
3
′′ ′
L(y ) − 6L(y ) + 5L(y) = . (8.3.7)
s−2
Now denote L(y) = Y (s) . Theorem 8.3.2 and the initial conditions in Equation 8.3.6 imply that
′
L(y ) = sY (s) − y(0) = sY (s) − 2
and
′′ 2 ′ 2
L(y ) = s Y (s) − y (0) − sy(0) = s Y (s) − 3 − 2s.
Substituting from the last two equations into Equation 8.3.7 yields
3
2
(s Y (s) − 3 − 2s) − 6 (sY (s) − 2) + 5Y (s) = .
s−2
Therefore
2
3
(s − 6s + 5)Y (s) = + (3 + 2s) + 6(−2), (8.3.8)
s−2
so
3 + (s − 2)(2s − 9)
(s − 5)(s − 1)Y (s) = ,
s−2
and
3 + (s − 2)(2s − 9)
Y (s) = .
(s − 2)(s − 5)(s − 1)
Heaviside’s method yields the partial fraction expansion

1 1 1 5 1
Y (s) = − + + ,
s−2 2 s−5 2 s−1
and taking the inverse transform of this yields

1 5
2t 5t t
y = −e + e + e
2 2
as the solution of Equation 8.3.6.

It isn’t necessary to write all the steps that we used to obtain Equation 8.3.8. To see how to avoid this, let’s apply the method
of Example 8.3.2 to the general initial value problem
′′ ′ ′
ay + b y + cy = f (t), y(0) = k0 , y (0) = k1 . (8.3.9)
Taking Laplace transforms of both sides of the differential equation in Equation 8.3.9 yields
′′ ′
aL(y ) + bL(y ) + cL(y) = F (s). (8.3.10)
Now let Y (s) = L(y) . Theorem 8.3.2 and the initial conditions in Equation 8.3.9 imply that
′ ′′ 2
L(y ) = sY (s) − k0 and L(y ) = s Y (s) − k1 − k0 s.
Substituting these into Equation 8.3.10 yields

2
a (s Y (s) − k1 − k0 s) + b (sY (s) − k0 ) + cY (s) = F (s). (8.3.11)
The coefficient of Y (s) on the left is the characteristic polynomial

2
p(s) = as + bs + c
of the complementary equation for Equation 8.3.9 . Using this and moving the terms involving k0 and k1 to the right side of
p(s)Y (s) = F (s) + a(k1 + k0 s) + b k0 . (8.3.12)
This equation corresponds to Equation 8.3.8 of Example 8.3.2. Having established the form of this equation in the general
case, it is preferable to go directly from the initial value problem to this equation. You may find it easier to remember Equation
8.3.12 rewritten as
′
p(s)Y (s) = F (s) + a (y (0) + sy(0)) + by(0). (8.3.13)
Example 8.3.3
Use the Laplace transform to solve the initial value problem
′′ ′ −2t ′
2y + 3y + y = 8e , y(0) = −4, y (0) = 2. (8.3.14)
Solution
The characteristic polynomial is
2
p(s) = 2 s + 3s + 1 = (2s + 1)(s + 1)
and
8
−2t
F (s) = L(8 e ) = ,
s+2

8
(2s + 1)(s + 1)Y (s) = + 2(2 − 4s) + 3(−4).
s+2

4 (1 − (s + 2)(s + 1))
Y (s) = .
(s + 1/2)(s + 1)(s + 2)
Heaviside’s method yields the partial fraction expansion

4 1 8 8 1
Y (s) = − + ,
3 s + 1/2 s+1 3 s+2
so the solution of Equation 8.3.14 is

−1
4 −t/2 −t
8 −2t
y =L (Y (s)) = e − 8e + e
3 3
(Figure 8.3.1).
Figure 8.3.1 : y = 4
3
e
−t/2
− 8e
−t
+
8
3
−2t
e
Example 8.3.4
′′ ′ ′
y + 2 y + 2y = 1, y(0) = −3, y (0) = 1. (8.3.15)
Solution
The characteristic polynomial is
2 2
p(s) = s + 2s + 2 = (s + 1 ) +1
and
1
F (s) = L(1) = ,
s

1
2
[(s + 1 ) + 1] Y (s) = + 1 ⋅ (1 − 3s) + 2(−3).
s

1 − s(5 + 3s)
Y (s) = .
2
s [(s + 1 ) + 1]
In Example 8.2.8 we found the inverse transform of this function to be

1 7 5
−t −t
y = − e cos t − e sin t
2 2 2
(Figure 8.3.2), which is therefore the solution of Equation 8.3.15.
Figure 8.3.2 : y = 1
2
−
7
2
e
−t
cos t −
5
2
e
−t
sin t
Note

In our examples we applied Theorems 8.3.1 and 8.3.2 without verifying that the unknown function y satisfies their
hypotheses. This is characteristic of the formal manipulative way in which the Laplace transform is used to solve
differential equations. Any doubts about the validity of the method for solving a given equation can be resolved by
verifying that the resulting function y is the solution of the given problem.

8.3E: Solution of Initial Value Problems (Exercises)
Q8.3.1
In Exercises 8.3.1-8.3.31 use the Laplace transform to solve the initial value problem.
1. y ′′ ′
+ 3 y + 2y = e ,
t
y(0) = 1,
′
y (0) = −6
2. y ′′ ′
− y − 6y = 2, y(0) = 1,
′
y (0) = 0
3. y ′′ ′
+ y − 2y = 2 e
3t
, y(0) = −1,
′
y (0) = 4
4. y ′′
− 4y = 2 e
3t
, y(0) = 1,
′
y (0) = −1
5. y ′′ ′
+ y − 2y = e
3t
, y(0) = 1,
′
y (0) = −1
6. y ′′ ′
+ 3 y + 2y = 6 e ,
t
y(0) = 1,
′
y (0) = −1
7. y ′′
+ y = sin 2t, y(0) = 0,
′
y (0) = 1
8. y ′′ ′
− 3 y + 2y = 2 e
3t
, y(0) = 1,
′
y (0) = −1
9. y ′′
− 3 y + 2y = e
′ 4t
, y(0) = 1,
′
y (0) = −2
10. y ′′
− 3 y + 2y = e
′ 3t
, y(0) = −1,
′
y (0) = −4
11. y ′′ ′
+ 3 y + 2y = 2 e ,
t
y(0) = 0,
′
y (0) = −1
12. y ′′ ′
+ y − 2y = −4, y(0) = 2,
′
y (0) = 3
13. y ′′
+ 4y = 4, y(0) = 0,
′
y (0) = 1
14. y ′′ ′
− y − 6y = 2, y(0) = 1,
′
y (0) = 0
15. y ′′ ′
+ 3 y + 2y = e ,
t
y(0) = 0,
′
y (0) = 1
16. y ′′
− y = 1, y(0) = 1,
′
y (0) = 0
17. y ′′
+ 4y = 3 sin t, y(0) = 1,
′
y (0) = −1
18. y ′′
+y
′
= 2e
3t
, y(0) = −1,
′
y (0) = 4
19. y ′′
+ y = 1, y(0) = 2,
′
y (0) = 0
20. y ′′
+ y = t, y(0) = 0,
′
y (0) = 2
21. y ′′
+ y = t − 3 sin 2t, y(0) = 1,
′
y (0) = −3
22. y ′′ ′
+ 5 y + 6y = 2 e
−t
, y(0) = 1,
′
y (0) = 3
23. y ′′ ′
+ 2 y + y = 6 sin t − 4 cos t, y(0) = −1, y (0) = 1
′
24. y ′′ ′
− 2 y − 3y = 10 cos t, y(0) = 2,
′
y (0) = 7
25. y ′′
+ y = 4 sin t + 6 cos t, y(0) = −6, y (0) = 2
′
26. y ′′
+ 4y = 8 sin 2t + 9 cos t, y(0) = 1,
′
y (0) = 0
27. y ′′ ′
− 5 y + 6y = 10 e cos t,
t
y(0) = 2,
′
y (0) = 1
28. y ′′ ′
+ 2 y + 2y = 2t, y(0) = 2,
′
y (0) = −7
29. y ′′ ′
− 2 y + 2y = 5 sin t + 10 cos t, y(0) = 1, y (0) = 2
′
30. y ′′ ′
+ 4 y + 13y = 10 e
−t
− 36 e ,
t
y(0) = 0, y (0) = −16
′
31. y ′′
+ 4 y + 5y = e
′ −t
(cos t + 3 sin t), y(0) = 0,
′
y (0) = 4
Q8.3.2
32. 2y ′′ ′
− 3 y − 2y = 4 e ,
t
y(0) = 1, y (0) = −2
′

33. 6y ′′ ′
− y − y = 3e
2t
,
′
y(0) = 0, y (0) = 0
34. 2y ′′ ′
+ 2 y + y = 2t,
′
y(0) = 1, y (0) = −1
35. 4y ′′ ′
− 4 y + 5y = 4 sin t − 4 cos t, y(0) = 0, y (0) = 11/17
′
36. 4y ′′ ′
+ 4 y + y = 3 sin t + cos t, y(0) = 2, y (0) = −1
′
37. 9y ′′ ′
+ 6y + y = 3e
3t
, y(0) = 0, y (0) = −3
′
38. Suppose a, b, and c are constants and a ≠ 0 . Let
−1
as + b −1
a
y1 = L ( ) and y2 = L ( ).
2 2
as + bs + c as + bs + c
Show that
′ ′
y1 (0) = 1, y (0) = 0 and y2 (0) = 0, y (0) = 1.
1 2
HINT: Use the Laplace transform to solve the initial value problems
′′ ′ ′
ay + b y + cy = 0, y(0) = 1, y (0) = 0
′′ ′ ′
ay + b y + cy = 0, y(0) = 0, y (0) = 1

8.4: A Brief Table of Laplace Transforms
Table 8.4.1
f (t) F (s)
1
1 s
(s > 0)
n n!
t (s > 0)
sn + 1
(n = integer > 0 )
p Γ(p + 1)
t , p > −1 (s > 0)
(p+ 1)
s
at 1
e (s > a)
s−a
n at n!
t e n+1
(s > 0)
(s−a)
(n = integer > 0 )
s
cosωt (s > 0)
2 2
s +ω
ω
sin ωt (s > 0)
s2 +ω2
λt s−λ
e cosωt 2
(s > λ)
(s−λ ) +ω2
λt ω
e sin ωt 2 2
(s > λ)
(s−λ ) +ω
s
cosh bt 2 (s > |b|)
s2 −b
b
sinh bt 2
(s > |b|)
2
s −b
2 2
s −ω
t cosωt (s > 0)
2
(s2 +ω2 )
2ωs
t sin ωt 2 (s > 0)
(s2 +ω2 )
3
2ω
sin ωt − ωt cosωt (s > 0)
2
(s2 +ω2 )
3
ω
ωt − sin ωt (s > 0)
2
s2 (s2 +ω2 )
1 ω
sin ωt arctan( ) (s > 0)
t s
at
e f (t) F (s − a)
k k (k)
t f (t) (−1 ) F (s)
1 s
f (ωt) F( ), ω > 0
ω ω
−τ s
e
u(t − τ ) (s > 0)
s
−τ s
u(t − τ )f (t − τ ) (τ > 0) e F (s)
∫ f (τ )g(t − τ ) dτ F (s) ⋅ G(s)

o
−as
δ(t − a) e (s > 0)

CHAPTER OVERVIEW
9: LINEAR HIGHER ORDER DIFFERENTIAL EQUATIONS
This chapter extend the results obtained in Chapter 5 for linear second order equations to linear
higher order equations.
9.1: INTRODUCTION TO LINEAR HIGHER ORDER EQUATIONS

This section presents a theoretical introduction to linear higher order equations. We will sketch the
general theory of linear n-th order equations.
9.1E: INTRODUCTION TO LINEAR HIGHER ORDER EQUATIONS (EXERCISES)

9.2: HIGHER ORDER CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS
In this section we consider the homogeneous constant coefficient equation of n-th order.
9.2E: HIGHER ORDER CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS (EXERCISES)

9.3: UNDETERMINED COEFFICIENTS FOR HIGHER ORDER EQUATIONS
This section presents the method of undetermined coefficients for higher order equations.
9.3E: UNDETERMINED COEFFICIENTS FOR HIGHER ORDER EQUATIONS (EXERCISES)

9.4: VARIATION OF PARAMETERS FOR HIGHER ORDER EQUATIONS
This section extends the method of variation of parameters to higher order equations. We’ll show how to use the method of variation
of parameters to find a particular solution of Ly=F, provided that we know a fundamental set of solutions of the homogeous equation:
Ly=0.
9.4E: VARIATION OF PARAMETERS FOR HIGHER ORDER EQUATIONS (EXERCISES)
1 9/16/2020
9.1: Introduction to Linear Higher Order Equations
An n th order differential equation is said to be linear if it can be written in the form
(n) (n−1)
y + p1 (x)y + ⋯ + pn (x)y = f (x). (9.1.1)
We considered equations of this form with n = 1 in Section 2.1 and with n = 2 in Chapter 5. In this chapter n is an arbitrary
positive integer. In this section we sketch the general theory of linear n th order equations. Since this theory has already been
discussed for n = 2 in Sections 5.1 and 5.3, we’ll omit proofs.
For convenience, we consider linear differential equations written as
(n) (n−1)
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = F (x), (9.1.2)
which can be rewritten as Equation 9.1.1 on any interval on which P has no zeros, with p = P /P , …, p = P 0 1 1 0 n n / P0 and
f = F /P . For simplicity, throughout this chapter we’ll abbreviate the left side of Equation 9.1.2 by Ly; that is,
0
(n) (n−1)
Ly = P0 y + P1 y + ⋯ + Pn y.
We say that the equation Ly = F is normal on (a, b) if P , P , …, P and F are continuous on (a, b) and P has no zeros on
0 1 n 0
(a, b) . If this is so then Ly = F can be written as Equation 9.1.1 with p , …, p and f continuous on (a, b) . 1 n
The next theorem is analogous to Theorem 5.3.1.
Theorem 9.1.1
Suppose Ly = F is normal on (a, b), let x be a point in (a, b), and let k , k , …, k
0 0 1 n−1 be arbitrary real numbers. Then
the initial value problem
′ (n−1)
Ly = F , y(x0 ) = k0 , y (x0 ) = k1 , … , y (x0 ) = kn−1
Homogeneous Equations
Equation 9.1.2 is said to be homogeneous if F ≡ 0 and nonhomogeneous otherwise. Since y ≡0 is obviously a solution of
Ly = 0 , we call it the trivial solution. Any other solution is nontrivial.
If y , y , …, y are defined on (a, b) and c , c , …, c are constants, then

1 2 n 1 2 n
y = c1 y1 + c2 y2 + ⋯ + cn yn (9.1.3)
is a linear combination of {y , y … , y }. It’s easy to show that if y , y , …, y are solutions of Ly = 0 on (a, b), then so is
1 2 n 1 2 n
any linear combination of {y , y , … , y }. (See the proof of Theorem 5.1.2.) We say that {y , y , … , y } is a fundamental
1 2 n 1 2 n
set of solutions of Ly = 0 on (a, b) if every solution of Ly = 0 on (a, b) can be written as a linear combination of
{ y , y , … , y }, as in Equation 9.1.3. In this case we say that Equation 9.1.3 is the general solution of Ly = 0 on (a, b) .
1 2 n
It can be shown (Exercises 9.1.14 and 9.1.15) that if the equation Ly = 0 is normal on (a, b) then it has infinitely many
fundamental sets of solutions on (a, b). The next definition will help to identify fundamental sets of solutions of Ly = 0 .
We say that {y 1, y2 , … , yn } is linearly independent on (a, b) if the only constants c , c , …, c such that 1 2 n
c1 y1 (x) + c2 y2 (x) + ⋯ + cn yn (x) = 0, a < x < b, (9.1.4)
are c1 = c2 = ⋯ = cn = 0 . If Equation 9.1.4 holds for some set of constants c1 , c2 , …, cn that are not all zero, then
{ y1 , y2 , … , yn } is linearly dependent on (a, b)
Theorem 9.1.1

If Ly = 0 is normal on (a, b), then a set {y 1, y2 , … , yn } of n solutions of Ly = 0 on (a, b) is a fundamental set if and
only if it is linearly independent on (a, b).
Example 9.1.1
The equation
3 ′′′ 2 ′′ ′
x y −x y − 2x y + 6y = 0 (9.1.5)
is normal and has the solutions y = x , y = x , and y = 1/x on (−∞, 0) and (0, ∞). Show that {y , y , y } is
1
2
2
3
3 1 2 3
linearly independent on (−∞, 0) and (0, ∞). Then find the general solution of Equation 9.1.5 on (−∞, 0) and (0, ∞).
Solution
Suppose
c3
2 3
c1 x + c2 x + =0 (9.1.6)
x
on (0, ∞). We must show that c 1 = c2 = c3 = 0 . Differentiating Equation 9.1.6 twice yields the system
c3
2 3
c1 x + c2 x + =0
x
c3
2
2 c1 x + 3 c2 x − =0 (9.1.7)
2
x
2c3
2 c1 + 6 c2 x + = 0.
3
x
If Equation 9.1.7 holds for all x in (0, ∞), then it certainly holds at x = 1 ; therefore,
c1 + c2 + c3 =0
2 c1 + 3 c2 − c3 =0 (9.1.8)
2 c1 + 6 c2 + 2 c3 = 0.
By solving this system directly, you can verify that it has only the trivial solution c = c = c = 0 ; however, for our 1 2 3
purposes it is more useful to recall from linear algebra that a homogeneous linear system of n equations in n unknowns
has only the trivial solution if its determinant is nonzero. Since the determinant of Equation 9.1.8 is
∣1 1 1∣ ∣1 0 0∣
∣ ∣ ∣ ∣
2 3 −1 = 2 1 −3 = 12,
∣ ∣ ∣ ∣
∣2 6 2∣ ∣2 4 0∣
it follows that Equation 9.1.8 has only the trivial solution, so { y1 , y2 , y3 } is linearly independent on (0, ∞) . Now
Theorem 9.1.2 implies that
2 3
c3
y = c1 x + c2 x +
x
is the general solution of Equation 9.1.5 on (0, ∞). To see that this is also true on (−∞, 0), assume that Equation 9.1.6
holds on (−∞, 0). Setting x = −1 in Equation 9.1.7 yields

c1 − c2 − c3 = 0
−2 c1 + 3 c2 − c3 = 0
2 c1 − 6 c2 − 2 c3 = 0.
Since the determinant of this system is

∣ 1 −1 −1 ∣ ∣ 1 0 0∣
∣ ∣ ∣ ∣
−2 3 −1 = −2 1 −3 = −12,
∣ ∣ ∣ ∣
∣ 2 −6 −2 ∣ ∣ 2 −4 0∣
it follows that c 1 = c2 = c3 = 0 ; that is, {y 1, y2 , y3 } is linearly independent on (−∞, 0).

Example 9.1.2
The equation
(4) ′′′ ′′ ′
y +y − 7y − y + 6y = 0 (9.1.9)
is normal and has the solutions y = e , y = e , y = e , and y = e

1
x
2
−x
on (−∞, ∞). (Verify.) Show that
3
2x
4
−3x
{ y , y , y , y } is linearly independent on (−∞, ∞) . Then find the general solution of Equation 9.1.9.
1 2 3 4
Solution
Suppose c , c , c , and c are constants such that
1 2 3 4
x −x 2x −3x
c1 e + c2 e + c3 e + c4 e =0 (9.1.10)
for all x. We must show that c 1 = c2 = c3 = c4 = 0 . Differentiating Equation 9.1.10 three times yields the system
x −x 2x −3x
c1 e + c2 e + c3 e + c4 e = 0
x −x 2x −3x
c1 e − c2 e + 2 c3 e − 3 c4 e = 0
(9.1.11)
x −x 2x −3x
c1 e + c2 e + 4 c3 e + 9 c4 e = 0
x −x 2x −3x
c1 e − c2 e + 8 c3 e − 27 c4 e = 0.
If Equation 9.1.11 holds for all x, then it certainly holds for x = 0 . Therefore
c1 + c2 + c3 + c4 = 0
c1 − c2 + 2 c3 − 3 c4 = 0
c1 + c2 + 4 c3 + 9 c4 = 0
c1 − c2 + 8 c3 − 27 c4 = 0.
The determinant of this system is

∣1 1 1 1∣ ∣1 1 1 1∣
∣ ∣ ∣ ∣ ∣ −2 1 −4 ∣
∣1 −1 2 −3 ∣ ∣0 −2 1 −4 ∣ ∣ ∣
= = 0 3 8
∣ ∣ ∣ ∣ ∣ ∣
1 1 4 9 0 0 3 8
∣ ∣ ∣ ∣ ∣ −2 7 −28 ∣
∣1 −1 8 −27 ∣ ∣0 −2 7 −28 ∣ (9.1.12)
∣ −2 1 −4 ∣
∣ ∣ ∣3 8∣
= 0 3 8 = −2 ∣ ∣ = 240,
∣ ∣ ∣6 −24 ∣
∣ 0 6 −24 ∣
so the system has only the trivial solution c 1 = c2 = c3 = c4 = 0 . Now Theorem 9.1.2 implies that
x −x 2x −3x
y = c1 e + c2 e + c3 e + c4 e
The Wronskian
We can use the method used in Examples 9.1.1 and 9.1.2 to test n solutions {y , y , … , y } of any n th order equation 1 2 n
Ly = 0 for linear independence on an interval (a, b) on which the equation is normal. Thus, if c , c ,…, c are constants such 1 2 n
that
c1 y1 + c2 y2 + ⋯ + cn yn = 0, a < b,
then differentiating n − 1 times leads to the n × n system of equations
c1 y1 (x) + c2 y2 (x)+ ⋯ +cn yn (x) = 0
′ ′ ′
c1 y (x) + c2 y (x)+ ⋯ +cn yn (x) = 0
1 2
(9.1.13)
⋮
(n−1) (n−1) (n−1)

c1 y (x) + c2 y (x)+ ⋯ +cn yn (x) = 0
1 2

for c , c , …, c . For a fixed x, the determinant of this system is
1 2 n
∣ y1 (x) y2 (x) ⋯ yn (x) ∣

∣ ∣
′ ′ ′
∣ y (x) y (x) ⋯ yn (x) ∣
1 2
∣ ∣
W (x) = .
∣ ∣
∣ ⋮ ⋮ ⋱ ⋮ ∣
∣ ∣
(n−1) (n−1) (n−1)
∣y (x) y (x) ⋯ yn (x) ∣
1 2
We call this determinant the Wronskian of {y , y , … , y }. If W (x) ≠ 0 for some x in (a, b) then the system Equation
1 2 n
9.1.13 has only the trivial solution c = c = ⋯ = c = 0 , and Theorem 9.1.2 implies that
1 2 n
y = c1 y1 + c2 y2 + ⋯ + cn yn
is the general solution of Ly = 0 on (a, b).

The next theorem generalizes Theorem 5.1.4. The proof is sketched in Exercises 9.1.17-9.1.20.
Theorem 9.1.3 Abel's Formula

Suppose the homogeneous linear n th order equation
(n) n−1
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = 0 (9.1.14)
is normal on (a, b), let y , 1 y2 ,…, y be solutions of Equation

n 9.1.14 on (a, b), and let x0 be in (a, b) . Then the
Wronskian of {y , y , … , y
1 2 n} is given by
x
P1 (t)
W (x) = W (x0 ) exp{− ∫ dt}, a < b. (9.1.15)
x0 P0 (t)
Therefore, either W has no zeros in (a, b) or W ≡0 on (a, b).
Formula Equation 9.1.15 is Abel’s formula.

Theorem 9.1.4
Suppose Ly = 0 is normal on (a, b) and let y , y , …, y be n solutions of 1 2 n Ly = 0 on (a, b) . Then the following
statements are equivalent; that is, they are either all true or all false:
a. The general solution of Ly = 0 on (a, b) is y = c y + c y + ⋯ + c y 1 1 2 2 n n.
b. {y , y , … , y } is a fundamental set of solutions of Ly = 0 on (a, b).

1 2 n
c. {y , y , … , y } is linearly independent on (a, b).

1 2 n
d. The Wronskian of {y , y , … , y } is nonzero at some point in (a, b).

1 2 n
e. The Wronskian of {y , y , … , y } is nonzero at all points in (a, b).

1 2 n
Example 9.1.3
In Example 9.1.1 we saw that the solutions y 1 =x
2
,y 2 =x
3
, and y 3 = 1/x of
3 ′′′ 2 ′′ ′
x y −x y − 2x y + 6y = 0
are linearly independent on (−∞, 0) and (0, ∞). Calculate the Wronskian of {y 1, y2 , y3 } .
Solution
If x ≠ 0 , then

1
∣ x2 x
3 1
∣ ∣1 x ∣
x x3
∣ ∣ ∣ ∣
2 1 3 1
W (x) = ∣ 2x 3x − 2
∣ = 2x ∣2 3x − 3
∣
x x
∣ ∣ ∣ ∣
2 1
∣ 2 6x 3
x
∣ ∣1 3x 3 ∣
x
where we factored x , x, and 2 out of the first, second, and third rows of
2
W (x) , respectively. Adding the second row of
the last determinant to the first and third rows yields
∣3 4x 0 ∣
∣ ∣ 1 ∣3 4x ∣
3 1 3
W (x) = 2 x ∣2 3x − 3
∣ = 2x ( )∣ ∣ = 12x
x 3
∣ ∣ x ∣3 6x ∣
∣3 6x 0 ∣
Therefore W (x) ≠ 0 on (−∞, 0) and (0, ∞).
Example 9.1.4
In Example 9.1.2 we saw that the solutions y 1 =e
x
,y 2 =e
−x
,y 3 =e
2x
, and y 4 =e
−3x
of
(4) ′′′ ′′ ′
y +y − 7y − y + 6y = 0
are linearly independent on every open interval. Calculate the Wronskian of {y 1, y2 , y3 , y4 } .

Solution
For all x,
∣ ex e
−x
e
2x
e
−3x
∣
∣ ∣
x −x 2x −3x
∣e −e 2e −3e ∣
W (x) = .
∣ x −x 2x −3x ∣
e e 4e 9e
∣ ∣
x −x 2x −3x
∣e −e 8e −27e ∣
Factoring the exponential common factor from each row yields

∣1 1 1 1∣
∣ ∣
−x ∣1 −1 2 −3 ∣ −x
W (x) = e = 240 e ,
∣ ∣
1 1 4 9
∣ ∣
∣1 −1 8 −27 ∣
from Equation 9.1.12.
Note
Under the assumptions of Theorem 9.1.4, it isn’t necessary to obtain a formula for W (x) . Just evaluate W (x) at a
convenient point in (a, b), as we did in Examples 9.1.1 and 9.1.2.
Theorem 9.1.5
Suppose c is in (a, b) and α , α , …, are real numbers, not all zero. Under the assumptions of Theorem 10.3.3, suppose
1 2
y and y are solutions of Equation 5.1.35 such that

1 2
(n−1)
′
α yi (c) + y (c) + ⋯ + y (c) = 0, 1 ≤ i ≤ n. (9.1.16)
i i
Then {y 1, y2 , … yn } isn’t linearly independent on (a, b).
Proof
Since α , α , …, α are not all zero, Equation 9.1.14implies that
1 2 n

(n−1)
∣ ′ ∣
y1 (c) y (c) ⋯ y (c)
1 1
∣ ∣
∣ (n−1) ∣
′
y2 (c) y (c) ⋯ y (c)
2 2
∣ ∣
= 0,
∣ ∣
∣ ⋮ ⋮ ⋱ ⋮ ∣
∣ ∣
(n−1)
′
∣ yn (c) yn (c) ⋯ yn (c) ∣
so
∣ y1 (c) y2 (c) ⋯ yn (c) ∣
∣ ∣
′ ′ ′
y (c) y (c) ⋯ yn (c)
∣ 1 2 ∣
∣ ∣ =0
∣ ⋮ ⋮ ⋱ ⋮ ∣
∣ ∣
(n−1) (n−1) (n−1)
∣y (c) y (c)(c) ⋯ yn (c)(c) ∣
1 2
and Theorem 9.1.4implies the stated conclusion.
General Solution of a Nonhomogeneous Equation

The next theorem is analogous to Theorem 5.3.2. It shows how to find the general solution of Ly = F if we know a particular
solution of Ly = F and a fundamental set of solutions of the complementary equation Ly = 0 .
Theorem 9.1.6
The probabilities assigned to events by a distribution function on a sample space are given by.
Proof
Suppose Ly = F is normal on (a, b). Let y be a particular solution of Ly = F on (a, b), and let {y , y , … , y } be
p 1 2 n
a fundamental set of solutions of the complementary equation Ly = 0 on (a, b) . Then y is a solution of Ly = F on

(a, b) if and only if
y = yp + c1 y1 + c2 y2 + ⋯ + cn yn ,
where c 1, c2 , … , cn are constants.
Theorem 9.1.7 The Principle of Superposition

Suppose for each i = 1, 2, …, r, the function y is a particular solution of Ly = F on (a, b). Then
pi i
yp = yp + yp + ⋯ + yp
1 2 r
Ly = F1 (x) + F2 (x) + ⋯ + Fr (x)
on (a, b).
We’ll apply Theorems 9.1.6 and 9.1.7 throughout the rest of this chapter.

9.1E: Introduction to Linear Higher Order Equations (Exercises)
Q9.1.1
1. Verify that the given function is the solution of the initial value problem.
−24
a. x y
3 ′′′
− 3x y
2 ′′
+ 6x y − 6y =
′
, y(−1) = 0 ,y ′
(−1) = 0, y
′′
(−1) = 0 ;y = −6x − 8x 2
− 3x
3
+
1
x
x
2
1 1 x −4 3 1 1
b. y ′′′
− y
′′
−y +
′
y =
4
, y(1) = ,
′
y (1) = ,y
′′
(1) = 1 ;y = x +
x x x 2 2 2x
c. xy
′′′
−y
′′
− xy + y = x ,
′ 2
y(1) = 2,
′
y (1) = 5, y
′′
(1) = −1 ;y = −x 2
− 2 + 2e
(x−1)
−e
−(x−1)
+ 4x
17
d. 4x 3
y
′′′
+ 4x y
2 ′′ ′
− 5x y + 2y = 30 x ,
2
y(1) = 5, y (1) =
′
;
2
63
′′ 2 1/2 −1/2 2
y (1) = ; y = 2x ln x − x + 2x + 4x
4
e. x y
4 (4)
− 4x y
3 ′′′
+ 12 x y
2 ′′ ′
− 24x y + 24y = 6 x ,
4
y(1) = −2 ;y ′
(1) = −9, y
′′
(1) = −27, y
′′′
(1) = −52 ;
4 2 3 4
y =x ln x + x − 2 x + 3x − 4x
f. x y (4)
−y
′′′
− 4x y
′′
+ 4y
′
= 96 x ,
2
y(1) = −5,
′
y (1) = −24 ;
′′ ′′′ 2 3
y (1) = −36; y (1) = −48; y = 9 − 12x + 6 x − 8x

3 ′′′ 2 ′′ ′ ′ ′′
x y −x y − 2x y + 6y = 0, y(−1) = −4, y (−1) = −14, y (−1) = −20.
HINT: See Example 9.1.1.

(4) ′′′ ′′ ′ ′ ′′ ′′′
y +y − 7y − y + 6y = 0, y(0) = 5, y (0) = −6, y (0) = 10, y (0) − 36.
HINT: See Example 9.1.2.

4. Find solutions y , y , …, y of the equation y 1 2 n
(n)
=0 that satisfy the initial conditions
0, j ≠ i − 1,
(j)
y (x0 ) = { 1 ≤ i ≤ n.
i
1, j = i − 1,
5.
a. Verify that the function
c3
3 2
y = c1 x + c2 x +
x
satisfies
3 ′′′ 2 ′′ ′
x y −x y − 2x y + 6y = 0 (A)
if c , c , and c are constants.

1 2 3
b. Use (a) to find solutions y , y , and y of (A) such that 1 2 3
′ ′′
y1 (1) = 1, y (1) = 0, y (1) = 0
1 1
′ ′′
y2 (1) = 0, y (1) = 1, y (1) = 0
2 2
′ ′′
y3 (1) = 0, y (1) = 0, y (1) = 1.
3 3
c. Use (b) to find the solution of (A) such that

′ ′′
y(1) = k0 , y (1) = k1 , y (1) = k2 .
6. Verify that the given functions are solutions of the given equation, and show that they form a fundamental set of solutions of
the equation on any interval on which the equation is normal.

a. y + y − y − y = 0; {e , e , x e }
′′′ ′′ ′ x −x −x
b. y − 3y + 7y − 5y = 0; {e , e cos 2x, e sin 2x} .

′′′ ′′ ′ x x x
c. x y − y − x y + y = 0; {e , e , x}
′′′ ′′ ′ x −x
d. x y + 2x y − (x + 2)y = 0; {e /x, e /x, 1}

2 ′′′ ′′ 2 x −x
e. (x − 2x + 2)y − x y + 2x y − 2y = 0; {x, x , e }
2 ′′′ 2 ′′ ′ 2 x
f. (2x − 1)y − 4x y + (5 − 2x)y + 4x y − 4y = 0; {x,

(4) ′′′ ′′ ′ x
e , e
−x
,e
2x
}
g. x y − y − 4x y + 4y = 0; {1, x , e , e }
(4) ′′′ ′ ′ 2 2x −2x
7. Find the Wronskian W of a set of three solutions of

′′′ ′′ x ′
y + 2x y + e y − y = 0,
given that W (0) = 2 .

8. Find the Wronskian W of a set of four solutions of
(4) ′′′ 2 ′′
y + (tan x)y +x y + 2xy = 0,
given that W (π/4) = K .

9.
a. Evaluate the Wronskian W {e , x x
xe , x e }
2 x
. Evaluate W (0).
b. Verify that y , y , and y satisfy 1 2 3
′′′ ′′ ′
y − 3y + 3 y − y = 0. (A)
c. Use W (0) from (a) and Abel’s formula to calculate W (x).

d. What is the general solution of (A)?
10. Compute the Wronskian of the given set of functions.
a. {1, e , e }x −x
b. {e , e sin x, e cos x}
x x x
c. {2, x + 1, x + 2} 2
d. x, x ln x, 1/x}
2 3 n
e. {1, x, , ,⋯,
x
2!
}
x
3!
x
n!
f. {e , e , x}
x −x
g. {e /x, e /x, 1}
x −x
h. {x, x , e } 2 x
i. {x, x , 1/x, 1/x }

3 2
j. {e , e , x, e }
x −x 2x
k. {e , e , 1, x }
2x −2x 2
11. Suppose Ly = 0 is normal on (a, b) and x is in (a, b). Use Theorem 9.1.1 to show that y ≡ 0 is the only solution of the
0

′ (n−1)
Ly = 0, y(x0 ) = 0, y (x0 ) = 0, … , y (x0 ) = 0,
on (a, b).
12. Prove: If y , y , …, y are solutions of Ly = 0 and the functions
1 2 n
zi = ∑ aij yj , 1 ≤ i ≤ n,
j=1
form a fundamental set of solutions of Ly = 0 , then so do y , y , …, y . 1 2 n
13. Prove: If
y = c1 y1 + c2 y2 + ⋯ + ck yk + yp

is a solution of a linear equation Ly = F for every choice of the constants c , c ,…, c , then Ly 1 2 k i =0 for 1 ≤ i ≤ k .
14. Suppose Ly = 0 is normal on (a, b) and let x0 be in (a, b) . For 1 ≤i ≤n , let yi be the solution of the initial value
problem
(j) 0, j ≠ i − 1,
Lyi = 0, y (x0 ) = { 1 ≤ i ≤ n,
i
1, j = i − 1,
where x is an arbitrary point in (a, b). Show that any solution of Ly = 0 on (a, b), can be written as
0
y = c1 y1 + c2 y2 + ⋯ + cn yn ,
with cj =y
(j−1)
(x0 ) .
15. Suppose {y 1, y2 , … , yn } is a fundamental set of solutions of
(n) (n−1)
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = 0
on (a, b), and let

z1 = a11 y1 + a12 y2 + ⋯ + a1n yn
z2 = a21 y1 + a22 y2 + ⋯ + a2n yn
⋮ ⋮ ⋮ ⋮
zn = an1 y1 + an2 y2 + ⋯ + ann yn ,
where the {a ij } are constants. Show that { z1 , z2 , … , zn } is a fundamental set of solutions of (A) if and only if the
determinant
∣ a11 a12 ⋯ a1n ∣
∣ ∣
∣ a21 a22 ⋯ a2n ∣
∣ ∣
∣ ∣
⋮ ⋮ ⋱ ⋮
∣ ∣
∣ an1 an2 ⋯ ann ∣
is nonzero. HINT: The determinant of a product of n × n matrices equals the product of the determinants.
16. Show that {y , y , … , y } is linearly dependent on (a, b) if and only if at least one of the functions y , y , …, y can be
1 2 n 1 2 n
written as a linear combination of the others on (a, b).
Q9.1.2
Take the following as a hint in Exercises 9.1.17-9.1.19:
By the definition of determinant,
∣ a11 a12 ⋯ a1n ∣
∣ ∣
∣ a21 a22 ⋯ a2n ∣
∣ ∣ = ∑ ±a1i a2i , ⋯ , ani ,
1 2 n
∣ ∣
⋮ ⋮ ⋱ ⋮
∣ ∣
∣ an1 an2 ⋯ ann ∣
where the sum is over all permutations (i , i i 2, ⋯ , in ) of (1, 2, ⋯ , n) and the choice of + or − in each term depends only on
the permutation associated with that term.
17. Prove: If

∣ a11 a12 ⋯ a1n ∣
∣ ∣
∣ a21 a22 ⋯ a2n ∣
∣ ∣
∣ ∣
A(u1 , u2 , … , un ) = ,
∣ ⋮ ⋮ ⋱ ⋮ ∣
∣ ∣
∣ an−1,1 an−1,2 ⋯ an−1,n ∣
∣ ∣
∣ u1 u2 ⋯ un ∣
then
A(u1 + v1 , u2 + v2 , … , un + vn ) = A(u1 , u2 , … , un ) + A(v1 , v2 , … , vn ).
18. Let
∣ f11 f12 ⋯ f1n ∣
∣ ∣
∣ f f22 ⋯ f2n ∣
21
∣ ∣
F = ,
∣ ∣
∣ ⋮ ⋮ ⋱ ⋮ ∣
∣ ∣
∣ fn1 fn2 ⋯ fnn ∣
where f ij (1 ≤ i, j ≤ n) is differentiable. Show that

′
F = F1 + F2 + ⋯ + Fn ,
where F is the determinant obtained by differentiating the ith row of F .

i
19. Use Exercise 9.1.18 to show that if W is the Wronskian of the n -times differentiable functions y , y , …, y , then 1 2 n
∣ y1 y2 ⋯ yn ∣
∣ ∣
′ ′ ′
∣ y y ⋯ yn ∣
1 2
∣ ∣
∣ ∣
′
W =∣ ⋮ ⋮ ⋱ ⋮ ∣.
∣ ∣
∣ (n−2) (n−2) (n−2) ∣
y y ⋯ yn
∣ 1 2 ∣
∣ ∣
(n) (n) (n)
∣ y y ⋯ yn ∣
1 2
Q9.1.3
20. Use Exercises 9.1.17 and 9.1.19 to show that if W is the Wronskian of solutions {y 1, y2 , … , yn } of the normal equation
(n) (n−1)
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = 0, (A)
then W ′
= −P1 W / P0 . Derive Abel’s formula (Equation 9.1.15) from this.
21. Prove Theorem 9.1.6.
22. Prove Theorem 9.1.7.
23. Show that if the Wronskian of the n -times continuously differentiable functions { y1 , y2 , … , yn } has no zeros in (a, b) ,
then the differential equation obtained by expanding the determinant
∣ y y1 y2 ⋯ yn ∣
∣ ∣
′ ′ ′ ′
∣ y y y ⋯ yn ∣
1 2
∣ ∣
= 0,
∣ ∣
∣ ⋮ ⋮ ⋮ ⋱ ⋮ ∣
∣ ∣
(n) (n) (n) (n)
∣y y
1
y
2
⋯ yn ∣
in cofactors of its first column is normal and has {y 1, y2 , … , yn } as a fundamental set of solutions on (a, b).

24. Use the method suggested by Exercise 9.1.23 to find a linear homogeneous equation such that the given set of functions is
a fundamental set of solutions on intervals on which the Wronskian of the set has no zeros.
a. {x, x − 1, x + 1}
2 2
b. {e , e , x}
x −x
c. {e , x e , 1}
x −x
d. {x, x , e }
2 x
e. {x, x , 1/x}
2
f. {x + 1, e , e }
x 3x
g. {x, x , 1/x, 1/x }

3 2
h. {x, x ln x, 1/x, x } 2
i. {e , e , x, e }
x −x 2x
j. {e , e , 1, x }
2x −2x 2

9.2: Higher Order Constant Coefficient Homogeneous Equations
If a , a , …, a are constants and a
0 1 n 0 ≠0 , then
(n) (n−1)
a0 y + a1 y + ⋯ + an y = F (x)
is said to be a constant coefficient equation. In this section we consider the homogeneous constant coefficient equation
(n) (n−1)
a0 y + a1 y + ⋯ + an y = 0. (9.2.1)
Since Equation 9.2.1 is normal on (−∞, ∞), the theorems in Section 9.1 all apply with (a, b) = (−∞, ∞) .
As in Section 5.2, we call
n n−1
p(r) = a0 r + a1 r + ⋯ + an (9.2.2)
the characteristic polynomial of Equation 9.2.1. We saw in Section 5.2 that when n = 2 the solutions of Equation 9.2.1 are
determined by the zeros of the characteristic polynomial. This is also true when n > 2 , but the situation is more complicated
in this case. Consequently, we take a different approach.
If k is a positive integer, let D stand for the k -th derivative operator; that is
k
k (k)
D y =y .
If
m m−1
q(r) = b0 r + b1 r + ⋯ + bm
is an arbitrary polynomial, define the operator

m m−1
q(D) = b0 D + b1 D + ⋯ + bm
such that
m m−1 (m) (m−1)
q(D)y = (b0 D + b1 D + ⋯ + bm )y = b0 y + b1 y + ⋯ + bm y
whenever y is a function with m derivatives. We call q(D) a polynomial operator.

With p as in Equation 9.2.2,
n n−1
p(D) = a0 D + a1 D + ⋯ + an ,
so Equation 9.2.1 can be written as p(D)y = 0. If r is a constant then

rx n rx n−1 rx rx
p(D)e = (a0 D e + a1 D e + ⋯ + an e )
n n−1 rx
= (a0 r + a1 r + ⋯ + an )e ;
that is
rx rx
p(D)(e ) = p(r)e .
This shows that y = e is a solution of Equation 9.2.1 if p(r) = 0 . In the simplest case, where p has n distinct real zeros r ,
rx
1
r ,…, r , this argument yields n solutions

2 n
r1 x r2 x rn x
y1 = e , y2 = e ,…, yn = e .
It can be shown (Exercise 9.2.39) that the Wronskian of {e , e , … , e } is nonzero if r , r , …, r are distinct; hence,
r1 x r2 x rn x
1 2 n
} is a fundamental set of solutions of p(D)y = 0 in this case.

r1 x r2 x rn x
{e ,e ,…,e
Example 9.2.1

′′′ ′′ ′
y − 6y + 11 y − 6y = 0. (9.2.3)
b. and solve the initial value problem

′′′ ′′ ′ ′ ′′
y − 6y + 11 y − 6y = 0, y(0) = 4, y (0) = 5, y (0) = 9. (9.2.4)
Solution a
The characteristic polynomial of Equation 9.2.3 is
3 2
p(r) = r − 6r + 11r − 6 = (r − 1)(r − 2)(r − 3).
Therefore {e x
,e
2x
,e
3x
} is a set of solutions of Equation 9.2.3. It is a fundamental set, since its Wronskian is
∣ ex e
2x
e
3x
∣ ∣1 1 1∣
∣ ∣
x 2x 3x 6x ∣ ∣ 6x
W (x) = ∣ e 2e 3e ∣ =e 1 2 3 = 2e ≠ 0.
∣ ∣ ∣ ∣
x 2x 3x
∣e 4e 9e ∣ ∣1 4 9∣

x 2x 3x
y = c1 e + c2 e + c3 e . (9.2.5)
Solution b
We must determine c , c and c in Equation
1 2 3 9.2.5 so that y satisfies the initial conditions in Equation 9.2.4 .
Differentiating Equation 9.2.5 twice yields
′ x 2x 3x
y = c1 e + 2 c2 e + 3 c3 e
(9.2.6)
′′ x 2x 3x
y = c1 e + 4 c2 e + 9 c3 e .
Setting x = 0 in Equation 9.2.5 and Equation 9.2.6 and imposing the initial conditions yields
c1 + c2 + c3 = 4
c1 + 2 c2 + 3 c3 = 5
c1 + 4 c2 + 9 c3 = 9.
The solution of this system is c 1 =4 ,c 2 = −1 ,c 3 =1 . Therefore the solution of Equation 9.2.4 is

x 2x 3x
y = 4e −e +e
(Figure 9.2.1).
Figure 9.2.1 : y = 4e x
−e
2x
+e
3x
Now we consider the case where the characteristic polynomial Equation 9.2.2 does not have n distinct real zeros. For this
purpose it is useful to define what we mean by a factorization of a polynomial operator. We begin with an example.
Example 9.2.2

Consider the polynomial
3 2
p(r) = r −r +r−1
and the associated polynomial operator

3 2
p(D) = D −D + D − 1.
Since p(r) can be factored as

2 2
p(r) = (r − 1)(r + 1) = (r + 1)(r − 1),
it is reasonable to expect that p(D) can be factored as

2 2
p(D) = (D − 1)(D + 1) = (D + 1)(D − 1). (9.2.7)
However, before we can make this assertion we must define what we mean by saying that two operators are equal, and
what we mean by the products of operators in Equation 9.2.7. We say that two operators are equal if they apply to the
same functions and always produce the same result. The definitions of the products in Equation 9.2.7 is this: if y is any
three-times differentiable function then
a. (D − 1)(D
2
+ 1)y is the function obtained by first applying D
2
+1 to y and then applying D−1 to the resulting
function
b. (D + 1)(D − 1)y is the function obtained by first applying
2
D−1 to y and then applying D
2
+1 to the resulting
function.
From (a),
2 2
(D − 1)(D + 1)y = (D − 1)[(D + 1)y]
′′ ′′ ′′
= (D − 1)(y + y) = D(y + y) − (y + y)
(9.2.8)
′′′ ′ ′′
= (y + y ) − (y + y)
′′′ ′′ ′ 3 2
= y −y + y − y = (D −D + D − 1)y.
This implies that

2 3 2
(D − 1)(D + 1) = (D −D + D − 1).
From (b),
2 2
(D + 1)(D − 1)y = (D + 1)[(D − 1)y]
2 ′ 2 ′ ′
= (D + 1)(y − y) = D (y − y) + (y − y)
(9.2.9)
′′′ ′′ ′
= (y −y ) + (y − y)
′′′ ′′ ′ 3 2
= y −y + y − y = (D −D + D − 1)y,
2 3 2
(D + 1)(D − 1) = (D −D + D − 1),
which completes the justification of Equation 9.2.7.
Example 9.2.3
Use the result of Example 9.2.2 to find the general solution of
′′′ ′′ ′
y −y + y − y = 0. (9.2.10)
Solution
From Equation 9.2.8, we can rewrite Equation 9.2.10 as
2
(D − 1)(D + 1)y = 0,
which implies that any solution of (D 2

+ 1)y = 0 is a solution of Equation 9.2.10. Therefore y 1 = cos x and y 2 = sin x
are solutions of Equation 9.2.10.

From Equation 9.2.9, we can rewrite Equation 9.2.10 as
2
(D + 1)(D − 1)y = 0,
which implies that any solution of (D − 1)y = 0 is a solution of Equation 9.2.10 . Therefore y3 = e
x
is solution of
Equation 9.2.10.
The Wronskian of {e x
, cos x, sin x} is
x
∣ cos x sin x e ∣
∣ x ∣
W (x) = − sin x cos x e .
∣ ∣
x
∣ − cos x − sin x e ∣
Since
∣ 1 0 1∣
∣ ∣
W (0) = 0 1 1 = 2,
∣ ∣
∣ −1 0 1∣
{cos x, sin x, e }
x
is linearly independent and
x
y = c1 cos x + c2 sin x + c3 e
Example 9.2.4
(4)
y − 16y = 0. (9.2.11)
Solution
4
p(r) = r − 16
2 2
= (r − 4)(r + 4)
2
= (r − 2)(r + 2)(r + 4).
By arguments similar to those used in Examples 9.2.2 and 9.2.4, it can be shown that Equation 9.2.11 can be written as
2
(D + 4)(D + 2)(D − 2)y = 0
or
2
(D + 4)(D − 2)(D + 2)y = 0
or
2
(D − 2)(D + 2)(D + 4)y = 0.
Therefore y is a solution of Equation 9.2.11 if it is a solution of any of the three equations

2
(D − 2)y = 0, (D + 2)y = 0, (D + 4)y = 0.
Hence, {e 2x
,e
−2x
, cos 2x, sin 2x} is a set of solutions of Equation 9.2.11. The Wronskian of this set is
2x −2x
∣ e e cos 2x sin 2x ∣
∣ ∣
2x −2x
∣ 2e −2e −2 sin 2x 2 cos 2x ∣
W (x) = .
∣ 2x −2x ∣
4e 4e −4 cos 2x −4 sin 2x
∣ ∣
2x −2x
∣ 8e −8e 8 sin 2x −8 cos 2x ∣

Since
∣1 1 1 0∣
∣ ∣
∣2 −2 0 2∣
W (0) = = −512,
∣ ∣
4 4 −4 0
∣ ∣
∣8 −8 0 −8 ∣
{e
2x
,e
−2x
, cos 2x, sin 2x} is linearly independent, and
2x −2x
y1 = c1 e + c2 e + c3 cos 2x + c4 sin 2x
It is known from algebra that every polynomial

n n−1
p(r) = a0 r + a1 r + ⋯ + an
with real coefficients can be factored as
p(r) = a0 p1 (r)p2 (r) ⋯ pk (r),
where no pair of the polynomials p , p , …, p has a commom factor and each is either of the form
1 2 k
mj
pj (r) = (r − rj ) , (9.2.12)
where r is real and m is a positive integer, or

j j
mj
2 2
pj (r) = [(r − λj ) +ω ] , (9.2.13)
j
where λ and ω are real, ω ≠ 0 , and m is a positive integer. If Equation 9.2.12 holds then r is a real zero of p, while if
j j j j j
Equation 9.2.13 holds then λ + iω and λ − iω are complex conjugate zeros of p. In either case, m is the multiplicity of the j
zero(s).
By arguments similar to those used in our examples, it can be shown that
p(D) = a0 p1 (D)p2 (D) ⋯ pk (D) (9.2.14)
and that the order of the factors on the right can be chosen arbitrarily. Therefore, if p (D)y = 0 for some j then p(D)y = 0. j
To see this, we simply rewrite Equation 9.2.14 so that p (D) is applied first. Therefore the problem of finding solutions of
j
p(D)y = 0 with p as in Equation 9.2.14 reduces to finding solutions of each of these equations
pj (D)y = 0, 1 ≤ j ≤ k,
where is a power of a first degree term or of an irreducible quadratic. To find a fundamental set of solutions
pj
of p(D)y = 0, we find fundamental set of solutions of each of the equations and take {y , y , … , y } to be
{ y1 , y2 , … , yn } 1 2 n
the set of all functions in these separate fundamental sets. In Exercise 9.2.40 we sketch the proof that {y , y , … , y } is 1 2 n
linearly independent, and therefore a fundamental set of solutions of p(D)y = 0.

To apply this procedure to general homogeneous constant coefficient equations, we must be able to find fundamental sets of
solutions of equations of the form
m
(D − a) y =0
and
m
2 2
[(D − λ ) +ω ] y = 0,
where m is an arbitrary positive integer. The next two theorems show how to do this.
Theorem 9.2.1
If m is a positive integer, then

ax ax m−1 ax
{e , xe ,…,x e } (9.2.15)
is a fundamental set of solutions of

m
(D − a) y = 0. (9.2.16)
Proof
We’ll show that if
m−1
f (x) = c1 + c2 x + ⋯ + cm x
is an arbitrary polynomial of degree ≤ m −1 , then y = e ax

f is a solution of Equation . First note that if
9.2.16 g is
any differentiable function then
ax ax ax ax ax ′ ax
(D − a)e g = De g − ae g = ae g+e g − ae g,
so
ax ax ′
(D − a)e g =e g . (9.2.17)
Therefore
ax ax ′
(D − a)e f = e f (from (9.2.17) with g = f )
2 ax ax ′ ax ′′ ′
(D − a) e f = (D − a)e f =e f (from (9.2.17) with g = f )
3 ax ax ′′ ax ′′′ ′′
(D − a) e f = (D − a)e f =e f (from (9.2.17) with g = f )
m ax ax (m−1) ax (m) (m−1)

(D − a) e f = (D − a)e f =e f (from (9.2.17) with g = f ).
Since f = 0 , the last equation implies that y = e

(m)
f is a solution of Equation 9.2.16 if f is any polynomial of
ax
degree ≤ m − 1 . In particular, each function in Equation 9.2.15is a solution of Equation 9.2.16. To see that Equation
9.2.15is linearly independent (and therefore a fundamental set of solutions of Equation 9.2.16), note that if
ax ax m−1 ax
c1 e + c2 x e + c ⋯ + cm−1 x e =0
for all x in some interval (a, b) , then

m−1
c1 + c2 x + c ⋯ + cm−1 x =0
for all x in (a, b) . However, we know from algebra that if this polynomial has more than m −1 zeros then
c1 = c2 = ⋯ = cn = 0 .
Example 9.2.5
′′′ ′′ ′
y + 3y + 3 y + y = 0. (9.2.18)
Solution
3 2 3
p(r) = r + 3r + 3r + 1 = (r + 1 ) .
Therefore Equation 9.2.18 can be written as

3
(D + 1 ) y = 0,
so Theorem 9.2.1 implies that the general solution of Equation 9.2.18 is

−x 2
y =e (c1 + c2 x + c3 x ).

The proof of the next theorem is sketched in Exercise 9.2.41.
Theorem 9.2.2
If ω ≠ 0 and m is a positive integer, then
λx λx m−1 λx
{e cos ωx, x e cos ωx, …,x e cos ωx,
λx λx m−1 λx
e sin ωx, x e sin ωx, …,x e sin ωx}
is a fundamental set of solutions of

2 2 m
[(D − λ ) +ω ] y = 0.
Example 9.2.6
2 3
(D + 4D + 13 ) y = 0. (9.2.19)
Solution
3
2 3 2
p(r) = (r + 4r + 13 ) = ((r + 2 ) + 9) .
Therefore Equation 9.2.19 can be be written as

2 3
[(D + 2 ) + 9 ] y = 0,
so Theorem 9.2.2 implies that the general solution of Equation 9.2.19 is

2 −2x 2 −2x
y = (a1 + a2 x + a3 x )e cos 3x + (b1 + b2 x + b3 x )e sin 3x.
Example 9.2.7
(4) ′′′ ′′ ′
y + 4y + 6y + 4y = 0. (9.2.20)
Solution
4 3 2
p(r) = r + 4r + 6r + 4r
3 2
= r(r + 4r + 6r + 4)
2
= r(r + 2)(r + 2r + 2)
2
= r(r + 2)[(r + 1 ) + 1].
Therefore Equation 9.2.20 can be written as

2
[(D + 1 ) + 1](D + 2)Dy = 0.
Fundamental sets of solutions of

2
[(D + 1 ) + 1] y = 0, (D + 2)y = 0, and Dy = 0.
are given by
−x −x −2x
{e cos x, e sin x}, {e }, and {1},
respectively. Therefore the general solution of Equation 9.2.20 is

−x −2x
y =e (c1 cos x + c2 sin x) + c3 e + c4 .
Example 9.2.8
Find a fundamental set of solutions of
2 2 3 2
[(D + 1 ) + 1 ] (D − 1 ) (D + 1)D y = 0. (9.2.21)
Solution
A fundamental set of solutions of Equation 9.2.21 can be obtained by combining fundamental sets of solutions of
2
2
[(D + 1 ) + 1] y =0
3
(D − 1 ) y = 0
(D + 1)y = 0
2
D y =0
Fundamental sets of solutions of these equations are given by

−x −x −x −x
{e cos x, x e cos x, e sin x, x e sin x}
x x 2 x
{e , x e , x e }
{e
−x
,
}
{1, x}
respectively. These ten functions form a fundamental set of solutions of Equation 9.2.21.

9.2E: Higher Order Constant Coefficient Homogeneous Equations (Exercises)
Q9.2.1
1. y ′′′
− 3y
′′
+ 3y − y = 0
′
2. y (4)
+ 8y
′′
− 9y = 0
3. y ′′′
−y
′′
+ 16 y − 16y = 0
′
4. 2y ′′′
+ 3y
′′
− 2 y − 3y = 0
′
5. y ′′′
+ 5y
′′
+ 9 y + 5y = 0
′
6. 4y ′′′
− 8y
′′
+ 5y − y = 0
′
7. 27y ′′′
+ 27 y
′′
+ 9y + y = 0
′
8. y (4)
+y
′′
=0
9. y (4)
− 16y = 0
10. y (4)
+ 12 y
′′
+ 36y = 0
11. 16y (4)

− 72 y
′′
+ 81y = 0
12. 6y (4)
+ 5y
′′′
+ 7y
′′ ′
+ 5y + y = 0
13. 4y (4)
+ 12 y
′′′
+ 3y
′′ ′
− 13 y − 6y = 0
14. y (4)
− 4y
′′′
+ 7y
′′ ′
− 6 y + 2y = 0
Q9.2.2
In Exercises 9.2.15-9.2.27 solve the initial value problem. Graph the solution for Exercises 9.2.17-9.2.19 and 9.2.27.
15. y ′′′
− 2y
′′
+ 4 y − 8y = 0,
′
y(0) = 2,
′
y (0) = −2, y
′′
(0) = 0
16. y ′′′
+ 3y
′′
− y − 3y = 0,
′
y(0) = 0,
′
y (0) = 14, y
′′
(0) = −40
17. y ′′′
−y
′′
− y + y = 0,
′
y(0) = −2,
′
y (0) = 9, y
′′
(0) = 4
18. y ′′′
− 2 y − 4y = 0,
′
y(0) = 6,
′
y (0) = 3, y
′′
(0) = 22
19. 3y ′′′
−y
′′
− 7 y + 5y = 0,
′
y(0) =
14
5
,
′
y (0) = 0, y
′′
(0) = 10
20. y ′′′
− 6y
′′
+ 12 y − 8y = 0,
′
y(0) = 1,
′
y (0) = −1, y
′′
(0) = −4
21. 2y ′′′
− 11 y
′′
+ 12 y + 9y = 0,
′
y(0) = 6,
′
y (0) = 3, y
′′
(0) = 13
22. 8y ′′′
− 4y
′′
− 2 y + y = 0,
′
y(0) = 4,
′
y (0) = −3, y
′′
(0) = −1
23. y (4)
− 16y = 0, y(0) = 2, y (0) = 2, y
′ ′′
(0) = −2, y
′′′
(0) = 0
24. y (4)
− 6y
′′′
+ 7y
′′ ′
+ 6 y − 8y = 0, y(0) = −2,
′
y (0) = −8, y
′′
(0) = −14, y
′′′
(0) = −62
25. 4y (4)
− 13 y
′′
+ 9y = 0, y(0) = 1,
′
y (0) = 3, y
′′
(0) = 1, y
′′′
(0) = 3
26. y (4)
+ 2y
′′′
− 2y
′′ ′
− 8 y − 8y = 0, y(0) = 5,
′
y (0) = −2, y
′′
(0) = 6, y
′′′
(0) = 8
27. 4y (4)
+ 8y
′′′
+ 19 y
′′ ′
+ 32 y + 12y = 0, y(0) = 3,
′
y (0) = −3, y
′′
(0) = −
7
2
, y
′′′
(0) =
31
Q9.2.3
28. Find a fundamental set of solutions of the given equation, and verify that it is a fundamental set by evaluating its
Wronskian at x = 0 .
a. 2
(D − 1 ) (D − 2)y = 0

b. (D + 4)(D − 3)y = 0
2
c. (D + 2D + 2)(D − 1)y = 0
2
d. D (D − 1)y = 0
3
e. (D − 1)(D + 1)y = 0
2 2
f. (D − 2D + 2)(D + 1)y = 0
2 2
Q9.2.4
In Exercises 9.2.29-9.2.38 find a fundamental set of solutions.
29. (D 2
+ 6D + 13)(D − 2 ) D y = 0
2 3
30. (D − 1) 2
(2D − 1 ) (D
3 2
+ 1)y = 0
31. (D 2 3
+ 9) D y = 0
2
32. (D − 2) 3
(D + 1 ) Dy = 0
2
33. (D 2
+ 1)(D
2 2
+ 9 ) (D − 2)y = 0
34. (D 4
− 16 ) y = 0
2
35. (4D 2
+ 4D + 9 ) y = 0
3
36. D 3
(D − 2 ) (D
2 2 2
+ 4) y = 0
37. (4D 2
+ 1 ) (9 D
2 2 3
+ 4) y = 0
38. [(D − 1) 4
− 16] y = 0
Q9.2.5
39 It can be shown that
∣ 1 1 ⋯ 1 ∣
∣ ∣
∣ a1 a2 ⋯ an ∣
∣ ∣
2 2 2
∣ a a ⋯ an ∣
1 2 = ∏ (aj − ai ), (A)
∣ ∣
1≤i<j≤n
∣ ∣
∣ ⋮ ⋮ ⋱ ⋮ ∣
∣ ∣
n−1 n−1 n−1
∣a a ⋯ an ∣
1 2
where the left side is the Vandermonde determinant and the right side is the product of all factors of the form (a j − ai ) with i
and j between 1 and n and i < j .>
a. Verify (A) for n = 2 and n = 3 .
b. Find the Wronskian of {e , e a1 x a2 x
,…,e
an x
} .
40. A theorem from algebra says that if P1 and P2 are polynomials with no common factors then there are polynomials Q1
and Q such that

2
Q1 P1 + Q2 P2 = 1.
This implies that
Q1 (D)P1 (D)y + Q2 (D)P2 (D)y = y
for every function y with enough derivatives for the left side to be defined.
a. Use this to show that if P and P have no common factors and
1 2
P1 (D)y = P2 (D)y = 0
then y = 0 .

b. Suppose and P are polynomials with no common factors. Let u , …,
P1 2 1 ur be linearly independent solutions of
P1 (D)y = 0 and let v , …, v be linearly independent solutions of
1 s P2 (D)y = 0 . Use (a) to show that
{ u , … , u , v , … , v } is a linearly independent set.
1 r 1 s
c. Suppose the characteristic polynomial of the constant coefficient equation

(n) (n−1)
a0 y + a1 y + ⋯ + an y = 0 (A)
has the factorization
p(r) = a0 p1 (r)p2 (r) ⋯ pk (r),
where each p is of the form

j
nj 2 2 mj
pj (r) = (r − rj ) or pj (r) = [(r − λj ) +w ] (ωj > 0)
j
and no two of the polynomials p , p , …, p have a common factor. Show that we can find a fundamental set of solutions
1 2 k
{ y , y , … , y } of (A) by finding a fundamental set of solutions of each of the equations

1 2 n
pj (D)y = 0, 1 ≤ j ≤ k,
and taking {y 1, y2 , … , yn } to be the set of all functions in these separate fundamental sets.
41.
a. Show that if
z = p(x) cos ωx + q(x) sin ωx, (A)
where p and q are polynomials of degree ≤ k , then

2 2
(D + ω )z = p1 (x) cos ωx + q1 (x) sin ωx,
where p and q are polynomials of degree ≤ k − 1 .

1 1
b. Apply (a) m times to show that if z is of the form (A) where p and q are polynomial of degree ≤ m − 1 , then
2 2 m
(D +ω ) z = 0. (B)
c. Use Equation 9.2.17 to show that if y = e λx

z then
2 2 m λx 2 2 m
[(D − λ ) +ω ] y =e (D +ω ) z.
d. Conclude from (b) and (c) that if p and q are arbitrary polynomials of degree ≤ m − 1 then
λx
y =e (p(x) cos ωx + q(x) sin ωx)
is a solution of
2 2 m
[(D − λ ) +ω ] y = 0. (C)
e. Conclude from (d) that the functions

λx λx m−1 λx
e cos ωx, x e cos ωx, …,x e cos ωx,
(D)
λx λx m−1 λx
e sin ωx, x e sin ωx, …,x e sin ωx
are all solutions of (C).

f. Complete the proof of Theorem 9.2.2 by showing that the functions in (D) are linearly independent.
42.
a. Use the trigonometric identities
sin(A + B) = cos A sin B + sin A cos B
to show that

(cos A + i sin A)(cos B + i sin B) = cos(A + B) + i sin(A + B).
b. Apply (a) repeatedly to show that if n is a positive integer then

n
∏(cos Ak + i sin Ak ) = cos(A1 + A2 + ⋯ + An ) + i sin(A1 + A2 + ⋯ + An ).
k=1
c. Infer from (b) that if n is a positive integer then

n
(cos θ + i sin θ) = cos nθ + i sin nθ. (A)
d. Show that (A) also holds if n = 0 or a negative integer. HINT: Verify by direct calculation that
−1
(cos θ + i sin θ) = (cos θ − i sin θ).
Then replace θ by −θ in (A).

e. Now suppose n is a positive integer. Infer from (A) that if
2kπ 2kπ
zk = cos( ) + i sin( ), k = 0, 1, … , n − 1,
n n
and
(2k + 1)π (2k + 1)π
ζk = cos( ) + i sin( ), k = 0, 1, … , n − 1,
n n
then
n n
z =1 and ζ = −1, k = 0, 1, … , n − 1.
k k
(Why don’t we also consider other integer values for k ?)

f. Let ρ be a positive number. Use (e) to show that
n 1/n 1/n 1/n
z − ρ = (z − ρ z0 )(z − ρ z1 ) ⋯ (z − ρ zn−1 )
and
n 1/n 1/n 1/n
z + ρ = (z − ρ ζ0 )(z − ρ ζ1 ) ⋯ (z − ρ ζn−1 ).
43. Use (e) of Exercise 9.2.42 to find a fundamental set of solutions of the given equation.
a. y − y = 0
′′′
b. y + y = 0
′′′
c. y + 64y = 0
(4)
d. y − y = 0
(6)
e. y + 64y = 0
(6)
f. [(D − 1) − 1] y = 0
6
g. y + y + y + y + y
(5) (4) ′′′ ′′ ′
+y = 0
44. An equation of the form

n (n) n−1 (n−1) ′
a0 x y + a1 x y + ⋯ + an−1 x y + an y = 0, x > 0, (A)
where a , a , …, a are constants, is an Euler or equidimensional equation. Show that if

0 1 n
t
x =e and Y (t) = y(x(t)), (B)
then

dy dY
x =
dx dt
2 2
d y d Y dY
2
x = −
2 2
dx dt dt
3 3 2
3
d y d Y d Y dY
x = −3 +2 .
dx3 dt3 dt2 dt
In general, it can be shown that if r is any integer ≥ 2 then

r r r−1
d y d Y d Y dY
r
x = + A1r + ⋯ + Ar−1,r
r r r−1
dx dt dt dt
where A , …, A
1r are integers. Use these results to show that the substitution (B) transforms (A) into a constant coefficient
r−1,r
equation for Y as a function of t .

45. Use Exercise 9.2.44 to show that a function y = y(x) satisfies the equation
3 ′′′ 2 ′′ ′
a0 x y + a1 x y + a2 x y + a3 y = 0, (A)
on (0, ∞) if and only if the function Y (t) = y(e t

) satisfies
3 2
d Y d Y dY
a0 + (a1 − 3 a0 ) + (a2 − a1 + 2 a0 ) + a3 Y = 0.
3 2
dt dt dt
Assuming that a , a , a , a are real and a

0 1 2 3 0 ≠0 , find the possible forms for the general solution of (A).

9.3: Undetermined Coefficients for Higher Order Equations
(n) (n−1)
a0 y + a1 y + ⋯ + an y = F (x), (9.3.1)
where n ≥ 3 and F is a linear combination of functions of the form

αx k
e (p0 + p1 x + ⋯ + pk x )
or
λx k k
e [(p0 + p1 x + ⋯ + pk x ) cos ωx + (q0 + q1 x + ⋯ + qk x ) sin ωx] .
From Theorem 9.1.5, the general solution of Equation 9.3.1 is y = y p + yc , where y is a particular solution of Equation 9.3.1
p
and y is the general solution of the complementary equation

c
(n) (n−1)
a0 y + a1 y + ⋯ + an y = 0.
In Section 9.2 we learned how to find y . Here we will learn how to find y when the forcing function has the form stated
c p
above. The procedure that we use is a generalization of the method that we used in Sections 5.4 and 5.5, and is again called
method of undetermined coefficients. Since the underlying ideas are the same as those in these section, we’ll give an informal
presentation based on examples.
Forcing Functions of the Form e αx

(p0 + p1 x+. . . +pk x )
k
We first consider equations of the form

(n) (n−1) αx k
a0 y + a1 y + ⋯ + an y = e (p0 + p1 x + ⋯ + pk x ) .
Example 9.3.1
′′′ ′′ ′ x 2 3
y + 3y + 2 y − y = e (21 + 24x + 28 x + 5 x ). (9.3.2)
Solution
Substituting
x
y = ue ,
′ x ′
y = e (u + u),
′′ x ′′ ′
y = e (u + 2 u + u),
′′′ x ′′′ ′′ ′
y = e (u + 3u + 3 u + u)
into Equation 9.3.2 and canceling e yields x
′′′ ′′ ′ ′′ ′ ′ 2 3
(u + 3u + 3 u + u) + 3(u + 2 u + u) + 2(u + u) − u = 21 + 24x + 28 x + 5x ,
or
′′′ ′′ ′ 2 3
u + 6u + 11 u + 5u = 21 + 24x + 28 x + 5x . (9.3.3)
Since the unknown u appears on the left, we can see that Equation 9.3.3 has a particular solution of the form
2 3
up = A + Bx + C x + Dx .
Then
′ 2
up = B + 2C x + 3Dx
′′
up = 2C + 6Dx
′′′
up = 6D.

Substituting from the last four equations into the left side of Equation 9.3.3 yields
′′′ ′′ ′ 2
up + 6 up + 11 up + 5 up = 6D + 6(2C + 6Dx) + 11(B + 2C x + 3Dx )
2 3
+ 5(A + Bx + C x + Dx )
= (5A + 11B + 12C + 6D) + (5B + 22C + 36D)x
2 3
+ (5C + 33D)x + 5Dx .
Comparing coefficients of like powers of x on the right sides of this equation and Equation 9.3.3 shows that up satisfies
Equation 9.3.3 if
5D =5
5C + 33D = 28
5B + 22C + 36D = 24
5A + 11B + 12C + 6D = 21.
Solving these equations successively yields D = 1 , C = −1 , B = 2 , A = 1 . Therefore

2 3
up = 1 + 2x − x +x
is a particular solution of Equation 9.3.3, so

x x 2 3
yp = e up = e (1 + 2x − x +x )
is a particular solution of Equation 9.3.2 (Figure 9.3.1).
Figure 9.3.1: y p
x
= e (1 + 2x − x
2
+x )
3
Example 9.3.2
(4) ′′′ ′′ ′ 2x 2
y −y − 6y + 4 y + 8y = e (4 + 19x + 6 x ). (9.3.4)
Solution
Substituting
2x
y = ue ,
′ 2x ′
y =e (u + 2u),
′′ 2x ′′ ′
y =e (u + 4 u + 4u),
′′′ 2x ′′′ ′′ ′
y =e (u + 6u + 12 u + 8u),
(4) 2x (4) ′′′ ′′ ′

y =e (u + 8u + 24 u + 32 u + 16u)
into Equation 9.3.4 and canceling e 2x

yields
(4) ′′′ ′′ ′ ′′′ ′′ ′
(u + 8u + 24 u + 32 u + 16u) − (u + 6u + 12 u + 8u)
′′ ′ ′ 2
−6(u + 4 u + 4u) + 4(u + 2u) + 8u = 4 + 19x + 6 x ,

or
(4) ′′′ ′′ 2
u + 7u + 12 u = 4 + 19x + 6 x . (9.3.5)
Since neither u nor u appear on the left, we can see that Equation 9.3.5 has a particular solution of the form
′
2 3 4
up = Ax + Bx + Cx . (9.3.6)
Then
′ 2 3
up = 2Ax + 3Bx + 4C x
′′ 2
up = 2A + 6Bx + 12C x
′′′
up = 6B + 24C x
(4)
up = 24C .
(4)
Substituting u , u , and u
′′
p
′′′
p p into the left side of Equation 9.3.5 yields
(4) 2
′′′ ′′
up + 7 up + 12 up = 24C + 7(6B + 24C x) + 12(2A + 6Bx + 12C x )
2
= (24A + 42B + 24C ) + (72B + 168C )x + 144C x .
Comparing coefficients of like powers of x on the right sides of this equation and Equation 9.3.5 shows that up satisfies
Equation 9.3.5 if
144C =6
72B + 168C = 19
24A + 42B + 24C = 4.
Solving these equations successively yields C = 1/24 , B = 1/6 , A = −1/6 . Substituting these into Equation 9.3.6
shows that
2
x
2
up = (−4 + 4x + x )
24

2 2x
2x
x e 2
yp = e up = (−4 + 4x + x )
24
2 2x
x e
Figure 9.3.2: y p = (−4 + 4x + x )
2
24
Forcing Functions of the Form e αx

(P (x) cos ωx + Q(x) sin ωx)
We now consider equations of the form

(n) (n−1) λx
a0 y + a1 y + ⋯ + an y = e (P (x) cos ωx + Q(x) sin ωx) ,

where P and Q are polynomials.
Example 9.3.3
′′′ ′′ ′ x
y +y − 4 y − 4y = e [(5 − 5x) cos x + (2 + 5x) sin x]. (9.3.7)
Solution
Substituting
x
y = ue ,
′ x ′
y = e (u + u),
′′ x ′′ ′
y = e (u + 2 u + u),
′′′ x ′′′ ′′ ′
y = e (u + 3u + 3 u + u)
into Equation 9.3.7 and canceling e yields x
′′′ ′′ ′ ′′ ′ ′
(u + 3u + 3 u + u) + (u + 2 u + u) − 4(u + u) − 4u = (5 − 5x) cos x + (2 + 5x) sin x,
or
′′′ ′′ ′
u + 4u + u − 6u = (5 − 5x) cos x + (2 + 5x) sin x. (9.3.8)
Since cos x and sin x are not solutions of the complementary equation
′′′ ′′ ′
u + 4u + u − 6u = 0,
a theorem analogous to Theorem 5.5.1 implies that Equation 9.3.8 has a particular solution of the form
up = (A0 + A1 x) cos x + (B0 + B1 x) sin x. (9.3.9)
Then
′
up = (A1 + B0 + B1 x) cos x + (B1 − A0 − A1 x) sin x,
′′
up = (2 B1 − A0 − A1 x) cos x − (2 A1 + B0 + B1 x) sin x,
′′′
up = −(3 A1 + B0 + B1 x) cos x − (3 B1 − A0 − A1 x) sin x,
so
′′′ ′′ ′
up + 4 up + up − 6 up = − [10 A0 + 2 A1 − 8 B1 + 10 A1 x] cos x − [10 B0 + 2 B1 + 8 A1 + 10 B1 x] sin x.
Comparing the coefficients of x cos x, x sin x, cos x, and sin x here with the corresponding coefficients in Equation 9.3.8
shows that u is a solution of Equation 9.3.8 if
p
−10A1 = −5
−10B1 = 5
−10 A0 − 2 A1 + 8 B1 =5
−10 B0 − 2 B1 − 8 A1 = 2.
Solving the first two equations yields A 1 = 1/2 ,B 1 = −1/2 . Substituting these into the last two equations yields
−10A0 = 5 + 2 A1 − 8 B1 = 10
−10B0 = 2 + 2 B1 + 8 A1 = 5,
so A 0 = −1 ,B 0 = −1/2 . Substituting A 0 = −1 ,A 1 = 1/2 ,B 0 = −1/2 ,B 1 = −1/2 into Equation 9.3.9 shows that
1
up = − [(2 − x) cos x + (1 + x) sin x]
2

x
x
e
yp = e up = − [(2 − x) cos x + (1 + x) sin x]
2
x
e
Figure 9.3.3: y p = {e up = −
x
[(2 − x) cos x + (1 + x) sin x]
2
Example 9.3.4
′′′ ′′ ′ −x
y + 4y + 6 y + 4y = e [(1 − 6x) cos x − (3 + 2x) sin x] . (9.3.10)
Solution
Substituting
−x
y = ue ,
′ −x ′
y =e (u − u),
′′ −x ′′ ′
y =e (u − 2 u + u),
′′′ −x ′′′ ′′ ′
y =e (u − 3u + 3 u − u)
into Equation 9.3.10 and canceling e −x

yields
′′′ ′′ ′ ′′ ′ ′
(u − 3u + 3 u − u) + 4(u − 2 u + u) + 6(u − u) + 4u = (1 − 6x) cos x − (3 + 2x) sin x,
or
′′′ ′′ ′
u +u + u + u = (1 − 6x) cos x − (3 + 2x) sin x. (9.3.11)
Since cos x and sin x are solutions of the complementary equation

′′′ ′′ ′
u +u + u + u = 0,
a theorem analogous to Theorem 5.5.1 implies that Equation 9.3.11 has a particular solution of the form
2 2
up = (A0 x + A1 x ) cos x + (B0 x + B1 x ) sin x. (9.3.12)
Then
′ 2 2
up = [ A0 + (2 A1 + B0 )x + B1 x ] cos x + [ B0 + (2 B1 − A0 )x − A1 x ] sin x,
′′ 2
up = [2 A1 + 2 B0 − (A0 − 4 B1 )x − A1 x ] cos x
2
+ [2 B1 − 2 A0 − (B0 + 4 A1 )x − B1 x ] sin x,
′′′ 2
up = − [3 A0 − 6 B1 + (6 A1 + B0 )x + B1 x ] cos x
2
− [3 B0 + 6 A1 + (6 B1 − A0 )x − A1 x ] sin x,
so

′′′ ′′ ′
up + up + up + up = − [2 A0 − 2 B0 − 2 A1 − 6 B1 + (4 A1 − 4 B1 )x] cos x
− [2 B0 + 2 A0 − 2 B1 + 6 A1 + (4 B1 + 4 A1 )x] sin x.
Comparing the coefficients of x cos x, x sin x, cos x, and sin x here with the corresponding coefficients in Equation
9.3.11 shows that u is a solution of Equation 9.3.11 if
p
−4 A1 + 4 B1 = −6
−4 A1 − 4 B1 = −2
−2 A0 + 2 B0 + 2 A1 + 6 B1 = 1
−2 A0 − 2 B0 − 6 A1 + 2 B1 = −3.
Solving the first two equations yields A 1 =1 ,B 1 = −1/2 . Substituting these into the last two equations yields
−2 A0 + 2 B0 = 1 − 2 A1 − 6 B1 = 2
−2 A0 − 2 B0 = −3 + 6 A1 − 2 B1 = 4,
−x
xe
Figure 9.3.4: y p = − [(3 − 2x) cos x + (1 + x) sin x]
2
so A = −3/2 and
0 B0 = −1/2 . Substituting A0 = −3/2 , A1 = 1 , B0 = −1/2 , B1 = −1/2 into Equation 9.3.12
shows that
x
up = − [(3 − 2x) cos x + (1 + x) sin x]
2

−x
−x
xe
yp = e up = − [(3 − 2x) cos x + (1 + x) sin x]
2
(Figure 9.3.4) is a particular solution of Equation 9.3.10.

9.3E: Undetermined Coefficients for Higher Order Equations (Exercises)
Q9.3.1
1. y ′′′
− 6y
′′
+ 11 y − 6y = −e
′ −x
(4 + 76x − 24 x )
2
2. y ′′′
− 2y
′′
− 5 y + 6y = e
′ −3x
(32 − 23x + 6 x )
2
3. 4y ′′′
+ 8y
′′ ′
− y − 2y = −e (4 + 45x + 9 x )
x 2
4. y ′′′
+ 3y
′′
− y − 3y = e
′ −2x
(2 − 17x + 3 x )
2
5. y ′′′
+ 3y
′′ ′
− y − 3y = e (−1 + 2x + 24 x
x 2
+ 16 x )
3
6. y ′′′
+y
′′
− 2y = e (14 + 34x + 15 x )
x 2
7. 4y ′′′
+ 8y
′′ ′
− y − 2y = −e
−2x
(1 − 15x)
8. y ′′′
−y
′′ ′
− y + y = e (7 + 6x)
x
9. 2y ′′′
− 7y
′′
+ 4 y + 4y = e
′ 2x
(17 + 30x)
10. y ′′′
− 5y
′′ ′
+ 3 y + 9y = 2 e
3x
(11 − 24 x )
2
11. y ′′′
− 7y
′′ ′
+ 8 y + 16y = 2 e
4x
(13 + 15x)
12. 8y ′′′
− 12 y
′′
+ 6y − y = e
′ x/2
(1 + 4x)
13. y (4)
+ 3y
′′′
− 3y
′′
− 7 y + 6y = −e
′ −x
(12 + 8x − 8 x )
2
14. y (4)
+ 3y
′′′
+y
′′
− 3 y − 2y = −3 e
′ 2x
(11 + 12x)
15. y (4)
+ 8y
′′′
+ 24 y
′′
+ 32 y
′
= −16 e
−2x
(1 + x + x
2
−x )
3
16. 4y (4)
− 11 y
′′ ′
− 9 y − 2y = −e (1 − 6x)
x
17. y (4)
− 2y
′′′ ′
+ 3 y − y = e (3 + 4x + x )
x 2
18. y (4)
− 4y
′′′
+ 6y
′′
− 4 y + 2y = e
′ 2x
(24 + x + x )
4
19. 2y (4)
+ 5y
′′′ ′
− 5 y − 2y = 18 e (5 + 2x)
x
20. y (4)
+y
′′′
− 2y
′′
− 6 y − 4y = −e
′ 2x
(4 + 28x + 15 x )
2
21. 2y (4)
+y
′′′
− 2y − y = 3e
′ −x/2
(1 − 6x)
22. y (4)
− 5y
′′
+ 4y = e (3 + x − 3 x )
x 2
23. y (4)
− 2y
′′′
− 3y
′′
+ 4 y + 4y = e
′ 2x
(13 + 33x + 18 x )
2
24. y (4)
− 3y
′′′
+ 4y
′
=e
2x
(15 + 26x + 12 x )
2
25. y (4)
− 2y
′′′ ′
+ 2 y − y = e (1 + x)
x
26. 2y (4)
− 5y
′′′
+ 3y
′′ ′
+ y − y = e (11 + 12x)
x
27. y (4)
+ 3y
′′′
+ 3y
′′
+y
′
=e
−x
(5 − 24x + 10 x )
2
28. y (4)
− 7y
′′′
+ 18 y
′′ ′
− 20 y + 8y = e
2x
(3 − 8x − 5 x )
2
29. y ′′′
−y
′′
− 4 y + 4y = e
′ −x
[(16 + 10x) cos x + (30 − 10x) sin x]
30. y ′′′
+y
′′
− 4 y − 4y = e
′ −x
[(1 − 22x) cos 2x − (1 + 6x) sin 2x]
31. y ′′′
−y
′′
+ 2 y − 2y = e
′ 2x 2
[(27 + 5x − x ) cos x + (2 + 13x + 9 x ) sin x]
2
32. y ′′′
− 2y
′′ ′ x
+ y − 2y = −e [(9 − 5x + 4 x ) cos 2x − (6 − 5x − 3 x ) sin 2x]
2 2
33. y ′′′
+ 3y
′′ ′
+ 4 y + 12y = 8 cos 2x − 16 sin 2x

34. y ′′′
−y
′′ x
+ 2y = e [(20 + 4x) cos x − (12 + 12x) sin x]
35. y ′′′
− 7y
′′ ′
+ 20 y − 24y = −e
2x
[(13 − 8x) cos 2x − (8 − 4x) sin 2x]
36. y ′′′
− 6y
′′
+ 18 y
′
= −e
3x
[(2 − 3x) cos 3x − (3 + 3x) sin 3x]
37. y (4)
+ 2y
′′′
− 2y
′′ ′
− 8 y − 8y = e (8 cos x + 16 sin x)
x
38. y (4)
− 3y
′′′
+ 2y
′′ ′
+ 2 y − 4y = e (2 cos 2x − sin 2x)
x
39. y (4)
− 8y
′′′
+ 24 y
′′ ′
− 32 y + 15y = e
2x
(15x cos 2x + 32 sin 2x)
40. y (4)
+ 6y
′′′
+ 13 y
′′ ′
+ 12 y + 4y = e
−x
[(4 − x) cos x − (5 + x) sin x]
41. y (4)
+ 3y
′′′
+ 2y
′′ ′
− 2 y − 4y = −e
−x
(cos x − sin x)
42. y (4)
− 5y
′′′
+ 13 y
′′ ′
− 19 y + 10y = e (cos 2x + sin 2x)
x
43. y (4)
+ 8y
′′′
+ 32 y
′′ ′
+ 64 y + 39y = e
−2x
[(4 − 15x) cos 3x − (4 + 15x) sin 3x]
44. y (4)
− 5y
′′′
+ 13 y
′′ ′ x
− 19 y + 10y = e [(7 + 8x) cos 2x + (8 − 4x) sin 2x]
45. y (4)
+ 4y
′′′
+ 8y
′′ ′
+ 8 y + 4y = −2 e
−x
(cos x − 2 sin x)
46. y (4)
− 8y
′′′
+ 32 y
′′ ′
− 64 y + 64y = e
2x
(cos 2x − sin 2x)
47. y (4)
− 8y
′′′
+ 26 y
′′ ′
− 40 y + 25y = e
2x
[3 cos x − (1 + 3x) sin x]
48. y ′′′
− 4y
′′ ′
+ 5 y − 2y = e
2x
− 4e
x
− 2 cos x + 4 sin x
49. y ′′′
−y
′′ ′
+ y − y = 5e
2x
+ 2e
x
− 4 cos x + 4 sin x
50. y ′′′
−y
′
= −2(1 + x) + 4 e
x
− 6e
−x
+ 96 e
3x
51. y ′′′
− 4y
′′ ′
+ 9 y − 10y = 10 e
2x
+ 20 e
x
sin 2x − 10
52. y ′′′
+ 3y
′′ ′
+ 3 y + y = 12 e
−x
+ 9 cos 2x − 13 sin 2x
53. y ′′′
+y
′′ ′
− y − y = 4e
−x
(1 − 6x) − 2x cos x + 2(1 + x) sin x
54. y (4)
− 5y
′′
+ 4y = −12 e
x
+ 6e
−x
+ 10 cos x
55. y (4)
− 4y
′′′
+ 11 y
′′ ′
− 14 y + 10y = −e (sin x + 2 cos 2x)
x
56. y (4)
+ 2y
′′′
− 3y
′′ ′
− 4 y + 4y = 2 e (1 + x) + e
x −2x
57. y (4)
+ 4y = sinh x cos x − cosh x sin x
58. y (4)
+ 5y
′′′
+ 9y
′′ ′
+ 7 y + 2y = e
−x
(30 + 24x) − e
−2x
59. y (4)
− 4y
′′′
+ 7y
′′ ′ x
− 6 y + 2y = e (12x − 2 cos x + 2 sin x)
Q9.3.2
60. y ′′′
−y
′′ ′
−y +y = e
2x
(10 + 3x)
61. y ′′′
+y
′′
− 2y = −e
3x
(9 + 67x + 17 x )
2
62. y ′′′
− 6y
′′
+ 11 y − 6y = e
′ 2x
(5 − 4x − 3 x )
2
63. y ′′′
+ 2y
′′
+y
′
= −2 e
−x
(7 − 18x + 6 x )
2
64. y ′′′
− 3y
′′ ′
+ 3 y − y = e (1 + x)
x
65. y (4)
− 2y
′′
+ y = −e
−x
(4 − 9x + 3 x )
2
66. y ′′′
+ 2y
′′ ′
− y − 2y = e
−2x
[(23 − 2x) cos x + (8 − 9x) sin x]
67. y (4)
− 3y
′′′
+ 4y
′′
− 2y
′
=e
x
[(28 + 6x) cos 2x + (11 − 12x) sin 2x]
68. y (4)
− 4y
′′′
+ 14 y
′′ ′
− 20 y + 25y = e
x
[(2 + 6x) cos 2x + 3 sin 2x]

Q9.3.3
69. y ′′′
− 2y
′′ ′
− 5 y + 6y = 2 e (1 − 6x),
x
y(0) = 2,
′
y (0) = 7, y
′′
(0) = 9
70. y ′′′
−y
′′ ′
− y + y = −e
−x
(4 − 8x), y(0) = 2,
′
y (0) = 0, y
′′
(0) = 0
71. 4y ′′′
− 3y − y = e
′ −x/2
(2 − 3x), y(0) = −1,
′
y (0) = 15, y
′′
(0) = −17
72. y (4)
+ 2y
′′′
+ 2y
′′ ′
+ 2y + y = e
−x
(20 − 12x), y(0) = 3, y (0) = −4, y
′ ′′
(0) = 7, y
′′′
(0) = −22
73. y ′′′
+ 2y
′′ ′
+ y + 2y = 30 cos x − 10 sin x, y(0) = 3,
′
y (0) = −4, y
′′
(0) = 16
74. y (4)
− 3y
′′′
+ 5y
′′
− 2y
′ x
= −2 e (cos x − sin x), y(0) = 2, y (0) = 0, y
′ ′′
(0) = − 1, y
′′′
(0) = −5
Q9.3.4
75. Prove: A function y is a solution of the constant coefficient nonhomogeneous equation
(n) (n−1) αx
a0 y + a1 y + ⋯ + an y = e G(x) (A)
if and only if y = ue αx
, where u satisfies the differential equation
(n−1) (n−2)
p (α) p (α)
(n) (n−1) (n−2)
a0 u + u + u + ⋯ + p(α)u = G(x) (B)
(n − 1)! (n − 2)!
and
n n−1
p(r) = a0 r + a1 r + ⋯ + an
is the characteristic polynomial of the complementary equation

(n) (n−1)
a0 y + a1 y + ⋯ + an y = 0.
76. Prove:
a. The equation
( n−1) ( n−2)
p (α) p (α)
(n) (n−1) (n−2)
a0 u + u + u + ⋯ + P (α)u
(n−1)! (n−2)!
k (A)
= (p0 + p1 x + ⋯ + pk x ) cos ωx
k
+ (q0 + q1 x + ⋯ + qk x ) sin ωx
has a particular solution of the form

m k k
up = x (u0 + u1 x + ⋯ + uk x ) cos ωx + (v0 + v1 x + ⋯ + vk x ) sin ωx.
b. If λ + iω is a zero of p with multiplicity m ≥ 1 , then (A) can be written as

′′ 2 k k
a(u + ω u) = (p0 + p1 x + ⋯ + pk x ) cos ωx + (q0 + q1 x + ⋯ + qk x ) sin ωx,
which has a particular solution of the form
up = U (x) cos ωx + V (x) sin ωx,
where
2 k+1 2 k+1
U (x) = u0 x + u1 x + ⋯ + uk x , V (x) = v0 x + v1 x + ⋯ + vk x
and
′′ ′ k
a(U (x) + 2ωV (x)) = p0 + p1 x + ⋯ + pk x
′′ ′ k
a(V (x) − 2ωU (x)) = q0 + q1 x + ⋯ + qk x .

9.4: Variation of Parameters for Higher Order Equations
Derivation of the method
We assume throughout this section that the nonhomogeneous linear equation
(n) (n−1)
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = F (x) (9.4.1)
is normal on an interval (a, b). We’ll abbreviate this equation as Ly = F , where

(n) (n−1)
Ly = P0 (x)y + P1 (x)y + ⋯ + Pn (x)y.
When we speak of solutions of this equation and its complementary equation Ly = 0 , we mean solutions on (a, b). We’ll
show how to use the method of variation of parameters to find a particular solution of Ly = F , provided that we know a
fundamental set of solutions {y , y , … , y } of Ly = 0 .
1 2 n
We seek a particular solution of Ly = F in the form
yp = u1 y1 + u2 y2 + ⋯ + un yn (9.4.2)
where {y 1, y2 , … , yn } is a known fundamental set of solutions of the complementary equation

(n) (n−1)
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = 0
and u , u , …, u are functions to be determined. We begin by imposing the following n − 1 conditions on u

1 2 n 1, :
u2 , … , un
′ ′ ′
u y1 + u y2 + ⋯ +un yn = 0
1 2
′ ′ ′ ′ ′ ′
u y +u y + ⋯ +un yn = 0
1 1 2 2
(9.4.3)
⋮
(n−2) (n−2) (n−2)

′ ′ ′
u y +u y + ⋯ +un yn = 0.
1 1 2 2
These conditions lead to simple formulas for the first n − 1 derivatives of y : p
(r) (r) (r) (r)

yp = u1 y + u2 y ⋯ + un yn , 0 ≤ r ≤ n − 1. (9.4.4)
1 2
These formulas are easy to remember, since they look as though we obtained them by differentiating Equation 9.4.2 n − 1
times while treating u , u , …, u as constants. To see that Equation 9.4.3 implies Equation 9.4.4, we first differentiate
1 2 n
Equation 9.4.2 to obtain

′ ′ ′ ′ ′ ′ ′
yp = u1 y + u2 y + ⋯ + un yn + u y1 + u y2 + ⋯ + un yn ,
1 2 1 2
which reduces to
′ ′ ′ ′
yp = u1 y + u2 y + ⋯ + un yn
1 2
because of the first equation in Equation 9.4.3. Differentiating this yields

′′ ′′ ′′ ′′ ′ ′ ′ ′ ′ ′
yp = u1 y + u2 y + ⋯ + un yn + u y +u y + ⋯ + un yn ,
1 2 1 1 2 2
which reduces to
′′ ′′ ′′ ′′
yp = u1 y + u2 y + ⋯ + un yn
1 2
because of the second equation in Equation 9.4.3. Continuing in this way yields Equation 9.4.4.
The last equation in Equation 9.4.4 is
(n−1) (n−1) (n−1) (n−1)
yp = u1 y + u2 y + ⋯ + un yn .
1 2

(n) (n) (n) (n) (n−1) (n−1) (n−1)
′ ′ ′
yp = u1 y + u2 y + ⋯ + un yn +u y +u y + ⋯ + un yn .
1 2 1 1 2 2

Substituting this and Equation 9.4.4 into Equation 9.4.1 yields
(n−1) (n−1) (n−1)
′ ′ ′
u1 Ly1 + u2 Ly2 + ⋯ + un Lyn + P0 (x) (u y +u y + ⋯ + un yn ) = F (x).
1 1 2 2
Since Ly i = 0 (1 ≤ i ≤ n) , this reduces to
(n−1) (n−1) (n−1)

F (x)
′ ′ ′
u y +u y + ⋯ + un yn = .
1 1 2 2
P0 (x)
Combining this equation with Equation 9.4.3 shows that
yp = u1 y1 + u2 y2 + ⋯ + un yn
is a solution of Equation 9.4.1 if

′ ′ ′
u y1 + u y2 + ⋯ +un yn = 0
1 2
′ ′ ′ ′ ′ ′
u y +u y + ⋯ +un yn = 0
1 1 2 2
(n−2) (n−2) (n−2)

′ ′ ′
u y +u y + ⋯ +un yn =0
1 1 2 2
(n−1) (n−1) (n−1)

′ ′ ′
u y +u y + ⋯ +un yn = F / P0 ,
1 1 2 2
which can be written in matrix form as

y1 y2 ⋯ yn
⎡ ⎤
′
u 0
⎢ ′ ′ ′ ⎥ ⎡ 1 ⎤ ⎡ ⎤
⎢ y y ⋯ yn ⎥
1 2
′
⎢ ⎥ ⎢ u ⎥ 0
⎢ ⎥ 2 ⎢ ⎥
⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ⎥ =⎢
⎢
⎥
⎥. (9.4.5)
⎢ ⋮ ⋮ ⋱ ⋮ ⎥ ⎢ ⋮ ⎥ ⋮
⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ′ ⎥
⎢ un−1 ⎢ 0 ⎥
⎢ (n−2) (n−2) (n−2)
⎥ ⎥
⎢ y1 y
2
⋯ yn ⎥ ⎣ ⎦
⎢ ⎥ ⎣ ′ ⎦ F /P0
un
(n−1) (n−1) (n−1)
⎣ ⎦
y y ⋯ yn
1 2
The determinant of this system is the Wronskian W of the fundamental set of solutions {y 1, y2 , … , yn } , which has no zeros
on (a, b), by Theorem 9.1.4. Solving Equation 9.4.5 by Cramer’s rule yields
F Wj
′ n−j
u = (−1 ) , 1 ≤ j ≤ n, (9.4.6)
j
P0 W
where W is the Wronskian of the set of functions obtained by deleting y from {y , y , … , y } and keeping the remaining
j j 1 2 n
functions in the same order. Equivalently, W is the determinant obtained by deleting the last row and j -th column of W .
j
Having obtained u , u , … ,u , we can integrate to obtain u , u , … , u . As in Section 5.7, we take the constants of
′
1
′
2
′
n 1 2 n
integration to be zero, and we drop any linear combination of {y , y , … , y } that may appear in y . 1 2 n p
Note
For efficiency, it's best to compute W 1, W2 , ⋯ , Wn first, and then compute W by expanding in cofactors of the last row;
thus,
n
n−j (n−1)
W = ∑(−1 ) y Wj .
j
j=1
Third Order Equations

If n = 3 , then

∣ y1 y2 y3 ∣
∣ ∣
∣ ′ ′ ′ ∣
W = y y y .
∣ 1 2 3 ∣
∣ ∣
′′ ′′ ′′
∣ y1 y
2
y
3
∣
Therefore
∣ y2 y3 ∣ ∣ y1 y3 ∣ ∣ y1 y2 ∣
W1 = ∣ ∣, W2 = ∣ ∣, W3 = ∣ ∣,
′ ′ ′ ′ ′ ′
∣ y2 y
3
∣ ∣ y1 y
3
∣ ∣ y1 y
2
∣

F W1 F W2 F W3
′ ′ ′
u = , u =− , u = . (9.4.7)
1 2 3
P0 W P0 W P0 W
Example 9.4.1
′′′ ′′ ′ 2 x
xy −y − xy + y = 8x e , (9.4.8)
given that y = x , y = e , and y = e

1 2
x
3
−x
form a fundamental set of solutions of the complementary equation. Then find
the general solution of Equation 9.4.8.
Solution
We seek a particular solution of Equation 9.4.8 of the form
x −x
yp = u1 x + u2 e + u3 e .
The Wronskian of {y 1, y2 , y3 } is
x −x
∣x e e ∣
∣ x −x ∣
W (x) = 1 e −e ,
∣ ∣
x −x
∣ 0 e e ∣
so
x −x
∣e e ∣
W1 = ∣ ∣ = −2,
x −x
∣e −e ∣
−x
∣x e ∣
−x
W2 = ∣ ∣ = −e (x + 1),
−x
∣ 1 −e ∣
x
∣x e ∣
x
W3 =∣ ∣ = e (x − 1).
x
∣ 1 e ∣
Expanding W by cofactors of the last row yields

x −x x −x −x x
W = 0 W1 − e W2 + e W3 = 0(−2) − e (−e (x + 1)) + e e (x − 1) = 2x.
Since F (x) = 8x 2
e
x
and P 0 (x) =x ,
2 x
F 8x e x
= = 4e .
P0 W x ⋅ 2x
Therefore, from Equation 9.4.7

′ x x x
u = 4 e W1 = 4 e (−2) = −8 e ,
1
′ x x −x
u = −4 e W2 = −4 e (−e (x + 1)) = 4(x + 1),
2
′ x x x 2x
u = 4 e W3 = 4e (e (x − 1)) = 4 e (x − 1).
3
Integrating and taking the constants of integration to be zero yields

x 2 2x
u1 = −8 e , u2 = 2(x + 1 ) , u3 = e (2x − 3).
Hence,
yp = u1 y1 + u2 y2 + u3 y3
x x 2 −x 2x
= (−8 e )x + e (2(x + 1 ) ) + e (e (2x − 3))
x 2
= e (2 x − 2x − 1).
Since −e is a solution of the complementary equation, we redefine

x
x
yp = 2x e (x − 1).

x x −x
y = 2x e (x − 1) + c1 x + c2 e + c3 e .
Fourth Order Equations

If n = 4 , then
∣ y1 y2 y3 y4 ∣
∣ ∣
∣ y′ y
′
y
′
y
′ ∣
1 2 3 4
∣ ∣
W = ,
∣ ′′ ′′ ′′ ′′ ∣
y y y y
∣ 1 2 3 4 ∣
∣ ∣
′′′ ′′′ ′′′ ′′′
∣ y1 y
2
y
3
y
4
∣
Therefore
∣ y2 y3 y4 ∣ ∣ y1 y3 y4 ∣
∣ ∣ ∣ ∣
∣ ′ ′ ′ ∣ ∣ ′ ′ ′ ∣
W1 = y y y , W2 = y y y ,
∣ 2 3 4 ∣ ∣ 1 3 4 ∣
∣ ∣ ∣ ∣
′′ ′′ ′′ ′′ ′′ ′′
∣ y2 y
3
y
4
∣ ∣ y1 y
3
y
4
∣
∣ y1 y2 y4 ∣ ∣ y1 y2 y3 ∣
∣ ∣ ∣ ∣
′ ′ ′ ∣ ′ ′ ′ ∣
W3 = ∣ y1 y
2
y
4
∣, W4 = y
1
y
2
y
3
,
∣ ∣
∣ ∣ ∣ ∣
′′ ′′ ′′ ′′ ′′ ′′
∣y y y ∣ ∣ y1 y y ∣
1 2 4 2 3

F W1 F W2 F W3 F W4
′ ′ ′ ′
u =− , u = , u =− , u = . (9.4.9)
1 2 3 4
P0 W P0 W P0 W P0 W
Example 9.4.2
4 (4) 3 ′′′ 2 ′′ ′ 2
x y + 6x y + 2x y − 4x y + 4y = 12 x , (9.4.10)
given that y = x , y = x , y = 1/x and y = 1/x form a fundamental set of solutions of the complementary
1 2
2
3 4
2
equation. Then find the general solution of Equation 9.4.10 on (−∞, 0) and (0, ∞).
Solution
We seek a particular solution of Equation 9.4.10 of the form
u3 u4
2
yp = u1 x + u2 x + + .
2
x x
The Wronskian of {y 1, y2 , y3 , y4 } is

2 2
∣x x 1/x −1/x ∣
∣ ∣
∣ 2 3 ∣
1 2x −1/x −2/x
∣ ∣
W (x) = ,
∣ 3 4
∣
∣ 0 2 2/x 6/x ∣
∣ ∣
4 5
∣ 0 0 −6/x −24/x ∣
so
2 2
∣ x 1/x 1/x ∣
∣ ∣
12
∣ 2 3 ∣
W1 = 2x −1/x −2/x =− ,
∣ ∣ 4
x
∣ ∣
3 4
∣ 2 2/x 6/x ∣
2
∣x 1/x 1/x ∣
∣ ∣
∣ ∣ 6
2 3
W2 = 1 −1/x −2/x =− ,
∣ ∣
x5
∣ ∣
3 4
∣ 0 2/x 6/x ∣
2 2
∣x x 1/x ∣
∣ ∣
∣ ∣ 12
3
W3 = 1 2x −2/x = ,
∣ ∣
x2
∣ ∣
4
∣ 0 2 6/x ∣
2
∣x x 1/x ∣
∣ ∣
6
∣ 2 ∣
W4 = 1 2x −1/x = .
∣ ∣
x
∣ ∣
3
∣ 0 2 2/x ∣
Expanding W by cofactors of the last row yields

6 24
W = −0 W1 + 0 W2 − (− ) W3 + (− ) W4
4 5
x x
6 12 24 6 72
= − =− .
4 2 5 6
x x x x x
Since F (x) = 12x and P

2
0 (x) =x
4
,
2 6 4
F 12x x x
= (− ) =− .
4
P0 W x 72 6
Therefore, from Equation 9.4.9,

4 4
x x 12
′
u = − (− ) W1 = (− ) = −2,
1 4
6 6 x
4 4
x x 6 1
′
u = − W2 = − (− ) = ,
2 5
6 6 x x
4 4
x x 12 2
′
u = − (− ) W3 = = 2x ,
3
6 6 x2
4 4
x x 6
′ 3
u = − W4 = − = −x .
4
6 6 x
Integrating these and taking the constants of integration to be zero yields

3 4
2x x
u1 = −2x, u2 = ln |x|, u3 = , u4 = − .
3 4
Hence,

yp = u1 y1 + u2 y2 + u3 y3 + u4 y4
3 4
2x 1 x 1
2
= (−2x)x + (ln |x|)x + + (− )
2
3 x 4 x
2
19x
2
=x ln |x| − .
12
Since −19x 2
/12 is a solution of the complementary equation, we redefine
2
yp = x ln |x|.
Therefore
2 2
c3 c4
y =x ln |x| + c1 x + c2 x + +
2
x x
is the general solution of Equation 9.4.10 on (−∞, 0) and (0, ∞).

9.4E: Variation of Parameters for Higher Order Equations (Exercises)
Q9.4.1
In Exercises 9.4.1-9.4.21 find a particular solution, given the fundamental set of solutions of the complementary equation.
1. x 3
y
′′′ 2
− x (x + 3)y
′′
+ 2x(x + 3)y − 2(x + 3)y = −4 x
′ 4
; {x, x , xe }
2 x
2. y ; {e
2 2 2 2
′′′ ′′ 2 ′ 3 1/2 −x −x −x 2 −x
+ 6x y + (6 + 12 x )y + (12x + 8 x )y = x e , xe , x e }
3. x 3
y
′′′
− 3x y
2 ′′
+ 6x y − 6y = 2x
′
; {x, x 2
,x }
3
4. x 2
y
′′′
+ 2x y
′′
− (x
2
+ 2)y
′
= 2x
2
;{1, x
e /x, e
−x
/x}
5. x 3
y
′′′ 2
− 3 x (x + 1)y
′′
+ 3x(x
2
+ 2x + 2)y − (x
′ 3
+ 3x
2
+ 6x + 6)y = x e
4 −3x
;{x e x 2
, x e , x e }
x 3 x
6. x(x 2
− 2)y
′′′
+ (x
2
− 6)y
′′
+ x(2 − x )y + (6 − x )y = 2(x
2 ′ 2 2
− 2)
2
; {e x
, e
−x
, 1/x}
7. x y ′′′
− (x − 3)y
′′
− (x + 2)y + (x − 1)y = −4 e
′ −x
; {e x x
, e /x, e
−x
/x}
− −
8. 4x 3
y
′′′
+ 4x y
2 ′′
− 5x y + 2y = 30 x
′ 2
; {√x, 2
1/ √x , x }
9. x(x 2
− 1)y
′′′
+ (5 x
2
+ 1)y
′′
+ 2x y − 2y = 12 x
′ 2
; {x, 1/(x − 1), 1/(x + 1)}
10. x(1 − x)y ′′′

+ (x
2
− 3x + 3)y
′′
+ x y − y = 2(x − 1 )
′ 2
; {x, 1/x, e /x}
x
11. x 3
y
′′′
+x y
2 ′′
− 2x y + 2y = x
′ 2
; {x, x , 1/x}
2
12. x y ′′′
−y
′′
− xy + y = x
′ 2
; {x, e , e
x −x
}
13. x y (4)
+ 4y
′′′
= 6 ln |x| ; {1, x, x , 1/x}
2
14. 16x 4
y
(4)
+ 96 x y
3 ′′′
+ 72 x y
2 ′′
− 24x y + 9y = 96 x
′ 5/2
; {√−
x,
−
1/ √x , x
3/2
, x
−3/2
}
15. x(x 2
− 6)y
(4)
+ 2(x
2
− 12)y
′′′
+ x(6 − x )y
2 ′′
+ 2(12 − x )y
2 ′
= 2(x
2
− 6)
2
;{1, 1/x, e , e
x −x
}
16. x 4
y
(4)
− 4x y
3 ′′′
+ 12 x y
2 ′′
− 24x y + 24y = x
′ 4
; {x, 2
x , x , x }
3 4
17. x 4
y
(4)
− 4x y
3 ′′′
+ 2 x (6 − x )y
2 2 ′′
+ 4x(x
2
− 6)y + (x
′ 4
− 4x
2
+ 24)y = 4 x e
5 x
;{x e x 2
, x e , xe
x −x 2
, x e
−x
}
18. x 4
y
(4)
+ 6x y
3 ′′′
+ 2x y
2 ′′
− 4x y + 4y = 12 x
′ 2
; {x, x 2
, 1/x, 1/ x }
2
19. x y (4)
+ 4y
′′′
− 2x y
′′
− 4 y + xy = 4 e
′ x
; {e x
, e
−x
, e /x, e
x −x
/x}
20. x y (4)
+ (4 − 6x)y
′′′
+ (13x − 18)y
′′
+ (26 − 12x)y + (4x − 12)y = 3 e
′ x
; {e x
, e
2x
, e /x, e
x 2x
/x}
21. x 4
y
(4)
− 4x y
3 ′′′ 2
+ x (12 − x )y
2 ′′
+ 2x(x
2 ′
− 12)y + 2(12 − x )y = 2 x
2 5
; {x, 2
x , xe , xe
x −x
}
Q9.4.2
In Exercises 9.4.22-9.4.33 solve the initial value problem, given the fundamental set of solutions of the complementary
equation. Graph the solution for Exercises 9.4.22, 9.4.26, 9.4.29, and 9.4.30.
22. x 3
y
′′′
− 2x y
2 ′′
+ 3x y − 3y = 4x,
′
y(1) = 4,
′
y (1) = 4, y
′′
(1) = 2 ; {x, 3
x , x ln x}
23. x 3
y
′′′
− 5x y
2 ′′
+ 14x y − 18y = x ,
′ 3
y(1) = 0,
′
y (1) = 1, y
′′
(1) = 7 ; {x 2
, x , x
3 3
ln x}
24. (5 − 6x)y ′′′

+ (12x − 4)y
′′
+ (6x − 23)y + (22 − 12x)y = −(6x − 5 ) e
′ 2 x
{y(0) = −4,
′
y (0) = −
3
2
,
′′ x 2x −x
y (0) = −19; { e , e , xe }
25. x 3
y
′′′
− 6x y
2 ′′
+ 16x y − 16y = 9 x ,
′ 4
y(1) = 2,
′
y (1) = 1, y
′′
(1) = 5 ;{x, 4
x , x
4
ln |x|}
26. (x 2
− 2x + 2)y
′′′
−x y
2 ′′
+ 2x y − 2y = (x
′ 2
− 2x + 2 ) ,
2
y(0) = 0,
′
y (0) = 5 ,y ′′
(0) = 0 {x, x , e } ; 2 x
27. x 3
y
′′′
+x y
2 ′′ ′
− 2x y + 2y = x(x + 1), y(−1) = −6,
′
y (−1) =
43
6
, y
′′
(−1) = −
5
2
; {x, x , 1/x}
2
28. (3x − 1)y ′′′

− (12x − 1)y
′′
+ 9(x + 1)y − 9y = 2 e (3x − 1 ) ,
′ x 2
y(0) =
3
4
,
′
y (0) =
5
4
,
′′ 1 x 3x
y (0) = ; {x + 1, e , e }
4

29. (x 2
− 2)y
′′′
− 2x y
′′
+ (2 − x )y + 2xy = 2(x
2 ′ 2
− 2) ,
2
y(0) = 1,
′
y (0) = −5 ,y ′′
;
(0) = 5 { x , e , e
2 x −x
}
30. x 4
y
(4)
+ 3x y
3 ′′′
−x y
2 ′′ ′
+ 2x y − 2y = 9 x ,
2
y(1) = −7,
′
y (1) = −11, y
′′
(1) = −5, y
′′′
(1) = 6;
2
{x, x , 1/x, x ln x}
31. (2x − 1)y (4)

− 4x y
′′′
+ (5 − 2x)y
′′ ′
+ 4x y − 4y = 6(2x − 1 ) ,
2
y(0) =
55
4
,
′
y (0) = 0, y
′′
(0) = 13,
′′′ x −x 2x
y (0) = 1; {x, e , e , e }
32. 4x 4
y
(4)
+ 24 x y
3 ′′′
+ 23 x y
2 ′′
− x y + y = 6x,
′
y(1) = 2,
′
y (1) = 0, y
′′
(1) = 4,
′′′ 37 − −
y (1) = − ; {x, √x , 1/x, 1/ √x }
4
33. x 4
y
4
+ 5x y
3 ′′′
− 3x y
2 ′′ ′
− 6x y + 6y = 40 x ,
3
y(−1) = −1, y (−1) = −7
′
,
y
′′
(−1) = −1, y
′′′
(−1) = −31 ; {x, 3
x , 1/x, 1/ x }
2
Q9.4.3
34. Suppose the equation
(n) (n−1)
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = F (x) (A)
is normal on an interval (a, b). Let {y , y , … , y } be a fundamental set of solutions of its complementary equation on (a, b),
1 2 n
let W be the Wronskian of {y , y , … , y }, and let W be the determinant obtained by deleting the last row and the j -th
1 2 n j
column of W . Suppose x is in (a, b), let 0
x
F (t)Wj (t)
(n−j)
uj (x) = (−1 ) ∫ dt, 1 ≤ j ≤ n,
x0 P0 (t)W (t)
and define
yp = u1 y1 + u2 y2 + ⋯ + un yn .
a. Show that y is a solution of (A) and that

p
(r) (r) (r) (r)

yp = u1 y + u2 y ⋯ + un yn , 1 ≤ r ≤ n − 1,
1 2
and
(n) (n) (n) (n) F

yp = u1 y + u2 y + ⋯ + un yn + .
1 2
P0
HINT: See the derivation of the method of variation of parameters at the beginning of the section.
b. Show that y is the solution of the initial value problem
p
(n) (n−1)
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = F (x),
′ (n−1)
y(x0 ) = 0, y (x0 ) = 0, … , y (x0 ) = 0.
c. Show that y can be written as p
yp (x) = ∫ G(x, t)F (t) dt,

x0
where
∣ y1 (t) y2 (t) ⋯ yn (t) ∣
∣ ∣
′ ′ ′
∣ y (t) y (t) ⋯ yn (t) ∣
1 2
∣ ∣
1 ∣ ∣
G(x, t) = ⋮ ⋮ ⋱ ⋮ ,
P0 (t)W (t) ∣ ∣
∣ ∣
(n−2) (n−2) (n−2)
∣y (t) y (t) ⋯ yn (t) ∣
1 2
∣ ∣
∣ y1 (x) y2 (x) ⋯ yn (x) ∣

which is called the Green’s function for (A).
d. Show that
∣ y1 (t) y2 (t) ⋯ yn (t) ∣
∣ ∣
∣ ′ ′ ′ ∣
y (t) y (t) ⋯ yn (t)
1 2
∣ ∣
j ∣ ∣
∂ G(x, t) 1
= ∣ ⋮ ⋮ ⋱ ⋮ ∣, 0 ≤ j ≤ n.
j
∂x P0 (t)W (t) ∣ ∣
∣ (n−2) (n−2) (n−2) ∣
y (t) y (t) ⋯ yn (t)
∣ 1 2 ∣
∣ ∣
(j) (j) (j)
∣ y (x) y (x) ⋯ yn (x) ∣
1 2
e. Show that if a < t < b then

j
∂ G(x, t) ∣ 0, 1 ≤ j ≤ n−2
∣ ={ 1
∂x j
∣ , j = n−1
x=t P0 (t)
f. Show that
j
x ∂ G(x,t)
⎧
⎪ ∫ j
F (t)dt, 1 ≤ j ≤ n − 1,
(j) x0 ∂x
yp (x) = ⎨ ( n)
F (x) x ∂ G(x,t)
⎩
⎪ +∫ n
F (t)dt, j = n.
P0 (x) x0 ∂x
Q9.4.4
In Exercises 9.4.35-9.4.42 use the method suggested by Exercise 9.4.34 to find a particular solution in the form
x
y =∫
p G(x, t)F (t)dt , given the indicated fundamental set of solutions. Assume that x and x are in an interval on which 0
x0
the equation is normal.

35. y ′′′ ′
+ 2 y − y − 2y = F (x);
′
{e , e
x −x
,e
−2x
}
36. x 3
y
′′′
+x y
2 ′′ ′
− 2x y + 2y = F (x); {x, x , 1/x}
2
37. x 3
y
′′′ 2
− x (x + 3)y
′′ ′
+ 2x(x + 3)y − 2(x + 3)y = F (x); {x, x , x e }
2 x
38. x(1 − x)y ′′′

+ (x
2
− 3x + 3)y
′′ ′
+ x y − y = F (x); {x, 1/x, e /x}
x
39. y (4)
− 5y
′′
+ 4y = F (x);
x
{e , e
−x
, e
2x
, e
−2x
}
40. x y (4)
+ 4y
′′′
= F (x); {1, x, x , 1/x}
2
41. x 4
y
(4)
+ 6x y
3 ′′′
+ 2x y
2 ′′ ′
− 4x y + 4y = F (x) ; {x, x 2
, 1/x, 1/ x }
2
42. x y (4)
−y
′′′
− 4x y + 4 y
′ ′
= F (x); {1, x , e
2 2x
,e
−2x
}

CHAPTER OVERVIEW
10: LINEAR SYSTEMS OF DIFFERENTIAL EQUATIONS
IN THIS CHAPTER we consider systems of differential equations involving more than one
unknown function. Such systems arise in many physical applications.SECTION 10.1 presents
examples of physical situations that lead to systems of differential equations. SECTION 10.2
discusses linear systems of differential equations. SECTION 10.3 deals with the basic theory of
homogeneous linear systems. SECTIONS 10.4, 10.5, AND 10.6 present the theory of constant
coefficient homogeneous systems. SECTION 10.7 presents the method of variation of parameters for
nonhomogeneous linear systems.
10.1: INTRODUCTION TO SYSTEMS OF DIFFERENTIAL EQUATIONS

Many physical situations are modeled by systems of n differential equations in n unknown functions, where n≥2 . This section
presents examples of physical situations that lead to systems of differential equations.
10.1E: INTRODUCTION TO SYSTEMS OF DIFFERENTIAL EQUATIONS (EXERCISES)

10.2: LINEAR SYSTEMS OF DIFFERENTIAL EQUATIONS
A first order system of differential equations are introduced.
10.2E: LINEAR SYSTEMS OF DIFFERENTIAL EQUATIONS (EXERCISES)

10.3: BASIC THEORY OF HOMOGENEOUS LINEAR SYSTEMS
In this section we consider homogeneous linear systems y′=A(t)y, where A=A(t) is a continuous n×n matrix function on an interval
(a,b). The theory of linear homogeneous systems has much in common with the theory of linear homogeneous scalar equations.
10.3E: BASIC THEORY OF HOMOGENEOUS LINEAR SYSTEMS (EXERCISES)

10.4: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS I
We’ll now begin our study of the homogeneous system y′=Ay, where A is an n×n constant matrix. . In this section we assume that all
the eigenvalues of A are real and that A has a set of n linearly independent eigenvectors. In the next two sections we consider the
cases where some of the eigenvalues of A are complex, or where A does not have n linearly independent eigenvectors.
10.4E: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS I (EXERCISES)

10.5: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS II
In this section we consider the case where A has n real eigenvalues, but does not have n linearly independent eigenvectors. It is shown
in linear algebra that this occurs if and only if A has at least one eigenvalue of multiplicity r>1 such that the associated eigenspace has
dimension less than r . In this case A is said to be defective. We will restrict our attention to some commonly occurring special cases.
10.5E: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS II (EXERCISES)

10.6: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS III
We now consider the system y′=Ay , where A has a complex eigenvalue λ=α+iβ with β≠0 . We continue to assume that A has real
entries, so the characteristic polynomial of A has real coefficients. This implies that λ=α−iβ is also an eigenvalue of A .
10.6E: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS III (EXERCISES)
1 9/16/2020
10.1: Introduction to Systems of Differential Equations
Many physical situations are modelled by systems of n differential equations in n unknown functions, where n ≥ 2 . The next
three examples illustrate physical problems that lead to systems of differential equations. In these examples and throughout
this chapter we’ll denote the independent variable by t .
Example 10.1.1 : Flow

Tanks T and T contain 100 gallons and 300 gallons of salt solutions, respectively. Salt solutions are simultaneously
1 2
added to both tanks from external sources, pumped from each tank to the other, and drained from both tanks (Figure
10.1.1 ). A solution with 1 pound of salt per gallon is pumped into T from an external source at 5 gal/min, and a solution
1
with 2 pounds of salt per gallon is pumped into T from an external source at 4 gal/min. The solution from T is pumped
2 1
into T at 2 gal/min, and the solution from T is pumped into T at 3 gal/min. T is drained at 6 gal/min and T is drained
2 2 1 1 2
at 3 gal/min. Let Q (t) and Q (t) be the number of pounds of salt in T and T , respectively, at time t > 0 . Derive a
1 2 1 2
system of differential equations for Q and Q . Assume that both mixtures are well stirred.
1 2
Figure 10.1.1
As in Section 4.2, let rate in and rate out denote the rates (lb/min) at which salt enters and leaves a tank; thus,
′
Q = (rate in)1 − (rate out)1 ,
1
′
Q = (rate in)2 − (rate out)2 .
2
Note that the volumes of the solutions in T and T remain constant at 100 gallons and 300 gallons, respectively.
1 2
T1 receives salt from the external source at the rate of
(1 lb/gal) × (5~gal/min) = 5 lb/min,
and from T at the rate of

2
1
(lb/gal in T2 ) × (3~gal/min) = Q2 × 3
300
1
= Q2 lb/min.
100
Therefore
1
(rate in) 1 = 5 + Q2 . (10.1.1)
100
Solution leaves T at the rate of 8 gal/min, since 6 gal/min are drained and 2 gal/min are pumped to T ; hence,
1 2
1 2
(rate out)1 = ( lb/gal in T1 ) × (8~gal/min) = Q1 × 8 = Q1 . (10.1.2)
100 25

Equations 10.1.1 and Equation 10.1.2 imply that
1 2
′
Q =5+ Q2 − Q1 . (10.1.3)
1
100 25
T2 receives salt from the external source at the rate of
(2 lb/gal) × (4~gal/min) = 8 lb/min,
and from T at the rate of

1
1
(lb/gal in T1 ) × (2~gal/min) = Q1 × 2
100
1
= Q1 lb/min.
50
Therefore
1
(rate in) =8+ Q1 . (10.1.4)
2
50
Solution leaves T at the rate of 6 gal/min, since 3 gal/min are drained and 3 gal/min are pumped to T ; hence,
2 1
1 1
(rate out)2 = ( lb/gal in T ) × (6~gal/min) = Q2 × 6 = Q2 . (10.1.5)
2
300 50
Equations 10.1.4 and 10.1.5 imply that

1 1
′
Q =8+ Q1 − Q2 . (10.1.6)
2
50 50
We say that Equations 10.1.3 and 10.1.6 form a system of two first order equations in two unknowns, and write them
together as
2 1
′
Q =5− Q1 + Q2
1
25 100
1 1
′
Q =8+ Q1 − Q2 .
2
50 50
Example 10.1.2 : Coupled Springs

A mass m is suspended from a rigid support on a spring S and a second mass m is suspended from the first on a
1 1 2
spring S (Figure 10.1.2). The springs obey Hooke’s law, with spring constants k and k . Internal friction causes the
2 1 2
springs to exert damping forces proportional to the rates of change of their lengths, with damping constants c and c . Let 1 2
y = y (t) and y = y (t) be the displacements of the two masses from their equilibrium positions at time t , measured
1 1 2 2
positive upward. Derive a system of differential equations for y and y , assuming that the masses of the springs are
1 2
negligible and that vertical external forces F and F also act on the objects.
1 2
Solution
In equilibrium, S supports both m and m and S supports only m . Therefore, if Δℓ and Δℓ are the elongations of
1 1 2 2 2 1 2
the springs in equilibrium then

(m1 + m2 )g = k1 Δℓ1 and m2 g = k2 Δℓ2 . (10.1.7)

Figure 10.1.2
Let H be the Hooke’s law force acting on m , and let D be the damping force on m . Similarly, let H and D be the
1 1 1 1 2 2
Hooke’s law and damping forces acting on m . According to Newton’s second law of motion,
2
′′
m1 y = −m1 g + H1 + D1 + F1 ,
1
(10.1.8)
′′
m2 y = −m2 g + H2 + D2 + F2 .
2
When the displacements are y and y , the change in length of S is −y + Δℓ and the change in length of S is
1 2 1 1 1 2
−y + y + Δℓ . Both springs exert Hooke’s law forces on m , while only S exerts a Hooke’s law force on m . These
2 1 2 1 2 2
forces are in directions that tend to restore the springs to their natural lengths. Therefore
H1 = k1 (−y1 + Δℓ1 ) − k2 (−y2 + y1 + Δℓ2 ) and H2 = k2 (−y2 + y1 + Δℓ2 ). (10.1.9)
When the velocities are y and y , S and S are changing length at the rates −y and −y + y , respectively. Both
′
1
′
2
1 2
′
1
′
2
′
1
springs exert damping forces on m , while only S exerts a damping force on m . Since the force due to damping exerted
1 2 2
by a spring is proportional to the rate of change of length of the spring and in a direction that opposes the change, it
follows that
′ ′ ′ ′ ′
D1 = −c1 y + c2 (y −y ) and D2 = −c2 (y − y ). (10.1.10)
1 2 1 2 1
From Equations 10.1.8, 10.1.9, and 10.1.10,

′ ′ ′
= −m1 g + k1 (−y1 + Δℓ1 ) − k2 (−y2 + y1 + Δℓ2 ) − c1 y + c2 (y − y ) + F1
1 2 1
(10.1.11)
′ ′ ′
= −(m1 g − k1 Δℓ1 + k2 Δℓ2 ) − k1 y1 + k2 (y2 − y1 ) − c1 y + c2 (y − y ) + F1
1 2 1
and
′′ ′ ′
m2 y = −m2 g + k2 (−y2 + y1 + Δℓ2 ) − c2 (y − y ) + F2
2 2 1
(10.1.12)
′ ′
= −(m2 g − k2 Δℓ2 ) − k2 (y2 − y1 ) − c2 (y − y ) + F2 .
2 1
m1 g − k1 Δℓ1 + k2 Δℓ2 = −m2 g + k2 Δℓ2 = 0.
Therefore we can rewrite Equation 10.1.11 and Equation 10.1.12 as

′′ ′ ′
m1 y = −(c1 + c2 )y + c2 y − (k1 + k2 )y1 + k2 y2 + F1
1 1 2
′′ ′ ′
m2 y = c2 y − c2 y + k2 y1 − k2 y2 + F2 .
2 1 2
Example 10.1.3 : Gravity

Let X = X(t) = x(t) i + y(t) j + z(t) k be the position vector at time t of an object with mass m , relative to a
rectangular coordinate system with origin at Earth’s center (Figure 10.1.3).
Figure 10.1.3
According to Newton’s law of gravitation, Earth’s gravitational force F = F(x, y, z) on the object is inversely
proportional to the square of the distance of the object from Earth’s center, and directed toward the center; thus,
K X xi+yj +z k
F = (− ) = −K , (10.1.13)
2 3/2
||X|| ||X|| (x2 + y2 + z2 )
where K is a constant. To determine K , we observe that the magnitude of F is

∥X∥ K K
∥F∥ = K = = .
3 2 2
∥X∥ ∥X∥ (x + y2 + z2 )
Let R be Earth’s radius. Since ∥F∥ = mg when the object is at Earth’s surface,
K
2
mg = , so K = mgR .
2
R
Therefore we can rewrite Equation 10.1.13 as
2
x i+y j +zk
F = −mgR .
2 2 2 3/2
(x +y +z )
Now suppose F is the only force acting on the object. According to Newton’s second law of motion, F = mX ; that is, ′′
x i+y j +zk
′′ ′′ ′′ 2
m(x i+y j +z k) = −mgR .
3/2
(x2 + y 2 + z 2 )
Cancelling the common factor m and equating components on the two sides of this equation yields the system
2
gR x
′′
x =−
2 2 2 3/2
(x +y +z )
2
gR y
′′
y =− (10.1.14)
2 2 2 3/2
(x +y +z )
2
gR z
′′
z =− .
2 2 2 3/2
(x +y +z )
Rewriting Higher Order Systems as First Order Systems

A system of the form

′
y = g1 (t, y1 , y2 , … , yn )
1
′
y = g2 (t, y1 , y2 , … , yn )
2
(10.1.15)
′
yn = gn (t, y1 , y2 , … , yn )
is called a first order system, since the only derivatives occurring in it are first derivatives. The derivative of each of the
unknowns may depend upon the independent variable and all the unknowns, but not on the derivatives of other unknowns.
When we wish to emphasize the number of unknown functions in Equation 10.1.15 we will say that Equation 10.1.15 is an
n × n system.
Systems involving higher order derivatives can often be reformulated as first order systems by introducing additional
unknowns. The next two examples illustrate this.
Example 10.1.4
Rewrite the system
′′ ′ ′
m1 y = −(c1 + c2 )y + c2 y − (k1 + k2 )y1 + k2 y2 + F1
1 1 2
(10.1.16)
′′ ′ ′
m2 y = c2 y − c2 y + k2 y1 − k2 y2 + F2 .
2 1 2
derived in Example 10.1.2 as a system of first order equations.

Solution
If we define v 1 =y
1
′
and v
2 =y
′
2
, then v ′
1
=y
′′
1
and v ′
2
=y
2
′′
, so Equation 10.1.16 becomes
′
m1 v = −(c1 + c2 )v1 + c2 v2 − (k1 + k2 )y1 + k2 y2 + F1
1
′
m2 v = c2 v1 − c2 v2 + k2 y1 − k2 y2 + F2 .
2
Therefore {y 1, y2 , v1 , v2 } satisfies the 4 × 4 first order system

′
y = v1
1
′
y = v2
2
1
v
′
= [−(c1 + c2 )v1 + c2 v2 − (k1 + k2 )y1 + k2 y2 + F1 ] (10.1.17)
1
m1
1
′
v = [ c2 v1 − c2 v2 + k2 y1 − k2 y2 + F2 ] .
2
m2
Note
The difference in form between Equation 10.1.15 and Equation 10.1.17, due to the way in which the unknowns are
denoted in the two systems, isn’t important; Equation 10.1.17 is a first order system, in that each equation in Equation
10.1.17 expresses the first derivative of one of the unknown functions in a way that does not involve derivatives of any of
the other unknowns.
Example 10.1.5
Rewrite the system
′′ ′ ′ ′′
x = f (t, x, x , y, y , y )
′′′ ′ ′ ′′
y = g(t, x, x , y, y y )
as a first order system.

Solution
We regard x, x , y , y , and y as unknown functions, and rename them
′ ′ ′′

′ ′ ′′
x = x1 , x = x2 , y = y1 , y = y2 , y = y3 .
These unknowns satisfy the system

′
x = x2
1
′
x = f (t, x1 , x2 , y1 , y2 , y3 )
2
′
y = y2
1
′
y = y3
2
′
y = g(t, x1 , x2 , y1 , y2 , y3 ).
3
Rewriting Scalar Differential Equations as Systems

In this chapter we’ll refer to differential equations involving only one unknown function as scalar differential equations.
Scalar differential equations can be rewritten as systems of first order equations by the method illustrated in the next two
examples.
Example 10.1.6
Rewrite the equation
(4) ′′′ ′′ ′
y + 4y + 6y + 4y + y = 0 (10.1.18)
as a 4 × 4 first order system.

Solution
We regard y , y , y , and y as unknowns and rename them
′ ′′ ′′′
′ ′′ ′′′
y = y1 , y = y2 , y = y3 , and y = y4 .
Then y (4)
=y
4
′
, so Equation 10.1.18 can be written as
′
y + 4 y4 + 6 y3 + 4 y2 + y1 = 0.
4
Therefore {y 1, y2 , y3 , y4 } satisfies the system

′
y = y2
1
′
y = y3
2
′
y = y4
3
′
y = −4 y4 − 6 y3 − 4 y2 − y1 .
4
Example 10.1.7
Rewrite
′′′ ′ ′′
x = f (t, x, x , x )
as a system of first order equations.

Solution
We regard x, x , and x as unknowns and rename them
′ ′′
′ ′′
x = y1 , x = y2 , and x = y3 .
Then
′ ′ ′ ′′ ′ ′′′
y = x = y2 , y =x = y3 , and y =x .
1 2 3
Therefore {y 1, y2 , y3 } satisfies the first order system

′
y = y2
1
′
y = y3
2
′
y = f (t, y1 , y2 , y3 ).
3
Since systems of differential equations involving higher derivatives can be rewritten as first order systems by the method used
in Examples 10.1.5-10.1.7 , we’ll consider only first order systems.
Numerical Solution of Systems

The numerical methods that we studied in Chapter 3 can be extended to systems, and most differential equation software
packages include programs to solve systems of equations. We will not go into detail on numerical methods for systems;
however, for illustrative purposes we’ll describe the Runge-Kutta method for the numerical solution of the initial value
problem
′
y = g1 (t, y1 , y2 ), y1 (t0 ) = y10 ,
1
′
y = g2 (t, y1 , y2 ), y2 (t0 ) = y20
2
at equally spaced points t , t , …, t

0 1 n =b in an interval [t 0, b]. Thus,
ti = t0 + ih, i = 0, 1, … , n,
where
b − t0
h = .
n
We’ll denote the approximate values of y and y at these points by y , y , … , y

1 2 10 11 1n and y
20 , . The Runge-Kutta
y21 , … , y2n
method computes these approximate values as follows: given y and y , compute 1i 2i
I1i = g1 (ti , y1i , y2i ),
J1i = g2 (ti , y1i , y2i ),
h h h
I2i = g1 ( ti + , y1i + I1i , y2i + J1i ) ,
2 2 2
h h h
J2i = g2 ( ti + , y1i + I1i , y2i + J1i ) ,
2 2 2
h h h
I3i = g1 ( ti + , y1i + I2i , y2i + J2i ) ,
2 2 2
h h h
J3i = g2 ( ti + , y1i + I2i , y2i + J2i ) ,
2 2 2
I4i = g1 (ti + h, y1i + h I3i , y2i + h J3i ),
J4i = g2 (ti + h, y1i + h I3i , y2i + h J3i ),
and
h
y1,i+1 = y1i + (I1i + 2 I2i + 2 I3i + I4i ),
6
h
y2,i+1 = y2i + (J1i + 2 J2i + 2 J3i + J4i )
6
for i = 0 , …, n − 1 . Under appropriate conditions on g and g , it can be shown that the global truncation error for the
1 2
Runge-Kutta method is O(h ), as in the scalar case considered in Section 3.3.

4

10.1E: Introduction to Systems of Differential Equations (Exercises)
Q10.1.1
1. Tanks T and T contain 50 gallons and 100 gallons of salt solutions, respectively. A solution with 2 pounds of salt per
1 2
gallon is pumped into T from an external source at 1 gal/min, and a solution with 3 pounds of salt per gallon is pumped into
1
T2 from an external source at 2 gal/min. The solution from T is pumped into T at 3 gal/min, and the solution from T is 1 2 2
pumped into T at 4 gal/min. T is drained at 2 gal/min and T is drained at 1 gal/min. Let Q (t) and Q (t) be the number of
1 1 2 1 2
pounds of salt in T and T , respectively, at time t > 0 . Derive a system of differential equations for Q and Q . Assume that
1 2 1 2
both mixtures are well stirred.

2. Two 500 gallon tanks T and T initially contain 100 gallons each of salt solution. A solution with 2 pounds of salt per
1 2
gallon is pumped into T from an external source at 6 gal/min, and a solution with 1 pound of salt per gallon is pumped into
1
T2 from an external source at 5 gal/min. The solution from T is pumped into T at 2 gal/min, and the solution from T is 1 2 2
pumped into T at 1 gal/min. Both tanks are drained at 3 gal/min. Let Q (t) and Q (t) be the number of pounds of salt in T
1 1 2 1
and T , respectively, at time t > 0 . Derive a system of differential equations for Q and Q that’s valid until a tank is about to
2 1 2
overflow. Assume that both mixtures are well stirred.

3. A mass m is suspended from a rigid support on a spring S with spring constant k and damping constant c . A second
1 1 1 1
mass m is suspended from the first on a spring S with spring constant k and damping constant c , and a third mass m is
2 2 2 2 3
suspended from the second on a spring S with spring constant k and damping constant c . Let y = y (t) , y = y (t) , and
3 3 3 1 1 2 2
y = y (t) be the displacements of the three masses from their equilibrium positions at time t , measured positive upward.
3 3
Derive a system of differential equations for y , y and y , assuming that the masses of the springs are negligible and that
1 2 3
vertical external forces F , F , and F also act on the masses. 1 2 3
4. Let X = x i + y j + z k be the position vector of an object with mass m, expressed in terms of a rectangular coordinate
system with origin at Earth’s center (Figure 10.1.3). Derive a system of differential equations for x, y , and z , assuming that the
object moves under Earth’s gravitational force (given by Newton’s law of gravitation, as in Example 10.1.3) and a resistive
force proportional to the speed of the object. Let α be the constant of proportionality.
5. Rewrite the given system as a first order system.
′′′ ′
x = f (t, x, y, y )
a.
′′ ′
y = g(t, y, y )
′ ′ ′
u = f (t, u, v, v , w )
b. v
′′
= g(t, u, v, v , w)
′
′′ ′ ′
w = h(t, u, v, v , w, w )
c. y ′′′
= f (t, y, y , y
′ ′′
)
d. y (4)
= f (t, y)
′′
x = f (t, x, y)
e.
′′
y = g(t, x, y)
6. Rewrite the system Equation 10.1.14 of differential equations derived in Example 10.1.3 as a first order system.
7. Formulate a version of Euler’s method (Section 3.1) for the numerical solution of the initial value problem
′
y = g1 (t, y1 , y2 ), y1 (t0 ) = y10 ,
1
′
y = g2 (t, y1 , y2 ), y2 (t0 ) = y20 ,
2
on an interval [t 0, b] .
8. Formulate a version of the improved Euler method (Section 3.2) for the numerical solution of the initial value problem
′
y = g1 (t, y1 , y2 ), y1 (t0 ) = y10 ,
1
′
y = g2 (t, y1 , y2 ), y2 (t0 ) = y20 ,
2

on an interval [t 0, .
b]

10.2: Linear Systems of Differential Equations
A first order system of differential equations that can be written in the form
′
y = a11 (t)y1 + a12 (t)y2 + ⋯ + a1n (t)yn + f1 (t)
1
′
y = a21 (t)y1 + a22 (t)y2 + ⋯ + a2n (t)yn + f2 (t)
2
(10.2.1)
′
yn = an1 (t)y1 + an2 (t)y2 + ⋯ + ann (t)yn + fn (t)
is called a linear system.

The linear system Equation 10.2.1 can be written in matrix form as
′
y a11 (t)12 (t) ⋯1n (t) f1 (t)
⎡ 1 ⎤ ⎡ ⎤ ⎡ y1 ⎤ ⎡ ⎤
′
⎢ y ⎥ ⎢ a21 (t)22 (t) ⋯2n (t) ⎥ ⎢ y2 ⎥ ⎢ f2 (t) ⎥
⎢ 2 ⎥ ⎢ ⎥ ⎢ ⎥+⎢ ⎥
⎢ ⎥ =⎢ ⎥ ⎢ ⎥,
⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⎥ ⎢ ⎢ ⋮ ⎥ ⎢ ⋮ ⎥
⋮ ⋮ ⋱ ⋮⎥
⎣ ′ ⎦ ⎣ ⎦⎣ ⎦ ⎣ ⎦
yn an1 (t)n2 (t) ⋯nn (t) yn fn (t)
or more briefly as
′
y = A(t)y + f (t), (10.2.2)
where
y1 a11 (t)12 (t) ⋯1n (t) f1 (t)

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢ y2 ⎥ ⎢ a21 (t)22 (t) ⋯2n (t) ⎥ ⎢ f2 (t) ⎥

⎢ ⎥ ⎢ ⎥
y = ⎢ ⎥, A(t) = ⎢ ⎥, and f (t) = ⎢ ⎥.
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⎥ ⎢ ⎢ ⋮ ⎥
⋮ ⋮ ⋱ ⋮⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
yn an1 (t)n2 (t) ⋯nn (t) fn (t)
We call A the coefficient matrix of Equation 10.2.2 and f the forcing function. We’ll say that A and f are continuous if their
entries are continuous. If f = 0 , then Equation 10.2.2 is homogeneous; otherwise, Equation 10.2.2 is nonhomogeneous.
An initial value problem for Equation 10.2.2 consists of finding a solution of Equation 10.2.2 that equals a given constant
vector
k1
⎡ ⎤
⎢ k2 ⎥
k = ⎢ ⎥.
⎢ ⎥
⎢ ⋮ ⎥
⎣ ⎦
kn
at some initial point t . We write this initial value problem as

0
′
y = A(t)y + f (t), y(t0 ) = k.
The next theorem gives sufficient conditions for the existence of solutions of initial value problems for Equation 10.2.2 . We
omit the proof.
Theorem 10.2.1 : Existence

Suppose the coefficient matrix A and the forcing function f are continuous on (a, b) , let t be in
0 (a, b) , and let k be an
arbitrary constant n -vector. Then the initial value problem
′
y = A(t)y + f (t), y(t0 ) = k

Example 10.2.1
a. Write the system
′ 4t
y = y1 + 2 y2 + 2 e
1
(10.2.3)
′ 4t
y = 2 y1 + y2 + e
2
in matrix form and conclude from Theorem 10.2.1 that every initial value problem for Equation 10.2.3 has a unique
solution on (−∞, ∞).
b. Verify that
1 8 4t
1 3t
1 −t
y = [ ]e + c1 [ ]e + c2 [ ]e (10.2.4)
5 7 1 −1
is a solution of Equation 10.2.3 for all values of the constants c and c . 1 2
c. Find the solution of the initial value problem
′
1 2 2 4t
1 3
y =[ ]y+[ ]e , y(0) = [ ]. (10.2.5)
2 1 1 5 22
Solution a
The system Equation 10.2.3 can be written in matrix form as
1 2
′ 4t
y =[ ]y+[ ]e .
2 1
An initial value problem for Equation 10.2.3 can be written as
′
1 2 2 4t
k1
y =[ ]y+[ ]e , y(t0 ) = [ ].
2 1 1 k2
Since the coefficient matrix and the forcing function are both continuous on (−∞, ∞), Theorem 10.2.1 implies that this
problem has a unique solution on (−∞, ∞).
Solution b
If y is given by Equation 10.2.4, then
1 1 2 8 1 2 1 1 2 1 2
4t 3t −t 4t
Ay + f = [ ][ ]e + c1 [ ][ ]e + c2 [ ][ ]e +[ ]e
5 2 1 7 2 1 1 2 1 −1 1
1 22 4t
3 3t
−1 −t
2 4t
= [ ]e + c1 [ ]e + c2 [ ]e +[ ]e
5 23 3 1 1
1 32 1 1
4t 3t −t
= [ ]e + 3 c1 [ ]e − c2 [ ]e
5 28 1 −1
′
=y .
Solution c
We must choose c and c in Equation 10.2.4 so that
1 2
1 8 1 1 1 3
[ ] + c1 [ ] + c2 [ ] = [ ],
5 7 1 −1 5 22
1 1 c1 −1
[ ][ ] =[ ].
1 −1 c2 3
Solving this system yields c 1 =1 ,c

2 = −2 , so

1 8 1 1
4t 3t −t
y = [ ]e +[ ]e −2 [ ]e
5 7 1 −1
is the solution of Equation 10.2.5.
Note
The theory of n × n linear systems of differential equations is analogous to the theory of the scalar n-th order equation
(n) (n−1)
P0 (t)y + P1 (t)y + ⋯ + Pn (t)y = F (t) (10.2.6)
as developed in Sections 9.1. For example by rewriting Equation 10.2.6 as an equivalent linear system it can be shown
that Theorem 10.2.1 implies Theorem 9.1.1 (Exercise 10.2.12).

10.2E: Linear Systems of Differential Equations (Exercises)
Q10.2.1
1. Rewrite the system in matrix form and verify that the given vector function satisfies the system for any choice of the
constants c and c .
1 2
′
y = 2 y1 + 4 y2 1 1
a. 1
′
y = c1 [ ]e
6t
+ c2 [ ]e
−2t
y = 4 y1 + 2 y2 ; 1 −1
2
′
y = −2 y1 − 2 y2 1 −2
b. 1
′
y = c1 [ ]e
−4t
+ c2 [ ]e
3t
y = −5 y1 + y2 ; 1 5
2
′
y = −4 y1 − 10 y2 −5 2
c. 1
′
y = c1 [ ]e
2t
+ c2 [ ]e
t
y = 3 y1 + 7 y2 ; 3 −1
2
′
y = 2 y1 + y2 1 1
d. 1
′
y = c1 [ ]e
3t
+ c2 [ ]e
t
y = y1 + 2 y2 ; 1 −1
2
constants c , c , and c .
1 2 3
′
y = −y1 + 2 y2 + 3 y3
1
a. y
2
′
= y2 + 6 y3
′
y = −2 y3 ;
3
1 1 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
t −t −2t
y = c1 ⎢ 1 ⎥ e + c2 ⎢ 0 ⎥ e + c3 ⎢ −2 ⎥ e
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
0 0 1
′
y = 2 y2 + 2 y3
1
b. y
2
′
= 2 y1 + 2 y3
′
y = 2 y1 + 2 y2 ;
3
−1 0 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
−2t −2t 4t
y = c1 ⎢ 0⎥e + c2 ⎢ −1 ⎥ e + c3 ⎢ 1 ⎥ e
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 1
′
y = −y1 + 2 y2 + 2 y3
1
c. y
2
′
= 2 y1 − y2 + 2 y3
′
y = 2 y1 + 2 y2 − y3 ;
3
−1 0 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
−3t −3t 3t
y = c1 ⎢ 0⎥e + c2 ⎢ −1 ⎥ e + c3 ⎢ 1 ⎥ e
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 1
′
y = 3 y1 − y2 − y3
1
d. y
2
′
= −2 y1 + 3 y2 + 2 y3
′
y = 4 y1 − y2 − 2 y3 ;
3
1 1 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
2t 3t −t
y = c1 ⎢ 0 ⎥ e + c2 ⎢ −1 ⎥ e + c3 ⎢ −3 ⎥ e
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 7
3. Rewrite the initial value problem in matrix form and verify that the given vector function is a solution.
′
y = y1 + y2 y1 (0) = 1 1 1
a. 1
′
y =2[ ]e
2t
−[ ]e
3t
y = −2 y1 + 4 y2 , y2 (0) = 0; 1 2
2
′
y = 5 y1 + 3 y2 y1 (0) = 12 1 3
b. 1
′
y =3[ ]e
2t
+3 [ ]e
4t
y = −y1 + y2 , y2 (0) = −6; −1 −1

2
4. Rewrite the initial value problem in matrix form and verify that the given vector function is a solution.

′
y = 6 y1 + 4 y2 + 4 y3 y1 (0) = 3
1
a. y
2
′
= −7 y1 − 2 y2 − y3 , , y2 (0) = −6
′
y = 7 y1 + 4 y2 + 3 y3 y3 (0) = 4
3
1 1 0
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
6t 2t −t
y = ⎢ −1 ⎥ e + 2 ⎢ −2 ⎥ e + ⎢ −1 ⎥ e
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 1
′
y = 8 y1 + 7 y2 + 7 y3 y1 (0) = 2
1
b. y
2
′
= −5 y1 − 6 y2 − 9 y3 , y2 (0) = −4
′
y = 5 y1 + 7 y2 + 10 y3 , y3 (0) = 3
3
1 0 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
8t 3t t
y = ⎢ −1 ⎥ e + ⎢ −1 ⎥ e + ⎢ −2 ⎥ e
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 1
constants c and c .
1 2
′
y = −3 y1 + 2 y2 + 3 − 2t
a. 1
′
y = −5 y1 + 3 y2 + 6 − 3t
2
2 cos t 2 sin t 1
y = c1 [ ] + c2 [ ]+[ ]
3 cos t − sin t 3 sin t + cos t t
′ t
y = 3 y1 + y2 − 5 e
b. 1
′ t
y = −y1 + y2 + e
2
−1 1 +t 1
2t 2t t
y = c1 [ ]e + c2 [ ]e +[ ]e
1 −t 3
′ t t
y = −y1 − 4 y2 + 4 e + 8te
c. 1
′ 3t t
y = −y1 − y2 + e + (4t + 2)e
2
3t
2 −3t
−2 t
e
y = c1 [ ]e + c2 [ ]e +[ ]
t
1 1 2te
′ 2t t
y = −6 y1 − 3 y2 + 14 e + 12 e
d. 1
′ 2t t
y = y1 − 2 y2 + 7 e − 12 e
2
2t t
−3 −1 e + 3e
−5t −3t
y = c1 [ ]e + c2 [ ]e +[ ]
2t t
1 1 2e − 3e
6. Convert the linear scalar equation

(n) (n−1)
P0 (t)y + P1 (t)y + ⋯ + Pn (t)y(t) = F (t) (A)
into an equivalent n × n system

′
y = A(t)y + f (t),
and show that A and f are continuous on an interval (a, b) if and only if (A) is normal on (a, b).
7. A matrix function
q11 (t) q12 (t) ⋯ q1s (t)
⎡ ⎤
⎢ q21 (t) q22 (t) ⋯ q2s (t) ⎥

⎢ ⎥
Q(t) = ⎢ ⎥
⎢ ⎥
⎢ ⋮ ⋮ ⋱ ⋮ ⎥
⎣ ⎦
qr1 (t) qr2 (t) ⋯ qrs (t)
is said to be differentiable if its entries {q ij } are differentiable. Then the derivative Q is defined by ′

′ ′ ′
q (t) q (t) ⋯ q (t)
⎡ 11 12 1s ⎤
′ ′ ′
⎢q (t) q (t) ⋯ q (t) ⎥
⎢ 21 22 2s
⎥
Q(t) = ⎢ ⎥
⎢ ⎥
⎢ ⋮ ⋮ ⋱ ⋮ ⎥
⎣ ′ ′ ′ ⎦
q (t) q (t) ⋯ qrs (t)
r1 r2
a. Prove: If P and Q are differentiable matrices such that P +Q is defined and if c and c are constants, then
1 2
′ ′ ′
(c1 P + c2 Q ) = c1 P + c2 Q .
b. Prove: If P and Q are differentiable matrices such that P Q is defined, then

′ ′ ′
(P Q ) = P Q + P Q .
8. Verify that Y ′
= AY .
6t −2t
e e 2 4
a. Y =[
6t −2t
], A =[ ]
e −e 4 2
−4t 3t
e −2e −2 −2
b. Y =[
−4t 3t
], A =[ ]
e 5e −5 1
2t t
−5e 2e −4 −10
c. Y =[
2t t
], A =[ ]
3e −e 3 7
3t t
e e 2 1
d. Y =[
3t t
], A =[ ]
e −e 1 2
t −t −2t
e e e −1 2 3
⎡ ⎤ ⎡ ⎤
e. Y =⎢e
t
0 −2e
−2t
⎥, A =⎢ 0 1 6 ⎥
⎣ −2t ⎦ ⎣ ⎦
0 0 e 0 0 −2
−2t −2t 4t
−e −e e 0 2 2
⎡ ⎤ ⎡ ⎤
f. Y =⎢ 0 e
−2t
e
4t
⎥, A =⎢2 0 2⎥
⎣ −2t 4t ⎦ ⎣ ⎦
e 0 e 2 2 0
3t −3t
e e 0 −9 6 6
⎡ ⎤ ⎡ ⎤
g. Y =⎢e
3t
0 −e
−3t
⎥, A = ⎢ −6 3 6⎥
⎣ 3t −3t −3t ⎦ ⎣ ⎦
e e e −6 6 3
2t 3t −t
e e e 3 −1 −1
⎡ ⎤ ⎡ ⎤
h. Y =⎢ 0 −e
3t
−3e
−t
⎥, A = ⎢ −2 3 2 ⎥
⎣ 2t 3t −t ⎦ ⎣ ⎦
e e 7e 4 −1 −2
9. Suppose
y11 y12
y1 = [ ] and y2 = [ ]
y21 y22
are solutions of the homogeneous system

′
y = A(t)y, (A)
and define
y11 y12
Y =[ ].
y21 y22
a. Show that Y = AY . ′
b. Show that if c is a constant vector then y = Y c is a solution of (A).

c. State generalizations of (a) and (b) for n × n systems.
10. Suppose Y is a differentiable square matrix.
a. Find a formula for the derivative of Y . 2
b. Find a formula for the derivative of Y , where n is any positive integer. n

c. State how the results obtained in (a) and (b) are analogous to results from calculus concerning scalar functions.
11. It can be shown that if Y is a differentiable and invertible square matrix function, then Y −1
is differentiable.
a. Show that (Y ) = −Y Y Y . (Hint: Differentiate the identity Y Y = I .)
−1 ′ −1 ′ −1 −1
n
b. Find the derivative of Y −n
= (Y
−1
) , where n is a positive integer.
c. State how the results obtained in (a) and (b) are analogous to results from calculus concerning scalar functions.
12. Show that Theorem 10.2.1 implies Theorem 9.1.1. HINT: Write the scalar function
(n) (n−1)
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = F (x)
as an n × n system of linear equations.

13. Suppose y is a solution of the n × n system y = A(t)y on (a, b), and that the n × n matrix P is invertible and
′
differentiable on (a, b). Find a matrix B such that the function x = P y is a solution of x = Bx on (a, b).
′

10.3: Basic Theory of Homogeneous Linear Systems
In this section we consider homogeneous linear systems y = A(t)y , where A = A(t) is a continuous n × n matrix function
′
on an interval (a, b). The theory of linear homogeneous systems has much in common with the theory of linear homogeneous
scalar equations, which we considered in Sections 2.1, 5.1, and 9.1.
Whenever we refer to solutions of y = A(t)y we’ll mean solutions on
′
(a, b) . Since y ≡0 is obviously a solution of
y = A(t)y , we call it the trivial solution. Any other solution is nontrivial.
′
If y , y , …, y are vector functions defined on an interval (a, b) and c , c , …, c are constants, then
1 2 n 1 2 n
y = c1 y1 + c2 y2 + ⋯ + cn yn (10.3.1)
is a linear combination of y , y , …,y . It’s easy show that if y , y , …,y are solutions of y = A(t)y on (a, b), then so is
1 2 n 1 2 n
′
any linear combination of y , y , …, y (Exercise 10.3.1). We say that {y , y , … , y } is a fundamental set of solutions of
1 2 n 1 2 n
y = A(t)y on (a, b) on if every solution of y = A(t)y on (a, b) can be written as a linear combination of y , y , …, y , as
′ ′
1 2 n
in Equation 10.3.1. In this case we say that Equation 10.3.1 is the general solution of y = A(t)y on (a, b). ′
It can be shown that if A is continuous on (a, b) then y = A(t)y has infinitely many fundamental sets of solutions on (a, b)
′
(Exercises 10.3.15 and 10.3.16). The next definition will help to characterize fundamental sets of solutions of y = A(t)y . ′
We say that a set {y 1

,y ,…,y }
2 n
of n -vector functions is linearly independent on (a, b) if the only constants c , 1 c2 , …, cn
such that
c1 y1 (t) + c2 y2 (t) + ⋯ + cn yn (t) = 0, a < t < b, (10.3.2)
are c1 = c2 = ⋯ = cn = 0 . If Equation 10.3.2 holds for some set of constants c1 , c2 , …, cn that are not all zero, then
{ y1 , y2 , … , yn } is linearly dependent on (a, b)
The next theorem is analogous to Theorems 5.1.3 and 9.1.2.
Theorem 10.3.1
Suppose the n × n matrix A = A(t) is continuous on (a, b). Then a set {y , y 1 2
, … , yn } of n solutions of ′
y = A(t)y
on (a, b) is a fundamental set if and only if it is linearly independent on (a, b).
Example 10.3.1
Show that the vector functions
t 2t
e 0 e
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
3t 3t
y1 = ⎢ 0 ⎥, y2 = ⎢ e ⎥, and y3 = ⎢ e ⎥
⎣ −t ⎦ ⎣ ⎦ ⎣ ⎦
e 1 0
are linearly independent on every interval (a, b).

Solution
Suppose
t 2t
e 0 e 0
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
3t 3t
c1 ⎢ 0 ⎥ + c2 ⎢ e ⎥ + c3 ⎢ e ⎥ = ⎢0⎥, a < t < b.
⎣ −t ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦
e 1 0 0
We must show that c 1 = c2 = c3 = 0 . Rewriting this equation in matrix form yields

t 2t
e 0 e c1 0
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
3t 3t
⎢ 0 e e ⎥ ⎢ c2 ⎥ = ⎢ 0 ⎥ , a < t < b.
⎣ −t ⎦ ⎣ ⎦ ⎣ ⎦
e 1 0 c3 0

Expanding the determinant of this system in cofactors of the entries of the first row yields
t 2t
∣ e 0 e ∣
3t 3t 3t 3t
∣ ∣e
∣ e ∣ 0∣ e ∣ ∣ 0 e ∣
3t 3t t 2t
0 e e =e ∣ ∣−0 ∣ ∣+e ∣ ∣
∣ ∣ ∣ 1
−t −t
0 ∣ ∣e 0 ∣ ∣e 1 ∣
−t
∣e 1 0 ∣
t 3t 2t 2t 4t
= e (−e )+e (−e ) = −2 e .
Since this determinant is never zero, c 1 = c2 = c3 = 0 .
We can use the method in Example 10.3.1 to test n solutions {y , y , … , y } of any n × n system y = A(t)y for linear
1 2 n
′
independence on an interval (a, b) on which A is continuous. To explain this (and for other purposes later), it is useful to write
a linear combination of y , y , …, y in a different way. We first write the vector functions in terms of their components as
1 2 n
y11 y12 y1n

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢ y21 ⎥ ⎢ y22 ⎥ ⎢ y2n ⎥

y1 =⎢ ⎥, y2 =⎢ ⎥,…, yn =⎢ ⎥.
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⎥ ⎢ ⋮ ⎥ ⎢ ⋮ ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
yn1 yn2 ynn
If
y = c1 y1 + c2 y2 + ⋯ + cn yn
then
y11 y12 y1n

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢ y21 ⎥ ⎢ y22 ⎥ ⎢ y2n ⎥

y = c1 ⎢ ⎥+c ⎢ ⎥+⋯ +c ⎢ ⎥
2 n
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⎥ ⎢ ⋮ ⎥ ⎢ ⋮ ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
yn1 yn2 ynn
y11 y12 ⋯ y1n c1

⎡ ⎤⎡ ⎤
⎢ y21 y22 ⋯ y2n ⎥ ⎢ c2 ⎥

⎢ ⎥ ⎢ ⎥.
=⎢ ⎥
⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⋮ ⋱ ⋮ ⎥ ⎢ ⋮ ⎥
⎣ ⎦⎣ ⎦
yn1 yn2 ⋯ ynn cn
This shows that

c1 y1 + c2 y2 + ⋯ + cn yn = Y c, (10.3.3)
where
c1
⎡ ⎤
⎢ c2 ⎥
c =⎢ ⎥
⎢ ⎥
⎢ ⋮ ⎥
⎣ ⎦
cn
and
y11 y12 ⋯ y1n
⎡ ⎤
⎢ y21 y22 ⋯ y2n ⎥

⎢ ⎥
Y = [ y1 y2 ⋯ yn ] = ⎢ ⎥; (10.3.4)
⎢ ⎥
⎢ ⋮ ⋮ ⋱ ⋮ ⎥
⎣ ⎦
yn1 yn2 ⋯ ynn
that is, the columns of Y are the vector functions y 1

, y2 , … , yn .
For reference below, note that

′ ′ ′ ′
Y = [y y ⋯ yn ]
1 2
= [Ay1 Ay2 ⋯ Ayn ]
= A[ y1 y2 ⋯ yn ] = AY ;
that is, Y satisfies the matrix differential equation

′
Y = AY .
The determinant of Y ,
∣ y11 y12 ⋯ y1n ∣
∣ ∣
∣ y21 y22 ⋯ y2n ∣
W =∣ ∣ (10.3.5)
∣ ⋮ ⋮ ⋱ ⋮ ∣
∣ ∣
∣ yn1 yn2 ⋯ ynn ∣
is called the Wronskian of {y , y , … , y }. It can be shown (Exercises 10.3.2 and 10.3.3) that this definition is analogous to
1 2 n
definitions of the Wronskian of scalar functions given in Sections 5.1 and 9.1. The next theorem is analogous to Theorems
5.1.4 and 9.1.3. The proof is sketched in Exercise 10.3.4 for n = 2 and in Exercise 10.3.5 for general n .
Theorem 10.3.2 : Abel’s Formula

Suppose the n × n matrix A = A(t) is continuous on (a, b), let y , y , …, y be solutions of y 1 2 n
′
= A(t)y on (a, b), and
let t be in (a, b). Then the Wronskian of {y , y , … , y } is given by
0 1 2 n
W (t) = W (t0 ) exp( ∫ [ a11 (s) + a22 (s) + ⋯ + ann (s)] ds), a < t < b. (10.3.6)
t0
Therefore, either W has no zeros in (a, b) or W ≡0 on (a, b).
Note
The sum of the diagonal entries of a square matrix A is called the trace of A , denoted by tr(A) . Thus, for an n×n
matrix A ,
tr(A) = a11 + a22 + ⋯ ann ,
and Equation 10.3.6 can be written as

t
W (t) = W (t0 )exp ( ∫ tr(A(s))ds) , a < t < b.

t0
The next theorem is analogous to Theorems 5.1.6 and 9.1.4.
Theorem 10.3.3
Suppose the n × n matrix A = A(t) is continuous on (a, b) and let y , y , …,y be solutions of 1 2 n
′
y = A(t)y on (a, b) .
Then the following statements are equivalent; that is, they are either all true or all false:
a. The general solution of y = A(t)y on (a, b) is y = c y + c y + ⋯ + c
′
1 1 2 2 n yn , where c1 , c2 , …, cn are arbitrary
constants.
b. {y , y , … , y } is a fundamental set of solutions of y = A(t)y on (a, b).
1 2 n
′
c. {y , y , … , y } is linearly independent on (a, b).

1 2 n
d. The Wronskian of {y , y , … , y } is nonzero at some point in (a, b).

1 2 n
e. The Wronskian of {y , y , … , y } is nonzero at all points in (a, b).

1 2 n

We say that Y in Equation 10.3.4 is a fundamental matrix for y = A(t)y if any (and therefore all) of the statements (a)-(e) of
′
Theorem 10.3.2 are true for the columns of Y . In this case, Equation 10.3.3 implies that the general solution of y = A(t)y ′
can be written as y = Y c , where c is an arbitrary constant n -vector.
Example 10.3.2
The vector functions
2t −t
−e −e
y1 = [ ] and y2 = [ ]
2t −t
2e e
are solutions of the constant coefficient system
′
−4 −3
y =[ ]y (10.3.7)
6 5
on (−∞, ∞). (Verify.)

a. Compute the Wronskian of {y , y } directly from the definition Equation 10.3.5
1 2
b. Verify Abel’s formula Equation 10.3.6 for the Wronskian of {y , y }. 1 2
c. Find the general solution of Equation 10.3.7.

′
−4 −3 4
y =[ ] y, y(0) = [ ]. (10.3.8)
6 5 −5
Solution a
From Equation 10.3.5
2t −t
∣ −e −e ∣ −1 −1
2t −t t
W (t) = ∣ ∣ =e e [ ] =e . (10.3.9)
2t −t
∣ 2e e ∣ 2 1
Solution b
Here
−4 −3
A =[ ],
6 5
so tr(A) = −4 + 5 = 1 . If t is an arbitrary real number then Equation 10.3.6 implies that

0
t 2t0 −t0
∣ −e −e ∣
(t−t0 ) t0 t−t0 t
W (t) = W (t0 ) exp ( ∫ 1 ds) = ∣ ∣e =e e =e ,
2t0 −t0
t0 ∣ 2e e ∣
which is consistent with Equation 10.3.9.

Solution c
Since W (t) ≠ 0 , Theorem 10.3.3 implies that {y 1
, y2 } is a fundamental set of solutions of Equation 10.3.7 and
2t −t
−e −e
Y =[ ]
2t −t
2e e
is a fundamental matrix for Equation 10.3.7. Therefore the general solution of Equation 10.3.7 is
2t −t 2t −t
−e −e −e −e c1
y = c1 y + c2 y = c1 [ ] + c2 [ ] =[ ][ ]. (10.3.10)
1 2 2t −t 2t −t
2e e 2e e c2
Solution d
Setting t = 0 in Equation 10.3.10 and imposing the initial condition in Equation 10.3.8 yields

−1 −1 4
c1 [ ] + c2 [ ] =[ ].
2 1 −5
Thus,
−c1 − c2 = 4
2 c1 + c2 = −5.
The solution of this system is c

1 = −1 ,c
2 = −3 . Substituting these values into Equation 10.3.10 yields
2t −t 2t −t
−e −e e + 3e
y = −[ ]−3 [ ] =[ ]
2t −t 2t −t
2e e −2 e − 3e
as the solution of Equation 10.3.8.

10.3E: Basic Theory of Homogeneous Linear Systems (Exercises)
Q10.3.1
1. Prove: If y , y , …, y are solutions of
1 2 n
′
y = A(t)y on (a, b) , then any linear combination of y1 , y2 , …, yn is also a
solution of y = A(t)y on (a, b).
′
2. In Section 5.1 the Wronskian of two solutions y and y of the scalar second order equation
1 2
′′ ′
P0 (x)y + P1 (x)y + P2 (x)y = 0 (A)
was defined to be
∣ y1 y2 ∣
W =∣ ∣.
′ ′
∣ y1 y ∣
2
a. Rewrite (A) as a system of first order equations and show that W is the Wronskian (as defined in this section) of two
solutions of this system.
b. Apply Equation 10.3.6 to the system derived in (a), and show that
x
P1 (s)
W (x) = W (x0 ) exp{− ∫ ds},
x0 P0 (s)
which is the form of Abel’s formula given in Theorem 9.1.3.

3. In Section 9.1 the Wronskian of n solutions y , y , …, y of the n−th order equation
1 2 n
(n) (n−1)
P0 (x)y + P1 (x)y + ⋯ + Pn (x)y = 0 (A)
was defined to be
∣ y1 y2 ⋯ yn ∣
∣ ∣
′ ′ ′
y y ⋯ yn
∣ 1 2 ∣
W =∣ ∣.
∣ ⋮ ⋮ ⋱ ⋮ ∣
∣ ∣
(n−1) (n−1) (n−1)
∣y y ⋯ yn ∣
1 2
a. Rewrite (A) as a system of first order equations and show that W is the Wronskian (as defined in this section) of n
solutions of this system.

b. Apply Equation 10.3.6 to the system derived in (a), and show that
x
P1 (s)
W (x) = W (x0 ) exp{− ∫ ds},
x0
P0 (s)
which is the form of Abel’s formula given in Theorem 9.1.3.

4. Suppose
y11 y12
y1 = [ ] and y2 = [ ]
y21 y22
are solutions of the 2 × 2 system y ′

= Ay on (a, b), and let
y11 y12 ∣ y11 y12 ∣
Y =[ ] and W =∣ ∣
y21 y22 ∣ y21 y22 ∣
thus, W is the Wronskian of {y 1

, y2 } .
a. Deduce from the definition of determinant that

′ ′
∣y y ∣ ∣ y11 y12 ∣
′ 11 12
W =∣ ∣+∣ ∣.
′ ′
∣ y21 y22 ∣ ∣ y21 y
22
∣
b. Use the equation Y ′

= A(t)Y and the definition of matrix multiplication to show that
′ ′
[y y ] = a11 [ y11 y12 ] + a12 [ y21 y22 ]
11 12
and
′ ′
[y y ] = a21 [ y11 y12 ] + a22 [ y21 y22 ].
21 22
c. Use properties of determinants to deduce from (a) and (a) that

∣ y′ y
′
∣ ∣ y11 y12 ∣
11 12
∣ ∣ = a11 W and ∣ ∣ = a22 W .
′ ′
∣ y21 y22 ∣ ∣y y ∣
21 22
d. Conclude from (c) that

′
W = (a11 + a22 )W ,
and use this to show that if a < t 0 <b then

t
W (t) = W (t0 ) exp( ∫ [ a11 (s) + a22 (s)] ds) a < t < b.
t0
5. Suppose the n × n matrix A = A(t) is continuous on (a, b). Let

y11 y12 ⋯ y1n
⎡ ⎤
⎢ y21 y22 ⋯ y2n ⎥

⎢ ⎥
Y =⎢ ⎥,
⎢ ⎥
⎢ ⋮ ⋮ ⋱ ⋮ ⎥
⎣ ⎦
yn1 yn2 ⋯ ynn
where the columns of Y are solutions of y ′

= A(t)y . Let
ri = [ yi1 yi2 … yin ]
be the ith row of Y , and let W be the determinant of Y .

a. Deduce from the definition of determinant that
′
W = W1 + W2 + ⋯ + Wn ,
where, for 1 ≤ m ≤ n , the ith row of W is r if i ≠ m , and r if i = m .

m i
′
m
b. Use the equation Y = AY and the definition of matrix multiplication to show that
′
′
rm = am1 r1 + am2 r2 + ⋯ + amn rn .
c. Use properties of determinants to deduce from (b) that
det(Wm ) = amm W .
d. Conclude from (a) and (c) that

′
W = (a11 + a22 + ⋯ + ann )W ,
and use this to show that if a < t 0 <b then

t
W (t) = W (t0 ) exp( ∫ [ a11 (s) + a22 (s) + ⋯ + ann (s)] ds), a < t < b.
t0
6. Suppose the n × n matrix A is continuous on (a, b) and t0 is a point in (a, b) . Let Y be a fundamental matrix for
y = A(t)y on (a, b) .
′
a. Show that Y (t 0) is invertible.

b. Show that if k is an arbitrary n -vector then the solution of the initial value problem
′
y = A(t)y, y(t0 ) = k
is
−1
y = Y (t)Y (t0 )k.
7. Let
6t −2t
2 4 e e −3
A =[ ], y1 = [ ], y2 = [ ], k =[ ].
6t −2t
4 2 e −e 9
a. Verify that {y , y } is a fundamental set of solutions for y

1 2
′
= Ay .
′
y = Ay, y(0) = k. (A)
c. Use the result of Exercise 10.3.6 (b) to find a formula for the solution of (A) for an arbitrary initial vector k.
8. Repeat Exercise 10.3.7 with
−4t 3t
−2 −2 e −2e 10
A =[ ], y1 = [ ], y2 = [ ], k =[ ].
−4t 3t
−5 1 e 5e −4

2t t
−4 −10 −5e 2e −19
A =[ ], y1 = [ ], y2 = [ ], k =[ ].
2t t
3 7 3e −e 11

3t t
2 1 e e 2
A =[ ], y1 = [ ], y2 = [ ], k =[ ].
3t t
1 2 e −e 8
11. Let
3 −1 −1
⎡ ⎤
A = ⎢ −2 3 2 ⎥,
⎣ ⎦
4 −1 −2
2t 3t −t
e e e 2
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
3t −t
y1 = ⎢ 0 ⎥, y2 = ⎢ −e ⎥, y3 = ⎢ −3e ⎥, k = ⎢ −7 ⎥ .
⎣ 2t ⎦ ⎣ 3t ⎦ ⎣ −t ⎦ ⎣ ⎦
e e 7e 20
a. Verify that {y , y , y } is a fundamental set of solutions for y

1 2 3
′
= Ay .
′
y = Ay, y(0) = k. (A)
c. Use the result of Exercise 10.3.6 (b) to find a formula for the solution of (A) for an arbitrary initial vector k.
0 2 2
⎡ ⎤
A =⎢2 0 2⎥,
⎣ ⎦
2 2 0
−2t −2t 4t
−e −e e 0
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
−2t 4t
y =⎢ 0 ⎥, y =⎢ e ⎥, y =⎢e ⎥, k = ⎢ −9 ⎥ .
1 2 3
⎣ −2t ⎦ ⎣ ⎦ ⎣ 4t ⎦ ⎣ ⎦
e 0 e 12

−1 2 3
⎡ ⎤
A =⎢ 0 1 6 ⎥,
⎣ ⎦
0 0 −2
t −t −2t
e e e 5
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
t −2t
y1 = ⎢e ⎥, y2 = ⎢ 0 ⎥, y3 = ⎢ −2e ⎥, k =⎢ 5⎥.
⎣ ⎦ ⎣ ⎦ ⎣ −2t ⎦ ⎣ ⎦
0 0 e −1
14. Suppose Y and Z are fundamental matrices for the n × n system y = A(t)y . Then some of the four matrices ′
Y Z
−1
,
Z, Z Y , ZY are necessarily constant. Identify them and prove that they are constant.
−1 −1 −1
Y
15. Suppose the columns of an n×n matrix Y are solutions of the n×n system ′
y = Ay and C is an n×n constant
matrix.
a. Show that the matrix Z = Y C satisfies the differential equation Z = AZ . ′
b. Show that Z is a fundamental matrix for y = A(t)y if and only if C is invertible and
′
Y is a fundamental matrix for
y = A(t)y .
′
16. Suppose the n × n matrix A = A(t) is continuous on (a, b) and t is in (a, b). For i = 1 , 2, …, n , let y be the solution 0 i
of the initial value problem y = A(t)y , y (t ) = e , where

′
i i i 0 i
1 0 0
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢0⎥ ⎢1⎥ ⎢0⎥

e1 =⎢ ⎥, e2 =⎢ ⎥, ⋯ en =⎢ ⎥;
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⎥ ⎢ ⋮ ⎥ ⎢ ⋮ ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
0 0 1
that is, the j th component of e is 1 if j = i , or 0 if j ≠ i .

i
a. Show that{y , y , … , y } is a fundamental set of solutions of y = A(t)y on (a, b).

1 2 n
′
b. Conclude from (a) and Exercise 10.3.15 that y = A(t)y has infinitely many fundamental sets of solutions on (a, b).
′
17. Show that Y is a fundamental matrix for the system y = A(t)y if and only if ′
Y
−1
is a fundamental matrix for
(t)y , where A denotes the transpose of A . HINT: See Exercise 10.3.11.
′ T T
y = −A
18. Let Z be the fundamental matrix for the constant coefficient system y ′
= Ay such that Z(0) = I .
a. Show that Z(t)Z(s) = Z(t + s) for all s and t . HINT: For fixed s let Γ (t) = Z(t)Z(s) and Γ (t) = Z(t + s) . Show 1 2
that Γ and Γ are both solutions of the matrix initial value problem Γ = AΓ, Γ(0) = Z(s) . Then conclude from
1 2
′
Theorem 10.2.1 that Γ = Γ . 1 2
b. Show that (Z(t)) = Z(−t) . −1
c. The matrix Z defined above is sometimes denoted by e . Discuss the motivation for this notation. tA

10.4: Constant Coefficient Homogeneous Systems I
We’ll now begin our study of the homogeneous system
′
y = Ay, (10.4.1)
where A is an n × n constant matrix. Since A is continuous on (−∞, ∞), Theorem 10.2.1 implies that all solutions of
Equation 10.4.1 are defined on (−∞, ∞). Therefore, when we speak of solutions of y = Ay , we’ll mean solutions on ′
(−∞, ∞) .
In this section we assume that all the eigenvalues of A are real and that A has a set of n linearly independent eigenvectors. In
the next two sections we consider the cases where some of the eigenvalues of A are complex, or where A does not have n
linearly independent eigenvectors.
In Example 10.3.2 we showed that the vector functions
2t −t
−e −e
y1 = [ ] and y2 = [ ]
2t −t
2e e
form a fundamental set of solutions of the system
′
−4 −3
y =[ ] y, (10.4.2)
6 5
but we did not show how we obtained y1 and y2 in the first place. To see how these solutions can be obtained we write
Equation 10.4.2 as
′
y = −4 y1 − 3 y2
1
(10.4.3)
′
y = 6 y1 + 5 y2
2
and look for solutions of the form

λt λt
y1 = x1 e and y2 = x2 e , (10.4.4)
where x , x , and λ are constants to be determined. Differentiating Equation 10.4.4 yields

1 2
′ λt ′ λt
y = λ x1 e and y = λ x2 e .
1 2
Substituting this and Equation 10.4.4 into Equation 10.4.3 and canceling the common factor e λt
yields
−4 x1 − 3 x2 = λx1
6 x1 + 5 x2 = λ x2 .
For a given λ , this is a homogeneous algebraic system, since it can be rewritten as

(−4 − λ)x1 − 3 x2 = 0
(10.4.5)
6 x1 + (5 − λ)x2 = 0.
The trivial solution x = x = 0 of this system isn’t useful, since it corresponds to the trivial solution y ≡ y ≡ 0 of
1 2 1 2
Equation 10.4.3, which can’t be part of a fundamental set of solutions of Equation 10.4.2. Therefore we consider only those
values of λ for which Equation 10.4.5 has nontrivial solutions. These are the values of λ for which the determinant of
Equation 10.4.5 is zero; that is,
∣ −4 − λ −3 ∣
∣ ∣ = (−4 − λ)(5 − λ) + 18
∣ 6 5 −λ ∣
2
=λ −λ −2
= (λ − 2)(λ + 1) = 0,
which has the solutions λ 1 =2 and λ

2 = −1 .
Taking λ = 2 in Equation 10.4.5 yields

−6 x1 − 3 x2 = 0
6 x1 + 3 x2 = 0,
which implies that x = −x /2 , where x can be chosen arbitrarily. Choosing

1 2 2 x2 = 2 yields the solution y1 = −e
2t
,
y = 2e
2 of Equation 10.4.3. We can write this solution in vector form as
2t
−1
2t
y1 = [ ]e . (10.4.6)
2
Taking λ = −1 in Equation 10.4.5 yields the system

−3 x1 − 3 x2 = 0
6 x1 + 6 x2 = 0,
so x = −x . Taking x
1 2 2 =1 here yields the solution y 1 = −e
−t
,y 2 =e
−t
of Equation 10.4.3. We can write this solution in
vector form as
−1 −t
y2 = [ ]e . (10.4.7)
1
In Equation 10.4.6 and Equation 10.4.7 the constant coefficients in the arguments of the exponential functions are the
eigenvalues of the coefficient matrix in Equation 10.4.2, and the vector coefficients of the exponential functions are associated
eigenvectors. This illustrates the next theorem.
Theorem 10.4.1
Suppose the n × n constant matrix A has n real eigenvalues λ , λ , … , λ (which need not be distinct) with associated
1 2 n
linearly independent eigenvectors x , x , … , x . Then the functions

1 2 n
λ1 t λ2 t λn t
y1 = x1 e , y2 = x2 e , … , yn = xn e
form a fundamental set of solutions of y ′

= Ay; that is, the general solution of this system is
λ1 t λ2 t λn t
y = c1 x1 e + c2 x2 e + ⋯ + cn xn e .
Proof
Differentiating y
i
= xi e
λi t
and recalling that Ax i = λi xi yields
′ λi t λi t
y = λi xi e = Axi e = Ayi .
i
This shows that y is a solution of y

i
′
= Ay .
The Wronskian of {y 1
, y2 , … , yn } is
∣ x11 eλ1 t x12 e
λ2 t
⋯ x1n e
λn t
∣ ∣ x11 x12 ⋯ x1n ∣
∣ ∣ ∣ ∣
λ1 t λ2 t λn t
∣ x21 e x22 e ⋯ x2n e ∣ ∣ x21 x22 ⋯ x2n ∣
λ1 t λ2 t λn t
∣ ∣ =e e ⋯e ∣ ∣.
∣ ∣ ∣ ∣
⋮ ⋮ ⋱ ⋮ ⋮ ⋮ ⋱ ⋮
∣ ∣ ∣ ∣
λ1 t λ2 t λxn t
∣ xn1 e xn2 e ⋯ xnn e ∣ ∣ xn1 xn2 ⋯ xnn ∣
Since the columns of the determinant on the right are x , x , …, x , which are assumed to be linearly independent,
1 2 n
the determinant is nonzero. Therefore Theorem 10.3.3 implies that {y , y , … , y } is a fundamental set of solutions 1 2 n
of y = Ay .
′
Example 10.4.1
′
2 4
y =[ ] y. (10.4.8)
4 2

′
2 4 5
y =[ ] y, y(0) = [ ]. (10.4.9)
4 2 −1
Solution a
The characteristic polynomial of the coefficient matrix A in Equation 10.4.8 is
∣ 2 −λ 4 ∣
2
∣ ∣ = (λ − 2 ) − 16
∣ 4 2 −λ ∣
= (λ − 2 − 4)(λ − 2 + 4)
= (λ − 6)(λ + 2).
Hence, λ 1 =6 and λ
2 = −2 are eigenvalues of A . To obtain the eigenvectors, we must solve the system
2 −λ 4 x1 0
[ ][ ] =[ ] (10.4.10)
4 2 −λ x2 0
with λ = 6 and λ = −2 . Setting λ = 6 in Equation 10.4.10 yields
−4 4 x1 0
[ ][ ] =[ ],
4 −4 x2 0
which implies that x 1 = x2 . Taking x2 =1 yields the eigenvector

1
x1 = [ ],
1
so
1
6t
y1 = [ ]e
1
is a solution of Equation 10.4.8. Setting λ = −2 in Equation 10.4.10 yields
4 4 x1 0
[ ][ ] =[ ],
4 4 x2 0
which implies that x 1 = −x2 . Taking x 2 =1 yields the eigenvector

−1
x2 = [ ],
1
so
−1
−2t
y2 = [ ]e
1
is a solution of Equation 10.4.8. From Theorem 10.4.1, the general solution of Equation 10.4.8 is
1 6t
−1 −2t
y = c1 y1 + c2 y2 = c1 [ ]e + c2 [ ]e . (10.4.11)
1 1
Solution b
To satisfy the initial condition in Equation 10.4.9, we must choose c and c in Equation 10.4.11 so that 1 2
1 −1 5
c1 [ ] + c2 [ ] =[ ].
1 1 −1
This is equivalent to the system

c1 − c2 = 5
c1 + c2 = −1,
so c1 = 2, c2 = −3 . Therefore the solution of Equation 10.4.9 is

1 −1
6t −2t
y =2[ ]e −3 [ ]e ,
1 1
or, in terms of components,

6t −2t 6t −2t
y1 = 2 e + 3e , y2 = 2 e − 3e .
Example 10.4.2
3 −1 −1
⎡ ⎤
′
y = ⎢ −2 3 2 ⎥ y. (10.4.12)
⎣ ⎦
4 −1 −2

3 −1 −1 2
⎡ ⎤ ⎡ ⎤
′
y = ⎢ −2 3 2 ⎥ y, y(0) = ⎢ −1 ⎥ . (10.4.13)
⎣ ⎦ ⎣ ⎦
4 −1 −2 8
Solution a
∣ 3 −λ −1 −1 ∣
∣ ∣
−2 3 −λ 2 = −(λ − 2)(λ − 3)(λ + 1).
∣ ∣
∣ 4 −1 −2 − λ ∣
Hence, the eigenvalues of A are λ 1 =2 ,λ 2 =3 , and λ 3 = −1 . To find the eigenvectors, we must solve the system
3 −λ −1 −1 x1 0
⎡ ⎤⎡ ⎤ ⎡ ⎤
⎢ −2 3 −λ 2 ⎥ ⎢ x2 ⎥ = ⎢ 0 ⎥ (10.4.14)
⎣ ⎦⎣ ⎦ ⎣ ⎦
4 −1 −2 − λ x3 0
with λ = 2 , 3, −1. With λ = 2 , the augmented matrix of Equation 10.4.14 is
⎡ 1 −1 −1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ −2 1 2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
4 −1 −4 ⋮ 0
which is row equivalent to
⎡1 0 −1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 1 0 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Hence, x 1 = x3 and x
2 =0 . Taking x 3 =1 yields
1
⎡ ⎤
2t
y1 = ⎢ 0 ⎥ e
⎣ ⎦
1

as a solution of Equation 10.4.12. With λ = 3 , the augmented matrix of Equation 10.4.14 is
⎡ 0 −1 −1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ −2 0 2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
4 −1 −5 ⋮ 0
⎡1 0 −1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 1 1 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Hence, x 1 = x3 and x2 = −x3 . Taking x 3 =1 yields

1
⎡ ⎤
3t
y2 = ⎢ −1 ⎥ e
⎣ ⎦
1
as a solution of Equation 10.4.12. With λ = −1 , the augmented matrix of Equation 10.4.14 is
⎡ 4 −1 −1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ −2 4 2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
4 −1 −1 ⋮ 0
1
⎡1 0 − ⋮ 0⎤
7
⎢ ⎥
⎢ 3 ⎥.
⎢0 1 ⋮ 0⎥
⎢ 7 ⎥
⎣ ⎦
0 0 0 ⋮ 0
Hence, x 1 = x3 /7 and x 2 = −3 x3 /7 . Taking x 3 =7 yields

1
⎡ ⎤
−t
y3 = ⎢ −3 ⎥ e
⎣ ⎦
7
as a solution of Equation 10.4.12. By Theorem 10.4.1, the general solution of Equation 10.4.12 is
1 1 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
2t 3t −t
y = c1 ⎢ 0 ⎥ e + c2 ⎢ −1 ⎥ e + c3 ⎢ −3 ⎥ e ,
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 7
which can also be written as

2t 3t −t
e e e c1
⎡ ⎤⎡ ⎤
3t −t
y =⎢ 0 −e −3e ⎥ ⎢ c2 ⎥ . (10.4.15)
⎣ 2t 3t −t ⎦⎣ ⎦
e e 7e c3
Solution b
To satisfy the initial condition in Equation 10.4.13 we must choose c , c , c in Equation 10.4.15 so that 1 2 3
1 1 1 c1 2
⎡ ⎤⎡ ⎤ ⎡ ⎤
⎢0 −1 −3 ⎥ ⎢ c2 ⎥ = ⎢ −1 ⎥ .
⎣ ⎦⎣ ⎦ ⎣ ⎦
1 1 7 c3 8

Solving this system yields c 1 =3 ,c 2 = −2 ,c 3 =1 . Hence, the solution of Equation 10.4.13 is
2t 3t −t
e e e 3
⎡ ⎤⎡ ⎤
3t −t
y =⎢ 0 −e −3e ⎥ ⎢ −2 ⎥
⎣ 2t 3t −t ⎦⎣ ⎦
e e 7e 1
1 1 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
2t 3t −t
= 3⎢0 ⎥e − 2 ⎢ −1 ⎥ e + ⎢ −3 ⎥ e .
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 7
Example 10.4.3
−3 2 2
⎡ ⎤
′
y =⎢ 2 −3 2 ⎥ y. (10.4.16)
⎣ ⎦
2 2 −3
Solution
∣ −3 − λ 2 2 ∣
∣ ∣ 2
2 −3 − λ 2 = −(λ − 1)(λ + 5 ) .
∣ ∣
∣ 2 2 −3 − λ ∣
Hence, λ = 1 is an eigenvalue of multiplicity 1, while λ = −5 is an eigenvalue of multiplicity

1 2 2 . Eigenvectors
associated with λ = 1 are solutions of the system with augmented matrix
1
⎡ −4 2 2 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ 2 −4 2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
2 2 −4 ⋮ 0
⎡1 0 −1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 1 −1 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Hence, x 1 = x2 = x3 , and we choose x 3 =1 to obtain the solution
1
⎡ ⎤
t
y1 = ⎢ 1 ⎥ e (10.4.17)
⎣ ⎦
1
of Equation 10.4.16. Eigenvectors associated with λ 2 = −5 are solutions of the system with augmented matrix
⎡2 2 2 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢2 2 2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
2 2 2 ⋮ 0
Hence, the components of these eigenvectors need only satisfy the single condition
x1 + x2 + x3 = 0.

Since there’s only one equation here, we can choose x and x arbitrarily. We obtain one eigenvector by choosing x
2 3 2 =0
and x = 1 , and another by choosing x = 1 and x = 0 . In both cases x = −1 . Therefore

3 2 3 1
−1 −1
⎡ ⎤ ⎡ ⎤
⎢ 0⎥ and ⎢ 1⎥
⎣ ⎦ ⎣ ⎦
1 0
are linearly independent eigenvectors associated with λ 2 = −5 , and the corresponding solutions of Equation 10.4.16 are
−1 −1
⎡ ⎤ ⎡ ⎤
−5t −5t
y2 = ⎢ 0⎥e and y3 = ⎢ 1⎥e .
⎣ ⎦ ⎣ ⎦
1 0
Because of this and Equation 10.4.17, Theorem 10.4.1 implies that the general solution of Equation 10.4.16 is
1 −1 −1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
t −5t −5t
y = c1 ⎢ 1 ⎥ e + c2 ⎢ 0⎥e + c3 ⎢ 1⎥e .
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 0
Geometric Properties of Solutions when n = 2

We’ll now consider the geometric properties of solutions of a 2 × 2 constant coefficient system
′
y a11 a12 y1
1
[ ] =[ ][ ]. (10.4.18)
′
y a21 a22 y2
2
y1
It is convenient to think of a “y -y plane," where a point is identified by rectangular coordinates (y
1 2 1, y2 ) . If y = [ ] is a
y2
non-constant solution of Equation 10.4.18, then the point (y (t), y (t)) moves along a curve C in the y -y plane as t varies
1 2 1 2
from −∞ to ∞. We call C the trajectory of y. (We also say that C is a trajectory of the system Equation 10.4.18.) I’s
important to note that C is the trajectory of infinitely many solutions of Equation 10.4.18, since if τ is any real number, then
y(t − τ ) is a solution of Equation 10.4.18 (Exercise 10.4.28b), and (y (t − τ ), y (t − τ )) also moves along C as t varies
1 2
from −∞ to ∞. Moreover, Exercise 10.4.28c implies that distinct trajectories of Equation 10.4.18 can’t intersect, and that two
solutions y and y of Equation 10.4.18 have the same trajectory if and only if y (t) = y (t − τ ) for some τ .
1 2 2 1
From Exercise 10.4.28a, a trajectory of a nontrivial solution of Equation 10.4.18 can’t contain (0, 0) , which we define to be
k1
the trajectory of the trivial solution y ≡0 . More generally, if y =[ ] ≠0 is a constant solution of Equation 10.4.18
k2
(which could occur if zero is an eigenvalue of the matrix of Equation 10.4.18 ), we define the trajectory of y to be the single
point (k , k ) .
1 2
To be specific, this is the question: What do the trajectories look like, and how are they traversed? In this section we’ll answer
this question, assuming that the matrix
a11 a12
A =[ ]
a21 a22
of Equation 10.4.18 has real eigenvalues λ1 and λ2 with associated linearly independent eigenvectors x1 and x2 . Then the
general solution of Equation 10.4.18 is
λ1 t λ2 t
y = c1 x1 e + c2 x2 e . (10.4.19)
We’ll consider other situations in the next two sections.

We leave it to you (Exercise 10.4.35) to classify the trajectories of Equation 10.4.18 if zero is an eigenvalue of A . We’ll
confine our attention here to the case where both eigenvalues are nonzero. In this case the simplest situation is where
λ = λ ≠ 0 , so Equation 10.4.19 becomes
1 2
λ1 t
y = (c1 x1 + c2 x2 )e .

Since x and x are linearly independent, an arbitrary vector x can be written as x = c x + c x . Therefore the general
1 2 1 1 2 2
solution of Equation 10.4.18 can be written as y = xe where x is an arbitrary 2-vector, and the trajectories of nontrivial
λ1 t
solutions of Equation 10.4.18 are half-lines through (but not including) the origin. The direction of motion is away from the
origin if λ > 0 (Figure 10.4.1), toward it if λ < 0 (Figure 10.4.2). (In these and the next figures an arrow through a point
1 1
indicates the direction of motion along the trajectory through the point.)
Figure 10.4.1 : Trajectories of a 2 × 2 system with a repeated positive eigenvalue.
Figure 10.4.2 : Trajectories of a 2 × 2 system with a repeated negative eigenvalue.

Now suppose λ > λ , and let L and L denote lines through the origin parallel to x and x , respectively. By a half-line of
2 1 1 2 1 2
L (or L ), we mean either of the rays obtained by removing the origin from L (or L ).
1 2 1 2
Letting c = 0 in Equation 10.4.19 yields y = c x e . If c ≠ 0 , the trajectory defined by this solution is a half-line of L .
2 1 1
λ1 t
1 1
The direction of motion is away from the origin if λ > 0 , toward the origin if λ < 0 . Similarly, the trajectory of
1 1
y =c x e
2 2 with c ≠ 0 is a half-line of L .
λ2 t
2 2
Henceforth, we assume that c and c in Equation 10.4.19 are both nonzero. In this case, the trajectory of Equation 10.4.19
1 2
can’t intersect L or L , since every point on these lines is on the trajectory of a solution for which either c = 0 or c = 0 .
1 2 1 2
(Remember: distinct trajectories can’t intersect!). Therefore the trajectory of Equation 10.4.19 must lie entirely in one of the
four open sectors bounded by L and L , but do not any point on L or L . Since the initial point (y (0), y (0)) defined by
1 2 1 2 1 2
y(0) = c1 x1 + c2 x2
is on the trajectory, we can determine which sector contains the trajectory from the signs of c1 and c2 , as shown in Figure
10.4.3.
Figure 10.4.3 : Four open sectors bounded by L and L . 1 2
The direction of y(t) in Equation 10.4.19 is the same as that of

−λ2 t −( λ2 −λ1 )t
e y(t) = c1 x1 e + c2 x2 (10.4.20)
and of

−λ1 t ( λ2 −λ1 )t
e y(t) = c1 x1 + c2 x2 e . (10.4.21)
Since the right side of Equation 10.4.20 approaches c x as t → ∞ , the trajectory is asymptotically parallel to L as t → ∞ .
2 2 2
Since the right side of Equation 10.4.21 approaches c x as t → −∞ , the trajectory is asymptotically parallel to L as
1 1 1
t → −∞ .
The shape and direction of traversal of the trajectory of Equation 10.4.19 depend upon whether λ1 and λ are both positive,
2
both negative, or of opposite signs. We’ll now analyze these three cases.
Henceforth ∥u∥ denote the length of the vector u .
Case 1: λ > λ > 0

2 1
Figure 10.4.4 shows some typical trajectories. In this case, lim ∥y(t)∥ = 0 , so the trajectory is not only asymptotically
t→−∞
parallel to L as t → −∞ , but is actually asymptotically tangent to L at the origin. On the other hand, lim
1 1 ∥y(t)∥ = ∞ t→∞
and
λ2 t
lim ∥ ∥ = lim ∥ c x eλ1 t ∥ = ∞,
∥y(t) − c2 x2 e ∥ 1 1
t→∞ t→∞
so, although the trajectory is asymptotically parallel to L2 as t → ∞ , it is not asymptotically tangent to L2 . The direction of
motion along each trajectory is away from the origin.
Figure 10.4.4 : Two positive eigenvalues; motion away from origin.
Case 2: 0 > λ > λ 2 1
Figure 10.4.5 shows some typical trajectories. In this case, lim ∥y(t)∥ = 0 , so the trajectory is asymptotically tangent to
t→∞
L at the origin as t → ∞ . On the other hand, lim

2 ∥y(t)∥ = ∞ and
t→−∞
λ1 t λ2 t
lim ∥y(t) − c x e ∥ = lim ∥ c2 x2 e ∥ = ∞,
∥ 1 1 ∥
t→−∞ t→−∞
so, although the trajectory is asymptotically parallel to L as t → −∞ , it is not asymptotically tangent to it. The direction of
1
motion along each trajectory is toward the origin.
Figure 10.4.5 : Two negative eigenvalues; motion toward the origin.
Case 3: λ > 0 > λ

2 1
Figure 10.4.6 shows some typical trajectories. In this case,

λ2 t
lim ∥y(t)∥ = ∞ and lim ∥ ∥ = lim ∥ c x eλ1 t ∥ = 0,
∥y(t) − c2 x2 e ∥ 1 1
t→∞ t→∞ t→∞
so the trajectory is asymptotically tangent to L as t → ∞ . Similarly,

2

λ1 t λ2 t
lim ∥y(t)∥ = ∞ and lim ∥y(t) − c x e ∥ = lim ∥ c2 x2 e ∥ = 0,
∥ 1 1 ∥
t→−∞ t→−∞ t→−∞
so the trajectory is asymptotically tangent to L as t → −∞ . The direction of motion is toward the origin on
1 L1 and away
from the origin on L . The direction of motion along any other trajectory is away from L , toward L .
2 1 2
Figure 10.4.6 : Eigenvalues of different signs.

10.4E: Constant Coefficient Homogeneous Systems I (Exercises)
Q10.4.1
1 2
1. y′
=[ ]y
2 1
−5 3
2. y′
=
1
4
[ ]y
3 −5
−4 3
3. y′
=
1
5
[ ]y
−2 −11
−1 −4
4. y′
=[ ]y
−1 −1
2 −4
5. y′
=[ ]y
−1 −1
4 −3
6. y′
=[ ]y
2 −1
−6 −3
7. y′
=[ ]y
1 −2
1 −1 −2
⎡ ⎤
8. y′
=⎢ 1 −2 −3 ⎥ y
⎣ ⎦
−4 1 −1
−6 −4 −8
⎡ ⎤
9. y′
= ⎢ −4 0 −4 ⎥ y
⎣ ⎦
−8 −4 −6
3 5 8
⎡ ⎤
10. y ′
=⎢ 1 −1 −2 ⎥ y
⎣ ⎦
−1 −1 −1
1 −1 2
⎡ ⎤
11. y ′
= ⎢ 12 −4 10 ⎥ y
⎣ ⎦
−6 1 −7
4 −1 −4
⎡ ⎤
12. y ′
= ⎢4 −3 −2 ⎥ y
⎣ ⎦
1 −1 −1
−2 2 −6
⎡ ⎤
13. y ′
=⎢ 2 6 2⎥y
⎣ ⎦
−2 −2 2
3 2 −2
⎡ ⎤
14. y ′
=⎢ −2 7 −2 ⎥ y
⎣ ⎦
−10 10 −5
3 1 −1
⎡ ⎤
15. y ′
=⎢ 3 5 1⎥y
⎣ ⎦
−6 2 4
Q10.4.2

−7 4 2
16. y′
=[ ] y, y(0) = [ ]
−6 7 −4
7 2 0
17. y′
=
1
6
[ ] y, y(0) = [ ]
−2 2 −3
21 −12 5
18. y′
=[ ] y, y(0) = [ ]
24 −15 3
−7 4 −1
19. y′
=[ ] y, y(0) = [ ]
−6 7 7
1 2 0 4
⎡ ⎤ ⎡ ⎤
20. y′
=
1
6
⎢4 −1 0 ⎥ y, y(0) = ⎢ 7 ⎥
⎣ ⎦ ⎣ ⎦
0 0 3 1
2 −2 3 1
⎡ ⎤ ⎡ ⎤
21. y′
=
1
3
⎢ −4 4 3 ⎥ y, y(0) = ⎢ 1 ⎥
⎣ ⎦ ⎣ ⎦
2 1 0 5
6 −3 −8 0
⎡ ⎤ ⎡ ⎤
22. y′
=⎢2 1 −2 ⎥ y, y(0) = ⎢ −1 ⎥
⎣ ⎦ ⎣ ⎦
3 −3 −5 −1
2 4 −7 4
⎡ ⎤ ⎡ ⎤
23. y′
=
1
3
⎢ 1 5 −5 ⎥ y, y(0) = ⎢ 1 ⎥
⎣ ⎦ ⎣ ⎦
−4 4 −1 3
3 0 1 2
⎡ ⎤ ⎡ ⎤
24. y′
= ⎢ 11 −2 7 ⎥ y, y(0) = ⎢ 7 ⎥
⎣ ⎦ ⎣ ⎦
1 0 3 6
−2 −5 −1 8
⎡ ⎤ ⎡ ⎤
25. y′
= ⎢ −4 −1 1 ⎥ y, y(0) = ⎢ −10 ⎥
⎣ ⎦ ⎣ ⎦
4 5 3 −4
3 −1 0 7
⎡ ⎤ ⎡ ⎤
26. y′
=⎢4 −2 0 ⎥ y, y(0) = ⎢ 10 ⎥
⎣ ⎦ ⎣ ⎦
4 −4 2 2
−2 2 6 6
⎡ ⎤ ⎡ ⎤
27. y′
=⎢ 2 6 2 ⎥ y, y(0) = ⎢ −10 ⎥
⎣ ⎦ ⎣ ⎦
−2 −2 2 7
Q10.4.3
28. Let A be an n × n constant matrix. Then Theorem 10.2.1 implies that the solutions of
′
y = Ay (A)
are all defined on (−∞, ∞).

a. Use Theorem 10.2.1 to show that the only solution of (A) that can ever equal the zero vector is y ≡ 0 .
b. Suppose y is a solution of (A) and y is defined by y (t) = y (t − τ ) , where τ is an arbitrary real number. Show that y
1 2 2 1 2
is also a solution of (A).

c. Suppose y and y are solutions of (A) and there are real numbers t and t such that y (t ) = y (t ) . Show that
1 2 1 2 1 1 2 2
y (t) = y (t − τ ) for all t , where τ = t − t .

2 1 2 1
Q10.4.4
In Exercises 10.4.29-10.4.34 describe and graph trajectories of the given system.

1 1
29. y′
=[ ]y
1 −1
−4 3
30. y′
=[ ]y
−2 −11
9 −3
31. y′
=[ ]y
−1 11
−1 −10
32. y′
=[ ]y
−5 4
5 −4
33. y′
=[ ]y
1 10
−7 1
34. y′
=[ ]y
3 −5
Q10.4.5
35. Suppose the eigenvalues of the 2 × 2 matrix A are λ = 0 and μ ≠ 0 , with corresponding eigenvectors x and x . Let L
1 2 1
be the line through the origin parallel to x .

1
a. Show that every point on L is the trajectory of a constant solution of y = Ay .

1
′
b. Show that the trajectories of nonconstant solutions of y = Ay are half-lines parallel to x and on either side of
′
2 L1 , and
that the direction of motion along these trajectories is away from L if μ > 0 , or toward L if μ < 0 .
1 1
Q10.4.6
The matrices of the systems in Exercises 10.4.36-10.4.41 are singular. Describe and graph the trajectories of nonconstant
solutions of the given systems.
−1 1
36. y′
=[ ]y
1 −1
−1 −3
37. y′
=[ ]y
2 6
1 −3
38. y′
=[ ]y
−1 3
1 −2
39. y′
=[ ]y
−1 2
−4 −4
40. y′
=[ ]y
1 1
3 −1
41. y′
=[ ]y
−3 1
Q10.4.6
42. Let P = P (t) and Q = Q(t) be the populations of two species at time t , and assume that each population would grow
exponentially if the other didn’t exist; that is, in the absence of competition,
′ ′
P = aP and Q = bQ, (A)
where a and b are positive constants. One way to model the effect of competition is to assume that the growth rate per
individual of each population is reduced by an amount proportional to the other population, so (A) is replaced by
′
P = aP − αQ
′
Q = −βP + bQ,
where α and β are positive constants. (Since negative population doesn’t make sense, this system holds only while P and Q
are both positive.) Now suppose P (0) = P > 0 and Q(0) = Q > 0 .
0 0

a. For several choices of a , b , α , and β, verify experimentally (by graphing trajectories of (A) in the P -Q plane) that there’s a
constant ρ > 0 (depending upon a , b , α , and β) with the following properties:
i. If Q0 > ρP0 , then P decreases monotonically to zero in finite time, during which Q remains positive.
ii. If Q0 < ρP0 , then Q decreases monotonically to zero in finite time, during which P remains positive.
b. Conclude from (a) that exactly one of the species becomes extinct in finite time if Q ≠ ρP . Determine experimentally
0 0
what happens if Q = ρP .
0 0
c. Confirm your experimental results and determine γ by expressing the eigenvalues and associated eigenvectors of
α −α
A =[ ]
−β b
in terms of a , b , α , and β, and applying the geometric arguments developed at the end of this section.

10.5: Constant Coefficient Homogeneous Systems II
We saw in Section 10.4 that if an n × n constant matrix A has n real eigenvalues λ , λ , …, λ (which need not be distinct) 1 2 n
with associated linearly independent eigenvectors x , x , …, x , then the general solution of y = Ay is

1 2 n
′
λ1 t λ2 t λn t
y = c1 x1 e + c2 x2 e + ⋯ + cn xn e .
In this section we consider the case where A has n real eigenvalues, but does not have n linearly independent eigenvectors. It
is shown in linear algebra that this occurs if and only if A has at least one eigenvalue of multiplicity r > 1 such that the
associated eigenspace has dimension less than r. In this case A is said to be defective. Since it is beyond the scope of this book
to give a complete analysis of systems with defective coefficient matrices, we will restrict our attention to some commonly
occurring special cases.
Example 10.5.1
Show that the system
′
11 −25
y =[ ]y (10.5.1)
4 −9
does not have a fundamental set of solutions of the form {x e , x e }, where λ and λ are eigenvalues of the
1
λ1 t
2
λ2 t
1 2
coefficient matrix A of Equation 10.5.1 and x , and x are associated linearly independent eigenvectors.
1 2
Solution
The characteristic polynomial of A is
∣ 11 − λ ∣
∣ ∣ = (λ − 11)(λ + 9) + 100
∣ 4 −λ ∣
2
=λ − 2λ + 1
2
= (λ − 1 ) .
Hence, λ = 1 is the only eigenvalue of A . The augmented matrix of the system (A − I )x = 0 is
⎡ 10 −25 ⋮ 0⎤
,
⎣ ⎦
4 −10 ⋮ 0
5
⎡1 − ⋮ 0⎤
2
⎣ ⎦
0 0 ⋮ 0
5
Hence, x 1 = 5 x2 /2 where x is arbitrary. Therefore all eigenvectors of A are scalar multiples of x
2 1 = [ ] , so A does
2
not have a set of two linearly independent eigenvectors.
5
From Example 10.5.1, we know that all scalar multiples of y
1
= [ ]e
t
are solutions of Equation 10.5.1 ; however, to find
2
the general solution we must find a second solution y such that {y

2 1
, y2 } is linearly independent. Based on your recollection
of the procedure for solving a constant coefficient scalar equation
′′ ′
ay + b y + cy = 0
in the case where the characteristic polynomial has a repeated root, you might expect to obtain a second solution of Equation
5
10.5.1 by multiplying the first solution by t . However, this yields y 2
=[ ] te
t
, which does not work, since
2

5 11 −25 5
′ t t t
y =[ ] (te + e ), while [ ] y2 = [ ] te .
2
2 4 −9 2
The next theorem shows what to do in this situation.
Theorem 10.5.1
Suppose the n × n matrix A has an eigenvalue λ of multiplicity ≥ 2 and the associated eigenspace has dimension 1;
1
that is, all λ -eigenvectors of A are scalar multiples of an eigenvector x. Then there are infinitely many vectors u such
1
that
(A − λ1 I )u = x. (10.5.2)
Moreover, if u is any such vector then

λ1 t λ1 t λ1 t
y1 = xe and y2 = u e + xte (10.5.3)
are linearly independent solutions of y ′

= Ay.
A complete proof of this theorem is beyond the scope of this book. The difficulty is in proving that there’s a vector u
satisfying Equation 10.5.2, since det(A − λ I ) = 0 . We’ll take this without proof and verify the other assertions of the
1
theorem. We already know that y in Equation 10.5.3 is a solution of y = Ay . To see that y is also a solution, we compute
1
′
2
′ λ1 t λ1 t λ1 t λ1 t λ1 t
y − Ay2 = λ1 u e + xe + λ1 xte − Au e − Axte
2
λ1 t λ1 t
= (λ1 u + x − Au)e + (λ1 x − Ax)te .
Since Ax = λ 1x , this can be written as

′ λ1 t
y − Ay2 = − ((A − λ1 I )u − x) e ,
2
and now Equation 10.5.2 implies that y

′
2
= Ay2 . To see that y1 and y2 are linearly independent, suppose c1 and c2 are
constants such that
λ1 t λ1 t λ1 t
c1 y + c2 y = c1 xe + c2 (u e + xte ) = 0. (10.5.4)
1 2
We must show that c 1 = c2 = 0 . Multiplying Equation 10.5.4 by e −λ1 t

shows that
c1 x + c2 (u + xt) = 0. (10.5.5)
By differentiating this with respect to t , we see that c 2x =0 , which implies c = 0 , because x ≠ 0 . Substituting c
2 2 =0 into
Equation 10.5.5 yields c x = 0 , which implies that c
1 1 =0 , again because x ≠ 0
Example 10.5.2
Use Theorem 10.5.1 to find the general solution of the system
′
11 −25
y =[ ]y (10.5.6)
4 −9
considered in Example 10.5.1.

Solution
In Example 10.5.1 we saw that λ = 1 is an eigenvalue of multiplicity 2 of the coefficient matrix A in Equation
1 10.5.6 ,
and that all of the eigenvectors of A are multiples of
5
x =[ ].
2
Therefore

5
t
y1 = [ ]e
2
is a solution of Equation 10.5.6. From Theorem 10.5.1, a second solution is given by t

y2 = u e + xte
t
, where
(A − I )u = x . The augmented matrix of this system is
⎡ 10 −25 ⋮ 5⎤
,
⎣ ⎦
4 −10 ⋮ 2
5 1
⎡1 − ⋮ ⎤
2 2
,
⎣ ⎦
0 0 ⋮ 0
Therefore the components of u must satisfy

5 1
u1 − u2 = ,
2 2
where u is arbitrary. We choose u

2 2 =0 , so that u 1 = 1/2 and
1
2
u =[ ].
0
Thus,
t
1 e 5 t
y2 = [ ] +[ ] te .
0 2 2
Since y1 and y are linearly independent by Theorem 10.5.1, they form a fundamental set of solutions of Equation
2
10.5.6 . Therefore the general solution of Equation 10.5.6 is

t
5 1 e 5
t t
y = c1 [ ] e + c2 ([ ] +[ ] te ) .
2 0 2 2
Note that choosing the arbitrary constant u2 to be nonzero is equivalent to adding a scalar multiple of y1 to the second
solution y (Exercise 10.5.33).
2
Example 10.5.3
3 4 −10
⎡ ⎤
′
y =⎢2 1 −2 ⎥ y. (10.5.7)
⎣ ⎦
2 2 −5
Solution
∣ 3 −λ 4 −10 ∣
∣ ∣ 2
2 1 −λ −2 = −(λ − 1)(λ + 1 ) .
∣ ∣
∣ 2 2 −5 − λ ∣
Hence, the eigenvalues are λ = 1 with multiplicity 1 and λ = −1 with multiplicity 2. Eigenvectors associated with
1 2
λ = 1 must satisfy (A − I )x = 0 . The augmented matrix of this system is

1

⎡2 4 −10 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢2 0 −2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
2 2 −6 ⋮ 0
⎡1 0 −1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 1 −2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Hence, x 1 = x3 and x 2 = 2 x3 , where x is arbitrary. Choosing x

3 3 =1 yields the eigenvector
1
⎡ ⎤
x1 = ⎢ 2 ⎥ .
⎣ ⎦
1
Therefore
1
⎡ ⎤
t
y1 = ⎢ 2 ⎥ e
⎣ ⎦
1
is a solution of Equation 10.5.7. Eigenvectors associated with λ 2 = −1 satisfy (A + I )x = 0 . The augmented matrix of
this system is
⎡4 4 −10 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢2 2 −2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
2 2 −4 ⋮ 0
⎡1 1 0 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 0 1 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Hence, x 3 =0 and x1 = −x2 , where x is arbitrary. Choosing x

2 2 =1 yields the eigenvector
−1
⎡ ⎤
x2 = ⎢ 1⎥,
⎣ ⎦
0
so
−1
⎡ ⎤
−t
y2 = ⎢ 1⎥e
⎣ ⎦
0
is a solution of Equation 10.5.7. Since all the eigenvectors of A associated with λ2 = −1 are multiples of x2 , we must
now use Theorem 10.5.1 to find a third solution of Equation 10.5.7 in the form
−1
⎡ ⎤
−t −t
y3 = u e +⎢ 1 ⎥ te , (10.5.8)
⎣ ⎦
0
where u is a solution of (A + I )u = x . The augmented matrix of this system is

2

⎡4 4 −10 ⋮ −1 ⎤
⎢ ⎥
⎢ ⎥,
⎢2 2 −2 ⋮ 1⎥
⎢ ⎥
⎣ ⎦
2 2 −4 ⋮ 0
⎡1 1 0 ⋮ 1⎤
⎢ ⎥
⎢ 1 ⎥.
⎢0 0 1 ⋮ ⎥
2
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Hence, u 3 = 1/2 and u 1 = 1 − u2 , where u is arbitrary. Choosing u

2 2 =0 yields
1
⎡ ⎤
u = ⎢ 0⎥,
⎣ 1 ⎦
2
and substituting this into Equation 10.5.8 yields the solution

2 −1
⎡ ⎤ −t ⎡ ⎤
e
−t
y3 = ⎢ 0 ⎥ +⎢ 1 ⎥ te
2
⎣ ⎦ ⎣ ⎦
1 0
of Equation 10.5.7. Since the Wronskian of {y 1

, y2 , y3 } at t = 0 is
∣1 −1 1∣
∣ ∣ 1
∣2 1 0∣ = ,
∣ ∣ 2
1
∣1 0 ∣
2
{ y1 , y2 , y3 } is a fundamental set of solutions of Equation 10.5.7. Therefore the general solution of Equation 10.5.7 is
1 −1 2 −1
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ −t ⎡ ⎤ ⎞
e
t −t −t
y = c1 ⎢ 2 ⎥ e + c2 ⎢ 1⎥e + c3 ⎜⎢ 0 ⎥ +⎢ 1 ⎥ te ⎟.
2
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
1 0 1 0
Theorem 10.5.2
Suppose the n × n matrix A has an eigenvalue λ of multiplicity ≥ 3 and the associated eigenspace is one–dimensional;
1
that is, all eigenvectors associated with λ are scalar multiples of the eigenvector x. Then there are infinitely many
1
vectors u such that

(A − λ1 I )u = x, (10.5.9)
and, if u is any such vector, there are infinitely many vectors v such that
(A − λ1 I )v = u. (10.5.10)
If u satisfies Equation 10.5.9 and v satisfies Equation 10.5.10, then

λ1 t
y1 = xe ,
λ1 t λ1 t
y2 = u e + xte , and
2 λ1 t
λ1 t λ1 t
t e
y3 = ve + ute +x
2

= Ay .

Again, it is beyond the scope of this book to prove that there are vectors u and v that satisfy Equation 10.5.9 and Equation
10.5.10. Theorem 10.5.1 implies that y and y are solutions of y = Ay . We leave the rest of the proof to you (Exercise
′
1 2
10.5.34).
Example 10.5.4
Use Theorem 10.5.2 to find the general solution of
1 1 1
⎡ ⎤
′
y =⎢1 3 −1 ⎥ y. (10.5.11)
⎣ ⎦
0 2 2
Solution:
∣ 1 −λ 1 1 ∣
∣ ∣ 3
1 3 −λ −1 = −(λ − 2 ) .
∣ ∣
∣ 0 2 2 −λ ∣
Hence, λ = 2 is an eigenvalue of multiplicity 3. The associated eigenvectors satisfy

1 (A − 2I )x = 0 . The augmented
matrix of this system is
⎡ −1 1 1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ 1 1 −1 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 2 0 ⋮ 0
⎡1 0 −1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 1 0 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Hence, x 1 = x3 and x
2 =0 , so the eigenvectors are all scalar multiples of
1
⎡ ⎤
x1 = ⎢ 0 ⎥ .
⎣ ⎦
1
Therefore
1
⎡ ⎤
2t
y1 = ⎢ 0 ⎥ e
⎣ ⎦
1
is a solution of Equation 10.5.11. We now find a second solution of Equation 10.5.11 in the form
1
⎡ ⎤
2t 2t
y2 = u e + ⎢ 0 ⎥ te ,
⎣ ⎦
1
where u satisfies (A − 2I )u = x . The augmented matrix of this system is

1
⎡ −1 1 1 ⋮ 1⎤
⎢ ⎥
⎢ ⎥
⎢ 1 1 −1 ⋮ 0⎥,
⎢ ⎥
⎣ ⎦
0 2 0 ⋮ 1

1
⎡1 0 −1 ⋮ − ⎤
2
⎢ ⎥
⎢ 1 ⎥.
⎢0 1 0 ⋮ ⎥
⎢ 2 ⎥
⎣ ⎦
0 0 0 ⋮ 0
Letting u 3 =0 yields u 1 = −1/2 and u 2 = 1/2 ; hence,

−1
⎡ ⎤
1
u = ⎢ 1⎥
2
⎣ ⎦
0
and
−1 1
⎡ ⎤ 2t ⎡ ⎤
e
2t
y2 = ⎢ 1⎥ + ⎢ 0 ⎥ te
2
⎣ ⎦ ⎣ ⎦
0 1
is a solution of Equation 10.5.11. We now find a third solution of Equation 10.5.11 in the form
−1 1
⎡ ⎤ 2t ⎡ ⎤ 2 2t
2t
te t e
y3 = ve +⎢ 1⎥ +⎢0 ⎥
2 2
⎣ ⎦ ⎣ ⎦
0 1
where v satisfies (A − 2I )v = u . The augmented matrix of this system is
1
⎡ −1 1 1 ⋮ − ⎤
2
⎢ ⎥
⎢ 1 ⎥,
⎢ 1 1 −1 ⋮ ⎥
2
⎢ ⎥
⎣ ⎦
0 2 0 ⋮ 0
1
⎡1 0 −1 ⋮ ⎤
2
⎢ ⎥
⎢ ⎥.
⎢0 1 0 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Letting v 3 =0 yields v 1 = 1/2 and v 2 =0 ; hence,

1
⎡ ⎤
1
v = ⎢0 ⎥.
2
⎣ ⎦
0
Therefore
1 −1 1
⎡ ⎤ 2t ⎡ ⎤ 2t ⎡ ⎤ 2 2t
e te t e
y3 = ⎢ 0 ⎥ +⎢ 1⎥ +⎢0 ⎥
2 2 2
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
0 0 1
is a solution of Equation 10.5.11. Since y , y , and y are linearly independent by Theorem 10.5.2, they form a
1 2 3
fundamental set of solutions of Equation 10.5.11. Therefore the general solution of Equation 10.5.11 is
1 −1 1 1 −1 1
⎡ ⎤ ⎛⎡ ⎤ 2t ⎡ ⎤ ⎞ ⎛⎡ ⎤ 2t ⎡ ⎤ 2t ⎡ ⎤ 2 2t ⎞
2t
e 2t
e te t e
y = c1 ⎢ 0 ⎥ e + c2 ⎜⎢ 1 ⎥ + ⎢ 0 ⎥ te ⎟ + c3 ⎜⎢ 0 ⎥ +⎢ 1 ⎥ +⎢ 0 ⎥ ⎟
2 2 2 2
⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠ ⎝⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎠
1 0 1 0 0 1

Theorem 10.5.3
Suppose the n × n matrix A has an eigenvalue λ of multiplicity ≥ 3 and the associated eigenspace is two–dimensional;
1
that is, all eigenvectors of A associated with λ are linear combinations of two linearly independent eigenvectors x and
1 1
x . Then there are constants α and β ( not both zero) such that if
2
x3 = α x1 + β x2 , (10.5.12)
then there are infinitely many vectors u such that

(A − λ1 I )u = x3 . (10.5.13)
If u satisfies Equation 10.5.13, then

λ1 t
y1 = x1 e
λ1 t
y2 = x2 e , and (10.5.14)
λ1 t λ1 t
y3 ue + x3 te ,

= Ay.
We omit the proof of this theorem.
Example 10.5.5
Use Theorem 10.5.3 to find the general solution of
0 0 1
⎡ ⎤
′
y = ⎢ −1 1 1 ⎥ y. (10.5.15)
⎣ ⎦
−1 0 2
Solution
∣ −λ 0 1 ∣
∣ ∣ 3
−1 1 −λ 1 = −(λ − 1 ) .
∣ ∣
∣ −1 0 2 −λ ∣
Hence, λ = 1 is an eigenvalue of multiplicity 3. The associated eigenvectors satisfy

1 (A − I )x = 0 . The augmented
⎡ −1 0 1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ −1 0 1 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
−1 0 1 ⋮ 0
⎡1 0 −1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 0 0 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Hence, x 1 = x3 and x is arbitrary, so the eigenvectors are of the form

2
x3 1 0
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
x1 = ⎢ x2 ⎥ = x3 ⎢ 0 ⎥ + x2 ⎢ 1 ⎥ .
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
x3 1 0
Therefore the vectors

1 0
⎡ ⎤ ⎡ ⎤
x1 = ⎢ 0 ⎥ and x2 = ⎢ 1 ⎥ (10.5.16)
⎣ ⎦ ⎣ ⎦
1 0
form a basis for the eigenspace, and

1 0
⎡ ⎤ ⎡ ⎤
t t
y1 = ⎢ 0 ⎥ e and y2 = ⎢ 1 ⎥ e
⎣ ⎦ ⎣ ⎦
1 0
are linearly independent solutions of Equation 10.5.15. To find a third linearly independent solution of Equation 10.5.15,
we must find constants α and β (not both zero) such that the system
(A − I )u = α x1 + β x2 (10.5.17)
has a solution u . The augmented matrix of this system is
⎡ −1 0 1 ⋮ α ⎤
⎢ ⎥
⎢ ⎥,
⎢ −1 0 1 ⋮ β ⎥
⎢ ⎥
⎣ ⎦
−1 0 1 ⋮ α
⎡1 0 −1 ⋮ −α ⎤
⎢ ⎥
⎢ ⎥. (10.5.18)
⎢0 0 0 ⋮ β −α ⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Therefore Equation 10.5.17 has a solution if and only if β = α , where α is arbitrary. If α = β = 1 then Equation 10.5.12
and Equation 10.5.16 yield
1 0 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
x3 = x1 + x2 = ⎢ 0 ⎥ + ⎢ 1 ⎥ = ⎢ 1 ⎥ ,
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 0 1
and the augmented matrix Equation 10.5.18 becomes
⎡1 0 −1 ⋮ −1 ⎤
⎢ ⎥
⎢ ⎥.
⎢0 0 0 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
This implies that u

1 = −1 + u3 , while u and u are arbitrary. Choosing u
2 3 2 = u3 = 0 yields
−1
⎡ ⎤
u =⎢ 0⎥.
⎣ ⎦
0
Therefore Equation 10.5.14 implies that
−1 1
⎡ ⎤ ⎡ ⎤
t t t t
y3 = u e + x3 te =⎢ 0 ⎥ e + ⎢ 1 ⎥ te
⎣ ⎦ ⎣ ⎦
0 1
is a solution of Equation 10.5.15. Since y , y , and y are linearly independent by Theorem [thmtype:10.5.3}, they form
1 2 3
a fundamental set of solutions for Equation 10.5.15. Therefore the general solution of Equation 10.5.15 is

1 0 −1 1
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
t t t t
y = c1 ⎢ 0 ⎥ e + c2 ⎢ 1 ⎥ e + c3 ⎜⎢ 0 ⎥ e + ⎢ 1 ⎥ te ⎟ . (10.5.19)
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
1 0 0 1

′
y a11 a12 y1
1
[ ] =[ ][ ] (10.5.20)
′
y a21 a22 y2
2
under the assumptions of this section; that is, when the matrix
a11 a12
A =[ ]
a21 a22
has a repeated eigenvalue λ and the associated eigenspace is one-dimensional. In this case we know from Theorem
1 10.5.1
that the general solution of Equation 10.5.20 is

λ1 t λ1 t λ1 t
y = c1 xe + c2 (u e + xte ), (10.5.21)
where x is an eigenvector of A and u is any one of the infinitely many solutions of

(A − λ1 I )u = x. (10.5.22)
We assume that λ 1 ≠0 .
Figure 10.5.1 : Positive and negative half-planes.

Let L denote the line through the origin parallel to x. By a half-line of L we mean either of the rays obtained by removing the
origin from L. Equation 10.5.21 is a parametric equation of the half-line of L in the direction of x if c > 0 , or of the half-line 1
of L in the direction of −x if c < 0 . The origin is the trajectory of the trivial solution y ≡ 0 .
1
Henceforth, we assume that c ≠ 0 . In this case, the trajectory of Equation 10.5.21 can’t intersect L, since every point of L is
2
on a trajectory obtained by setting c = 0 . Therefore the trajectory of Equation 10.5.21 must lie entirely in one of the open
2
half-planes bounded by L, but does not contain any point on L. Since the initial point (y (0), y (0)) defined by 1 2
y(0) = c x + c u is on the trajectory, we can determine which half-plane contains the trajectory from the sign of c , as
1 1 2 2
shown in Figure . For convenience we’ll call the half-plane where c > 0 the positive half-plane. Similarly, the-half plane2
where c < 0 is the negative half-plane. You should convince yourself (Exercise 10.5.35) that even though there are infinitely
2
many vectors u that satisfy Equation 10.5.22, they all define the same positive and negative half-planes. In the figures simply
regard u as an arrow pointing to the positive half-plane, since wen’t attempted to give u its proper length or direction in
comparison with x. For our purposes here, only the relative orientation of x and u is important; that is, whether the positive
half-plane is to the right of an observer facing the direction of x (as in Figures 10.5.2 and 10.5.5), or to the left of the observer
(as in Figures 10.5.3 and 10.5.4).
Multiplying Equation 10.5.21 by e −λ1 t
yields
−λ1 t
e y(t) = c1 x + c2 u + c2 tx.

Since the last term on the right is dominant when |t| is large, this provides the following information on the direction of y(t) :
a. Along trajectories in the positive half-plane (c 2 >0 ), the direction of y(t) approaches the direction of x as t → ∞ and the
direction of −x as t → −∞ .
b. Along trajectories in the negative half-plane (c 2 <0 ), the direction of y(t) approaches the direction of −x as t → ∞ and
the direction of x as t → −∞ .
Since
lim ∥y(t)∥ = ∞ and lim y(t) = 0 if λ1 > 0,

t→∞ t→−∞
or
lim ∥y(t)∥ = ∞ and lim y(t) = 0 if λ1 < 0,

t−→∞ t→∞
there are four possible patterns for the trajectories of Equation 10.5.20, depending upon the signs of c and λ . Figures 10.5.2
2 1
- 10.5.5 illustrate these patterns, and reveal the following principle:

If λ and c have the same sign then the direction of the traectory approaches the direction of −x as ∥y∥ → 0 and the
1 2
direction of x as ∥y∥ → ∞. If λ and c have opposite signs then the direction of the trajectory approaches the direction of x
1 2
as ∥y∥ → 0 and the direction of −x as ∥y∥ → ∞.
Figure 10.5.2 : Positive eigenvalue; motion away from the origin.
Figure 10.5.3 : Positive eigenvalue; motion away from the origin.
Figure 10.5.4 : Negative eigenvalue; motion toward the origin.

Figure 10.5.5 : Negative eigenvalue; motion toward the origin.

10.5E: Constant Coefficient Homogeneous Systems II (Exercises)
Q10.5.1
3 4
1. y′
=[ ]y
−1 7
0 −1
2. y′
=[ ]y
1 −2
−7 4
3. y′
=[ ]y
−1 −11
3 1
4. y′
=[ ]y
−1 1
4 12
5. y′
=[ ]y
−3 −8
−10 9
6. y′
=[ ]y
−4 2
−13 16
7. y′
=[ ]y
−9 11
0 2 1
⎡ ⎤
8. y′
= ⎢ −4 6 1⎥y
⎣ ⎦
0 4 2
1 1 −3
⎡ ⎤
9. y′
=
1
3
⎢ −4 −4 3 ⎥y
⎣ ⎦
−2 1 0
−1 1 −1
⎡ ⎤
10. y ′
= ⎢ −2 0 2 ⎥y
⎣ ⎦
−1 3 −1
4 −2 −2
⎡ ⎤
11. y ′
= ⎢ −2 3 −1 ⎥ y
⎣ ⎦
2 −1 3
6 −5 3
⎡ ⎤
12. y ′
=⎢2 −1 3⎥y
⎣ ⎦
2 1 1
Q10.5.2
−11 8 6
13. y ′
=[ ] y, y(0) = [ ]
−2 −3 2
15 −9 5
14. y ′
=[ ] y, y(0) = [ ]
16 −9 8
−3 −4 2
15. y ′
=[ ] y, y(0) = [ ]
1 −7 3
−7 24 3
16. y ′
=[ ] y, y(0) = [ ]
−6 17 1

−7 3 0
17. y′
=[ ] y, y(0) = [ ]
−3 −1 2
−1 1 0 6
⎡ ⎤ ⎡ ⎤
18. y′
=⎢ 1 −1 −2 ⎥ y, y(0) = ⎢ 5 ⎥
⎣ ⎦ ⎣ ⎦
−1 −1 −1 −7
−2 2 1 −6
⎡ ⎤ ⎡ ⎤
19. y′
= ⎢ −2 2 1 ⎥ y, y(0) = ⎢ −2 ⎥
⎣ ⎦ ⎣ ⎦
−3 3 2 0
−7 −4 4 −6
⎡ ⎤ ⎡ ⎤
20. y′
= ⎢ −1 0 1 ⎥ y, y(0) = ⎢ 9 ⎥
⎣ ⎦ ⎣ ⎦
−9 −5 6 −1
−1 −4 −1 −2
⎡ ⎤ ⎡ ⎤
21. y′
=⎢ 3 6 1 ⎥ y, y(0) = ⎢ 1 ⎥
⎣ ⎦ ⎣ ⎦
−3 −2 3 3
4 −8 −4 −4
⎡ ⎤ ⎡ ⎤
22. y′
= ⎢ −3 −1 −3 ⎥ y, y(0) = ⎢ 1 ⎥
⎣ ⎦ ⎣ ⎦
1 −1 9 −3
−5 −1 11 0
⎡ ⎤ ⎡ ⎤
23. y′
= ⎢ −7 1 13 ⎥ y, y(0) = ⎢ 2 ⎥
⎣ ⎦ ⎣ ⎦
−4 0 8 2
Q10.5.3
The coefficient matrices in Exercises 10.5.24-10.5.32 have eigenvalues of multiplicity 3. Find the general solution.
5 −1 1
⎡ ⎤
24. y′
= ⎢ −1 9 −3 ⎥ y
⎣ ⎦
−2 2 4
1 10 −12
⎡ ⎤
25. y′
=⎢2 2 3 ⎥y
⎣ ⎦
2 −1 6
−6 −4 −4
⎡ ⎤
26. y′
=⎢ 2 −1 1 ⎥y
⎣ ⎦
2 3 1
0 2 −2
⎡ ⎤
27. y′
= ⎢ −1 5 −3 ⎥ y
⎣ ⎦
1 1 1
−2 −12 10
⎡ ⎤
28. y′
=⎢ 2 −24 11 ⎥ y
⎣ ⎦
2 −24 8
−1 −12 8
⎡ ⎤
29. y′
=⎢ 1 −9 4⎥y
⎣ ⎦
1 −6 1
−4 0 −1
⎡ ⎤
30. y′
= ⎢ −1 −3 −1 ⎥ y
⎣ ⎦
1 0 −2

−3 −3 4
⎡ ⎤
31. y′
=⎢ 4 5 −8 ⎥ y
⎣ ⎦
2 3 −5
−3 −1 0
⎡ ⎤
32. y′
=⎢ 1 −1 0 ⎥y
⎣ ⎦
−1 −1 −2
Q10.5.4
33. Under the assumptions of Theorem 10.5.1, suppose u and u
^ are vectors such that
(A − λ1 I )u = x and ^ = x,
(A − λ1 I )u
and let
λ1 t λ1 t λ1 t λ1 t
y = ue + xte and y
^ =u
ê + xte .
2 2
Show that y 2
−y
^
2
is a scalar multiple of y 1
= xe
λ1 t
.
34. Under the assumptions of Theorem 10.5.2, let
λ1 t
y1 = xe ,
λ1 t λ1 t
y2 = u e + xte , and
2 λ1 t
t e
λ1 t λ1 t
y3 = ve + ute +x .
2
Complete the proof of Theorem 10.5.2 by showing that y3 is a solution of ′

y = Ay and that { y1 , y2 , y3 } is linearly
independent.
35. Suppose the matrix
a11 a12
A =[ ]
a21 a22
has a repeated eigenvalue λ and the associated eigenspace is one-dimensional. Let x be a λ -eigenvector of A . Show that if
1 1
(A − λ I )u = x and (A − λ I )u = x , then u − u
1 1 1 2 is parallel to x. Conclude from this that all vectors u such that
2 1
(A − λ I )u = x define the same positive and negative half-planes with respect to the line L through the origin parallel to x .
1
Q10.5.5
In Exercises 10.5.36-10.5.45 plot trajectories of the given system.
−3 −1
36. y′
=[ ]y
4 1
2 −1
37. y′
=[ ]y
1 0
−1 −3
38. y′
=[ ]y
3 5
−5 3
39. y′
=[ ]y
−3 1
−2 −3
40. y′
=[ ]y
3 4
−4 −3
41. y′
=[ ]y
3 2
0 −1
42. y′
=[ ]y
1 −2

0 1
43. y′
=[ ]y
−1 2
−2 1
44. y′
=[ ]y
−1 0
0 −4
45. y′
=[ ]y
1 −4

10.6: Constant Coefficient Homogeneous Systems III
We now consider the system y = Ay , where A has a complex eigenvalue λ = α + iβ with β ≠ 0 . We continue to assume
′
that A has real entries, so the characteristic polynomial of A has real coefficients. This implies that λ = α − iβ is also an ¯¯
¯
eigenvalue of A .
An eigenvector x of A associated with λ = α + iβ will have complex entries, so we’ll write
x = u + iv
where u and v have real entries; that is, u and v are the real and imaginary parts of x. Since Ax = λx ,
A(u + iv) = (α + iβ)(u + iv). (10.6.1)
Taking complex conjugates here and recalling that A has real entries yields
A(u − iv) = (α − iβ)(u − iv),
¯¯
¯ ¯¯
¯
which shows that x = u − iv is an eigenvector associated with λ = α − iβ . The complex conjugate eigenvalues λ and λ can
be separately associated with linearly independent solutions y = Ay ; however, we will not pursue this approach, since
′
solutions obtained in this way turn out to be complex–valued. Instead, we’ll obtain solutions of y = Ay in the form ′
y = f1 u + f2 v (10.6.2)
where f and f are real–valued scalar functions. The next theorem shows how to do this.
1 2
Theorem 10.6.1
Let A be an n × n matrix with real entries. Let λ = α + iβ (β ≠ 0 ) be a complex eigenvalue of A and let x = u + iv
be an associated eigenvector, where u and v have real components. Then u and v are both nonzero and
αt αt
y1 = e (u cos βt − v sin βt) and y2 = e (u sin βt + v cos βt),
which are the real and imaginary parts of

αt
e (cos βt + i sin βt)(u + iv), (10.6.3)

= Ay .
Proof
A function of the form Equation 10.6.2is a solution of y ′
= Ay if and only if
′ ′
f u + f v = f1 Au + f2 Av. (10.6.4)
1 2
Carrying out the multiplication indicated on the right side of Equation 10.6.1 and collecting the real and imaginary
parts of the result yields
A(u + iv) = (αu − βv) + i(αv + βu).
Equating real and imaginary parts on the two sides of this equation yields
Au = αu − βv
Av = αv + βu.
We leave it to you (Exercise 10.6.25) to show from this that u and v are both nonzero. Substituting from these
equations into Equation 10.6.4yields
′ ′
f u +f v = f1 (αu − βv) + f2 (αv + βu)
1 2
= (α f1 + β f2 )u + (−β f1 + α f2 )v.
This is true if

′ ′
f = α f1 + β f2 f − α f1 = βf2
1 1
or equivalently
′ ′
f = −β f1 + α f2 , f − α f2 = −β f1 .
2 2
If we let f 1 = g1 e
αt
and f 2 = g2 e
αt
, where g and g are to be determined, then the last two equations become
1 2
′
g = βg2
1
′
g = −β g1 ,
2
which implies that

′′ ′ 2
g = βg = −β g1 ,
1 2
so
′′ 2
g + β g1 = 0.
1
The general solution of this equation is
g1 = c1 cos βt + c2 sin βt.
Moreover, since g 2
′
= g /β
1
,
g2 = −c1 sin βt + c2 cos βt.
Multiplying g and g by e
1 2
αt
shows that
αt
f1 = e ( c1 cos βt + c2 sin βt),
αt
f2 = e (−c1 sin βt + c2 cos βt).
Substituting these into Equation 10.6.2shows that

αt
y = e [(c1 cos βt + c2 sin βt)u + (−c1 sin βt + c2 cos βt)v]
(10.6.5)
αt αt
= c1 e (u cos βt − v sin βt) + c2 e (u sin βt + v cos βt)
is a solution of y = Ay for any choice of the constants c and c . In particular, by first taking c = 1 and c = 0 and
′
1 2 1 2
then taking c = 0 and c = 1 , we see that y and y are solutions of y = Ay . We leave it to you to verify that they
1 2 1 2
′
are, respectively, the real and imaginary parts of Equation 10.6.3 (Exercise 10.6.26), and that they are linearly
independent (Exercise 10.6.27).
Example 10.6.1
4 −5
′
y =[ ] y. (10.6.6)
5 −2
Solution
∣ 4 −λ −5 ∣
2
∣ ∣ = (λ − 1 ) + 16.
∣ 5 −2 − λ ∣
Hence, λ = 1 + 4i is an eigenvalue of A . The associated eigenvectors satisfy (A − (1 + 4i) I ) x = 0 . The augmented

⎡ 3 − 4i −5 ⋮ 0⎤
,
⎣ ⎦
5 −3 − 4i ⋮ 0

3+4i
⎡1 − ⋮ 0⎤
5
⎢ ⎥.
⎣ ⎦
0 0 ⋮ 0
Therefore x 1 = (3 + 4i)x2 /5 . Taking x 2 =5 yields x 1 = 3 + 4i , so
3 + 4i
x =[ ]
5
is an eigenvector. The real and imaginary parts of
t
3 + 4i
e (cos 4t + i sin 4t) [ ]
5
are
t
3 cos 4t − 4 sin 4t t
3 sin 4t + 4 cos 4t
y1 = e [ ] and y2 = e [ ],
5 cos 4t 5 sin 4t
which are linearly independent solutions of Equation 10.6.6. The general solution of Equation 10.6.6 is
t
3 cos 4t − 4 sin 4t t
3 sin 4t + 4 cos 4t
y = c1 e [ ] + c2 e [ ].
5 cos 4t 5 sin 4t
Example 10.6.2
′
−14 39
y =[ ] y. (10.6.7)
−6 16
Solution
∣ −14 − λ 39 ∣
2
∣ ∣ = (λ − 1 ) + 9.
∣ −6 16 − λ ∣
Hence, λ = 1 + 3i is an eigenvalue of A . The associated eigenvectors satisfy (A − (1 + 3i)I ) x = 0 . The augmented

augmented matrix of this system is
⎡ −15 − 3i 39 ⋮ 0⎤
,
⎣ ⎦
−6 15 − 3i ⋮ 0
−5+i
⎡1 ⋮ 0⎤
2
.
⎣ ⎦
0 0 ⋮ 0
Therefore x 1 = (5 − i)/2 . Taking x2 =2 yields x 1 = 5 −i , so
5 −i
x =[ ]
2
is an eigenvector. The real and imaginary parts of
t
5 −i
e (cos 3t + i sin 3t) [ ]
2
are

sin 3t + 5 cos 3t − cos 3t + 5 sin 3t
t t
y1 = e [ ] and y2 = e [ ],
2 cos 3t 2 sin 3t
which are linearly independent solutions of Equation 10.6.7. The general solution of Equation 10.6.7 is
t
sin 3t + 5 cos 3t t
− cos 3t + 5 sin 3t
y = c1 e [ ] + c2 e [ ].
2 cos 3t 2 sin 3t
Example 10.6.3
−5 5 4
⎡ ⎤
′
y = ⎢ −8 7 6 ⎥ y. (10.6.8)
⎣ ⎦
1 0 0
Solution
∣ −5 − λ 5 4 ∣
∣ ∣ 2
−8 7 −λ 6 = −(λ − 2)(λ + 1).
∣ ∣
∣ 1 0 −λ ∣
Hence, the eigenvalues of A are λ 1 =2 ,λ 2 =i , and λ 3 = −i . The augmented matrix of (A − 2I )x = 0 is
⎡ −7 5 4 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ −8 5 6 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
1 0 −2 ⋮ 0
⎡1 0 −2 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 1 −2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Therefore x 1 = x2 = 2 x3 . Taking x 3 =1 yields

2
⎡ ⎤
x1 = ⎢ 2 ⎥ ,
⎣ ⎦
1
so
2
⎡ ⎤
2t
y1 = ⎢ 2 ⎥ e
⎣ ⎦
1
is a solution of Equation 10.6.8.

The augmented matrix of (A − iI )x = 0 is
⎡ −5 − i 5 4 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ −8 7 −i 6 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
1 0 −i ⋮ 0

⎡1 0 −i ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 1 1 −i ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Therefore x 1 = i x3 and x
2 = −(1 − i)x3 . Taking x 3 =1 yields the eigenvector
i
⎡ ⎤
x2 = ⎢ −1 + i ⎥ .
⎣ ⎦
1
The real and imaginary parts of

i
⎡ ⎤
(cos t + i sin t) ⎢ −1 + i ⎥
⎣ ⎦
1
are
− sin t cos t
⎡ ⎤ ⎡ ⎤
y2 = ⎢ − cos t − sin t ⎥ and y3 = ⎢ cos t − sin t ⎥ ,
⎣ ⎦ ⎣ ⎦
cos t sin t
which are solutions of Equation 10.6.8. Since the Wronskian of {y 1

, y2 , y3 } at t = 0 is
∣2 0 1∣
∣ ∣
2 −1 1 = 1,
∣ ∣
∣1 1 0∣
{ y1 , y2 , y3 } is a fundamental set of solutions of Equation 10.6.8. The general solution of Equation 10.6.8 is
2 − sin t cos t
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
2t
y = c1 ⎢ 2 ⎥ e + c2 ⎢ − cos t − sin t ⎥ + c3 ⎢ cos t − sin t ⎥ .
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 cos t sin t
Example 10.6.4
1 −1 −2
⎡ ⎤
′
y =⎢1 3 2 ⎥ y. (10.6.9)
⎣ ⎦
1 −1 2
Solution
∣ 1 −λ −1 −2 ∣
∣ ∣ 2
1 3 −λ 2 = −(λ − 2) ((λ − 2 ) + 4) .
∣ ∣
∣ 1 −1 2 −λ ∣
Hence, the eigenvalues of A are λ 1 =2 ,λ 2 = 2 + 2i , and λ 3 = 2 − 2i . The augmented matrix of (A − 2I )x = 0 is
⎡ −1 −1 −2 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ 1 1 2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
1 −1 0 ⋮ 0

⎡1 0 1 ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 1 1 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Therefore x 1 = x2 = −x3 . Taking x3 =1 yields

−1
⎡ ⎤
x1 = ⎢ −1 ⎥ ,
⎣ ⎦
1
so
−1
⎡ ⎤
2t
y1 = ⎢ −1 ⎥ e
⎣ ⎦
1
is a solution of Equation 10.6.9. The augmented matrix of (A − (2 + 2i)I ) x = 0 is
⎡ −1 − 2i −1 −2 ⋮ 0⎤
⎢ ⎥
⎢ ⎥,
⎢ 1 1 − 2i 2 ⋮ 0⎥
⎢ ⎥
⎣ ⎦
1 −1 −2i ⋮ 0
⎡1 0 −i ⋮ 0⎤
⎢ ⎥
⎢ ⎥.
⎢0 1 i ⋮ 0⎥
⎢ ⎥
⎣ ⎦
0 0 0 ⋮ 0
Therefore x 1 = i x3 and x
2 = −i x3 . Taking x 3 =1 yields the eigenvector
i
⎡ ⎤
x2 = ⎢ −i ⎥
⎣ ⎦
1
The real and imaginary parts of

i
⎡ ⎤
2t
e (cos 2t + i sin 2t) ⎢ −i ⎥
⎣ ⎦
1
are
− sin 2t cos 2t
⎡ ⎤ ⎡ ⎤
2t 2t
y2 = e ⎢ sin 2t ⎥ and y2 = e ⎢ − cos 2t ⎥ ,
⎣ ⎦ ⎣ ⎦
cos 2t sin 2t
which are solutions of Equation 10.6.9. Since the Wronskian of {y 1

,y ,y }
2 3
at t = 0 is
∣ −1 0 1∣
∣ ∣
−1 0 −1 = −2,
∣ ∣
∣ 1 1 0∣
{ y1 , y2 , y3 } is a fundamental set of solutions of Equation 10.6.9. The general solution of Equation 10.6.9 is

−1 − sin 2t cos 2t
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
2t 2t 2t
y = c1 ⎢ −1 ⎥ e + c2 e ⎢ sin 2t ⎥ + c3 e ⎢ − cos 2t ⎥ .
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 cos 2t sin 2t

′
y a11 a12 y1
1
[ ] =[ ][ ] (10.6.10)
′
y a21 a22 y2
2
under the assumptions of this section; that is, when the matrix
a11 a12
A =[ ]
a21 a22
has a complex eigenvalue λ = α + iβ (β ≠ 0 ) and x = u + iv is an associated eigenvector, where u and v have real
components. To describe the trajectories accurately it is necessary to introduce a new rectangular coordinate system in the y - 1
y plane. This raises a point that hasn’t come up before: It is always possible to choose x so that (u, v) = 0 . A special effort is
2
required to do this, since not every eigenvector has this property. However, if we know an eigenvector that doesn’t, we can
multiply it by a suitable complex constant to obtain one that does. To see this, note that if x is a λ -eigenvector of A and k is
an arbitrary real number, then
x1 = (1 + ik)x = (1 + ik)(u + iv) = (u − kv) + i(v + ku)
is also a λ -eigenvector of A , since
Ax1 = A((1 + ik)x) = (1 + ik)Ax = (1 + ik)λx = λ((1 + ik)x) = λ x1 .
The real and imaginary parts of x are 1
u1 = u − kv and v1 = v + ku, (10.6.11)
so
2 2 2
(u1 , v1 ) = (u − kv, v + ku) = − [(u, v)k + (∥v∥ − ∥u ∥ )k − (u, v)] .
Therefore (u 1, v1 ) = 0 if
2 2 2
(u, v)k + (∥v∥ − ∥u ∥ )k − (u, v) = 0. (10.6.12)
If (u, v) ≠ 0 we can use the quadratic formula to find two real values of k such that (u 1, v1 ) = 0 (Exercise 10.6.28).
Example 10.6.5
In Example 10.6.1 we found the eigenvector
3 + 4i 3 4
x =[ ] =[ ]+i [ ]
5 5 0
3 4
for the matrix of the system Equation 10.6.6. Here u = [ ] and v = [ ] are not orthogonal, since (u, v) = 12 . Since
5 0
∥v∥
2
− ∥u ∥
2
= −18 , Equation 10.6.12 is equivalent to
2
2k − 3k − 2 = 0.
The zeros of this equation are k 1 =2 and k 2 = −1/2 . Letting k = 2 in Equation 10.6.11 yields
−5 10
u1 = u − 2v = [ ] and v1 = v + 2u = [ ],
5 10
and (u 1, v1 ) = 0 . Letting k = −1/2 in Equation 10.6.11 yields

v 5 u 1 −5
u1 = u + =[ ] and v1 = v − = [ ],
2 5 2 2 5
and again (u 1, v1 ) = 0 .
(The numbers don’t always work out as nicely as in this example. You’ll need a calculator or computer to do Exercises
10.6.29-10.6.40.) Henceforth, we’ll assume that (u, v) = 0 . Let U and V be unit vectors in the directions of u and v,
respectively; that is, U = u/∥u∥ and V = v/∥v∥ . The new rectangular coordinate system will have the same origin as the
y -y system. The coordinates of a point in this system will be denoted by (z , z ), where z and z are the displacements in
1 2 1 2 1 2
the directions of U and V, respectively. From Equation 10.6.5, the solutions of Equation 10.6.10 are given by
αt
y =e [(c1 cos βt + c2 sin βt)u + (−c1 sin βt + c2 cos βt)v] . (10.6.13)
For convenience, let’s call the curve traversed by e

−αt
y(t) a shadow trajectory of Equation 10.6.10 . Multiplying Equation
10.6.13 by e yields
−αt
−αt
e y(t) = z1 (t)U + z2 (t)V,
where
z1 (t) = ∥u∥(c1 cos βt + c2 sin βt)
z2 (t) = ∥v∥(−c1 sin βt + c2 cos βt).
Therefore
2 2
(z1 (t)) (z2 (t))
2 2
+ =c +c
2 2 1 2
∥u∥ ∥v∥
(verify!), which means that the shadow trajectories of Equation 10.6.10 are ellipses centered at the origin, with axes of
symmetry parallel to U and V. Since
β∥u∥ β∥v∥
′ ′
z = z2 and z =− z1 ,
1 2
∥v∥ ∥u∥
the vector from the origin to a point on the shadow ellipse rotates in the same direction that V would have to be rotated by π/2
radians to bring it into coincidence with U (Figures 10.6.1 and 10.6.2).
Figure 10.6.1 : Shadow trajectories traversed clockwise.
Figure 10.6.2 : Shadow trajectories traversed counterclockwise

If α = 0 , then any trajectory of Equation 10.6.10 is a shadow trajectory of Equation 10.6.10; therefore, if λ is purely
imaginary, then the trajectories of Equation 10.6.10 are ellipses traversed periodically as indicated in Figures 10.6.1 and

10.6.2 . If α > 0 , then
lim ∥y(t)∥ = ∞ and lim y(t) = 0,

t→∞ t→−∞
so the trajectory spirals away from the origin as t varies from −∞ to ∞. The direction of the spiral depends upon the relative
orientation of U and V, as shown in Figures 10.6.3 and 10.6.4. If α < 0 , then
lim ∥y(t)∥ = ∞ and lim y(t) = 0,

t→−∞ t→∞
so the trajectory spirals toward the origin as t varies from −∞ to ∞. Again, the direction of the spiral depends upon the
relative orientation of U and V, as shown in Figures 10.6.5 and 10.6.6.
Figure 10.6.3 : α > 0 ; shadow trajectory spiraling outward.
Figure 10.6.4 : α > 0 ; shadow trajectory spiraling outward.
Figure 10.6.5 : α < 0 ; shadow trajectory spiraling inward.
Figure 10.6.6 : α < 0 ; shadow trajectory spiraling inward.

10.6E: Constant Coefficient Homogeneous Systems III (Exercises)
Q10.6.1
−1 2
1. y′
=[ ]y
−5 5
−11 4
2. y′
=[ ]y
−26 9
1 2
3. y′
=[ ]y
−4 5
5 −6
4. y′
=[ ]y
3 −1
3 −3 1
⎡ ⎤
5. y′
=⎢0 2 2⎥y
⎣ ⎦
5 1 1
−3 3 1
⎡ ⎤
6. y′
=⎢ 1 −5 −3 ⎥ y
⎣ ⎦
−3 7 3
2 1 −1
⎡ ⎤
7. y′
=⎢0 1 1 ⎥y
⎣ ⎦
1 0 1
−3 1 −3
⎡ ⎤
8. y′
=⎢ 4 −1 2 ⎥y
⎣ ⎦
4 −2 3
5 −4
9. y′
=[ ]y
10 1
7 −5
10. y ′
=
1
3
[ ]y
2 5
3 2
11. y ′
=[ ]y
−5 1
34 52
12. y ′
=[ ]y
−20 −30
1 1 2
⎡ ⎤
13. y ′
=⎢ 1 0 −1 ⎥ y
⎣ ⎦
−1 −2 −1
3 −4 −2
⎡ ⎤
14. y ′
= ⎢ −5 7 −8 ⎥ y
⎣ ⎦
−10 13 −8
6 0 −3
⎡ ⎤
15. y ′
= ⎢ −3 3 3 ⎥y
⎣ ⎦
1 −2 6
1 2 −2
⎡ ⎤
16. y ′
=⎢0 2 −1 ⎥ y
⎣ ⎦
1 0 0

Q10.6.2
4 −6 5
17. y′
=[ ] y, y(0) = [ ]
3 −2 2
7 15 5
18. y′
=[ ] y, y(0) = [ ]
−3 1 1
7 −15 17
19. y′
=[ ] y, y(0) = [ ]
3 −5 7
4 −2 1
20. y′
=
1
6
[ ] y, y(0) = [ ]
5 2 −1
5 2 −1 4
⎡ ⎤ ⎡ ⎤
21. y′
= ⎢ −3 2 2 ⎥ y, y(0) = ⎢ 0 ⎥
⎣ ⎦ ⎣ ⎦
1 3 2 6
4 4 0 8
⎡ ⎤ ⎡ ⎤
22. y′
=⎢8 10 −20 ⎥ y, y(0) = ⎢ 6 ⎥
⎣ ⎦ ⎣ ⎦
2 3 −2 5
1 15 −15 15
⎡ ⎤ ⎡ ⎤
23. y′
= ⎢ −6 18 −22 ⎥ y, y(0) = ⎢ 17 ⎥
⎣ ⎦ ⎣ ⎦
−3 11 −15 10
4 −4 4 16
⎡ ⎤ ⎡ ⎤
24. y′
= ⎢ −10 3 15 ⎥ y, y(0) = ⎢ 14 ⎥
⎣ ⎦ ⎣ ⎦
2 −3 1 6
Q10.6.3
25. Suppose an n × n matrix A with real entries has a complex eigenvalue λ = α + iβ (β ≠ 0 ) with associated eigenvector
x = u + iv , where u and v have real components. Show that u and v are both nonzero.
26. Verify that

αt αt
y1 = e (u cos βt − v sin βt) and y2 = e (u sin βt + v cos βt),
are the real and imaginary parts of

αt
e (cos βt + i sin βt)(u + iv).
27. Show that if the vectors u and v are not both 0 and β ≠ 0 then the vector functions
αt αt
y1 = e (u cos βt − v sin βt) and y2 = e (u sin βt + v cos βt)
are linearly independent on every interval.

u1 v1
28. Suppose u = [ ] and v = [ ] are not orthogonal; that is, (u, v) ≠ 0 .
u2 v2
1. Show that the quadratic equation

2 2 2
(u, v)k + (∥v∥ − ∥u ∥ )k − (u, v) = 0
has a positive root k and a negative root k

1 2 = −1/ k1 .
(1) (1) (2) (2) (1) (1) (2) (2)
2. Let u = u − k v , v = v + k u , u
1 1 1 1 1
= u − k2 v , and v
1
= v + k2 u , so that (u
1
,v
1
) = (u
1
,v
1
) =0 ,
from the discussion given above. Show that
(1) (1)
v u
(2) 1 (2) 1
u = and v =− .
1 1
k1 k1

(1) (1) (2) (2)
3. Let U , V , U , and V be unit vectors in the directions of u , v , u , and v , respectively. Conclude from (a) that
1 1 2 2 1 1 1 1
U =V2 and V = −U , and that therefore the counterclockwise angles from U to V and from U to V are both
1 2 1 1 1 2 2
π/2 or both −π/2.
Q10.6.4
In Exercises 10.6.29-10.6.32 find vectors U and V parallel to the axes of symmetry of the trajectories, and plot some typical
trajectories.
3 −5
29. y′
=[ ]y
5 −3
−15 10
30. y′
=[ ]y
−25 15
−4 8
31. y′
=[ ]y
−4 4
−3 −15
32. y′
=[ ]y
3 3
Q10.6.5
In Exercises 10.6.33-10.6.40 find vectors U and V parallel to the axes of symmetry of the shadow trajectories, and plot a
typical trajectory.
−5 6
33. y′
=[ ]y
−12 7
5 −12
34. y′
=[ ]y
6 −7
4 −5
35. y′
=[ ]y
9 −2
−4 9
36. y′
=[ ]y
−5 2
−1 10
37. y′
=[ ]y
−10 −1
−1 −5
38. y′
=[ ]y
20 −1
−7 10
39. y′
=[ ]y
−10 9
−7 6
40. y′
=[ ]y
−12 5

CHAPTER OVERVIEW
11: SYSTEMS OF EQUATIONS
11.1: GEOMETRY
11.2: ELEMENTARY OPERATIONS
We have taken an in depth look at graphical representations of systems of equations, as well as
how to find possible solutions graphically. Our attention now turns to working with systems
algebraically.
11.3: GAUSSIAN ELIMINATION

The work we did in the previous section will always find the solution to the system. In this section,
we will explore a less cumbersome way to find the solutions. First, we will represent a linear
system with an augmented matrix. A matrix is simply a rectangular array of numbers. The size or
dimension of a matrix is defined as m×n where m is the number of rows and n is the number of columns.
11.4: EXERCISES
1 9/16/2020
11.1: Geometry
Learning Objectives
1. Relate the types of solution sets of a system of two (three) variables to the intersections of lines in a plane (the
intersection of planes in three space)
As you may remember, linear equations like 2x + 3y = 6 can be graphed as straight lines in the coordinate plane. We say that
this equation is in two variables, in this case x and y . Suppose you have two such equations, each of which can be graphed as a
straight line, and consider the resulting graph of two lines. What would it mean if there exists a point of intersection between
the two lines? This point, which lies on both graphs, gives x and y values for which both equations are true. In other words,
this point gives the ordered pair (x, y) that satisfy both equations. If the point (x, y) is a point of intersection, we say that
(x, y) is a solution to the two equations. In linear algebra, we often are concerned with finding the solution(s) to a system of
equations, if such solutions exist. First, we consider graphical representations of solutions and later we will consider the
algebraic methods for finding solutions.
When looking for the intersection of two lines in a graph, several situations may arise. The following picture demonstrates the
possible situations when considering two equations (two lines in the graph) involving two variables.
In the first diagram, there is a unique point of intersection, which means that there is only one (unique) solution to the two
equations. In the second, there are no points of intersection and no solution. When no solution exists, this means that the two
lines are parallel and they never intersect. The third situation which can occur, as demonstrated in diagram three, is that the
two lines are really the same line. For example, x + y = 1 and 2x + 2y = 2 are equations which when graphed yield the same
line. In this case there are infinitely many points which are solutions of these two equations, as every ordered pair which is on
the graph of the line satisfies both equations. When considering linear systems of equations, there are always three types of
solutions possible; exactly one (unique) solution, infinitely many solutions, or no solution.
Example 11.1.1 : A Graphical Solution

Use a graph to find the solution to the following system of equations
x +y = 3
(11.1.1)
y −x = 5
Solution
Through graphing the above equations and identifying the point of intersection, we can find the solution(s). Remember
that we must have either one solution, infinitely many, or no solutions at all. The following graph shows the two
equations, as well as the intersection. Remember, the point of intersection represents the solution of the two equations, or
the (x, y) which satisfy both equations. In this case, there is one point of intersection at (−1, 4) which means we have one
unique solution, x = −1, y = 4 .
In the above example, we investigated the intersection point of two equations in two variables, x and y . Now we will consider
the graphical solutions of three equations in two variables.
Consider a system of three equations in two variables. Again, these equations can be graphed as straight lines in the plane, so
that the resulting graph contains three straight lines. Recall the three possible types of solutions; no solution, one solution, and
infinitely many solutions. There are now more complex ways of achieving these situations, due to the presence of the third
line. For example, you can imagine the case of three intersecting lines having no common point of intersection. Perhaps you
can also imagine three intersecting lines which do intersect at a single point. These two situations are illustrated below.
Consider the first picture above. While all three lines intersect with one another, there is no common point of intersection
where all three lines meet at one point. Hence, there is no solution to the three equations. Remember, a solution is a point
(x, y) which satisfies all three equations. In the case of the second picture, the lines intersect at a common point. This means
Ken Kuttler 9/11/2020 11.1.1 CC-BY https://math.libretexts.org/@go/page/44310

that there is one solution to the three equations whose graphs are the given lines. You should take a moment now to draw the
graph of a system which results in three parallel lines. Next, try the graph of three identical lines. Which type of solution is
represented in each of these graphs?
We have now considered the graphical solutions of systems of two equations in two variables, as well as three equations in two
variables. However, there is no reason to limit our investigation to equations in two variables. We will now consider equations
in three variables.
You may recall that equations in three variables, such as 2x + 4y − 5z = 8 , form a plane. Above, we were looking for
intersections of lines in order to identify any possible solutions. When graphically solving systems of equations in three
variables, we look for intersections of planes. These points of intersection give the (x, y, z) that satisfy all the equations in the
system. What types of solutions are possible when working with three variables? Consider the following picture involving two
planes, which are given by two equations in three variables.
Notice how these two planes intersect in a line. This means that the points (x, y, z) on this line satisfy both equations in the
system. Since the line contains infinitely many points, this system has infinitely many solutions.
It could also happen that the two planes fail to intersect. However, is it possible to have two planes intersect at a single point?
Take a moment to attempt drawing this situation, and convince yourself that it is not possible! This means that when we have
only two equations in three variables, there is no way to have a unique solution! Hence, the types of solutions possible for two
equations in three variables are no solution or infinitely many solutions.
Now imagine adding a third plane. In other words, consider three equations in three variables. What types of solutions are now
possible? Consider the following diagram.
In this diagram, there is no point which lies in all three planes. There is no intersection between all planes so there is no
solution. The picture illustrates the situation in which the line of intersection of the new plane with one of the original planes
forms a line parallel to the line of intersection of the first two planes. However, in three dimensions, it is possible for two lines
to fail to intersect even though they are not parallel. Such lines are called skew lines.
Recall that when working with two equations in three variables, it was not possible to have a unique solution. Is it possible
when considering three equations in three variables? In fact, it is possible, and we demonstrate this situation in the following
picture.
In this case, the three planes have a single point of intersection. Can you think of other types of solutions possible? Another is
that the three planes could intersect in a line, resulting in infinitely many solutions, as in the following diagram.
We have now seen how three equations in three variables can have no solution, a unique solution, or intersect in a line
resulting in infinitely many solutions. It is also possible that the three equations graph the same plane, which also leads to
infinitely many solutions.
You can see that when working with equations in three variables, there are many more ways to achieve the different types of
solutions than when working with two variables. It may prove enlightening to spend time imagining (and drawing) many
possible scenarios, and you should take some time to try a few.
You should also take some time to imagine (and draw) graphs of systems in more than three variables. Equations like
x + y − 2z + 4w = 8 with more than three variables are often called hyper-planes. You may soon realize that it is tricky to
draw the graphs of hyper-planes! Through the tools of linear algebra, we can algebraically examine these types of systems
which are difficult to graph. In the following section, we will consider these algebraic tools.
Exercise 11.1.1
Graphically, find the point (x 1, y1 ) which lies on both lines, x + 3y = 1 and 4x − y = 3 . That is, graph each line and see
where they intersect.
Exercise 11.1.2
Graphically, find the point of intersection of the two lines, 3x + y = 3 and x + 2y = 1 . That is, graph each line and see
where they intersect

Exercise 11.1.3
You have a system of k equations in two variables, k ≥ 2 . Explain the geometric significance of
1. No solution.
2. A unique solution.
3. An infinite number of solutions.

11.2: Elementary Operations
Learning Objectives
1. Use elementary operations to find the solution to a linear system of equations.
2. Find the and of a matrix.
3. Determine whether a system of linear equations has no solution, a unique solution or an infinite number of solutions
from its .
4. Solve a system of equations using Gaussian Elimination and Gauss-Jordan Elimination.
5. Model a physical system with linear equations and then solve.
We have taken an in depth look at graphical representations of systems of equations, as well as how to find possible solutions
graphically. Our attention now turns to working with systems algebraically.
Definition 1.2: System of Linear Equations

A system of linear equations is a list of equations,
a11 x1 + a12 x2 + ⋯ + a1n xn = b1
a21 x1 + a22 x2 + ⋯ + a2n xn = b2

(11.2.1)
⋮
am1 x1 + am2 x2 + ⋯ + amn xn = bm
where a and b are real numbers. The above is a system of m equations in the n variables, x
ij j 1, x2 ⋯ , xn . Written more
simply in terms of summation notation, the above can be written in the form
n
∑ aij xj = bi , i = 1, 2, 3, ⋯ , m (11.2.2)
j=1
The relative size of m and n is not important here. Notice that we have allowed a and b to be any real number. We can also
ij j
call these numbers scalars . We will use this term throughout the text, so keep in mind that the term scalar just means that we
are working with real numbers.
Now, suppose we have a system where bi = 0 for all i. In other words every equation equals 0. This is a special type of
system.
Definition 1.3: Homogeneous System of Equations

A system of equations is called homogeneous if each equation in the system is equal to 0. A homogeneous system has the
form
a11 x1 + a12 x2 + ⋯ + a1n xn = 0
a21 x1 + a22 x2 + ⋯ + a2n xn = 0

(11.2.3)
⋮
am1 x1 + am2 x2 + ⋯ + amn xn = 0
where a are scalars and x are variables.

ij i
Recall from the previous section that our goal when working with systems of linear equations was to find the point of
intersection of the equations when graphed. In other words, we looked for the solutions to the system. We now wish to find
these solutions algebraically. We want to find values for x , ⋯ , x which solve all of the equations. If such a set of values
1 n
exists, we call (x , ⋯ , x ) the solution set.

1 n

Recall the above discussions about the types of solutions possible. We will see that systems of linear equations will have one
unique solution, infinitely many solutions, or no solution. Consider the following definition.
Definition 1.4: Consistent and Inconsistent Systems

A system of linear equations is a list of equations,
a11 x1 + a12 x2 + ⋯ + a1n xn = b1
a21 x1 + a22 x2 + ⋯ + a2n xn = b2

(11.2.4)
am1 x1 + am2 x2 + ⋯ + amn xn = bm
where a and b are real numbers. The above is a system of m equations in the n variables, x
ij j 1, x2 ⋯ , xn . Written more
simply in terms of summation notation, the above can be written in the form
n
∑ aij xj = bi , i = 1, 2, 3, ⋯ , m (11.2.5)
j=1
A system of linear equations is called consistent if there exists at least one solution. It is called inconsistent if there is no
solution.
If you think of each equation as a condition which must be satisfied by the variables, consistent would mean there is some
choice of variables which can satisfy all the conditions. Inconsistent would mean there is no choice of the variables which can
satisfy all of the conditions.
The following sections provide methods for determining if a system is consistent or inconsistent, and finding solutions if they
exist.
Elementary Operations
We begin this section with an example. Recall from Example [graphicalsoln] that the solution to the given system was
(x, y) = (−1, 4).
Example 11.2.5 : Verifying an Ordered Pair is a Solution

Algebraically verify that (x, y) = (−1, 4) is a solution to the following system of equations.
x +y = 3
(11.2.6)
y −x = 5
Solution
By graphing these two equations and identifying the point of intersection, we previously found that (x, y) = (−1, 4) is
the unique solution.
We can verify algebraically by substituting these values into the original equations, and ensuring that the equations hold.
First, we substitute the values into the first equation and check that it equals 3.
x + y = (−1) + (4) = 3 (11.2.7)
This equals 3 as needed, so we see that (−1, 4) is a solution to the first equation. Substituting the values into the second
equation yields
y − x = (4) − (−1) = 4 + 1 = 5 (11.2.8)
which is true. For (x, y) = (−1, 4) each equation is true and therefore, this is a solution to the system.
Now, the interesting question is this: If you were not given these numbers to verify, how could you algebraically determine the
solution? Linear algebra gives us the tools needed to answer this question. The following basic operations are important tools

that we will utilize.
Definition 11.2.6 : Elementary Operations

Elementary operations are those operations consisting of the following.
1. Interchange the order in which the equations are listed.
2. Multiply any equation by a nonzero number.
3. Replace any equation with itself added to a multiple of another equation.
It is important to note that none of these operations will change the set of solutions of the system of equations. In fact,
elementary operations are the key tool we use in linear algebra to find solutions to systems of equations.
Consider the following example.
Example 11.2.7 : Effects of an Elementary Operation

Show that the system
x +y = 7
(11.2.9)
2x − y = 8
has the same solution as the system

x +y = 7
(11.2.10)
−3y = −6
Solution
Notice that the second system has been obtained by taking the second equation of the first system and adding -2 times the
first equation, as follows:
2x − y + (−2)(x + y) = 8 + (−2)(7) (11.2.11)
By simplifying, we obtain
−3y = −6 (11.2.12)
which is the second equation in the second system. Now, from here we can solve for y and see that y =2 . Next, we
substitute this value into the first equation as follows
x +y = x +2 = 7 (11.2.13)
Hence x = 5 and so (x, y) = (5, 2) is a solution to the second system. We want to check if (5, 2) is also a solution to the
first system. We check this by substituting (x, y) = (5, 2) into the system and ensuring the equations are true.
x + y = (5) + (2) = 7
(11.2.14)
2x − y = 2 (5) − (2) = 8
Hence, (5, 2) is also a solution to the first system.
This example illustrates how an elementary operation applied to a system of two equations in two variables does not affect the
solution set. However, a linear system may involve many equations and many variables and there is no reason to limit our
study to small systems. For any size of system in any number of variables, the solution set is still the collection of solutions to
the equations. In every case, the above operations of Definition [def:elementaryoperations] do not change the set of solutions
to the system of linear equations.
In the following theorem, we use the notation E to represent an equation, while b denotes a constant.
i i
Theorem 11.2.8 : Elementary Operations and Solutions

Suppose you have a system of two linear equations
E1 = b1
(11.2.15)
E2 = b2
Then the following systems have the same solution set as [system]:
1. E2 = b2
(11.2.16)
E1 = b1
2. E1 = b1
(11.2.17)
kE2 = kb2
for any scalar k , provided k ≠ 0 .
3. E1 = b1
(11.2.18)
E2 + kE1 = b2 + kb1
for any scalar k (including k = 0 ).

Before we proceed with the proof of Theorem [thm:elementaryoperationsandsolns], let us consider this theorem in context
of Example [exa:effectsofelementaryoperation]. Then,
E1 = x + y, b1 = 7
(11.2.19)
E2 = 2x − y, b2 = 8
Recall the elementary operations that we used to modify the system in the solution to the example. First, we added (−2)
times the first equation to the second equation. In terms of Theorem [thm:elementaryoperationsandsolns], this action is
given by
E2 + (−2) E1 = b2 + (−2) b1 (11.2.20)
or
2x − y + (−2) (x + y) = 8 + (−2) 7 (11.2.21)
This gave us the second system in Example [exa:effectsofelementaryoperation], given by

E1 = b1
(11.2.22)
E2 + (−2) E1 = b2 + (−2) b1
From this point, we were able to find the solution to the system. Theorem [thm:elementaryoperationsandsolns] tells us
that the solution we found is in fact a solution to the original system.
We will now prove Theorem [thm:elementaryoperationsandsolns].
Proof
The proof that the systems [system] and [thm1.9.1] have the same solution set is as follows. Suppose that
(x , ⋯ , x ) is a solution to E = b , E = b
1 n 1 1 2 . We want to show that this is a solution to the system in [thm1.9.1]
2
above. This is clear, because the system in [thm1.9.1] is the original system, but listed in a different order. Changing
the order does not effect the solution set, so (x , ⋯ , x ) is a solution to [thm1.9.1].
1 n
Next we want to prove that the systems [system] and [thm1.9.2] have the same solution set. That is E = b , E = b 1 1 2 2
has the same solution set as the system E = b , kE = kb provided k ≠ 0 . Let (x , ⋯ , x ) be a solution of
1 1 2 2 1 n
E = b , E = b , . We want to show that it is a solution to E = b , kE = kb

1 1 2 2 . Notice that the only difference
1 1 2 2
between these two systems is that the second involves multiplying the equation, E = b by the scalar k . Recall that
2 2
when you multiply both sides of an equation by the same number, the sides are still equal to each other. Hence if
(x , ⋯ , x ) is a solution to E = b , then it will also be a solution to kE = kb . Hence, (x , ⋯ , x ) is also a
1 n 2 2 2 2 1 n
solution to [thm1.9.2].
Similarly, let (x , ⋯ , x ) be a solution of E = b , kE = kb . Then we can multiply the equation kE = kb by
1 n 1 1 2 2 2 2
the scalar 1/k, which is possible only because we have required that k ≠ 0 . Just as above, this action preserves

equality and we obtain the equation E 2 = b2 . Hence (x 1, ⋯ , xn ) is also a solution to E 1 = b1 , E2 = b2 .
Finally, we will prove that the systems [system] and [thm1.9.3] have the same solution set. We will show that any
solution of E = b , E = b is also a solution of [thm1.9.3]. Then, we will show that any solution of [thm1.9.3] is
1 1 2 2
also a solution of E = b , E = b . Let (x , ⋯ , x ) be a solution to E = b , E = b . Then in particular it solves

1 1 2 2 1 n 1 1 2 2
E = b . Hence, it solves the first equation in [thm1.9.3]. Similarly, it also solves E = b . By our proof of
1 1 2 2
[thm1.9.2], it also solves kE = kb . Notice that if we add E and kE , this is equal to b + kb . Therefore, if
1 1 2 1 2 1
(x , ⋯ , x ) solves E = b , E = b
1 n 1 1 it must also solve E + kE = b + kb .
2 2 2 1 2 1
Now suppose (x , ⋯ , x ) solves the system E = b , E + kE = b + kb . Then in particular it is a solution of

1 n 1 1 2 1 2 1
E = b . Again by our proof of [thm1.9.2], it is also a solution to kE = kb . Now if we subtract these equal
1 1 1 1
quantities from both sides of E + kE = b + kb we obtain E = b , which shows that the solution also satisfies
2 1 2 1 2 2
E1 = b1 , E2 = b2 .
Stated simply, the above theorem shows that the elementary operations do not change the solution set of a system of equations.
We will now look at an example of a system of three equations and three variables. Similarly to the previous examples, the
goal is to find values for x, y, z such that each of the given equations are satisfied when these values are substituted in.
Example 11.2.9 : Solving a System of Equations with Elementary Operations

Find the solutions to the system,
x + 3y + 6z = 25
2x + 7y + 14z = 58 (11.2.23)
2y + 5z = 19
Solution
We can relate this system to Theorem [thm:elementaryoperationsandsolns] above. In this case, we have
E1 = x + 3y + 6z, b1 = 25
E2 = 2x + 7y + 14z, b2 = 58 (11.2.24)
E3 = 2y + 5z, b3 = 19
Theorem [thm:elementaryoperationsandsolns] claims that if we do elementary operations on this system, we will not
change the solution set. Therefore, we can solve this system using the elementary operations given in Definition
[def:elementaryoperations]. First, replace the second equation by (−2) times the first equation added to the second. This
yields the system
x + 3y + 6z = 25
y + 2z = 8 (11.2.25)
2y + 5z = 19
Now, replace the third equation with (−2) times the second added to the third. This yields the system
x + 3y + 6z = 25
y + 2z = 8 (11.2.26)
z =3
At this point, we can easily find the solution. Simply take z =3 and substitute this back into the previous equation to
solve for y , and similarly to solve for x.
x + 3y + 6 (3) = x + 3y + 18 = 25
y + 2 (3) = y + 6 = 8 (11.2.27)
z =3
The second equation is now

y +6 = 8 (11.2.28)
You can see from this equation that y = 2 . Therefore, we can substitute this value into the first equation as follows:
x + 3 (2) + 18 = 25 (11.2.29)
By simplifying this equation, we find that x = 1 . Hence, the solution to this system is (x, y, z) = (1, 2, 3). This process is
called back substitution.
Alternatively, in [solvingasystem3] you could have continued as follows. Add (−2) times the third equation to the second
and then add (−6) times the second to the first. This yields
x + 3y = 7
y =2 (11.2.30)
z =3
Now add (−3) times the second to the first. This yields
x =1
y =2 (11.2.31)
z =3
a system which has the same solution set as the original system. This avoided back substitution and led to the same
solution set. It is your decision which you prefer to use, as both methods lead to the correct solution, (x, y, z) = (1, 2, 3).

11.3: Gaussian Elimination
The work we did in the previous section will always find the solution to the system. In this section, we will explore a less
cumbersome way to find the solutions. First, we will represent a linear system with an augmented matrix. A matrix is simply
a rectangular array of numbers. The size or dimension of a matrix is defined as m × n where m is the number of rows and n
is the number of columns. In order to construct an augmented matrix from a linear system, we create a coefficient matrix
from the coefficients of the variables in the system, as well as a constant matrix from the constants. The coefficients from one
equation of the system create one row of the augmented matrix.
For example, consider the linear system in Example [exa:solvingasystemwithelementaryops]
x + 3y + 6z = 25
2x + 7y + 14z = 58 (11.3.1)
2y + 5z = 19
This system can be written as an augmented matrix, as follows
⎡ 1 3 6 25 ⎤
⎢ 2 7 14 58 ⎥ (11.3.2)
⎢ ⎥
⎣ 0 2 5 19 ⎦
Notice that it has exactly the same information as the original system. Here it is understood that the first column contains the
1
⎡ ⎤
coefficients from x in each equation, in order, ⎢2 ⎥. Similarly, we create a column from the coefficients on y in each
⎣ ⎦
0
3 6
⎡ ⎤ ⎡ ⎤
equation, ⎢ 7 ⎥ and a column from the coefficients on z in each equation, ⎢ 14 ⎥ . For a system of more than three variables,
⎣ ⎦ ⎣ ⎦
2 5
we would continue in this way constructing a column for each variable. Similarly, for a system of less than three variables, we
simply construct a column for each variable.
25
⎡ ⎤
Finally, we construct a column from the constants of the equations, ⎢ 58 ⎥ .
⎣ ⎦
19
The rows of the augmented matrix correspond to the equations in the system. For example, the top row in the augmented
matrix, [ 1 3 6 | 25 ] corresponds to the equation
x + 3y + 6z = 25. (11.3.3)
Consider the following definition.
Definition 11.3.1 : Augmented Matrix of a Linear System

For a linear system of the form
a11 x1 + ⋯ + a1n xn = b1
(11.3.4)
⋮
am1 x1 + ⋯ + amn xn = bm
where the x are variables and the a and b are constants, the augmented matrix of this system is given by
i ij i
⎡ a11 ⋯ a1n b1 ⎤
⎢ ⎥
⎢ ⎥ (11.3.5)
⎢ ⋮ ⋮ ⋮ ⎥
⎣ am1 ⋯ amn bm ⎦

Now, consider elementary operations in the context of the augmented matrix. The elementary operations in Definition
[def:elementaryoperations] can be used on the rows just as we used them on equations previously. Changes to a system of
equations in as a result of an elementary operation are equivalent to changes in the augmented matrix resulting from the
corresponding row operation. Note that Theorem [thm:elementaryoperationsandsolns] implies that any elementary row
operations used on an augmented matrix will not change the solution to the corresponding system of equations. We now
formally define elementary row operations. These are the key tool we will use to find solutions to systems of equations.
Definition 11.3.2 : Elementary Row Operations

The elementary row operations (also known as row operations) consist of the following
1. Switch two rows.
2. Multiply a row by a nonzero number.
3. Replace a row by any multiple of another row added to it.
Recall how we solved Example [exa:solvingasystemwithelementaryops]. We can do the exact same steps as above, except now
in the context of an augmented matrix and using row operations. The augmented matrix of this system is
⎡ 1 3 6 25 ⎤
⎢ 2 7 14 58 ⎥ (11.3.6)
⎢ ⎥
⎣ 0 2 5 19 ⎦
Thus the first step in solving the system given by [solvingasystem1] would be to take (−2) times the first row of the
augmented matrix and add it to the second row,
⎡ 1 3 6 25 ⎤
⎢ 0 1 2 8 ⎥ (11.3.7)
⎢ ⎥
⎣ 0 2 5 19 ⎦
Note how this corresponds to [solvingasystem2]. Next take (−2) times the second row and add to the third,
⎡ 1 3 6 25 ⎤
⎢ 0 1 2 8 ⎥ (11.3.8)
⎢ ⎥
⎣ 0 0 1 3 ⎦
This augmented matrix corresponds to the system
x + 3y + 6z = 25
y + 2z = 8 (11.3.9)
z =3
which is the same as [solvingasystem3]. By back substitution you obtain the solution x = 1, y = 2, and z = 3.
Through a systematic procedure of row operations, we can simplify an augmented matrix and carry it to row-echelon form or
reduced row-echelon form, which we define next. These forms are used to find the solutions of the system of equations
corresponding to the augmented matrix.
In the following definitions, the term leading entry refers to the first nonzero entry of a row when scanning the row from left
to right.
Definition 11.3.3 : Row-Echelon Form

An augmented matrix is in row-echelon form if
1. All nonzero rows are above any rows of zeros.
2. Each leading entry of a row is in a column to the right of the leading entries of any row above it.
3. Each leading entry of a row is equal to 1.

We also consider another reduced form of the augmented matrix which has one further condition.
Definition 11.3.4 : Reduced Row-Echelon Form

An augmented matrix is in reduced row-echelon form if
1. All nonzero rows are above any rows of zeros.
2. Each leading entry of a row is in a column to the right of the leading entries of any rows above it.
3. Each leading entry of a row is equal to 1.
4. All entries in a column above and below a leading entry are zero.
Notice that the first three conditions on a reduced row-echelon form matrix are the same as those for row-echelon form.
Hence, every reduced row-echelon form matrix is also in row-echelon form. The converse is not necessarily true; we cannot
assume that every matrix in row-echelon form is also in reduced row-echelon form. However, it often happens that the row-
echelon form is sufficient to provide information about the solution of a system.
The following examples describe matrices in these various forms. As an exercise, take the time to carefully verify that they are
in the specified form.
Example 11.3.5 : Not in Row-Echelon Form

The following augmented matrices are not in row-echelon form (and therefore also not in reduced row-echelon form).
⎡ 0 0 0 0 ⎤
⎡ 0 2 3 3 ⎤
⎢ 1 2 3 3 ⎥ ⎡ 1 2 3 ⎤
⎢ ⎥ ⎢ ⎥
⎢ ⎥ 1 5 0 2
⎢ 0 1 0 2 ⎥,⎢
⎢ 2 4 −6 ⎥,⎢
⎥ ⎢
⎥ (11.3.10)
⎢ ⎥ ⎥
⎢ 7 5 0 1 ⎥
⎢ ⎥ ⎣ ⎦
⎢ 0 0 0 1 ⎥ 4 0 7
⎣ 0 0 1 0 ⎦
⎣ 0 0 0 0 ⎦
Example 11.3.6 : Matrices in Row-Echelon Form

The following augmented matrices are in row-echelon form, but not in reduced row-echelon form.
⎡ 1 3 5 4 ⎤
⎡ 1 0 6 5 8 2 ⎤ ⎡ 1 0 6 0 ⎤
⎢ 0 1 0 7 ⎥
⎢ 0 0 1 2 7 3 ⎥ ⎢ ⎥ ⎢
0 1 4 0 ⎥
⎢ ⎥,⎢
⎢ 0 0 1 0
⎥ ⎢
⎥, ⎥ (11.3.11)
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ 0 0 0 0 1 1 ⎥ ⎢ 0 0 1 0
⎥ ⎢ ⎥
⎢ 0 0 0 1 ⎥
⎣ 0 0 0 0 0 0 ⎦ ⎣ 0 0 0 0 ⎦
⎣ 0 0 0 0 ⎦
Notice that we could apply further row operations to these matrices to carry them to reduced row-echelon form. Take the time
to try that on your own. Consider the following matrices, which are in reduced row-echelon form.
Example 11.3.7 : Matrices in Reduced Row-Echelon Form

The following augmented matrices are in reduced row-echelon form.
⎡ 1 0 0 0 ⎤
⎡ 1 0 0 5 0 0 ⎤
⎢ 0 1 0 0 ⎥ ⎡ 1 0 0 4 ⎤
⎢ 0 0 1 2 0 0 ⎥ ⎢
⎢
⎥
⎥
⎢ ⎥,⎢ 0 0 1 0 ⎥,⎢ 0 1 0 3 ⎥ (11.3.12)
⎢ ⎥ ⎢ ⎢ ⎥
0 0 0 0 1 1 ⎥
⎢ ⎥ ⎢ ⎥ ⎣
0 0 0 1 0 0 1 2 ⎦
⎢ ⎥
⎣ 0 0 0 0 0 0 ⎦
⎣ 0 0 0 0 ⎦

One way in which the row-echelon form of a matrix is useful is in identifying the pivot positions and pivot columns of the
matrix.
Definition 11.3.8 : Pivot Position and Pivot Column

A pivot position in a matrix is the location of a leading entry in the row-echelon form of a matrix.
A pivot column is a column that contains a pivot position.
For example consider the following.
Example 11.3.9 : Pivot Position

Let
⎡ 1 2 3 4 ⎤
A =⎢
⎢ 3 2 1 6 ⎥
⎥ (11.3.13)
⎣ 4 4 4 10 ⎦
Where are the pivot positions and pivot columns of the augmented matrix A ?
Solution
The row-echelon form of this matrix is
⎡ 1 2 3 4 ⎤
3
⎢ 0 1 2 ⎥ (11.3.14)
⎢ 2 ⎥
⎣ 0 0 0 0 ⎦
This is all we need in this example, but note that this matrix is not in reduced row-echelon form.
In order to identify the pivot positions in the original matrix, we look for the leading entries in the row-echelon form of
the matrix. Here, the entry in the first row and first column, as well as the entry in the second row and second column are
the leading entries. Hence, these locations are the pivot positions. We identify the pivot positions in the original matrix, as
in the following:
⎡ 1 2 3 4 ⎤
⎢ ⎥ (11.3.15)
⎢ 3 2 1 6 ⎥
⎣ 4 4 4 10 ⎦
Thus the pivot columns in the matrix are the first two columns.
The following is an algorithm for carrying a matrix to row-echelon form and reduced row-echelon form. You may wish to use
this algorithm to carry the above matrix to row-echelon form or reduced row-echelon form yourself for practice.
Algorithm 11.3.10 : Reduced Row-Echelon Form Algorithm

This algorithm provides a method for using row operations to take a matrix to its reduced row-echelon form. We begin
with the matrix in its original form.
1. Starting from the left, find the first nonzero column. This is the first pivot column, and the position at the top of this
column is the first pivot position. Switch rows if necessary to place a nonzero number in the first pivot position.
2. Use row operations to make the entries below the first pivot position (in the first pivot column) equal to zero.
3. Ignoring the row containing the first pivot position, repeat steps 1 and 2 with the remaining rows. Repeat the process
until there are no more rows to modify.

4. Divide each nonzero row by the value of the leading entry, so that the leading entry becomes 1. The matrix will then
be in row-echelon form.
The following step will carry the matrix from row-echelon form to reduced row-echelon form.
5. Moving from right to left, use row operations to create zeros in the entries of the pivot columns which are above the
pivot positions. The result will be a matrix in reduced row-echelon form.
Most often we will apply this algorithm to an augmented matrix in order to find the solution to a system of linear equations.
However, we can use this algorithm to compute the reduced row-echelon form of any matrix which could be useful in other
applications.
Consider the following example of Algorithm [algo:rrefalgorithm].
Example 11.3.11 Finding row-echelon form and

reduced row-echelon form of a Matrix
Let
0 −5 −4
⎡ ⎤
A = ⎢1 4 3⎥ (11.3.16)
⎣ ⎦
5 10 7
Find the row-echelon form of A . Then complete the process until A is in reduced row-echelon form.
Solution
In working through this example, we will use the steps outlined in Algorithm [algo:rrefalgorithm].
1. The first pivot column is the first column of the matrix, as this is the first nonzero column from the left. Hence the
first pivot position is the one in the first row and first column. Switch the first two rows to obtain a nonzero entry in
the first pivot position, outlined in a box below.
⎡ 1 4 3⎤
⎢ 0 −5 −4 ⎥ (11.3.17)
⎣ ⎦
5 10 7
2. Step two involves creating zeros in the entries below the first pivot position. The first entry of the second row is
already a zero. All we need to do is subtract 5 times the first row from the third row. The resulting matrix is
1 4 3
⎡ ⎤
⎢0 −5 −4 ⎥ (11.3.18)
⎣ ⎦
0 10 8
3. Now ignore the top row. Apply steps 1 and 2 to the smaller matrix
−5 −4
[ ] (11.3.19)
10 8
In this matrix, the first column is a pivot column, and −5 is in the first pivot position. Therefore, we need to create a
zero below it. To do this, add 2 times the first row (of this matrix) to the second. The resulting matrix is
−5 −4
[ ] (11.3.20)
0 0
Our original matrix now looks like

1 4 3
⎡ ⎤
⎢0 −5 −4 ⎥ (11.3.21)
⎣ ⎦
0 0 0
We can see that there are no more rows to modify.

4. Now, we need to create leading 1s in each row. The first row already has a leading 1 so no work is needed here.
Divide the second row by −5 to create a leading 1. The resulting matrix is
1 4 3
⎡ ⎤
4
⎢0 1 ⎥ (11.3.22)
5
⎣ ⎦
0 0 0
This matrix is now in row-echelon form.

5. Now create zeros in the entries above pivot positions in each column, in order to carry this matrix all the way to
reduced row-echelon form. Notice that there is no pivot position in the third column so we do not need to create any
zeros in this column! The column in which we need to create zeros is the second. To do so, subtract 4 times the
second row from the first row. The resulting matrix is
1
1 0 −
⎡ 5 ⎤
⎢ 4 ⎥ (11.3.23)
⎢0 1
5
⎥
⎣ ⎦
0 0 0
This matrix is now in reduced row-echelon form.

The above algorithm gives you a simple way to obtain the row-echelon form and reduced row-echelon form of a matrix.
The main idea is to do row operations in such a way as to end up with a matrix in row-echelon form or reduced row-
echelon form. This process is important because the resulting matrix will allow you to describe the solutions to the
corresponding linear system of equations in a meaningful way.
In the next example, we look at how to solve a system of equations using the corresponding augmented matrix.
Example 11.3.12 : Finding the Solution to a System

Give the complete solution to the following system of equations
2x + 4y − 3z = −1
5x + 10y − 7z = −2 (11.3.24)
3x + 6y + 5z = 9
Solution
The augmented matrix for this system is
⎡ 2 4 −3 −1 ⎤
⎢ 5 10 −7 −2 ⎥ (11.3.25)
⎢ ⎥
⎣ 3 6 5 9 ⎦
In order to find the solution to this system, we wish to carry the augmented matrix to . We will do so using Algorithm
[algo:rrefalgorithm]. Notice that the first column is nonzero, so this is our first pivot column. The first entry in the
first row, 2, is the first leading entry and it is in the first pivot position. We will use row operations to create zeros in
the entries below the 2. First, replace the second row with −5 times the first row plus 2 times the second row. This
yields
⎡ 2 4 −3 −1 ⎤
⎢ 0 0 1 1 ⎥ (11.3.26)
⎢ ⎥
⎣ 3 6 5 9 ⎦
Now, replace the third row with −3 times the first row plus to 2 times the third row. This yields
⎡ 2 4 −3 −1 ⎤
⎢ 0 0 1 1 ⎥ (11.3.27)
⎢ ⎥
⎣ 0 0 1 21 ⎦

Now the entries in the first column below the pivot position are zeros. We now look for the second pivot column,
which in this case is column three. Here, the 1 in the second row and third column is in the pivot position. We need
to do just one row operation to create a zero below the 1.
Taking −1 times the second row and adding it to the third row yields
⎡ 2 4 −3 −1 ⎤
⎢ 0 0 1 1 ⎥ (11.3.28)
⎢ ⎥
⎣ 0 0 0 20 ⎦
We could proceed with the algorithm to carry this matrix to row-echelon form or reduced row-echelon form.
However, remember that we are looking for the solutions to the system of equations. Take another look at the third
row of the matrix. Notice that it corresponds to the equation
0x + 0y + 0z = 20 (11.3.29)
There is no solution to this equation because for all x, y, z, the left side will equal 0 and 0 ≠ 20. This shows there is
no solution to the given system of equations. In other words, this system is inconsistent.
The following is another example of how to find the solution to a system of equations by carrying the corresponding
augmented matrix to reduced row-echelon form.
Example 11.3.13 : An Infinite Set of Solutions

Give the complete solution to the system of equations
3x − y − 5z = 9
y − 10z = 0 (11.3.30)
−2x + y = −6
Solution
The augmented matrix of this system is
⎡ 3 −1 −5 9 ⎤
⎢ 0 1 −10 0 ⎥ (11.3.31)
⎢ ⎥
⎣ −2 1 0 −6 ⎦
In order to find the solution to this system, we will carry the augmented matrix to reduced row-echelon form, using
Algorithm [algo:rrefalgorithm]. The first column is the first pivot column. We want to use row operations to create
zeros beneath the first entry in this column, which is in the first pivot position. Replace the third row with 2 times the
first row added to 3 times the third row. This gives
⎡ 3 −1 −5 9 ⎤
⎢ 0 1 −10 0 ⎥ (11.3.32)
⎢ ⎥
⎣ 0 1 −10 0 ⎦
Now, we have created zeros beneath the 3 in the first column, so we move on to the second pivot column (which is
the second column) and repeat the procedure. Take −1 times the second row and add to the third row.
⎡ 3 −1 −5 9 ⎤
⎢ 0 1 −10 0 ⎥ (11.3.33)
⎢ ⎥
⎣ 0 0 0 0 ⎦
The entry below the pivot position in the second column is now a zero. Notice that we have no more pivot columns
because we have only two leading entries.
At this stage, we also want the leading entries to be equal to one. To do so, divide the first row by 3.

1 5
⎡ 1 − − 3 ⎤
3 3
⎢ ⎥ (11.3.34)
⎢ 0 1 −10 0 ⎥
⎣ 0 0 0 0 ⎦
This matrix is now in row-echelon form.

Let’s continue with row operations until the matrix is in reduced row-echelon form. This involves creating zeros
above the pivot positions in each pivot column. This requires only one step, which is to add times the second row
1
to the first row.
⎡ 1 0 −5 3 ⎤
⎢ 0 1 −10 0 ⎥ (11.3.35)
⎢ ⎥
⎣ 0 0 0 0 ⎦
This is in reduced row-echelon form, which you should verify using Definition [def:rref]. The equations
corresponding to this reduced row-echelon form are
x − 5z = 3
(11.3.36)
y − 10z = 0
or
x = 3 + 5z
(11.3.37)
y = 10z
Observe that z is not restrained by any equation. In fact, z can equal any number. For example, we can let z = t ,
where we can choose t to be any number. In this context t is called a parameter . Therefore, the solution set of this
system is
x = 3 + 5t
y = 10t (11.3.38)
z =t
where t is arbitrary. The system has an infinite set of solutions which are given by these equations. For any value of
t we select, x, y, and z will be given by the above equations. For example, if we choose t = 4 then the
corresponding solution would be
x = 3 + 5(4) = 23
y = 10(4) = 40 (11.3.39)
z =4
In Example [exa:infinitesetofsoln] the solution involved one parameter. It may happen that the solution to a system
involves more than one parameter, as shown in the following example.
Example 11.3.1 : A Two Parameter Set of Solutions

Find the solution to the system
x + 2y − z + w = 3
x +y −z+w = 1 (11.3.40)
x + 3y − z + w = 5
Solution
The augmented matrix is

⎡ 1 2 −1 1 3 ⎤
⎢ 1 1 −1 1 1 ⎥ (11.3.41)
⎢ ⎥
⎣ ⎦
1 3 −1 1 5
We wish to carry this matrix to row-echelon form. Here, we will outline the row operations used. However, make
sure that you understand the steps in terms of Algorithm [algo:rrefalgorithm].
Take −1 times the first row and add to the second. Then take −1 times the first row and add to the third. This yields
⎡ 1 2 −1 1 3 ⎤
⎢ 0 −1 0 0 −2 ⎥ (11.3.42)
⎢ ⎥
⎣ 0 1 0 0 2 ⎦
Now add the second row to the third row and divide the second row by −1.
⎡ 1 2 −1 1 3 ⎤
⎢ 0 1 0 0 2 ⎥ (11.3.43)
⎢ ⎥
⎣ 0 0 0 0 0 ⎦
This matrix is in row-echelon form and we can see that x and y correspond to pivot columns, while z and w do not.
Therefore, we will assign parameters to the variables z and w. Assign the parameter s to z and the parameter t to w.
Then the first row yields the equation x + 2y − s + t = 3 , while the second row yields the equation y = 2 . Since
y = 2 , the first equation becomes x + 4 − s + t = 3 showing that the solution is given by
x = −1 + s − t
y =2
(11.3.44)
z =s
w =t
It is customary to write this solution in the form
x −1 + s − t
⎡ ⎤ ⎡ ⎤
⎢ y ⎥ ⎢ 2 ⎥
⎢ ⎥ =⎢ ⎥ (11.3.45)
⎢ z ⎥ ⎢ s ⎥
⎣ ⎦ ⎣ ⎦
w t
This example shows a system of equations with an infinite solution set which depends on two parameters. It can be less
confusing in the case of an infinite solution set to first place the augmented matrix in reduced row-echelon form rather
than just row-echelon form before seeking to write down the description of the solution.
In the above steps, this means we don’t stop with the row-echelon form in equation [twoparameters1]. Instead we first
place it in reduced row-echelon form as follows.
⎡ 1 0 −1 1 −1 ⎤
⎢ 0 1 0 0 2 ⎥ (11.3.46)
⎢ ⎥
⎣ 0 0 0 0 0 ⎦
Then the solution is y = 2 from the second row and x = −1 + z − w from the first. Thus letting z =s and w = t, the
solution is given by [twoparameters2].
You can see here that there are two paths to the correct answer, which both yield the same answer. Hence, either approach
may be used. The process which we first used in the above solution is called Gaussian Elimination This process
involves carrying the matrix to row-echelon form, converting back to equations, and using back substitution to find the
solution. When you do row operations until you obtain reduced row-echelon form, the process is called Gauss-Jordan
Elimination.

We have now found solutions for systems of equations with no solution and infinitely many solutions, with one parameter
as well as two parameters. Recall the three types of solution sets which we discussed in the previous section; no solution,
one solution, and infinitely many solutions. Each of these types of solutions could be identified from the graph of the
system. It turns out that we can also identify the type of solution from the reduced row-echelon form of the augmented
matrix.
No Solution: In the case where the system of equations has no solution, the row-echelon form of the augmented matrix
will have a row of the form
[0 0 0 | 1] (11.3.47)
This row indicates that the system is inconsistent and has no solution.
One Solution: In the case where the system of equations has one solution, every column of the coefficient matrix is a
pivot column. The following is an example of an augmented matrix in for a row-echelon form system of equations
with one solution.
⎡ 1 0 0 5 ⎤
⎢ 0 1 0 0 ⎥ (11.3.48)
⎢ ⎥
⎣ 0 0 1 2 ⎦
Infinitely Many Solutions: In the case where the system of equations has infinitely many solutions, the solution
contains parameters. There will be columns of the coefficient matrix which are not pivot columns. The following are
examples of augmented matrices in reduced row-echelon form for systems of equations with infinitely many
solutions.
⎡ 1 0 0 5 ⎤
⎢ 0 1 2 −3 ⎥ (11.3.49)
⎢ ⎥
⎣ 0 0 0 0 ⎦
or
1 0 0 5
[ ] (11.3.50)
0 1 0 −3

11.4: Exercises
Exercise 11.4.1
Graphically, find the point (x1, y1 ) which lies on both lines, x + 3y = 1 and 4x − y = 3. That is, graph each line and see
Answer
x + 3y = 1
, Solution is: [x = 10
13
,y =
1
13
] .
4x − y = 3
Exercise 11.4.2
Graphically, find the point of intersection of the two lines 3x + y = 3 and x + 2y = 1. That is, graph each line and see
Answer
3x + y = 3
, Solution is: [x = 1, y = 0]
x + 2y = 1
Exercise 11.4.3
You have a system of k equations in two variables, k ≥ 2 . Explain the geometric significance of
1. No solution.
2. A unique solution.
3. An infinite number of solutions.
Exercise 11.4.4
Consider the following augmented matrix in which ∗ denotes an arbitrary number and ■ denotes a nonzero number.
Determine whether the given augmented matrix is consistent. If consistent, is the solution unique?
⎡ ■ ∗ ∗ ∗ ∗ ∗ ⎤
⎢ 0 ■ ∗ ∗ 0 ∗ ⎥
⎢ ⎥ (11.4.1)
⎢ ⎥
⎢ 0 0 ■ ∗ ∗ ∗ ⎥
⎣ ⎦
0 0 0 0 ■ ∗
Answer
The solution exists but is not unique.
Exercise 11.4.5
⎡ ■ ∗ ∗ ∗ ⎤
⎢ 0 ■ ∗ ∗ ⎥ (11.4.2)
⎢ ⎥
⎣ 0 0 ■ ∗ ⎦

Answer
A solution exists and is unique.
Exercise 11.4.6
⎡ ■ ∗ ∗ ∗ ∗ ∗ ⎤
⎢ 0 ■ 0 ∗ 0 ∗ ⎥
⎢ ⎥ (11.4.3)
⎢ ⎥
⎢ 0 0 0 ■ ∗ ∗ ⎥
⎣ 0 0 0 0 ■ ∗ ⎦
Exercise 11.4.7
⎡ ■ ∗ ∗ ∗ ∗ ∗ ⎤
⎢ 0 ■ ∗ ∗ 0 ∗ ⎥
⎢ ⎥ (11.4.4)
⎢ ⎥
⎢ 0 0 0 0 ■ 0 ⎥
⎣ 0 0 0 0 ∗ ■ ⎦
Answer
There might be a solution. If so, there are infinitely many.
Exercise 11.4.8
Suppose a system of equations has fewer equations than variables. Will such a system necessarily be consistent? If so,
explain why and if not, give an example which is not consistent.
Answer
No. Consider x + y + z = 2 and x + y + z = 1.
Exercise 11.4.9
If a system of equations has more equations than variables, can it have a solution? If so, give an example and if not, tell
why not.
Answer
These can have a solution. For example, x + y = 1, 2x + 2y = 2, 3x + 3y = 3 even has an infinite set of solutions.
Exercise 11.4.10
Find h such that
2 h 4
[ ] (11.4.5)
3 6 7
is the augmented matrix of an inconsistent system.

Answer
h =4
Exercise 11.4.11
Find h such that
1 h 3
[ ] (11.4.6)
2 4 6
is the augmented matrix of a consistent system.
Answer
Any h will work.
Exercise 11.4.12
Find h such that
1 1 4
[ ] (11.4.7)
3 h 12
is the augmented matrix of a consistent system.
Answer
Any h will work.
Exercise 11.4.13
Choose h and k such that the augmented matrix shown has each of the following:
1. one solution
2. no solution
3. infinitely many solutions
1 h 2
[ ] (11.4.8)
2 4 k
Answer
If h ≠ 2 there will be a unique solution for any k . If h = 2 and there are no solutions. If h = 2 and k = 4, then there
are infinitely many solutions.
Exercise 11.4.14
Choose h and k such that the augmented matrix shown has each of the following:
1. one solution
2. no solution
3. infinitely many solutions
1 2 2
[ ] (11.4.9)
2 h k

Answer
If h ≠ 4, then there is exactly one solution. If h = 4 and k ≠ 4, then there are no solutions. If h = 4 and k = 4, then
there are infinitely many solutions.
Exercise 11.4.15
Determine if the system is consistent. If so, is the solution unique?
x + 2y + z − w = 2
x −y +z+w = 1
(11.4.10)
2x + y − z = 1
4x + 2y + z = 5
Answer
There is no solution. The system is inconsistent. You can see this from the augmented matrix.
1
1 2 1 −1 2 ⎡1 0 0 \vspace0.05in
3
0⎤
⎡ ⎤
⎢ 2 ⎥
⎢1 −1 1 1 1⎥ ⎢0 1 0 −\vspace0.05in 0⎥
⎢ ⎥ ,: ⎢ 3
⎥.
⎢2 1 −1 0 1⎥ ⎢ ⎥
⎢0 0 1 0 0⎥
⎣ ⎦
4 2 1 0 5 ⎣ ⎦
0 0 0 0 1
Exercise 11.4.16
Determine if the system is consistent. If so, is the solution unique?
x + 2y + z − w = 2
x −y +z+w = 0
(11.4.11)
2x + y − z = 1
4x + 2y + z = 3
Answer
Solution is: [w = 3
2
y − 1, x =
2
3
−
1
2
y, z =
1
3
]
Exercise 11.4.17
Determine which matrices are in reduced row-echelon form.
1 2 0
1. [ ]
0 1 7
1 0 0 0
⎡ ⎤
2. ⎢ 0 0 1 2⎥
⎣ ⎦
0 0 0 0
1 1 0 0 0 5
⎡ ⎤
3. ⎢ 0 0 1 2 0 4⎥
⎣ ⎦
0 0 0 0 1 3
Answer
1. This one is not.
2. This one is.
3. This one is.

Exercise 11.4.18
Row reduce the following matrix to obtain the row-echelon form. Then continue to obtain the reduced row-echelon form.
2 −1 3 −1
⎡ ⎤
⎢1 0 2 1⎥ (11.4.12)
⎣ ⎦
1 −1 1 −2
Exercise 11.4.19
0 0 −1 −1
⎡ ⎤
⎢1 1 1 0⎥ (11.4.13)
⎣ ⎦
1 1 0 −1
Exercise 11.4.20
3 −6 −7 −8
⎡ ⎤
⎢1 −2 −2 −2 ⎥ (11.4.14)
⎣ ⎦
1 −2 −3 −4
Exercise 11.4.21
2 4 5 15
⎡ ⎤
⎢1 2 3 9⎥ (11.4.15)
⎣ ⎦
1 2 2 6
Exercise 11.4.22
4 −1 7 10
⎡ ⎤
⎢1 0 3 3⎥ (11.4.16)
⎣ ⎦
1 −1 −2 1
Exercise 11.4.23
3 5 −4 2
⎡ ⎤
⎢1 2 −1 1⎥ (11.4.17)
⎣ ⎦
1 1 −2 0
Exercise 11.4.24

−2 3 −8 7
⎡ ⎤
⎢ 1 −2 5 −5 ⎥ (11.4.18)
⎣ ⎦
1 −3 7 −8
Exercise 11.4.25
Find the solution of the system whose augmented matrix is
⎡ 1 2 0 2 ⎤
⎢ 1 3 4 2 ⎥ (11.4.19)
⎢ ⎥
⎣ 1 0 2 1 ⎦
Exercise 11.4.26
⎡ 1 2 0 2 ⎤
⎢ 2 0 1 1 ⎥ (11.4.20)
⎢ ⎥
⎣ 3 2 1 3 ⎦
Answer
1 1
⎡ 1 0 \vspace0.05in \vspace0.05in ⎤
2 2
The reduced row-echelon form is ⎢

⎢ 0 1 −\vspace0.05in
1
\vspace0.05in
3 ⎥
⎥. Therefore, the solution is of
⎢ 4 4 ⎥
⎣ 0 0 0 0 ⎦
the form z = t, y = 3
4
+t (
1
4
),x =
1
2
−
1
2
t where t ∈ R .
Exercise 11.4.27
1 1 0 1
[ ] (11.4.21)
1 0 4 2
Answer
1 0 4 2
The reduced row-echelon form is [ ] and so the solution is z = t, y = 4t, x = 2 − 4t.
0 1 −4 −1
Exercise 11.4.28
⎡ 1 0 2 1 1 2 ⎤
⎢ 0 1 0 1 2 1 ⎥
⎢ ⎥ (11.4.22)
⎢ ⎥
⎢ 1 2 0 0 1 3 ⎥
⎣ 1 0 1 0 2 2 ⎦
Answer

⎡ 1 0 0 0 9 3 ⎤
⎢ 0 1 0 0 −4 0 ⎥
⎢
⎥
⎥
and so
⎢ 0 0 1 0 −7 −1 ⎥
⎣ 0 0 0 1 6 1 ⎦
x5 = t, x4 = 1 − 6t, x3 = −1 + 7t, x2 = 4t, x1 = 3 − 9t .
Exercise 11.4.29
⎡ 1 0 2 1 1 2 ⎤
⎢ 0 1 0 1 2 1 ⎥
⎢ ⎥ (11.4.23)
⎢ ⎥
⎢ 0 2 0 0 1 3 ⎥
⎣ 1 −1 2 2 2 0 ⎦
Answer
1 5
⎡ 1 0 2 0 −\vspace0.05in \vspace0.05in ⎤
2 2
⎢ 1 3 ⎥
⎢ 0 1 0 0 \vspace0.05in \vspace0.05in ⎥
⎢
2 2 ⎥
⎥
. Therefore, let
3 1
⎢ 0 0 0 1 \vspace0.05in −\vspace0.05in ⎥
⎢ 2 2 ⎥
⎣ 0 0 0 0 0 0 ⎦
x5 = t, x3 = s. Then the other variables are given by x 4 =−

1
2
−
3
2
t, x2 =
3
2
−t
1
2
, , x1 =
5
2
+
1
2
t − 2s.
Exercise 11.4.30
Find the solution to the system of equations, 7x + 14y + 15z = 22, 2x + 4y + 3z = 5, and 3x + 6y + 10z = 13.
Answer
Solution is: [x = 1 − 2t, z = 1, y = t]
Exercise 11.4.31
Find the solution to the system of equations, 3x − y + 4z = 6, y + 8z = 0, and −2x + y = −4.
Answer
Solution is: [x = 2 − 4t, y = −8t, z = t]
Exercise 11.4.32
Find the solution to the system of equations, 9x − 2y + 4z = −17, 13x − 3y + 6z = −25, and −2x − z = 3.
Answer
Solution is: [x = −1, y = 2, z = −1]
Exercise 11.4.33
Find the solution to the system of equations, 65x + 84y + 16z = 546, 81x + 105y + 20z = 682, and
84x + 110y + 21z = 713.

Answer
Solution is: [x = 2, y = 4, z = 5]
Exercise 11.4.34
Find the solution to the system of equations, 8x + 2y + 3z = −3, 8x + 3y + 3z = −1, and 4x + y + 3z = −9.
Answer
Solution is: [x = 1, y = 2, z = −5]
Exercise 11.4.35
Find the solution to the system of equations, −8x + 2y + 5z = 18, −8x + 3y + 5z = 13, and −4x + y + 5z = 19.
Answer
Solution is: [x = −1, y = −5, z = 4]
Exercise 11.4.36
Find the solution to the system of equations, 3x − y − 2z = 3, y − 4z = 0, and −2x + y = −2.
Answer
Solution is: [x = 2t + 1, y = 4t, z = t]
Exercise 11.4.37
Find the solution to the system of equations, −9x + 15y = 66, −11x + 18y = 79 , −x + y = 4 , and z = 3 .
Answer
Solution is: [x = 1, y = 5, z = 3]
Exercise 11.4.38
Find the solution to the system of equations, −19x + 8y = −108, −71x + 30y = −404, −2x + y = −12,
4x + z = 14.
Answer
Solution is: [x = 4, y = −4, z = −2]
Exercise 11.4.39
Suppose a system of equations has fewer equations than variables and you have found a solution to this system of
equations. Is it possible that your solution is the only one? Explain.
Answer
No. Consider x + y + z = 2 and x + y + z = 1.
Exercise 11.4.40

Suppose a system of linear equations has a 2 × 4 augmented matrix and the last column is a pivot column. Could the
system of linear equations be consistent? Explain.
Answer
No. This would lead to 0 = 1.
Exercise 11.4.41
Suppose the coefficient matrix of a system of n equations with n variables has the property that every column is a pivot
column. Does it follow that the system of equations must have a solution? If so, must the solution be unique? Explain.
Answer
Yes. It has a unique solution.
Exercise 11.4.42
Suppose there is a unique solution to a system of linear equations. What must be true of the pivot columns in the
augmented matrix?
Answer
The last column must not be a pivot column. The remaining columns must each be pivot columns.
Exercise 11.4.43
The steady state temperature, u, of a plate solves Laplace’s equation, Δu = 0. One way to approximate the solution is to
divide the plate into a square mesh and require the temperature at each node to equal the average of the temperature at the
four adjacent nodes. In the following picture, the numbers represent the observed temperature at the indicated nodes. Find
the temperature at the interior nodes, indicated by x, y, z, and w. One of the equations is
z = \vspace0.05in (10 + 0 + w + x) .
1
(1,60) (100,0)
(1,1) (0,30)(1,0)90 (0,60)(1,0)90 (30,0)(0,1)90 (60,0)(0,1)90 (60,0) (30,0) (0,60) (0,30) (30,30) (30,60) (60,30) (60,60)
(60,90) (30,90) (90,30) (90,60) (63,-4)10 (33,-4)10 (-10,60)20 (-10,30)20 (34,34)x (34,64)y (64,34)z (64,64)w (64,90)30
(34,90)30 (94,30)0 (94,60)0
Answer
1
(20 + 30 + w + x) − y = 0
4
1
(y + 30 + 0 + z) − w = 0
You need 4
1
, Solution is: [w = 15, x = 15, y = 20, z = 10] .
(20 + y + z + 10) − x = 0
4
1
(x + w + 0 + 10) − z = 0
4
Exercise 11.4.44
Find the rank of the following matrix.
4 −16 −1 −5
⎡ ⎤
⎢1 −4 0 −1 ⎥ (11.4.24)
⎣ ⎦
1 −4 −1 −2

Exercise 11.4.45
3 6 5 12
⎡ ⎤
⎢1 2 2 5⎥ (11.4.25)
⎣ ⎦
1 2 1 2
Exercise 11.4.46
0 0 −1 0 3
⎡ ⎤
⎢ 1 4 1 0 −8 ⎥
⎢ ⎥ (11.4.26)
⎢ 1 4 0 1 2⎥
⎣ ⎦
−1 −4 0 −1 −2
Exercise 11.4.47
4 −4 3 −9
⎡ ⎤
⎢1 −1 1 −2 ⎥ (11.4.27)
⎣ ⎦
1 −1 0 −3
Exercise 11.4.48
2 0 1 0 1
⎡ ⎤
⎢1 0 1 0 0⎥
⎢ ⎥ (11.4.28)
⎢1 0 0 1 7⎥
⎣ ⎦
1 0 0 1 7
Exercise 11.4.49
4 15 29
⎡ ⎤
⎢1 4 8⎥
⎢ ⎥ (11.4.29)
⎢1 3 5⎥
⎣ ⎦
3 9 15
Exercise 11.4.50
0 0 −1 0 1
⎡ ⎤
⎢ 1 2 3 −2 −18 ⎥
⎢ ⎥ (11.4.30)
⎢ 1 2 2 −1 −11 ⎥
⎣ ⎦
−1 −2 −2 1 11

Exercise 11.4.51
1 −2 0 3 11
⎡ ⎤
⎢1 −2 0 4 15 ⎥
⎢ ⎥ (11.4.31)
⎢1 −2 0 3 11 ⎥
⎣ ⎦
0 0 0 0 0
Exercise 11.4.52
−2 −3 −2
⎡ ⎤
⎢ 1 1 1⎥
⎢ ⎥ (11.4.32)
⎢ 1 0 1⎥
⎣ ⎦
−3 0 −3
Exercise 11.4.53
4 4 20 −1 17
⎡ ⎤
⎢1 1 5 0 5⎥
⎢ ⎥ (11.4.33)
⎢1 1 5 −1 2⎥
⎣ ⎦
3 3 15 −3 6
Exercise 11.4.54
−1 3 4 −3 8
⎡ ⎤
⎢ 1 −3 −4 2 −5 ⎥
⎢ ⎥ (11.4.34)
⎢ 1 −3 −4 1 −2 ⎥
⎣ ⎦
−2 6 8 −2 4
Exercise 11.4.55
Suppose A is an m × n matrix. Explain why the rank of A is always no larger than min (m, n) .
Answer
It is because you cannot have more than min (m, n) nonzero rows in the reduced row-echelon form. Recall that the
number of pivot columns is the same as the number of nonzero rows from the description of this reduced row-echelon
form.
Exercise 11.4.56
State whether each of the following sets of data are possible for the matrix equation AX = B . If possible, describe the
solution set. That is, tell whether there exists a unique solution, no solution or infinitely many solutions. Here, [A|B]
denotes the augmented matrix.
1. A is a 5 × 6 matrix, rank (A) = 4 and rank [A|B] = 4.

5. A is a 4 × 2 matrix, rank (A) = 2 and rank [A|B] = 2 .
Answer
This says B is in the span of four of the columns. Thus the columns are not independent. Infinite solution set.
This surely can’t happen. If you add in another column, the rank does not get smaller.
This says B is in the span of the columns and the columns must be independent. You can’t have the rank equal 4 if you
only have two columns.
This says B is not in the span of the columns. In this case, there is no solution to the system of equations represented
by the augmented matrix.
In this case, there is a unique solution since the columns of A are independent.
Exercise 11.4.57
Consider the system −5x + 2y − z = 0 and −5x − 2y − z = 0. Both equations equal zero and so
−5x + 2y − z = −5x − 2y − z which is equivalent to y = 0. Does it follow that x and z can equal anything? Notice
that when x = 1 , z = −4, and y = 0 are plugged in to the equations, the equations do not equal 0. Why?
Answer
These are not legitimate row operations. They do not preserve the solution set of the system.
Exercise 11.4.58
Balance the following chemical reactions.
1. KN O + H C O → K C O + H N O
3 2 3 2 3 3
2. AgI + N a S → Ag S + N aI
2 2
3. Ba N + H O → Ba(OH ) + N H
3 2 2 2 3
4. C aC l + N a P O → C a (P O ) + N aC l
2 3 4 3 4 2
Exercise 11.4.59
In the section on dimensionless variables it was observed that ρV AB has the units of force. Describe a systematic way
2
to obtain such combinations of the variables which will yield something which has the units of force.
Exercise 11.4.60
Consider the following diagram of four circuits.
The current in amps in the four circuits is denoted by I , I , I , I and it is understood that the motion is in the counter
1 2 3 4
clockwise direction. If I ends up being negative, then it just means the current flows in the clockwise direction.
k
In the above diagram, the top left circuit should give the equation
2 I2 − 2 I1 + 5 I2 − 5 I3 + 3 I2 = 5 (11.4.35)
For the circuit on the lower left, you should have

4 I1 + I1 − I4 + 2 I1 − 2 I2 = −10 (11.4.36)
Write equations for each of the other two circuits and then give a solution to the resulting system of equations.
Answer

The other two equations are
6 I3 − 6 I4 + I3 + I3 + 5 I3 − 5 I2 = −20
2 I4 + 3 I4 + 6 I4 − 6 I3 + I4 − I1 = 0
Then the system is
2 I2 − 2 I1 + 5 I2 − 5 I3 + 3 I2 = 5
4 I1 + I1 − I4 + 2 I1 − 2 I2 = −10
(11.4.37)
6 I3 − 6 I4 + I3 + I3 + 5 I3 − 5 I2 = −20
2 I4 + 3 I4 + 6 I4 − 6 I3 + I4 − I1 = 0
The solution is:

750
I1 = −
373
1421
I2 = −
1119
3061
I3 = −
1119
1718
I4 = −
1119
Exercise 11.4.61
Consider the following diagram of three circuits.
The current in amps in the four circuits is denoted by I , I , I and it is understood that the motion is in the counter
1 2 3
clockwise direction. If I ends up being negative, then it just means the current flows in the clockwise direction.
k
Find I 1, I2 , I3 .
Answer
You have
2 I1 + 5 I1 + 3 I1 − 5 I2 = 10
I2 − I3 + 3 I2 + 7 I2 + 5 I2 − 5 I1 = −12
2 I3 + 4 I3 + 4 I3 + I3 − I2 = 0
Simplifying this yields
10 I1 − 5 I2 = 10
−5 I1 + 16 I2 − I3 = −12
−I2 + 11 I3 = 0
The solution is given by

218 154 14
I1 = , I2 = − , I3 = − (11.4.38)
295 295 295

CHAPTER OVERVIEW
12: MATRICES
12.1: MATRIX ARITHMETIC

You have now solved systems of equations by writing them in terms of an augmented matrix and
then doing row operations on this augmented matrix. It turns out that matrices are important not
only for systems of equations but also in many applications.
12.2: MULTIPLICATION OF MATRICES

The next important matrix operation we will explore is multiplication of matrices. The operation
of matrix multiplication is one of the most important and useful of the matrix operations.
12.3: THE IJTH ENTRY OF A PRODUCT

In previous sections, we used the entries of a matrix to describe the action of matrix addition and scalar multiplication. We can also
study matrix multiplication using the entries of matrices.
12.4: PROPERTIES OF MATRIX MULTIPLICATION

As pointed out above, it is sometimes possible to multiply matrices in one order but not in the other order. However, even if both AB
and BA are defined, they may not be equal.
12.5: THE TRANSPOSE

Another important operation on matrices is that of taking the transpose.
12.6: THE IDENTITY AND INVERSES

There is a special matrix, denoted I , which is called to as the identity matrix
12.7: FINDING THE INVERSE OF A MATRIX

In Example 2.6.1, we were given A^−1 and asked to verify that this matrix was in fact the inverse of A. In this section, we explore
how to find A 1.
−
12.E: EXERCISES
1 9/16/2020
12.1: Matrix Arithmetic
Learning Objectives
1. Perform the matrix operations of matrix addition, scalar multiplication, transposition and matrix multiplication.
Identify when these operations are not defined. Represent these operations in terms of the entries of a matrix.
2. Prove algebraic properties for matrix addition, scalar multiplication, transposition, and matrix multiplication. Apply
these properties to manipulate an algebraic expression involving matrices.
3. Compute the inverse of a matrix using row operations, and prove identities involving matrix inverses.
4. Solve a linear system using matrix algebra.
5. Use multiplication by an elementary matrix to apply row operations.
6. Write a matrix as a product of elementary matrices.
You have now solved systems of equations by writing them in terms of an augmented matrix and then doing row operations on
this augmented matrix. It turns out that matrices are important not only for systems of equations but also in many applications.
Recall that a matrix is a rectangular array of numbers. Several of them are referred to as matrices. For example, here is a
matrix.
1 2 3 4
⎡ ⎤
⎢5 2 8 7⎥ (12.1.1)
⎣ ⎦
6 −9 1 2
Recall that the size or dimension of a matrix is defined as m × n where m is the number of rows and n is the number of
columns. The above matrix is a 3 × 4 matrix because there are three rows and four columns. You can remember the columns
are like columns in a Greek temple. They stand upright while the rows lay flat like rows made by a tractor in a plowed field.
When specifying the size of a matrix, you always list the number of rows before the number of columns.You might remember
that you always list the rows before the columns by using the phrase Rowman Catholic.
Definition 12.1.1 : Square Matrix

A matrix A which has size n × n is called a square matrix . In other words, A is a square matrix if it has the same
number of rows and columns.
There is some notation specific to matrices which we now introduce. We denote the columns of a matrix A by A as follows j
A = [ A1 A2 ⋯ An ] (12.1.2)
Therefore, A is the j
j
th
column of A , when counted from left to right.
The individual elements of the matrix are called entries or components of A . Elements of the matrix are identified according
to their position. The (i, j) -entry of a matrix is the entry in the i row and j column. For example, in the matrix in Equation
th th
12.1.1, 8 is in position (2, 3) (and is called the (2, 3)-entry) because it is in the second row and the third column.
In order to remember which matrix we are speaking of, we will denote the entry in the i row and the j column of matrix A
th th
by a . Then, we can write A in terms of its entries, as A = [a ] . Using this notation on the matrix in Equation 12.1.1,
ij ij
a23 = 8, a = −9, a
32 = 2, etc.
12
There are various operations which are done on matrices of appropriate sizes. Matrices can be added to and subtracted from
other matrices, multiplied by a scalar, and multiplied by other matrices. We will never divide a matrix by another matrix, but
we will see later how matrix inverses play a similar role.
In doing arithmetic with matrices, we often define the action by what happens in terms of the entries (or components) of the
matrices. Before looking at these operations in depth, consider a few general definitions.

Definition 12.1.2 : The Zero Matrix
The m × n zero matrix is the m × n matrix having every entry equal to zero. It is denoted by 0.
One possible zero matrix is shown in the following example.
Example 12.1.1 : The Zero Matrix

0 0 0
The 2 × 3 zero matrix is 0 = [ ] .
0 0 0
Note there is a 2 × 3 zero matrix, a 3 × 4 zero matrix, etc. In fact there is a zero matrix for every size!
Definition 12.1.3 :Equality of Matrices

The m × n zero matrix is the m × n matrix having every entry equal to zero. It is denoted by 0.
Let A and B be two m × n matrices. Then A = B means that for A = [a ij ] and B = [b

ij ] , aij = bij for all 1 ≤i ≤m and
1 ≤ j ≤ n.
In other words, two matrices are equal exactly when they are the same size and the corresponding entries are identical. Thus
0 0
⎡ ⎤
0 0
⎢0 0⎥ ≠[ ] (12.1.3)
⎣ ⎦ 0 0
0 0
because they are different sizes. Also,

0 1 1 0
[ ] ≠[ ] (12.1.4)
3 2 2 3
because, although they are the same size, their corresponding entries are not identical.
In the following section, we explore addition of matrices.
Addition of Matrices
When adding matrices, all matrices in the sum need have the same size. For example,
1 2
⎡ ⎤
⎢3 4⎥ (12.1.5)
⎣ ⎦
5 2
and
−1 4 8
[ ] (12.1.6)
2 8 5
cannot be added, as one has size 3 × 2 while the other has size 2 × 3 .
However, the addition
4 6 3 0 5 0
⎡ ⎤ ⎡ ⎤
⎢ 5 0 4 ⎥+⎢4 −4 14 ⎥ (12.1.7)
⎣ ⎦ ⎣ ⎦
11 −2 3 1 2 6
is possible.
The formal definition is as follows.

Definition 12.1.4 : Addition of Matrices
Let A = [a ij ] and B = [b
ij ] be two m ×n matrices. Then A+B = C where C is the m ×n matrix C = [ cij ] defined
by
cij = aij + bij (12.1.8)
This definition tells us that when adding matrices, we simply add corresponding entries of the matrices. This is demonstrated
in the next example.
Example 12.1.2 : Addition of Matrices of Same Size

Add the following matrices, if possible.
1 2 3 5 2 3
A =[ ],B = [ ] (12.1.9)
1 0 4 −6 2 1
Solution
Notice that both A and B are of size 2 × 3 . Since A and B are of the same size, the addition is possible. Using Definition
12.1.4, the addition is done as follows.
1 2 3 5 2 3 1 +5 2 +2 3 +3 6 4 6
A+B = [ ]+[ ] = [ ] =[ ] (12.1.10)
1 0 4 −6 2 1 1 + −6 0 +2 4 +1 −5 2 5
Addition of matrices obeys very much the same properties as normal addition with numbers. Note that when we write for
example A + B then we assume that both matrices are of equal size so that the operation is indeed possible.
Theorem 12.1.1
Properties of Matrix Additionpropertiesofaddition Let A, B and C be matrices. Then, the following properties hold.
Commutative Law of Addition
A+B = B+A (12.1.11)
Associative Law of Addition

(A + B) + C = A + (B + C ) (12.1.12)
Existence of an Additive Identity

There exists a zero matrix 0 such that
(12.1.13)
A+0 = A
Existence of an Additive Inverse

There exists a matrix −A such that
(12.1.14)
A + (−A) = 0
Proof
Consider the Commutative Law of Addition given in [mat1]. Let A, B, C , and D be matrices such that A + B = C
and B + A = D. We want to show that D = C . To do so, we will use the definition of matrix addition given in
Definition 12.1.4. Now,
cij = aij + bij = bij + aij = dij (12.1.15)
Therefore, C = D because the ij entries are the same for all i and j . Note that the conclusion follows from the
th
commutative law of addition of numbers, which says that if a and b are two numbers, then a + b = b + a . The proof
of the other results are similar, and are left as an exercise.

We call the zero matrix in [mat3] the additive identity. Similarly, we call the matrix −A in [mat4] the additive inverse. −A
is defined to equal (−1) A = [−a ]. In other words, every entry of A is multiplied by −1.
ij
In the next section we will study scalar multiplication in more depth to understand what is meant by (−1) A.
Scalar Multiplication of Matrices

Recall that we use the word scalar when referring to numbers. Therefore, scalar multiplication of a matrix is the
multiplication of a matrix by a number. To illustrate this concept, consider the following example in which a matrix is
multiplied by the scalar 3.
1 2 3 4 3 6 9 12
⎡ ⎤ ⎡ ⎤
3⎢5 2 8 7 ⎥ = ⎢ 15 6 24 21 ⎥ (12.1.16)
⎣ ⎦ ⎣ ⎦
6 −9 1 2 18 −27 3 6
The new matrix is obtained by multiplying every entry of the original matrix by the given scalar.
The formal definition of scalar multiplication is as follows.
Definition 12.1.5 : Scalar Multiplication of Matrices

If A = [a ij ] and k is a scalar, then kA = [ka ij ] .
Example 12.1.3 : Effect of Multiplication by a Scalar

Find the result of multiplying the following matrix A by 7.
2 0
A =[ ] (12.1.17)
1 −4
Solution
By Definition [def:scalarmultofmatrices], we multiply each element of A by 7. Therefore,
2 0 7(2) 7(0) 14 0
7A = 7 [ ] =[ ] =[ ] (12.1.18)
1 −4 7(1) 7(−4) 7 −28
Similarly to addition of matrices, there are several properties of scalar multiplication which hold.
Theorem 12.1.2 : Properties of Scalar Multiplication

Let A, B be matrices, and k, p be scalars. Then, the following properties hold.
Distributive Law over Matrix Addition
k (A + B) = kA + kB (12.1.19)
Distributive Law over Scalar Addition

(k + p) A = kA + pA (12.1.20)
Associative Law for Scalar Multiplication

k (pA) = (kp) A (12.1.21)
Rule for Multiplication by 1

1A = A (12.1.22)

Proof
The proof of this proposition is similar to the proof of Proposition [prop:propertiesofaddition] and is left an exercise to
the reader.

12.2: Multiplication of Matrices
The next important matrix operation we will explore is multiplication of matrices. The operation of matrix multiplication is
one of the most important and useful of the matrix operations. Throughout this section, we will also demonstrate how matrix
multiplication relates to linear systems of equations.
First, we provide a formal definition of row and column vectors.
Definition 12.2.1 : Row and Column Vectors

Matrices of size n × 1 or 1 × n are called vectors. If X is such a matrix, then we write x to denote the entry of X in the
i
i row of a column matrix, or the i column of a row matrix.

th th
The n × 1 matrix
x1
⎡ ⎤
X =⎢
⎢
⎥
⎥ (12.2.1)
⋮
⎣ ⎦
xn
is called a column vector. The 1 × n matrix

X = [ x1 ⋯ xn ] (12.2.2)
is called a row vector.
We may simply use the term vector throughout this text to refer to either a column or row vector. If we do so, the context will
make it clear which we are referring to.
In this chapter, we will again use the notion of linear combination of vectors as in Definition [def:linearcombination]. In this
context, a linear combination is a sum consisting of vectors multiplied by scalars. For example,
50 1 2 3
[ ] =7[ ]+8 [ ]+9 [ ] (12.2.3)
122 4 5 6
is a linear combination of three vectors.

It turns out that we can express any system of linear equations as a linear combination of vectors. In fact, the vectors that we
will use are just the columns of the corresponding augmented matrix!
Definition 12.2.2 : The Vector Form of a System of Linear Equations

Suppose we have a system of equations given by
a11 x1 + ⋯ + a1n xn = b1
(12.2.4)
⋮
We can express this system in vector form which is as follows:

a11 a12 a1n b1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢ a21 ⎥ ⎢ a22 ⎥ ⎢ a2n ⎥ ⎢ b2 ⎥

x1 ⎢ ⎥+x ⎢ ⎥+⋯ +x ⎢ ⎥ =⎢ ⎥ (12.2.5)
⎢ ⎥ 2 ⎢ ⎥ n ⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⎥ ⎢ ⋮ ⎥ ⎢ ⋮ ⎥ ⎢ ⋮ ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦
am1 am2 amn bm
Notice that each vector used here is one column from the corresponding augmented matrix. There is one vector for each
variable in the system, along with the constant vector.

The first important form of matrix multiplication is multiplying a matrix by a vector. Consider the product given by
7
⎡ ⎤
1 2 3
[ ]⎢8 ⎥ (12.2.6)
4 5 6
⎣ ⎦
9
We will soon see that this equals

1 2 3 50
7[ ]+8 [ ]+9 [ ] =[ ] (12.2.7)
4 5 6 122
In general terms,
x1
⎡ ⎤
a11 a12 a13 a11 a12 a13
[ ] ⎢ x2 ⎥ = x1 [ ] + x2 [ ] + x3 [ ]
a21 a22 a23 a21 a22 a23
⎣ ⎦
x3
a11 x1 + a12 x2 + a13 x3

= [ ]
a21 x1 + a22 x2 + a23 x3
Thus you take x times the first column, add to x times the second column, and finally x times the third column. The above
1 2 3
sum is a linear combination of the columns of the matrix. When you multiply a matrix on the left by a vector on the right, the
numbers making up the vector are just the scalars to be used in the linear combination of the columns as illustrated above.
Here is the formal definition of how to multiply an m × n matrix by an n × 1 column vector.
Definition 12.2.3 : Multiplication of Vector by Matrix

Let A = [a ij ] be an m × n matrix and let X be an n × 1 matrix given by
x1
⎡ ⎤
A = [ A1 ⋯ An ] , X = ⎢ ⎥ (12.2.8)
⎢ ⋮⎥
⎣ ⎦
xn
Then the product AX is the m × 1 column vector which equals the following linear combination of the columns of A :
n
x1 A1 + x2 A2 + ⋯ + xn An = ∑ xj Aj (12.2.9)
j=1
If we write the columns of A in terms of their entries, they are of the form
a1j
⎡ ⎤
⎢ a2j ⎥
⎢ ⎥
Aj =⎢ ⎥ (12.2.10)
⎢ ⎥
⎢ ⋮ ⎥
⎣ ⎦
amj
Then, we can write the product AX as

a11 a12 a1n
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢ a21 ⎥ ⎢ a22 ⎥ ⎢ a2n ⎥

AX = x1 ⎢ ⎥+x ⎢ ⎥+⋯ +x ⎢ ⎥ (12.2.11)
2 n
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⎥ ⎢ ⋮ ⎥ ⎢ ⋮ ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
am1 am2 amn
Note that multiplication of an m × n matrix and an n × 1 vector produces an m × 1 vector.

Here is an example.

Example 12.2.1 : A Vector Multiplied by a Matrix
Compute the product AX for
1
⎡ ⎤
1 2 1 3
⎡ ⎤
⎢2 ⎥
A = ⎢0 2 1 −2 ⎥ , X = ⎢ ⎥ (12.2.12)
⎢0 ⎥
⎣ ⎦
2 1 4 1 ⎣ ⎦
1
Solution
We will use Definition [def:multiplicationvectormatrix] to compute the product. Therefore, we compute the product AX
as follows.
1 2 1 3
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
1 ⎢ 0 ⎥ + 2 ⎢ 2 ⎥ + 0 ⎢ 1 ⎥ + 1 ⎢ −2 ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 1 4 1
1 4 0 3
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
= ⎢ 0 ⎥ + ⎢ 4 ⎥ + ⎢ 0 ⎥ + ⎢ −2 ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 2 0 1
8
⎡ ⎤
= ⎢2 ⎥
⎣ ⎦
5
Using the above operation, we can also write a system of linear equations in matrix form. In this form, we express the system
as a matrix multiplied by a vector. Consider the following definition.
Definition 12.2.4 : The Matrix Form of a System of Linear Equations

Suppose we have a system of equations given by
a11 x1 + ⋯ + a1n xn = b1
a21 x1 + ⋯ + a2n xn = b2
(12.2.13)
⋮
Then we can express this system in matrix form as follows.

a11 a12 ⋯ a1n
⎡ ⎤ ⎡ x1 ⎤ ⎡
b1
⎤
⎢ a21 a22 ⋯ a2n ⎥ ⎢ x2 ⎥ ⎢ b2 ⎥
⎢ ⎥ ⎢ ⎥ =⎢ ⎥
⎢ ⎥ (12.2.14)
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⋮ ⋱ ⋮ ⎥ ⎢ ⋮ ⎥ ⎢ ⋮ ⎥
⎣ ⎦⎣ ⎦ ⎣ ⎦
am1 am2 ⋯ amn xn bm
The expression AX = B is also known as the Matrix Form of the corresponding system of linear equations. The matrix A is
simply the coefficient matrix of the system, the vector X is the column vector constructed from the variables of the system,
and finally the vector B is the column vector constructed from the constants of the system. It is important to note that any
system of linear equations can be written in this form.
Notice that if we write a homogeneous system of equations in matrix form, it would have the form AX = 0 , for the zero
vector 0.
You can see from this definition that a vector

x1
⎡ ⎤
⎢ x2 ⎥
X =⎢ ⎥ (12.2.15)
⎢ ⎥
⎢ ⋮ ⎥
⎣ ⎦
xn
will satisfy the equation AX = B only when the entries x 1, x2 , ⋯ , xn of the vector X are solutions to the original system.
Now that we have examined how to multiply a matrix by a vector, we wish to consider the case where we multiply two
matrices of more general sizes, although these sizes still need to be appropriate as we will see. For example, in Example
[exa:vectormultbymatrix], we multiplied a 3 × 4 matrix by a 4 × 1 vector. We want to investigate how to multiply other sizes
of matrices.
We have not yet given any conditions on when matrix multiplication is possible! For matrices A and B , in order to form the
product AB, the number of columns of A must equal the number of rows of B. Consider a product AB where A has size
m × n and B has size n × p . Then, the product in terms of size of matrices is given by
these must match!
ˆ
(m × n) (n × p ) = m ×p (12.2.16)
Note the two outside numbers give the size of the product. One of the most important rules regarding matrix multiplication is
the following. If the two middle numbers don’t match, you can’t multiply the matrices!
When the number of columns of A equals the number of rows of B the two matrices are said to be conformable and the
product AB is obtained as follows.
Definition 12.2.4 : Multiplication of Two Matrices

Let A be an m × n matrix and let B be an n × p matrix of the form
B = [ B1 ⋯ Bp ] (12.2.17)
where B 1, . . . , Bp are the n × 1 columns of B . Then the m × p matrix AB is defined as follows:
AB = A [ B1 ⋯ Bp ] = [(AB)1 ⋯ (AB)p ] (12.2.18)
where (AB) is an m × 1 matrix or column vector which gives the k

k
th
column of AB.
Example 12.2.2 : Multiplying Two Matrices

Find AB if possible.
1 2 0
⎡ ⎤
1 2 1
A =[ ],B = ⎢ 0 3 1⎥ (12.2.19)
0 2 1 ⎣ ⎦
−2 1 1
Solution
The first thing you need to verify when calculating a product is whether the multiplication is possible. The first matrix has
size 2 × 3 and the second matrix has size 3 × 3 . The inside numbers are equal, so A and B are conformable matrices.
According to the above discussion AB will be a 2 × 3 matrix. Definition [def:multiplicationoftwomatrices] gives us a
way to calculate each column of AB, as follows.
First column Second column Third column
⎡    ⎤
⎢ 1 2 0 ⎥
⎢ 1 ⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎥
2 1 1 2 1 1 2 1
⎢[ ]⎢ 0 ⎥, [ ] ⎢ 3 ⎥, [ ] ⎢ 1 ⎥⎥ (12.2.20)
⎢ ⎥
⎢ 0 2 1 ⎣ ⎦ 0 2 1 ⎣ ⎦ 0 2 1 ⎣ ⎦⎥
⎢ −2 1 1 ⎥
⎣ ⎦

You know how to multiply a matrix times a vector, using Definition [def:multiplicationvectormatrix] for each of the three
columns. Thus
1 2 0
⎡ ⎤
1 2 1 −1 9 3
[ ]⎢ 0 3 1 ⎥ = [ ] (12.2.21)
0 2 1 −2 7 3
⎣ ⎦
−2 1 1
Since vectors are simply n × 1 or 1 × m matrices, we can also multiply a vector by another vector.
Example 12.2.3 : Vector Times Vector Multiplication

1
⎡ ⎤
Multiply if possible ⎢ 2 ⎥ [ 1 2 1 0].
⎣ ⎦
1
Solution
In this case we are multiplying a matrix of size 3 × 1 by a matrix of size 1 × 4. The inside numbers match so the product
is defined. Note that the product will be a matrix of size 3 × 4 . Using Definition [def:multiplicationoftwomatrices], we
can compute this product as follows
First column Second column Third column Fourth column
⎡     ⎤
1 ⎢ 1 1 1 1 ⎥
⎡ ⎤ ⎢⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎥
⎢2 ⎥[1 2 1 0 ] = ⎢⎢ 2 ⎥ [ 1 ], ⎢2 ⎥[2 ] , ⎢2 ⎥[1 ] , ⎢2 ⎥[0 ] ⎥ (12.2.22)
⎢ ⎥
⎣ ⎦ ⎢⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎥
1 ⎢ 1 1 1 1 ⎥
⎣ ⎦
You can use Definition [def:multiplicationvectormatrix] to verify that this product is

1 2 1 0
⎡ ⎤
⎢2 4 2 0⎥ (12.2.23)
⎣ ⎦
1 2 1 0
Example 12.2.4 : A Multiplication Which is Not Defined

Find BA if possible.
1 2 0
⎡ ⎤
1 2 1
B =⎢ 0 3 1⎥,A = [ ] (12.2.24)
0 2 1
⎣ ⎦
−2 1 1
Solution
First check if it is possible. This product is of the form (3 × 3) (2 × 3) . The inside numbers do not match and so you
can’t do this multiplication.
In this case, we say that the multiplication is not defined. Notice that these are the same matrices which we used in Example
[exa:multiplicationoftwomatrices]. In this example, we tried to calculate BA instead of AB. This demonstrates another
property of matrix multiplication. While the product AB maybe be defined, we cannot assume that the product BA will be
possible. Therefore, it is important to always check that the product is defined before carrying out any calculations.
Earlier, we defined the zero matrix 0 to be the matrix (of appropriate size) containing zeros in all entries. Consider the
following example for multiplication by the zero matrix.
Example 12.2.5 : Multiplication by the Zero Matrix

Compute the product A0 for the matrix
1 2
A =[ ] (12.2.25)
3 4
and the 2 × 2 zero matrix given by

0 0
0 =[ ] (12.2.26)
0 0
Solution
In this product, we compute
1 2 0 0 0 0
[ ][ ] =[ ] (12.2.27)
3 4 0 0 0 0
Hence, A0 = 0 .
0
Notice that we could also multiply A by the 2 ×1 zero vector given by [ ] . The result would be the 2 ×1 zero vector.
0
Therefore, it is always the case that A0 = 0 , for an appropriately sized zero matrix or vector.

12.3: The ijth Entry of a Product
In previous sections, we used the entries of a matrix to describe the action of matrix addition and scalar multiplication. We can
also study matrix multiplication using the entries of matrices.
What is the ij entry of AB? It is the entry in the i
th th
row and the j th
column of the product AB.
Now if A is m × n and B is n × p , then we know that the product AB has the form
a11 a12 ⋯ a1n b11 b12 ⋯ b1j ⋯ b1p

⎡ ⎤⎡ ⎤
⎢ a21 a22 ⋯ a2n ⎥ ⎢ b21 b22 ⋯ b2j ⋯ b2p ⎥

⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ⎥ (12.3.1)
⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⋮ ⋱ ⋮ ⎥ ⎢ ⋮ ⋮ ⋮ ⋮ ⎥
⎣ ⎦⎣ ⎦
am1 am2 ⋯ amn bn1 bn2 ⋯ bnj ⋯ bnp
The j th
column of AB is of the form
a11 a12 ⋯ a1n b1j

⎡ ⎤⎡ ⎤
⎢ a21 a22 ⋯ a2n ⎥ ⎢ b2j ⎥

⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ⎥ (12.3.2)
⎢ ⎥ ⎢ ⎥
⎢ ⋮ ⋮ ⋱ ⋮ ⎥ ⎢ ⋮ ⎥
⎣ ⎦⎣ ⎦
am1 am2 ⋯ amn bnj
which is an m × 1 column vector. It is calculated by

a11 a12 a1n
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢ a21 ⎥ ⎢ a22 ⎥ ⎢ a2n ⎥

b1j ⎢ ⎥+b ⎢ ⎥+⋯ +b ⎢ ⎥ (12.3.3)
⎢ ⎥ 2j ⎢ ⎥ nj ⎢ ⎥
⎢ ⋮ ⎥ ⎢ ⋮ ⎥ ⎢ ⋮ ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
am1 am2 amn
Therefore, the ij entry is the entry in row i of this vector. This is computed by
th
ai1 b1j + ai2 b2j + ⋯ + ain bnj = ∑ aik bkj (12.3.4)
k=1
The following is the formal definition for the ij entry of a product of matrices.
th
Definition 12.3.1 : The ij th

Entry of a Product
Let A = [a ] be an m × n matrix and let B = [b
ij ij ] be an n × p matrix. Then AB is an m × p matrix and the (i, j) -entry
of AB is defined as
n
(AB)ij = ∑ aik bkj (12.3.5)
k=1
Another way to write this is

b1j
⎡ ⎤
⎢ b2j ⎥
⎢ ⎥
(AB)ij = [ ai1 ai2 ⋯ ain ]⎢ ⎥ = ai1 b1j + ai2 b2j + ⋯ + ain bnj (12.3.6)
⎢ ⎥
⎢ ⋮ ⎥
⎣ ⎦
bnj
In other words, to find the (i, j) -entry of the product AB, or (AB) , you multiply the ij i
th
row of A, on the left by the j
th
column of B . To express AB in terms of its entries, we write AB = [(AB) ] . ij

Example 12.3.1 : The Entries of a Product
Compute AB if possible. If it is, find the (3, 2)-entry of AB using Definition [def:ijentryofproduct].
1 2
⎡ ⎤
2 3 1
A =⎢3 1⎥,B = [ ] (12.3.7)
7 6 2
⎣ ⎦
2 6
Solution
First check if the product is possible. It is of the form (3 × 2) (2 × 3) and since the inside numbers match, it is possible to
do the multiplication. The result should be a 3 × 3 matrix. We can first compute AB:
1 2 1 2 1 2
⎡⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎤
2 3 1
⎢⎢ 3 1⎥[ ],⎢3 1⎥[ ],⎢3 1⎥[ ]⎥ (12.3.8)
⎣⎣ ⎦ 7 ⎣ ⎦ 6 ⎣ ⎦ 2 ⎦
2 6 2 6 2 6
where the commas separate the columns in the resulting product. Thus the above product equals
16 15 5
⎡ ⎤
⎢ 13 15 5⎥ (12.3.9)
⎣ ⎦
46 42 14
which is a 3 × 3 matrix as desired. Thus, the (3, 2)-entry equals 42.

Now using Definition [def:ijentryofproduct], we can find that the (3, 2)-entry equals
2
∑ a3k bk2 = a31 b12 + a32 b22
k=1
= 2 × 3 + 6 × 6 = 42
Consulting our result for AB above, this is correct!

You may wish to use this method to verify that the rest of the entries in AB are correct.
Here is another example.
Example 12.3.1 : Finding the Entries of a Product

Determine if the product AB is defined. If it is, find the (2, 1)-entry of the product.
2 3 1 1 2
⎡ ⎤ ⎡ ⎤
A = ⎢7 6 2 ⎥,B = ⎢3 1⎥ (12.3.10)
⎣ ⎦ ⎣ ⎦
0 0 0 2 6
Solution
This product is of the form (3 × 3) (3 × 2) . The middle numbers match so the matrices are conformable and it is possible
to compute the product.
We want to find the (2, 1)-entry of AB, that is, the entry in the second row and first column of the product. We will use
Definition [def:ijentryofproduct], which states
n
(AB)ij = ∑ aik bkj (12.3.11)
k=1
In this case, n = 3 , i = 2 and j = 1 . Hence the (2, 1)-entry is found by computing

3
b11
⎡ ⎤
(AB)21 = ∑ a2k bk1 = [ a21 a22 a23 ] ⎢ b21 ⎥ (12.3.12)
k=1 ⎣ ⎦
b31
Substituting in the appropriate values, this product becomes
b11 1
⎡ ⎤ ⎡ ⎤
[ a21 a22 a23 ] ⎢ b21 ⎥ = [ 7 6 2 ] ⎢ 3 ⎥ = 1 × 7 + 3 × 6 + 2 × 2 = 29 (12.3.13)

⎣ ⎦ ⎣ ⎦
b31 2
Hence, (AB) 21 = 29 .
You should take a moment to find a few other entries of AB . You can multiply the matrices to check that your answers
are correct. The product AB is given by
13 13
⎡ ⎤
AB = ⎢ 29 32 ⎥ (12.3.14)
⎣ ⎦
0 0

12.4: Properties of Matrix Multiplication
As pointed out above, it is sometimes possible to multiply matrices in one order but not in the other order. However, even if
both AB and BA are defined, they may not be equal.
Example 12.4.1 : Matrix Multiplication is Not Commutative

1 2 0 1
Compare the products AB and BA, for matrices A = [ ],B = [ ]
3 4 1 0
Solution
First, notice that A and B are both of size 2 ×2 . Therefore, both products AB and BA are defined. The first product,
AB is
1 2 0 1 2 1
AB = [ ][ ] =[ ]
3 4 1 0 4 3
The second product, BA is

0 1 1 2 3 4
[ ][ ] =[ ]
1 0 3 4 1 2
Therefore, AB ≠ BA .
This example illustrates that you cannot assume AB = BA even when multiplication is defined in both orders. If for some
matrices A and B it is true that AB = BA , then we say that A and B commute. This is one important property of matrix
multiplication.
The following are other important properties of matrix multiplication. Notice that these properties hold only when the size of
matrices are such that the products are defined.
Proposition 12.4.1 : Properties of Matrix Multiplication

The following hold for matrices A, B, and C and for scalars r and s ,
A (rB + sC ) = r (AB) + s (AC ) (12.4.1)
(B + C ) A = BA + C A (12.4.2)
A (BC ) = (AB) C (12.4.3)
Proof
First we will prove 12.4.1. We will use Definition [def:ijentryofproduct] and prove this statement using the ij entries th
of a matrix. Therefore,
(A (rB + sC )) = ∑ aik (rB + sC )
ij kj
= ∑ aik (rbkj + sckj )
= r ∑ aik bkj + s ∑ aik ckj
k k
= r(AB) + s(AC )
ij ij
= (r (AB) + s (AC )) ij
Thus

A (rB + sC ) = r(AB) + s(AC )
as claimed.
The proof of Equation 12.4.2follows the same pattern and is left as an exercise.
Statement Equation 12.4.3is the associative law of multiplication. Using Definition [def:ijentryofproduct],
(A (BC )) = ∑ aik (BC )

ij kj
= ∑ aik ∑ bkl clj
k l
= ∑ (AB) clj = ((AB) C ) .

il ij
This proves [matrixproperties3].

12.5: The Transpose
Another important operation on matrices is that of taking the transpose. For a matrix A , we denote the transpose of A by A . T
Before formally defining the transpose, we explore this operation on the following matrix.
T
1 4
⎡ ⎤
1 3 2
⎢3 1⎥ = [ ]
⎣ ⎦ 4 1 6
2 6
What happened? The first column became the first row and the second column became the second row. Thus the 3 × 2 matrix
became a 2 × 3 matrix. The number 4 was in the first row and the second column and it ended up in the second row and first
column.
The definition of the transpose is as follows.
Definition 12.5.1 : The Transpose of a Matrix

Let A be an m × n matrix. Then A , the transpose of A , denotes the n × m matrix given by
T
T T
A = [ aij ] = [ aji ] (12.5.1)
The (i, j) -entry of A becomes the (j, i) -entry of A . T
Example 12.5.1 : The Transpose of a Matrix

Calculate A for the following matrix
T
1 2 −6
A =[ ]
3 5 4
Solution
By Definition 12.5.1, we know that for A = [a ] , A
ij
T
= [ aji ] . In other words, we switch the row and column location of
each entry. The (1, 2)-entry becomes the (2, 1)-entry.
Thus,
1 3
⎡ ⎤
T
A =⎢ 2 5⎥
⎣ ⎦
−6 4
Notice that A is a 2 × 3 matrix, while A is a 3 × 2 matrix.

T
The transpose of a matrix has the following important properties.
Lemma 12.5.1 : Properties of the Transpose of a Matrix

Let A be an m × n matrix, B an n × p matrix, and r and s scalars. Then
1. (A
T
)
T
=A (12.5.2)
2. (AB)
T
=B
T
A
T
(12.5.3)
3. (rA + sB)
T
= rA
T
+ sB
T
(12.5.4)
Proof
Ken Kuttler 9/11/2020 12.5.1 https://math.libretexts.org/@go/page/44319

First we prove 12.5.3. From Definition [def:matrixtranspose],
T T
(AB) = [(AB)ij ] = [(AB)ji ] = ∑ ajk bki = ∑ bki ajk
k k
T T T T T T
= ∑ [ bik ] [ akj ] = [ bij ] [ aij ] =B A
The proof of Formula 12.5.4is left as an exercise.
The transpose of a matrix is related to other important topics. Consider the following definition.
Definition 12.5.2 : Symmetric and Skew Symmetric Matrices

An n × n matrix A is said to be symmetric if A = A T
. It is said to be skew symmetric if A = −A T
.
We will explore these definitions in the following examples.
Example 12.5.2 : Symmetric Matrices

Let
2 1 3
⎡ ⎤
A = ⎢1 5 −3 ⎥
⎣ ⎦
3 −3 7
Use Definition 12.5.2 to show that A is symmetric.

Solution
By Definition 12.5.2, we need to show that A = A . Now, using Definition 12.5.1,
T
2 1 3
⎡ ⎤
T
A = ⎢1 5 −3 ⎥
⎣ ⎦
3 −3 7
Hence, A = A , so A is symmetric.
T
Example 12.5.3 : A Skew Symmetric Matrix

Let
0 1 3
⎡ ⎤
A = ⎢ −1 0 2⎥
⎣ ⎦
−3 −2 0
Show that A is skew symmetric.

Solution
By Definition 12.5.1,
0 −1 −3
⎡ ⎤
T
A = ⎢1 0 −2 ⎥
⎣ ⎦
3 2 0
You can see that each entry of A

T
is equal to −1 times the same entry of A . Hence, A
T
= −A and so by Definition
12.5.2, A is skew symmetric.

12.6: The Identity and Inverses
There is a special matrix, denoted I , which is called to as the identity matrix. The identity matrix is always a square matrix,
and it has the property that there are ones down the main diagonal and zeroes elsewhere. Here are some identity matrices of
various sizes.
1 0 0 0
⎡ ⎤
1 0 0
⎡ ⎤
1 0 ⎢0 1 0 0⎥
[1] , [ ],⎢0 1 0⎥,⎢ ⎥ (12.6.1)
0 1 ⎢0 0 1 0⎥
⎣ ⎦
0 0 1 ⎣ ⎦
0 0 0 1
The first is the 1 × 1 identity matrix, the second is the 2 × 2 identity matrix, and so on. By extension, you can likely see what
the n × n identity matrix would be. When it is necessary to distinguish which size of identity matrix is being discussed, we
will use the notation I for the n × n identity matrix.
n
The identity matrix is so important that there is a special symbol to denote the ij entry of the identity matrix. This symbol is
th
given by I = δ where δ is the Kronecker symbol defined by

ij ij ij
1 if i = j
δij = { (12.6.2)
0 if i ≠ j
In is called the identity matrix because it is a multiplicative identity in the following sense.
Lemma 12.6.1 : Multiplication by the Identity Matrix

Suppose A is an m × n matrix and I is the n × n identity matrix. Then AI
n n = A. If I
m is the m ×m identity matrix,
it also follows that I A = A.
m
Proof
The (i, j)-entry of AI is given by:
n
∑ aik δkj = aij (12.6.3)
and so AI n = A. The other case is left as an exercise for you.
We now define the matrix operation which in some ways plays the role of division.
Definition 12.6.1 : The Inverse of a Matrix

A square n × n matrix A is said to have an inverse A −1
if and only if
−1 −1
AA =A A = In (12.6.4)
In this case, the matrix A is called invertible.
Such a matrix A will have the same size as the matrix A . It is very important to observe that the inverse of a matrix, if it
−1
exists, is unique. Another way to think of this is that if it acts like the inverse, then it is the inverse.
Theorem 12.6.1 : Uniqueness of Inverse

Suppose A is an n × n matrix such that an inverse A −1
exists. Then there is only one such inverse matrix. That is, given
any matrix B such that AB = BA = I , B = A . −1
Proof

In this proof, it is assumed that I is the n × n identity matrix. Let A, B be n × n matrices such that A −1
exists and
AB = BA = I . We want to show that A = B . Now using properties we have seen, we get:
−1
−1 −1 −1 −1
A =A I =A (AB) = (A A) B = I B = B (12.6.5)
Hence, A −1
=B which tells us that the inverse is unique.
The next example demonstrates how to check the inverse of a matrix.
Example 12.6.1 : Verifying the Inverse of a Matrix

1 1 2 −1
Let A = [ ]. Show [ ] is the inverse of A.
1 2 −1 1
Solution
To check this, multiply
1 1 2 −1 1 0
[ ][ ] = [ ] =I (12.6.6)
1 2 −1 1 0 1
and
2 −1 1 1 1 0
[ ][ ] = [ ] =I (12.6.7)
−1 1 1 2 0 1
showing that this matrix is indeed the inverse of A.
Unlike ordinary multiplication of numbers, it can happen that A ≠0 but A may fail to have an inverse. This is illustrated in
the following example.
Example 12.6.2 : A Nonzero Matrix With No Inverse

1 1
Let A = [ ]. Show that A does not have an inverse.
1 1
Solution
One might think A would have an inverse because it does not equal zero. However, note that
1 1 −1 0
[ ][ ] =[ ] (12.6.8)
1 1 1 0
If A −1
existed, we would have the following
0 −1
0
[ ] = A ([ ])
0 0
−1
−1
= A (A [ ])
1
−1
−1
= (A A) [ ]
1
−1
= I [ ]
1
−1
= [ ]
1
This says that

0 −1
[ ] =[ ] (12.6.9)
0 1
which is impossible! Therefore, A does not have an inverse.
In the next section, we will explore how to find the inverse of a matrix, if it exists.

12.7: Finding the Inverse of a Matrix
In Example [exa:verifyinginverse], we were given A −1
and asked to verify that this matrix was in fact the inverse of A . In this
section, we explore how to find A .
−1
Let
1 1
A =[ ] (12.7.1)
1 2
x z
as in Example [exa:verifyinginverse]. In order to find A −1
, we need to find a matrix [ ] such that
y w
1 1 x z 1 0
[ ][ ] =[ ] (12.7.2)
1 2 y w 0 1
We can multiply these two matrices, and see that in order for this equation to be true, we must find the solution to the systems
of equations,
x +y = 1
(12.7.3)
x + 2y = 0
and
z+w = 0
(12.7.4)
z + 2w = 1
Writing the augmented matrix for these two systems gives
1 1 1
[ ] (12.7.5)
1 2 0
for the first system and
1 1 0
[ ] (12.7.6)
1 2 1
for the second.

Let’s solve the first system. Take −1 times the first row and add to the second to get
1 1 1
[ ] (12.7.7)
0 1 −1
Now take −1 times the second row and add to the first to get
1 0 2
[ ] (12.7.8)
0 1 −1
Writing in terms of variables, this says x = 2 and y = −1.

Now solve the second system, [inverse2a] to find z and w. You will find that z = −1 and w = 1 .
If we take the values found for x, y, z, and w and put them into our inverse matrix, we see that the inverse is
−1
x z 2 −1
A =[ ] =[ ] (12.7.9)
y w −1 1
After taking the time to solve the second system, you may have noticed that exactly the same row operations were used to
solve both systems. In each case, the end result was something of the form [I |X] where I is the identity and X gave a column

of the inverse. In the above,
x
[ ] (12.7.10)
y
the first column of the inverse was obtained by solving the first system and then the second column
z
[ ] (12.7.11)
w
To simplify this procedure, we could have solved both systems at once! To do so, we could have written
1 1 1 0
[ ] (12.7.12)
1 2 0 1
and row reduced until we obtained
1 0 2 −1
[ ] (12.7.13)
0 1 −1 1
and read off the inverse as the 2 × 2 matrix on the right side.
This exploration motivates the following important algorithm.
Algorithm 12.7.1 : Matrix Inverse Algorithm

Suppose A is an n × n matrix. To find A −1
if it exists, form the augmented n × 2n matrix
[A|I ] (12.7.14)
If possible do row operations until you obtain an n × 2n matrix of the form
[I |B] (12.7.15)
When this has been done, B = A . In this case, we say that A is invertible. If it is impossible to row reduce to a matrix
−1
of the form [I |B] , then A has no inverse.
This algorithm shows how to find the inverse if it exists. It will also tell you if A does not have an inverse.
Example 12.7.1 : Finding the Inverse

1 2 2
⎡ ⎤
Let A = ⎢ 1 0 2⎥ . Find A
−1
if it exists.
⎣ ⎦
3 1 −1
Solution
Set up the augmented matrix
⎡ 1 2 2 1 0 0 ⎤
[A|I ] = ⎢
⎢ 1 0 2 0 1 0 ⎥
⎥ (12.7.16)
⎣ 3 1 −1 0 0 1 ⎦
Now we row reduce, with the goal of obtaining the 3 × 3 identity matrix on the left hand side. First, take −1 times the
first row and add to the second followed by −3 times the first row added to the third row. This yields

⎡ 1 2 2 1 0 0 ⎤
⎢ 0 −2 0 −1 1 0 ⎥ (12.7.17)
⎢ ⎥
⎣ ⎦
0 −5 −7 −3 0 1
Then take 5 times the second row and add to -2 times the third row.
⎡ 1 2 2 1 0 0 ⎤
⎢ 0 −10 0 −5 5 0 ⎥ (12.7.18)
⎢ ⎥
⎣ 0 0 14 1 5 −2 ⎦
Next take the third row and add to −7 times the first row. This yields
⎡ −7 −14 0 −6 5 −2 ⎤
⎢ 0 −10 0 −5 5 0 ⎥ (12.7.19)
⎢ ⎥
⎣ 0 0 14 1 5 −2 ⎦
Now take − times the second row and add to the first row.
7
⎡ −7 0 0 1 −2 −2 ⎤
⎢ 0 −10 0 −5 5 0 ⎥ (12.7.20)
⎢ ⎥
⎣ 0 0 14 1 5 −2 ⎦
Finally divide the first row by -7, the second row by -10 and the third row by 14 which yields
1 2 2
⎡ 1 0 0 − ⎤
7 7 7
⎢ 1 1 ⎥
⎢ 0 1 0 − 0 ⎥ (12.7.21)
⎢ 2 2 ⎥
1 5 1
⎣ 0 0 1 − ⎦
14 14 7
Notice that the left hand side of this matrix is now the 3 × 3 identity matrix I . Therefore, the inverse is the 3 × 3 matrix 3
on the right hand side, given by

1 2 2
−
⎡ 7 7 7 ⎤
1 1
⎢ − 0⎥ (12.7.22)
⎢ 2 2 ⎥
1 5 1
⎣ − ⎦
14 14 7
It may happen that through this algorithm, you discover that the left hand side cannot be row reduced to the identity matrix.
Consider the following example of this situation.
Example 12.7.2 : A Matrix Which Has No Inverse

1 2 2
⎡ ⎤
Let A = ⎢ 1 0 2⎥ . Find A
−1
if it exists.
⎣ ⎦
2 2 4
Solution
Write the augmented matrix [A|I ]
⎡ 1 2 2 1 0 0 ⎤
⎢ 1 0 2 0 1 0 ⎥ (12.7.23)
⎢ ⎥
⎣ 2 2 4 0 0 1 ⎦
and proceed to do row operations attempting to obtain [I |A −1

]. Take −1 times the first row and add to the second. Then
take −2 times the first row and add to the third row.

⎡ 1 2 2 1 0 0 ⎤
⎢ 0 −2 0 −1 1 0 ⎥ (12.7.24)
⎢ ⎥
⎣ ⎦
0 −2 0 −2 0 1
Next add −1 times the second row to the third row.
⎡ 1 2 2 1 0 0 ⎤
⎢ 0 −2 0 −1 1 0 ⎥ (12.7.25)
⎢ ⎥
⎣ 0 0 0 −1 −1 1 ⎦
At this point, you can see there will be no way to obtain I on the left side of this augmented matrix. Hence, there is no
way to complete this algorithm, and therefore the inverse of A does not exist. In this case, we say that A is not invertible.
If the algorithm provides an inverse for the original matrix, it is always possible to check your answer. To do so, use the
method demonstrated in Example [exa:verifyinginverse]. Check that the products AA and A A both equal the identity −1 −1
matrix. Through this method, you can always be sure that you have calculated A properly! −1
One way in which the inverse of a matrix is useful is to find the solution of a system of linear equations. Recall from
Definition [def:matrixform] that we can write a system of equations in matrix form, which is of the form AX = B . Suppose
you find the inverse of the matrix A . Then you could multiply both sides of this equation on the left by A and simplify to
−1 −1
obtain
−1 −1
(A ) AX = A B
−1 −1
(A A) X = A B
(12.7.26)
−1
IX = A B
−1
X =A B
Therefore we can find X, the solution to the system, by computing X = A B . Note that once you have found A
−1 −1
, you can
easily get the solution for different right hand sides (different B ). It is always just A B . −1
We will explore this method of finding the solution to a system in the following example.
Example 12.7.3 : Using the Inverse to Solve a System of Equations

Consider the following system of equations. Use the inverse of a suitable matrix to give the solutions to this system.
x +z = 1
x −y +z = 3 (12.7.27)
x +y −z = 2
Solution
First, we can write the system of equations in matrix form
1 0 1 x 1
⎡ ⎤⎡ ⎤ ⎡ ⎤
AX = ⎢ 1 −1 1 ⎥⎢ y ⎥ = ⎢3 ⎥ = B (12.7.28)
⎣ ⎦⎣ ⎦ ⎣ ⎦
1 1 −1 z 2
The inverse of the matrix
1 0 1
⎡ ⎤
A = ⎢1 −1 1⎥ (12.7.29)
⎣ ⎦
1 1 −1
is

1 1
0
⎡ 2 2 ⎤
−1
A = ⎢1 −1 0⎥ (12.7.30)
⎢ ⎥
⎣1 1 1 ⎦
− −
2 2
Verifying this inverse is left as an exercise.

From here, the solution to the given system [inversesystem1] is found by
1 1 5
x 0
⎡ ⎤ ⎡ 2 2 ⎤⎡1 ⎤ ⎡ 2 ⎤
−1
⎢ y ⎥ =A B =⎢
⎢1 −1 0⎥ ⎢
⎥⎢3 ⎥ = ⎢ −2 ⎥
⎥ (12.7.31)
⎣ ⎦ 1 1 ⎣ ⎦ 3
z ⎣1 − − ⎦ 2 ⎣ − ⎦
2 2 2
0
⎡ ⎤
What if the right side, B , of [inversesystem1] had been ⎢ 1 ⎥? In other words, what would be the solution to
⎣ ⎦
3
1 0 1 x 0
⎡ ⎤⎡ ⎤ ⎡ ⎤
⎢1 −1 1 ⎥ ⎢ y ⎥ = ⎢ 1 ⎥? (12.7.32)
⎣ ⎦⎣ ⎦ ⎣ ⎦
1 1 −1 z 3
By the above discussion, the solution is given by

1 1
x 0
⎡ ⎤ ⎡ 2 2 ⎤⎡0 ⎤ ⎡
2
⎤
−1
⎢ y ⎥ =A B =⎢
⎢1 −1 0⎥
⎥ ⎢ 1 ⎥ = ⎢ −1 ⎥ (12.7.33)
⎣ ⎦ 1 1 ⎣ ⎦ ⎣ ⎦
z ⎣1 − − ⎦ 3 −2
2 2
This illustrates that for a system AX = B where A −1

exists, it is easy to find the solution when the vector B is changed.
We conclude this section with some important properties of the inverse.
Theorem 12.7.1 : Inverses of Transposes and Products

Let A, B , and A for i = 1, . . . , k be n × n matrices.
i
1. If A is an invertible matrix, then (A ) = (A ) T −1 −1 T
2. If A and B are invertible matrices, then AB is invertible and (AB) = B −1 −1

A
−1
3. If A ,A ,...,A
1 2 kare invertible, then the product A 1 A2 ⋯ Ak is invertible, and
−1 −1 −1 −1 −1
(A1 A2 ⋯ Ak ) =A A ⋯A A
k k−1 2 1
Consider the following theorem.
Theorem 12.7.1 : Properties of the Inverse

Let A be an n × n matrix and I the usual identity matrix.
1. I is invertible and I = I−1
2. If A is invertible then so is A , and (A ) = A

−1 −1 −1
3. If A is invertible then so is A , and (A ) = (A )

k k −1 −1 k
4. If A is invertible and p is a nonzero real number, then pA is invertible and (pA) −1

=
1
p
−1
A

12.E: Exercises
Exercise 12.E . 1
For the following pairs of matrices, determine if the sum A + B is defined. If so, find the sum.
1 0 0 1
1. A = [ ],B = [ ]
0 1 1 0
2 1 2 −1 0 3
2. A = [ ],B = [ ]
1 1 0 0 1 4
1 0
⎡ ⎤
2 7 −1
3. A = ⎢ −2 3⎥,B = [ ]
0 3 4
⎣ ⎦
4 2
Exercise 12.E . 2
For each matrix A , find the matrix −A such that A + (−A) = 0 .
1 2
1. A = [ ]
2 1
−2 3
2. A = [ ]
0 2
0 1 2
⎡ ⎤
3. A = ⎢ 1 −1 3⎥
⎣ ⎦
4 2 0
Exercise 12.E . 3
In the context of Proposition [prop:propertiesofaddition], describe −A and 0.
Answer
To get −A, just replace every entry of A with its additive inverse. The 0 matrix is the one which has all zeros in it.
2.1.2: Scalar Multiplication of Matrices
Exercise 12.E . 4
For each matrix A , find the product (−2)A, 0A, and 3A.
1 2
1. A = [ ]
2 1
−2 3
2. A = [ ]
0 2
0 1 2
⎡ ⎤
3. A = ⎢ 1 −1 3⎥
⎣ ⎦
4 2 0
Exercise 12.E . 5
Using only the properties given in Proposition [prop:propertiesofaddition] and Proposition [prop:propertiesscalarmult],
show −A is unique.
Ken Kuttler 9/11/2020 12.E.1 CC-BY https://math.libretexts.org/@go/page/44322

Answer
Suppose B also works. Then
−A = −A + (A + B) = (−A + A) + B = 0 + B = B (12.E.1)
Exercise 12.E . 6
Using only the properties given in Proposition [prop:propertiesofaddition] and Proposition [prop:propertiesscalarmult]
show 0 is unique.
Answer
Suppose 0 also works. Then 0
′ ′ ′
= 0 + 0 = 0.
Exercise 12.E . 7
Using only the properties given in Proposition [prop:propertiesofaddition] and Proposition [prop:propertiesscalarmult]
show 0A = 0. Here the 0 on the left is the scalar 0 and the 0 on the right is the zero matrix of appropriate size.
Answer
0A = (0 + 0) A = 0A + 0A. Now add − (0A) to both sides. Then 0 = 0A.
Exercise 12.E . 8
Using only the properties given in Proposition [prop:propertiesofaddition] and Proposition [prop:propertiesscalarmult], as
well as previous problems show (−1) A = −A.
Answer
A + (−1) A = (1 + (−1)) A = 0A = 0. Therefore, from the uniqueness of the additive inverse proved in the above
Problem [addinvrstunique], it follows that −A = (−1) A .
2.2
Exercise 12.E . 9
1 2 3 3 −1 2 1 2
Consider the matrices A = [ ],B = [ ],C = [ ], .
2 1 7 −3 2 1 3 1
−1 2 2
D =[ ],E = [ ]
2 −3 3
Exercise 12.E . 10
Find the following if possible. If it is not possible explain why.
1. −3A
2. 3B − A
3. AC
4. C B
5. AE
6. EA
Answer

−3 −6 −9
1. [ ]
−6 −3 −21
8 −5 3
2. [ ]
−11 5 −4
3. Not possible
−3 3 4
4. [ ]
6 −1 7
5. Not possible
6. Not possible
Exercise 12.E . 11
1 2
⎡ ⎤
2 −5 2 1 2
Consider the matrices A = ⎢ 3 2⎥,B = [ ],C = [ ],
⎣ ⎦ −3 2 1 5 0
1 −1
−1 1 1
D =[ ],E = [ ]
4 −3 3
Exercise 12.E . 12
1. −3A
2. 3B − A
3. AC
4. C A
5. AE
6. EA
7. BE
8. DE
Answer
−3 −6
⎡ ⎤
1. ⎢ −9 −6 ⎥
⎣ ⎦
−3 3
2. Not possible.
11 2
⎡ ⎤
3. ⎢ 13 6⎥
⎣ ⎦
−4 2
4. Not possible.
7
⎡ ⎤
5. ⎢ 9⎥
⎣ ⎦
−2
6. Not possible.
7. Not possible.
2
8. [ ]
−5
Exercise 12.E . 13

1 1 1 1 −3
⎡ ⎤ ⎡ ⎤
1 −1 −2
Let A = ⎢ −2 −1 ⎥ ,B=[ ], and C = ⎢ −1 2 0⎥. Find the following if possible.
2 1 −2
⎣ ⎦ ⎣ ⎦
1 2 −3 −1 0
1. AB
2. BA
3. AC
4. C A
5. C B
6. BC
Answer
3 0 −4
⎡ ⎤
1. ⎢ −4 1 6⎥
⎣ ⎦
5 1 −6
1 −2
2. [ ]
−2 −3
3. Not possible
−4 −6
⎡ ⎤
4. ⎢ −5 −3 ⎥
⎣ ⎦
−1 −2
8 1 −3
5. [ ]
7 6 −6
Exercise 12.E . 14
−1 −1
Let A = [ ] . Find all 2 × 2 matrices, B such that AB = 0.
3 3
Answer
−1 −1 x y −x − z −w − y
[ ][ ] = [ ]
3 3 z w 3x + 3z 3w + 3y
0 0
= [ ]
0 0
x y
Solution is: w = −y, x = −z so the matrices are of the form [ ].
−x −y
Exercise 12.E . 15
Let X = [ −1 −1 1] and Y =[0 1 2]. Find X T
Y and XY T
if possible.
Answer
0 −1 −2
⎡ ⎤
T T
X Y = ⎢0 −1 −2 ⎥ , X Y =1
⎣ ⎦
0 1 2
Exercise 12.E . 16
1 2 1 2
Let A = [ ],B = [ ]. Is it possible to choose k such that AB = BA? If so, what should k equal?
3 4 3 k

Answer
1 2 1 2 7 2k + 2
[ ][ ] = [ ]
3 4 3 k 15 4k + 6
1 2 1 2 7 10
[ ][ ] = [ ]
3 k 3 4 3k + 3 4k + 6
3k + 3 = 15
Thus you must have , Solution is: [k = 4]
2k + 2 = 10
Exercise 12.E . 17
1 2 1 2
Let A = [ ],B = [ ]. Is it possible to choose k such that AB = BA? If so, what should k equal?
3 4 1 k
Answer
1 2 1 2 3 2k + 2
[ ][ ] = [ ]
3 4 1 k 7 4k + 6
1 2 1 2 7 10
[ ][ ] = [ ]
1 k 3 4 3k + 1 4k + 2
However, 7 ≠ 3 and so there is no possible choice of k which will make these matrices commute.
Exercise 12.E . 18
Find 2 × 2 matrices, A , B, and C such that A ≠ 0, C ≠ B, but AC = AB.
Exercise 12.E . 19
1 −1 1 1 2 2
Let A = [ ],B = [ ],C = [ ] .
−1 1 1 1 2 2
Answer
1 −1 1 1 0 0
[ ][ ] = [ ]
−1 1 1 1 0 0
1 −1 2 2 0 0
[ ][ ] = [ ]
−1 1 2 2 0 0
Exercise 12.E . 20
Give an example of matrices (of any size), A, B, C such that B ≠ C , A ≠ 0, and yet AB = AC .
Exercise 12.E . 21
Find 2 × 2 matrices A and B such that A ≠ 0 and B ≠ 0 but AB = 0 .
Exercise 12.E . 22
1 −1 1 1
Let A = [ ],B = [ ].
−1 1 1 1

1 −1 1 1 0 0
[ ][ ] = [ ] (12.E.2)
−1 1 1 1 0 0
Exercise 12.E . 23
Give an example of matrices (of any size), A, B such that A ≠ 0 and B ≠ 0 but AB = 0.
Exercise 12.E . 24
Find 2 × 2 matrices A and B such that A ≠ 0 and B ≠ 0 with AB ≠ BA .
Exercise 12.E . 25
0 1 1 2
Let A = [ ],B = [ ] .
1 0 3 4
0 1 1 2 3 4
[ ][ ] = [ ]
1 0 3 4 1 2
1 2 0 1 2 1
[ ][ ] = [ ]
3 4 1 0 4 3
Exercise 12.E . 26
Write the system
x1 − x2 + 2 x3
2 x3 + x1
(12.E.3)
3x3
3 x4 + 3 x2 + x1
x1
⎡ ⎤
⎢ x2 ⎥
in the form A⎢ ⎥ where A is an appropriate matrix.
⎢x ⎥
3
⎣ ⎦
x4
Answer
1 −1 2 0
⎡ ⎤
⎢1 0 2 0⎥
A =⎢ ⎥
⎢0 0 3 0⎥
⎣ ⎦
1 3 0 3
Exercise 12.E . 27
Write the system
x1 + 3 x2 + 2 x3
2 x3 + x1
(12.E.4)
6x3
x4 + 3 x2 + x1

x1
⎡ ⎤
⎢ x2 ⎥
⎢x ⎥
3
⎣ ⎦
x4
Answer
1 3 2 0
⎡ ⎤
⎢1 0 2 0⎥
A =⎢ ⎥
⎢0 0 6 0⎥
⎣ ⎦
1 3 0 1
Exercise 12.E . 28
Write the system
x1 + x2 + x3
2 x3 + x1 + x2
(12.E.5)
x3 − x1
3 x4 + x1
x1
⎡ ⎤
⎢ x2 ⎥
⎢x ⎥
3
⎣ ⎦
x4
Answer
1 1 1 0
⎡ ⎤
⎢ 1 1 2 0⎥
A =⎢ ⎥
⎢ −1 0 1 0⎥
⎣ ⎦
1 0 0 3
Exercise 12.E . 29
A matrix A is called idempotent if A 2
= A. Let
2 0 2
⎡ ⎤
A =⎢ 1 1 2⎥ (12.E.6)
⎣ ⎦
−1 0 −1
and show that A is idempotent.
2.3
Exercise 12.E . 30
For each pair of matrices, find the (1, 2)-entry and (2, 3)-entry of the product AB.
1 2 −1 4 6 −2
⎡ ⎤ ⎡ ⎤
1. A = ⎢ 3 4 0⎥,B = ⎢ 7 2 1⎥
⎣ ⎦ ⎣ ⎦
2 5 1 −1 0 0
1 3 1 2 3 0
⎡ ⎤ ⎡ ⎤
2. A = ⎢ 0 2 4 ⎥ , B = ⎢ −4 16 1⎥
⎣ ⎦ ⎣ ⎦
1 0 5 0 2 2

2.4
Exercise 12.E . 31
Suppose A and B are square matrices of the same size. Which of the following are necessarily true?
1. (A − B) = A − 2AB + B
2 2 2
2
2. (AB) = A B 2 2
3. (A + B) = A + 2AB + B
2 2 2
4. (A + B) = A + AB + BA + B
2 2 2
5. A B = A (AB) B
2 2
3
6. (A + B) = A + 3A B + 3AB + B 3 2 2 3
7. (A + B) (A − B) = A − B 2 2
Answer
1. Not necessarily true.
4. Necessarily true.
5. Necessarily true.
2.5
Exercise 12.E . 1
1 2
⎡ ⎤
2 −5 2 1 2
Consider the matrices A = ⎢ 3 2⎥,B = [ ],C = [ ],
−3 2 1 5 0
⎣ ⎦
1 −1
−1 1 1
D =[ ],E = [ ]
4 −3 3
Exercise 12.E . 1
1. −3A T
2. 3B − A T
3. E B
T
4. EE T
5. B B
T
6. C A T
7. D BE
T
Answer
−3 −9 −3
1. [ ]
−6 −6 3
5 −18 5
2. [ ]
−11 4 4
3. [ −7 1 5]

1 3
4. [ ]
3 9
13 −16 1
⎡ ⎤
5. ⎢ −16 29 −8 ⎥
⎣ ⎦
1 −8 5
5 7 −1
6. [ ]
5 15 5
7. Not possible.
Exercise 12.E . 1
Let A be an n × n matrix. Show A equals the sum of a symmetric and a skew symmetric matrix.
Hint
Show that 1
2
(A
T
+ A) is symmetric and then consider using this as one of the matrices.
Exercise 12.E . 1
T T
Show that 1
2
(A
T
+ A) is symmetric and then consider using this as one of the matrices. A = A+A
2
+
A−A
2
.
Exercise 12.E . 1
Show that the main diagonal of every skew symmetric matrix consists of only zeros. Recall that the main diagonal
consists of every entry of the matrix which is of the form a . ii
Answer
If A is symmetric then A = −A T
. It follows that a
ii = −aii and so each a
ii =0 .
Exercise 12.E . 1
Prove [matrixtranspose2]. That is, show that for an m × n matrix A , an m × n matrix B , and scalars r, s , the following
holds:
T T T
(rA + sB) = rA + sB (12.E.7)
2.6
Exercise 12.E . 1
Prove that I mA =A where A is an m × n matrix.
Answer
(Im A) ≡∑ δik Akj = Aij
ij j
Exercise 12.E . 1
Suppose AB = AC and A is an invertible n × n matrix. Does it follow that B = C ? Explain why or why not.
Answer
Yes B = C . Multiply AB = AC on the left by A −1
.

Exercise 12.E . 1
Suppose AB = AC and A is a non invertible n × n matrix. Does it follow that B = C ? Explain why or why not.
Exercise 12.E . 1
Give an example of a matrix A such that A 2
=I and yet A ≠ I and A ≠ −I .
Answer
1 0 0
⎡ ⎤
A = ⎢0 −1 0⎥
⎣ ⎦
0 0 1
2.7
Exercise 12.E . 1
Let
2 1
A =[ ] (12.E.8)
−1 3
Find A −1
if possible. If A −1
does not exist, explain why.
Answer
−1 3 1
2 1 \vspace0.05in −\vspace0.05in
7 7
[ ] =[ ]
1 2
−1 3 \vspace0.05in \vspace0.05in
7 7
Exercise 12.E . 1
Let
0 1
A =[ ] (12.E.9)
5 3
Find A −1
Answer
−1 3 1
0 1 −\vspace0.05in \vspace0.05in
5 5
[ ] =[ ]
5 3 1 0
Exercise 12.E . 1
Add exercises text here.Let
2 1
A =[ ] (12.E.10)
3 0
Find A −1
Answer

−1 1
2 1 0 \vspace0.05in
3
[ ] =[ ]
2
3 0 1 −\vspace0.05in
3
Exercise 12.E . 1
Let
2 1
A =[ ] (12.E.11)
4 2
Find A −1
Answer
−1 1
2 1 1 \vspace0.05in
[ ] does not exist. The of this matrix is [ 2
]
4 2 0 0
Exercise 12.E . 1
a b
Let A be a 2 × 2 invertible matrix, with A = [ ]. Find a formula for A −1
in terms of a, b, c, d.
c d
Answer
−1 d b
a b −
ad−bc ad−bc
[ ] =[ ]
c a
c d −
ad−bc ad−bc
Exercise 12.E . 1
Let
1 2 3
⎡ ⎤
A = ⎢2 1 4⎥ (12.E.12)
⎣ ⎦
1 0 2
Find A −1
Answer
−1
1 2 3 −2 4 −5
⎡ ⎤ ⎡ ⎤
⎢2 1 4⎥ =⎢ 0 1 −2 ⎥
⎣ ⎦ ⎣ ⎦
1 0 2 1 −2 3
Exercise 12.E . 1
Let
1 0 3
⎡ ⎤
A = ⎢2 3 4⎥ (12.E.13)
⎣ ⎦
1 0 2
Find A −1
Answer

−1
1 0 3 −2 0 3
⎡ ⎤ ⎡ ⎤
1 2
⎢2 3 4⎥ =⎢ 0 \vspace0.05in −\vspace0.05in ⎥
3 3
⎣ ⎦ ⎣ ⎦
1 0 2 1 0 −1
Exercise 12.E . 1
Let
1 2 3
⎡ ⎤
A = ⎢2 1 4⎥ (12.E.14)
⎣ ⎦
4 5 10
Find A −1
Answer
5
⎡1 0 \vspace0.05in
3 ⎤
The is ⎢
⎢0 1 \vspace0.05in
2 ⎥
⎥
. There is no inverse.
3
⎣ ⎦
0 0 0
Exercise 12.E . 1
Let
1 2 0 2
⎡ ⎤
⎢1 1 2 0⎥
A =⎢ ⎥ (12.E.15)
⎢2 1 −3 2⎥
⎣ ⎦
1 2 1 2
Find A −1
Answer
1 1 1
−1
1 2 0 2 ⎡ −1 \vspace0.05in
2
\vspace0.05in
2
\vspace0.05in
2 ⎤
⎡ ⎤
⎢ 1 1 5⎥
⎢1 1 2 0⎥ ⎢ 3 \vspace0.05in −\vspace0.05in −\vspace0.05in ⎥
2 2 2
⎢ ⎥ =⎢ ⎥
⎢2 1 −3 2⎥ ⎢ ⎥
⎢ −1 0 0 1⎥
⎣ ⎦
1 2 1 2 ⎣ 3 1 9 ⎦
−2 −\vspace0.05in \vspace0.05in \vspace0.05in
4 4 4
Exercise 12.E . 1
Using the inverse of the matrix, find the solution to the systems:
1. 2 4 x 1
[ ][ ] =[ ] (12.E.16)
1 1 y 2
2. 2 4 x 2
[ ][ ] =[ ] (12.E.17)
1 1 y 0
Now give the solution in terms of a and b to

2 4 x a
[ ][ ] =[ ] (12.E.18)
1 1 y b

Exercise 12.E . 1
Using the inverse of the matrix, find the solution to the systems:
1. ⎡
1 0 3
⎤⎡
x
⎤ ⎡
1
⎤
⎢2 3 4 ⎥⎢ y ⎥ = ⎢0 ⎥ (12.E.19)
⎣ ⎦⎣ ⎦ ⎣ ⎦
1 0 2 z 1
2. ⎡
1 0 3
⎤⎡
x
⎤ ⎡
3
⎤
⎢2 3 4 ⎥ ⎢ y ⎥ = ⎢ −1 ⎥ (12.E.20)
⎣ ⎦⎣ ⎦ ⎣ ⎦
1 0 2 z −2
Now give the solution in terms of a, b, and c to the following:

1 0 3 x a
⎡ ⎤⎡ ⎤ ⎡ ⎤
⎢2 3 4⎥⎢ y ⎥ = ⎢ b ⎥ (12.E.21)
⎣ ⎦⎣ ⎦ ⎣ ⎦
1 0 2 z c
Answer
x 1
⎡ ⎤ ⎡ ⎤
1. ⎢ y ⎥ = ⎢ −\vspace0.05in
2
3
⎥
⎣ ⎦ ⎣ ⎦
z 0
x −12
⎡ ⎤ ⎡ ⎤
2. ⎢ y ⎥ =⎢ 1⎥
⎣ ⎦ ⎣ ⎦
z 5
x 3c − 2a
⎡ ⎤ ⎡ ⎤
3. ⎢ y ⎥ =⎢
1
3
b−
2
3
c ⎥
⎣ ⎦ ⎣ ⎦
z a−c
Exercise 12.E . 1
Show that if A is an n×n invertible matrix and X is a n×1 matrix such that AX = B for B an n×1 matrix, then
B.
−1
X =A
Answer
Multiply both sides of AX = B on the left by A −1
.
Exercise 12.E . 1
Prove that if A −1
exists and AX = 0 then X = 0 .
Answer
Multiply on both sides on the left by A −1
. Thus
−1 −1 −1
0 =A 0 =A (AX) = (A A) X = I X = X (12.E.22)
Exercise 12.E . 1 : Inverse Product

Show that if A −1
exists for an n × n matrix, then it is unique. That is, if BA = I and AB = I , then B = A −1
.
Answer
−1 −1 −1 −1
A =A I =A (AB) = (A A) B = I B = B.

Exercise 12.E . 1
−1 T
Show that if A is an invertible n × n matrix, then so is A and (A T T
) = (A
−1
) .
Answer
T
You need to show that (A ) acts like the inverse of A because from uniqueness in the above problem, this will
−1 T
imply it is the inverse. From properties of the transpose,

T T
T −1 −1 T
A (A ) = (A A) =I =I
T T
−1 T −1 T
(A ) A = (AA ) =I =I
T −1
Hence (A −1
) = (A
T
) and this last matrix exists.
Exercise 12.E . 1
Show (AB) −1
=B
−1 −1
A by verifying that
−1 −1
AB (B A ) =I (12.E.23)
and
−1 −1
B A (AB) = I (12.E.24)
Hint: Use Problem [exerinverseprod].
Answer
−1 −1 −1 −1 −1 −1 −1 −1 −1 −1 −1
(AB) B A = A (BB )A = AA =I B A (AB) = B (A A) B = B IB = B B =I
Exercise 12.E . 1
Show that (ABC ) −1
=C
−1
B
−1
A
−1
by verifying that
−1 −1 −1
(ABC ) (C B A ) =I (12.E.25)
and
−1 −1 −1
(C B A ) (ABC ) = I (12.E.26)
Hint: Use Problem [exerinverseprod].
Answer
The proof of this exercise follows from the previous one.
Exercise 12.E . 1
−1 2
If A is invertible, show (A 2
) = (A
−1
) . Hint: Use Problem [exerinverseprod].
Answer
2 2
2 −1 −1 −1 −1 −1 −1 2 −1 −1 −1 −1
A (A ) = AAA A = AI A = AA =I (A ) A =A A AA = A IA = A A =I
Exercise 12.E . 1

−1
If A is invertible, show (A −1
) = A. Hint: Use Problem [exerinverseprod].
Answer
−1
−1
A A = AA
−1
=I and so by uniqueness, (A −1
) =A .
2.8
Exercise 12.E . 1
2 3 1 2
Let A = [ ] . Suppose a row operation is applied to A and the result is B = [ ] . Find the elementary matrix E
1 2 2 3
that represents this row operation.
Exercise 12.E . 1
4 0 8 0
Let A = [ ] . Suppose a row operation is applied to A and the result is B = [ ] . Find the elementary matrix E
2 1 2 1
that represents this row operation.
Exercise 12.E . 1
1 −3 1 −3
Let A =[ ] . Suppose a row operation is applied to A and the result is B =[ ] . Find the elementary
0 5 2 −1
matrix E that represents this row operation.
Exercise 12.E . 1
1 2 1 1 2 1
⎡ ⎤ ⎡ ⎤
Let A = ⎢ 0 5 1⎥ . Suppose a row operation is applied to A and the result is B = ⎢ 2 −1 4⎥ .
⎣ ⎦ ⎣ ⎦
2 −1 4 0 5 1
1. Find the elementary matrix E such that EA = B .

2. Find the inverse of E , E , such that E B = A .
−1 −1
Exercise 12.E . 1
1 2 1 1 2 1
⎡ ⎤ ⎡ ⎤
Let A = ⎢ 0 5 1⎥ . Suppose a row operation is applied to A and the result is B = ⎢ 0 10 2⎥ .
⎣ ⎦ ⎣ ⎦
2 −1 4 2 −1 4

−1 −1
Exercise 12.E . 1
1 2 1 1 2 1
⎡ ⎤ ⎡ ⎤
Let A = ⎢ 0 5 1 ⎥ . Suppose a row operation is applied to A and the result is B = ⎢ 0 5 1⎥ .

⎣ ⎦ ⎣ 1 ⎦
2 −1 4 1 − 2
2

−1 −1

Exercise 12.E . 1
1 2 1 1 2 1
⎡ ⎤ ⎡ ⎤
Let A = ⎢ 0 5 1 ⎥ . Suppose a row operation is applied to A and the result is B = ⎢ 2 4 5⎥ .
⎣ ⎦ ⎣ ⎦
2 −1 4 2 −1 4

−1 −1
2.10
Exercise 12.E . 1
1 2 0
⎡ ⎤
Find an LU factorization of ⎢ 2 1 3⎥.
⎣ ⎦
1 2 3
Answer
1 2 0 1 0 0 1 2 0
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢2 1 3⎥ =⎢2 1 0⎥⎢0 −3 3⎥ (12.E.27)

⎣ ⎦ ⎣ ⎦⎣ ⎦
1 2 3 1 0 1 0 0 3
Exercise 12.E . 1
1 2 3 2
⎡ ⎤
Find an LU factorization of ⎢ 1 3 2 1⎥.
⎣ ⎦
5 0 1 3
Answer
1 2 3 2 1 0 0 1 2 3 2
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢1 3 2 1⎥ = ⎢1 1 0 ⎥⎢0 1 −1 −1 ⎥ (12.E.28)
⎣ ⎦ ⎣ ⎦⎣ ⎦
5 0 1 3 5 −10 1 0 0 −24 −17
Exercise 12.E . 1
1 −2 −5 0
⎡ ⎤
Find an LU factorization of the matrix ⎢ −2 5 11 3⎥.
⎣ ⎦
3 −6 −15 1
Answer
1 −2 −5 0 1 0 0 1 −2 −5 0
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢ −2 5 11 3 ⎥ = ⎢ −2 1 0 ⎥⎢0 1 1 3⎥ (12.E.29)
⎣ ⎦ ⎣ ⎦⎣ ⎦
3 −6 −15 1 3 0 1 0 0 0 1
Exercise 12.E . 1
1 −1 −3 −1
⎡ ⎤
Find an LU factorization of the matrix ⎢ −1 2 4 3⎥.

⎣ ⎦
2 −3 −7 −3
Answer

1 −1 −3 −1 1 0 0 1 −1 −3 −1
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢ −1 2 4 3 ⎥ = ⎢ −1 1 0 ⎥⎢0 1 1 2⎥ (12.E.30)
⎣ ⎦ ⎣ ⎦⎣ ⎦
2 −3 −7 −3 2 −1 1 0 0 0 1
Exercise 12.E . 1
1 −3 −4 −3
⎡ ⎤
Find an LU factorization of the matrix ⎢ −3 10 10 10 ⎥ .

⎣ ⎦
1 −6 2 −5
Answer
1 −3 −4 −3 1 0 0 1 −3 −4 −3
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢ −3 10 10 10 ⎥ = ⎢ −3 1 0 ⎥⎢0 1 −2 1⎥ (12.E.31)
⎣ ⎦ ⎣ ⎦⎣ ⎦
1 −6 2 −5 1 −3 1 0 0 0 1
Exercise 12.E . 1
1 3 1 −1
⎡ ⎤
Find an LU factorization of the matrix ⎢ 3 10 8 −1 ⎥ .
⎣ ⎦
2 5 −3 −3
Answer
1 3 1 −1 1 0 0 1 3 1 −1
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢3 10 8 −1 ⎥ = ⎢ 3 1 0 ⎥⎢0 1 5 2⎥ (12.E.32)
⎣ ⎦ ⎣ ⎦⎣ ⎦
2 5 −3 −3 2 −1 1 0 0 0 1
Exercise 12.E . 1
3 −2 1
⎡ ⎤
⎢ 9 −8 6⎥
Find an LU factorization of the matrix ⎢ ⎥.
⎢ −6 2 2⎥
⎣ ⎦
3 2 −7
Answer
3 −2 1 1 0 0 0 3 −2 1
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢ 9 −8 6⎥ ⎢ 3 1 0 0 ⎥⎢0 −2 3⎥
⎢ ⎥ =⎢ ⎥⎢ ⎥ (12.E.33)
⎢ −6 2 2⎥ ⎢ −2 1 1 0 ⎥⎢0 0 1⎥
⎣ ⎦ ⎣ ⎦⎣ ⎦
3 2 −7 1 −2 −2 1 0 0 0
Exercise 12.E . 1
−3 −1 3
⎡ ⎤
⎢ 9 9 −12 ⎥
⎢ 3 19 −16 ⎥
⎣ ⎦
12 40 −26
Exercise 12.E . 1

−1 −3 −1
⎡ ⎤
⎢ 1 3 0⎥
⎢ 3 9 0⎥
⎣ ⎦
4 12 16
Answer
−1 −3 −1 1 0 0 0 −1 −3 −1
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢ 1 3 0⎥ ⎢ −1 1 0 0⎥⎢ 0 0 −1 ⎥
⎢ ⎥ =⎢ ⎥⎢ ⎥ (12.E.34)
⎢ 3 9 0 ⎥ ⎢ −3 0 1 0⎥⎢ 0 0 −3 ⎥
⎣ ⎦ ⎣ ⎦⎣ ⎦
4 12 16 −4 0 −4 1 0 0 0
Exercise 12.E . 1
Find the LU factorization of the coefficient matrix using Dolittle’s method and use it to solve the system of equations.
x + 2y = 5
(12.E.35)
2x + 3y = 6
Answer
An LU factorization of the coefficient matrix is
1 2 1 0 1 2
[ ] =[ ][ ] (12.E.36)
2 3 2 1 0 −1
First solve
1 0 u 5
[ ][ ] =[ ] (12.E.37)
2 1 v 6
u 5
which gives [ ] = [ ]. Then solve
v −4
1 2 x 5
[ ][ ] =[ ] (12.E.38)
0 −1 y −4
which says that y = 4 and x = −3.
Exercise 12.E . 1
x + 2y + z = 1
y + 3z = 2 (12.E.39)
2x + 3y = 6
Answer
1 2 1 1 0 0 1 2 1
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢0 1 3 ⎥ = ⎢0 1 0 ⎥⎢0 1 3⎥ (12.E.40)
⎣ ⎦ ⎣ ⎦⎣ ⎦
2 3 0 2 −1 1 0 0 1
First solve

1 0 0 u 1
⎡ ⎤⎡ ⎤ ⎡ ⎤
⎢0 1 0⎥⎢ v ⎥ = ⎢2⎥ (12.E.41)

⎣ ⎦⎣ ⎦ ⎣ ⎦
2 −1 1 w 6
which yields u = 1, v = 2, w = 6 . Next solve
1 2 1 x 1
⎡ ⎤⎡ ⎤ ⎡ ⎤
⎢0 1 3⎥⎢ y ⎥ = ⎢2⎥ (12.E.42)
⎣ ⎦⎣ ⎦ ⎣ ⎦
0 0 1 z 6
This yields z = 6, y = −16, x = 27.
Exercise 12.E . 1
x + 2y + 3z = 5
2x + 3y + z = 6 (12.E.43)
x −y +z = 2
Exercise 12.E . 1
x + 2y + 3z = 5
2x + 3y + z = 6 (12.E.44)
3x + 5y + 4z = 11
Answer
1 2 3 1 0 0 1 2 3
⎡ ⎤ ⎡ ⎤⎡ ⎤
⎢2 3 1 ⎥ = ⎢2 1 0 ⎥⎢0 −1 −5 ⎥ (12.E.45)
⎣ ⎦ ⎣ ⎦⎣ ⎦
3 5 4 3 1 1 0 0 0
First solve
1 0 0 u 5
⎡ ⎤⎡ ⎤ ⎡ ⎤
⎢2 1 0⎥⎢ v ⎥ = ⎢ 6 ⎥ (12.E.46)
⎣ ⎦⎣ ⎦ ⎣ ⎦
3 1 1 w 11
u 5
⎡ ⎤ ⎡ ⎤
Solution is: ⎢ v ⎥ = ⎢ −4 ⎥ . Next solve
⎣ ⎦ ⎣ ⎦
w 0
1 2 3 x 5
⎡ ⎤⎡ ⎤ ⎡ ⎤
⎢0 −1 −5 ⎥ ⎢ y ⎥ = ⎢ −4 ⎥ (12.E.47)
⎣ ⎦⎣ ⎦ ⎣ ⎦
0 0 0 z 0
x 7t − 3
⎡ ⎤ ⎡ ⎤
Solution is: ⎢ y ⎥ = ⎢ 4 − 5t ⎥ , t ∈ R .
⎣ ⎦ ⎣ ⎦
z t
Exercise 12.E . 1

Is there only one LU factorization for a given matrix? Hint: Consider the equation
0 1 1 0 0 1
[ ] =[ ][ ]. (12.E.48)
0 1 1 1 0 0
Look for all possible LU factorizations.
Answer
Sometimes there is more than one LU factorization as is the case in this example. The given equation clearly gives an
LU factorization. However, it appears that the following equation gives another LU factorization.
0 1 1 0 0 1
[ ] =[ ][ ] (12.E.49)
0 1 0 1 0 1

CHAPTER OVERVIEW
13: DETERMINANTS
13.1: BASIC TECHNIQUES

Let A be an n×n matrix. That is, let A be a square matrix. The determinant of A, denoted by det(A)
is a very important number which we will explore throughout this section.
13.2: PROPERTIES OF DETERMINANTS

There are many important properties of determinants. Since many of these properties involve the
row operations discussed in Chapter 1, we recall that definition now. We will now consider the
effect of row operations on the determinant of a matrix. In future sections, we will see that using
the following properties can greatly assist in finding determinants. This section will use the
theorems as motivation to provide various examples of the usefulness of the properties.
13.3: FINDING DETERMINANTS USING ROW OPERATIONS

In this section, we look at two examples where row operations are used to find the determinant of a large matrix.
13.4: APPLICATIONS OF THE DETERMINANT

The determinant of a matrix also provides a way to find the inverse of a matrix.
13.E: EXERCISES
1 9/16/2020
13.1: Basic Techniques
Learning Objectives
Evaluate the determinant of a square matrix using either Laplace Expansion or row operations.
Demonstrate the effects that row operations have on determinants.
Verify the following:
The determinant of a product of matrices is the product of the determinants.
The determinant of a matrix is equal to the determinant of its transpose.
Let A be an n × n matrix. That is, let A be a square matrix. The determinant of A , denoted by det (A) is a very important
number which we will explore throughout this section.
If A is a 2×2 matrix, the determinant is given by the following formula.
Definition 13.1.1 : Determinant of a Two By Two Matrix

a b
Let A = [ ]. Then
c d
det (A) = ad − cb (13.1.1)
The determinant is also often denoted by enclosing the matrix with two vertical lines. Thus
a b ∣a b∣
det [ ] =∣ ∣ = ad − bc (13.1.2)
c d ∣ c d∣
The following is an example of finding the determinant of a 2 × 2 matrix.
Example 13.1.1 : A Two by Two Determinant

2 4
Find det (A) for the matrix A = [ ].
−1 6
Solution
From Equation 13.1.1:
det (A) = (2) (6) − (−1) (4) = 12 + 4 = 16 (13.1.3)
The 2 × 2 determinant can be used to find the determinant of larger matrices. We will now explore how to find the
determinant of a 3 × 3 matrix, using several tools including the 2 × 2 determinant.
We begin with the following definition.

Minor of a Matrix
Let A be a 3 × 3 matrix. The ij minor of A , denoted as
th
minor(A)
ij
, is the determinant of the 2 ×2 matrix which
results from deleting the i row and the j column of A .
th th
In general, if A is an n × n matrix, then the ij minor of th

A is the determinant of the n−1 ×n−1 matrix which results
from deleting the i row and the j column of A .
th th
Hence, there is a minor associated with each entry of A . Consider the following example which demonstrates this definition.

Example 13.1.2 : Finding Minors of a Matrix
Let
1 2 3
⎡ ⎤
A = ⎢4 3 2⎥ (13.1.4)
⎣ ⎦
3 2 1
Find minor(A) 12
and minor(A) . 23
Solution
First we will find minor(A) . By Definition [def:ijthminor], this is the determinant of the
12
2 ×2 matrix which results
when you delete the first row and the second column. This minor is given by
4 2
minor(A)12 = det [ ] (13.1.5)
3 1
Using Definition [def:twobytwodeterminant], we see that

4 2
det [ ] = (4) (1) − (3) (2) = 4 − 6 = −2 (13.1.6)
3 1
Therefore minor(A) 12
= −2 .
Similarly, minor(A) is the determinant of the 2 × 2 matrix which results when you delete the second row and the third
23
column. This minor is therefore

1 2
minor(A) = det [ ] = −4 (13.1.7)
23
3 2
Finding the other minors of A is left as an exercise.
The ij minor of a matrix A is used in another important definition, given next.

th

Cofactor of a Matrix
Suppose A is an n × n matrix. The ij cofactor, denoted by cof (A) is defined to be
th
ij
i+j
cof (A) = (−1) minor(A) (13.1.8)
ij ij
It is also convenient to refer to the cofactor of an entry of a matrix as follows. If aij is the ij
th
entry of the matrix, then its
cofactor is just cof (A) . ij
Example 13.1.3 : Finding Cofactors of a Matrix

Consider the matrix
1 2 3
⎡ ⎤
A = ⎢4 3 2⎥ (13.1.9)
⎣ ⎦
3 2 1
Find cof (A) 12

and cof (A) . 23
Solution
We will use Definition [def:ijthcofactor] to compute these cofactors.
First, we will compute cof (A) . Therefore, we need to find minor(A) . This is the determinant of the
12 12
2 ×2 matrix
which results when you delete the first row and the second column. Thus minor(A) is given by 12

4 2
det [ ] = −2
3 1
Then,
1+2 1+2
cof (A)12 = (−1) minor(A)12 = (−1) (−2) = 2
Hence, cof (A) 12

=2 .
Similarly, we can find cof (A) . First, find minor(A) , which is the determinant of the
23 23
2 ×2 matrix which results
when you delete the second row and the third column. This minor is therefore
1 2
det [ ] = −4
3 2
Hence,
2+3 2+3
cof (A)23 = (−1) minor(A)23 = (−1) (−4) = 4 (13.1.10)
You may wish to find the remaining cofactors for the above matrix. Remember that there is a cofactor for every entry in the
matrix.
We have now established the tools we need to find the determinant of a 3 × 3 matrix.
Definition 13.1.4 : The Determinant of a Three By Three Matrix

Let A be a 3 × 3 matrix. Then, det (A) is calculated by picking a row (or column) and taking the product of each entry in
that row (column) with its cofactor and adding these products together.
This process when applied to the i th
row (column) is known as expanding along the i th
row (column) as is given by
det (A) = ai1 cof (A)i1 + ai2 cof (A)i2 + ai3 cof (A)i3
When calculating the determinant, you can choose to expand any row or any column. Regardless of your choice, you will
always get the same number which is the determinant of the matrix A. This method of evaluating a determinant by expanding
along a row or a column is called Laplace Expansion or Cofactor Expansion.
Example 13.1.4 : Finding the Determinant of a Three by Three Matrix

Let
1 2 3
⎡ ⎤
A = ⎢4 3 2⎥ (13.1.11)
⎣ ⎦
3 2 1
Find det (A) using the method of Laplace Expansion.

Solution
First, we will calculate det (A) by expanding along the first column. Using Definition [def:threebythreedeterminant], we
take the 1 in the first column and multiply it by its cofactor,
1+1
∣3 2∣
1 (−1) ∣ ∣ = (1)(1)(−1) = −1
∣2 1∣
Similarly, we take the 4 in the first column and multiply it by its cofactor, as well as with the 3 in the first column. Finally,
we add these numbers together, as given in the following equation.

cof (A) cof (A) cof (A)
11 21 31
  
1+1 ∣3 2∣ 2+1 ∣ 2 3∣ 3+1 ∣ 2 3∣

det (A) = 1 (−1) ∣ ∣ + 4 (−1) ∣ ∣ + 3 (−1) ∣ ∣
∣2 1∣ ∣2 1∣ ∣3 2∣
Calculating each of these, we obtain
det (A) = 1 (1) (−1) + 4 (−1) (−4) + 3 (1) (−5) = −1 + 16 + −15 = 0
Hence, det (A) = 0 .

As mentioned in Definition [def:threebythreedeterminant], we can choose to expand along any row or column. Let’s try
now by expanding along the second row. Here, we take the 4 in the second row and multiply it to its cofactor, then add
this to the 3 in the second row multiplied by its cofactor, and the 2 in the second row multiplied by its cofactor. The
calculation is as follows.
cof (A)21 cof (A)22 cof (A)23
  
2+1
∣2 3∣ 2+2
∣1 3∣ 2+3
∣1 2∣
det (A) = 4 (−1) ∣ ∣ + 3 (−1) ∣ ∣ + 2 (−1) ∣ ∣
∣2 1∣ ∣3 1∣ ∣3 2∣
Calculating each of these products, we obtain

det (A) = 4 (−1) (−2) + 3 (1) (−8) + 2 (−1) (−4) = 0 (13.1.12)
You can see that for both methods, we obtained det (A) = 0 .
As mentioned above, we will always come up with the same value for det (A) regardless of the row or column we choose to
expand along. You should try to compute the above determinant by expanding along other rows and columns. This is a good
way to check your work, because you should come up with the same number each time!
We present this idea formally in the following theorem.
Theorem 13.1.1 : The Determinant is Well Defined

Expanding the n × n matrix along any row or column always gives the same answer, which is the determinant.
We have now looked at the determinant of 2 × 2 and 3 × 3 matrices. It turns out that the method used to calculate the
determinant of a 3 × 3 matrix can be used to calculate the determinant of any sized matrix. Notice that Definition
[def:ijthminor], Definition [def:ijthcofactor] and Definition [def:threebythreedeterminant] can all be applied to a matrix of any
size.
For example, the ij minor of a 4 × 4 matrix is the determinant of the 3 × 3 matrix you obtain when you delete the i row
th th
and the j column. Just as with the 3 × 3 determinant, we can compute the determinant of a 4 × 4 matrix by Laplace
th
Expansion, along any row or column

Example 13.1.5 : Determinant of a Four by Four Matrix

Find det (A) where
1 2 3 4
⎡ ⎤
⎢5 4 2 3⎥
A =⎢ ⎥ (13.1.13)
⎢1 3 4 5⎥
⎣ ⎦
3 4 3 2
Solution
As in the case of a 3 × 3 matrix, you can expand this along any row or column. Lets pick the third column. Then, using
Laplace Expansion,

∣5 4 3∣ ∣1 2 4∣
1+3 ∣ ∣ 2+3 ∣ ∣
det (A) = 3 (−1) 1 3 5 + 2 (−1) 1 3 5 + (13.1.14)
∣ ∣ ∣ ∣
∣3 4 2∣ ∣3 4 2∣
∣1 2 4∣ ∣1 2 4∣
3+3 ∣ ∣ 4+3 ∣ ∣
4 (−1) 5 4 3 + 3 (−1) 5 4 3 (13.1.15)
∣ ∣ ∣ ∣
∣3 4 2∣ ∣1 3 5∣
Now, you can calculate each 3 × 3 determinant using Laplace Expansion, as we did above. You should complete these as
an exercise and verify that det (A) = −12 .
The following provides a formal definition for the determinant of an n × n matrix. You may wish to take a moment and
consider the above definitions for 2 × 2 and 3 × 3 determinants in context of this definition.
Definition 13.1.5 : The Determinant of an n × n Matrix

Let A be an n × n matrix where n ≥ 2 and suppose the determinant of an (n − 1) × (n − 1) has been defined. Then
n n
det (A) = ∑ aij cof (A)ij = ∑ aij cof (A)ij (13.1.16)
j=1 i=1
The first formula consists of expanding the determinant along the i th

row and the second expands the determinant along
the j column.
th
In the following sections, we will explore some important properties and characteristics of the determinant.
The Determinant of a Triangular Matrix

There is a certain type of matrix for which finding the determinant is a very simple procedure. Consider the following
definition.
Definition 13.1.6 : Triangular Matrices

A matrix A is upper triangular if a = 0 whenever i > j . Thus the entries of such a matrix below the main diagonal
ij
equal 0, as shown. Here, ∗ refers to any nonzero number.

∗ ∗ ⋯ ∗
⎡ ⎤
⎢ ⎥
⎢0 ∗ ⋯ ⋮ ⎥
⎢ ⎥ (13.1.17)
⎢ ⎥
⎢ ⎥
⎢ ⋮ ⋮ ⋱ ∗⎥
⎣ ⎦
0 ⋯ 0 ∗
A lower triangular matrix is defined similarly as a matrix for which all entries above the main diagonal are equal to zero.
The following theorem provides a useful way to calculate the determinant of a triangular matrix.
Theorem 13.1.1 : Determinant of a Triangular Matrix

Let A be an upper or lower triangular matrix. Then det (A) is obtained by taking the product of the entries on the main
diagonal.
The verification of this Theorem can be done by computing the determinant using Laplace Expansion along the first row or
column.

Example 13.1.1 : Determinant of a Triangular Matrix
Let
1 2 3 77
⎡ ⎤
⎢0 2 6 7⎥
A =⎢ ⎥ (13.1.18)
⎢0 0 3 33.7 ⎥
⎣ ⎦
0 0 0 −1
Find det (A) .

Solution
From Theorem [thm:determinantoftriangularmatrix], it suffices to take the product of the elements on the main diagonal.
Thus det (A) = 1 × 2 × 3 × (−1) = −6.
Without using Theorem [thm:determinantoftriangularmatrix], you could use Laplace Expansion. We will expand along the
first column. This gives
∣2 6 7∣ ∣2 3 77 ∣
∣ ∣ 2+1 ∣ ∣
det (A) = 1 0 3 33.7 + 0 (−1) 0 3 33.7 +
∣ ∣ ∣ ∣
∣0 0 −1 ∣ ∣0 0 −1 ∣
∣2 3 77 ∣ ∣2 3 77 ∣
3+1 ∣ ∣ 4+1 ∣ ∣
0 (−1) 2 6 7 + 0 (−1) 2 6 7
∣ ∣ ∣ ∣
∣0 0 −1 ∣ ∣0 3 33.7 ∣
and the only nonzero term in the expansion is

∣2 6 7∣
∣ ∣
1 0 3 33.7 (13.1.19)
∣ ∣
∣0 0 −1 ∣
Now find the determinant of this 3 × 3 matrix, by expanding along the first column to obtain
∣3 33.7 ∣ 2+1
∣6 7∣ 3+1
∣6 7∣
det (A) = 1 × (2 × ∣ ∣ + 0 (−1) ∣ ∣ + 0 (−1) ∣ ∣) (13.1.20)
∣0 −1 ∣ ∣0 −1 ∣ ∣3 33.7 ∣
∣3 33.7 ∣
= 1 ×2 ×∣ ∣ (13.1.21)
∣0 −1 ∣
Next use Definition [def:twobytwodeterminant] to find the determinant of this 2 ×2 matrix, which is just
3 × −1 − 0 × 33.7 = −3 . Putting all these steps together, we have
det (A) = 1 × 2 × 3 × (−1) = −6 (13.1.22)
which is just the product of the entries down the main diagonal of the original matrix!
You can see that while both methods result in the same answer, Theorem [thm:determinantoftriangularmatrix] provides a much
quicker method.
In the next section, we explore some important properties of determinants.

13.2: Properties of Determinants
There are many important properties of determinants. Since many of these properties involve the row operations discussed in
Chapter 1, we recall that definition now.
Definition 13.2.1 : Row Operations

The row operations consist of the following
1. Switch two rows.
2. Multiply a row by a nonzero number.
3. Replace a row by a multiple of another row added to itself.
We will now consider the effect of row operations on the determinant of a matrix. In future sections, we will see that using the
following properties can greatly assist in finding determinants. This section will use the theorems as motivation to provide
various examples of the usefulness of the properties.
The first theorem explains the effect on the determinant of a matrix when two rows are switched.
Theorem 13.2.1 : Switching Rows

Let A be an n×n matrix and let B be a matrix which results from switching two rows of A. Then
det (B) = − det (A) .
When we switch two rows of a matrix, the determinant is multiplied by −1. Consider the following example.
Example 13.2.1 : Switching Two Rows

1 2 3 4
Let A = [ ] and let B = [ ] . Knowing that det (A) = −2 , find det (B) .
3 4 1 2
Solution
By Definition [def:twobytwodeterminant], det (A) = 1 × 4 − 3 × 2 = −2 . Notice that the rows of B are the rows of A
but switched. By Theorem 13.2.1 since two rows of A have been switched, det (B) = − det (A) = − (−2) = 2 . You
can verify this using Definition [def:twobytwodeterminant].
The next theorem demonstrates the effect on the determinant of a matrix when we multiply a row by a scalar.
Theorem 13.2.2 : Multiplying a Row by a Scalar

Let A be an n×n matrix and let B be a matrix which results from multiplying some row of A by a scalar k . Then
det (B) = k det (A) .
Notice that this theorem is true when we multiply one row of the matrix by k . If we were to multiply two rows of A by k to
obtain B , we would have det (B) = k det (A) . Suppose we were to multiply all n rows of A by k to obtain the matrix B , so
2
that B = kA . Then, det (B) = k det (A) . This gives the next theorem.
n
Theorem 13.2.3 : Scalar Multiplication

Let A and B be n × n matrices and k a scalar, such that B = kA . Then det(B) = k n
det(A) .

Example 13.2.2 : Multiplying a Row by 5
1 2 5 10
Let A = [ ] , B = [ ]. Knowing that det (A) = −2 , find det (B) .
3 4 3 4
Solution
By Definition [def:twobytwodeterminant], det (A) = −2. We can also compute det (B) using Definition
[def:twobytwodeterminant], and we see that det (B) = −10 .
Now, let’s compute det (B) using Theorem [thm:multiplyingrowbyscalar] and see if we obtain the same answer. Notice
that the first row of B is 5 times the first row of A , while the second row of B is equal to the second row of A . By
Theorem [thm:multiplyingrowbyscalar], det (B) = 5 × det (A) = 5 × −2 = −10.
You can see that this matches our answer above.
Finally, consider the next theorem for the last row operation, that of adding a multiple of a row to another row.
Theorem 13.2.3 : Adding a Multiple of a Row to Another Row

Let A be an n×n matrix and let B be a matrix which results from adding a multiple of a row to another row. Then
det (A) = det (B) .
Therefore, when we add a multiple of a row to another row, the determinant of the matrix is unchanged. Note that if a matrix
A contains a row which is a multiple of another row, det (A) will equal 0 . To see this, suppose the first row of A is equal to
−1 times the second row. By Theorem [thm:addingmultipleofrow], we can add the first row to the second row, and the
determinant will be unchanged. However, this row operation will result in a row of zeros. Using Laplace Expansion along the
row of zeros, we find that the determinant is 0.
Example 13.2.3 : Adding a Row to Another Row

1 2 1 2
Let A = [ ] and let B = [ ]. Find det (B) .
3 4 5 8
Solution
By Definition [def:twobytwodeterminant], det (A) = −2 . Notice that the second row of B is two times the first row of A
added to the second row. By Theorem [thm:switchingrows], det (B) = det (A) = −2 . As usual, you can verify this
answer using Definition [def:twobytwodeterminant].
Example 13.2.4 : Multiple of a Row

1 2
Let A = [ ] . Show that det (A) = 0 .
2 4
Solution
Using Definition [def:twobytwodeterminant], the determinant is given by
det (A) = 1 × 4 − 2 × 2 = 0
However notice that the second row is equal to 2 times the first row. Then by the discussion above following Theorem
[thm:addingmultipleofrow] the determinant will equal 0.
Until now, our focus has primarily been on row operations. However, we can carry out the same operations with columns,
rather than rows. The three operations outlined in Definition [def:operations] can be done with columns instead of rows. In this

case, in Theorems [thm:switchingrows], [thm:multiplyingrowbyscalar], and [thm:addingmultipleofrow] you can replace the
word, "row" with the word "column".
There are several other major properties of determinants which do not involve row (or column) operations. The first is the
determinant of a product of matrices.
Theorem 13.2.4 : Determinant of a Product

Let A and B be two n × n matrices. Then
det (AB) = det (A) det (B) (13.2.1)
In order to find the determinant of a product of matrices, we can simply take the product of the determinants.
Example 13.2.4 : The Determinant of a Product

Compare det (AB) and det (A) det (B) for
1 2 3 2
A =[ ],B = [ ] (13.2.2)
−3 2 4 1
Solution
First compute AB, which is given by
1 2 3 2 11 4
AB = [ ][ ] =[ ]
−3 2 4 1 −1 −4
and so by Definition [def:twobytwodeterminant]
11 4
det (AB) = det [ ] = −40
−1 −4
Now
1 2
det (A) = det [ ] =8
−3 2
and
3 2
det (B) = det [ ] = −5
4 1
Computing det (A) × det (B) we have 8 × −5 = −40 . This is the same answer as above and you can see that
det (A) det (B) = 8 × (−5) = −40 = det (AB) .
Consider the next important property.
Theorem 13.2.5 : Determinant of the Transpose

Let A be a matrix where A is the transpose of A . Then,
T
T
det (A ) = det (A) (13.2.3)
This theorem is illustrated in the following example.
Example 13.2.5 : Determinant of the Transpose

Let
2 5
A =[ ] (13.2.4)
4 3
Find det (A T
.
)
Solution
First, note that
2 4
T
A =[ ]
5 3
Using Definition [def:twobytwodeterminant], we can compute det (A) and det (A ). It follows T
that
det (A) = 2 × 3 − 4 × 5 = −14 and det (A ) = 2 × 3 − 5 × 4 = −14 . Hence, det (A) = det (A ) .
T T
The following provides an essential property of the determinant, as well as a useful way to determine if a matrix is invertible.
Theorem 13.2.6 : Determinant of the Inverse

Let A be an n × n matrix. Then A is invertible if and only if det(A) ≠ 0 . If this is true, it follows that
−1
1
det(A ) =
det(A)
Example 13.2.6 : Determinant of an Invertible Matrix

3 6 2 3
Let A = [ ],B = [ ] . For each matrix, determine if it is invertible. If so, find the determinant of the inverse.
2 4 5 1
Solution
Consider the matrix A first. Using Definition [def:twobytwodeterminant] we can find the determinant as follows:
det (A) = 3 × 4 − 2 × 6 = 12 − 12 = 0 (13.2.5)
By Theorem [thm:detinverse] A is not invertible.

Now consider the matrix B . Again by Definition [def:twobytwodeterminant] we have
det (B) = 2 × 1 − 5 × 3 = 2 − 15 = −13 (13.2.6)
By Theorem [thm:detinverse] B is invertible and the determinant of the inverse is given by

1
−1
det (A ) =
det(A)
1
=
−13
1
= −
13
Properties of Determinants II: Some Important Proofs

This section includes some important proofs on determinants and cofactors.
First we recall the definition of a determinant. If A = [ aij ] is an n×n matrix, then det A is defined by computing the
expansion along the first row:

n
det A = ∑ a1,i cof (A)1,i . (13.2.7)
i=1
If n = 1 then det A = a 1,1 .

The following example is straightforward and strongly recommended as a means for getting used to definitions.
Example 13.2.7 :
1. Let E be the elementary matrix obtained by interchanging ith and j th rows of I . Then det E = −1 .
ij ij
2. Let E be the elementary matrix obtained by multiplying the ith row of I by k . Then det E = k .
ik ik
3. Let E be the elementary matrix obtained by multiplying ith row of I by k and adding it to its j th row. Then
ijk
det E =1.
ijk
4. If C and B are such that C B is defined and the ith row of C consists of zeros, then the ith row of C B consists of
zeros.
5. If E is an elementary matrix, then det E = det E . T
Many of the proofs in section use the Principle of Mathematical Induction. This concept is discussed in Appendix A.2 and is
reviewed here for convenience. First we check that the assertion is true for n = 2 (the case n = 1 is either completely trivial
or meaningless).
Next, we assume that the assertion is true for n − 1 (where n ≥ 3 ) and prove it for n . Once this is accomplished, by the
Principle of Mathematical Induction we can conclude that the statement is true for all n × n matrices for every n ≥ 2 .
If A is an n×n matrix and 1 ≤ j ≤ n , then the matrix obtained by removing 1st column and j th row from A is an
n−1 ×n−1 matrix (we shall denote this matrix by A(j) below). Since these matrices are used in computation of cofactors
cof (A)1,i , for 1 ≤ i ≠ n , the inductive assumption applies to these matrices.
Consider the following lemma.
Lemma 13.2.1 :
If A is an n × n matrix such that one of its rows consists of zeros, then det A = 0 .
Proof
We will prove this lemma using Mathematical Induction.
If n = 2 this is easy (check!).
Let n ≥ 3 be such that every matrix of size n − 1 × n − 1 with a row consisting of zeros has determinant equal to
zero. Let i be such that the ith row of A consists of zeros. Then we have a = 0 for 1 ≤ j ≤ n .
ij
Fix j ∈ {1, 2, … , n} such that j ≠ i . Then matrix A(j) used in computation of cof (A)1,j has a row consisting of
zeros, and by our inductive assumption cof (A) = 0 .
1,j
On the other hand, if j = i then a1,j =0 . Therefore a 1,j cof (A)1,j =0 for all j and by [E1] we have
n
det A = ∑ a1,j cof (A)1,j = 0 (13.2.8)
j=1
as each of the summands is equal to 0.
Lemma 13.2.2 :
Assume A , B and C are n × n matrices that for some 1 ≤ i ≤ n satisfy the following.
1. j th rows of all three matrices are identical, for j ≠ i .
2. Each entry in the j th row of A is the sum of the corresponding entries in j th rows of B and C .

Then det A = det B + det C .
Proof
This is not difficult to check for n = 2 (do check it!).
Now assume that the statement of Lemma is true for n − 1 × n − 1 matrices and fix A, B and C as in the statement.
The assumptions state that we have a = b = c l,j for j ≠ i and for 1 ≤ l ≤ n and a = b + c
l,j l,j for all l,i l,i l,i
1 ≤ l ≤ n . Therefore A(i) = B(i) = C (i) , and A(j) has the property that its ith row is the sum of ith rows of B(j)
and C (j) for j ≠ i while the other rows of all three matrices are identical. Therefore by our inductive assumption we
have cof (A) = cof (B) + cof (C ) for j ≠ i .
1j 1j 1j
By [E1] we have (using all equalities established above)

n
det A = ∑ a1,l cof (A)1,l
l=1
= ∑ a1,l (cof (B)1,l + cof (C )1,l ) + (b1,i + c1,i )cof (A)1,i
l≠i
= det B + det C
This proves that the assertion is true for all n and completes the proof.
Theorem 13.2.7 :
Let A and B be n × n matrices.
1. If A is obtained by interchanging ith and j th rows of B (with i ≠ j ), then det A = − det B .
2. If A is obtained by multiplying ith row of B by k then det A = k det B .
3. If two rows of A are identical then det A = 0 .
4. If A is obtained by multiplying ith row of B by k and adding it to j th row of B (i ≠ j ) then det A = det B .
Proof
We prove all statements by induction. The case n = 2 is easily checked directly (and it is strongly suggested that you
do check it).
We assume n ≥ 3 and (1)–(4) are true for all matrices of size n − 1 × n − 1 .
(1) We prove the case when j = i + 1 , i.e., we are interchanging two consecutive rows.
Let l ∈ {1, … , n} ∖ {i, j}. Then A(l) is obtained from B(l) by interchanging two of its rows (draw a picture) and by
our assumption
cof (A)1,l = −cof (B)1,l . (13.2.9)
Now consider a 1,i cof (A)1,l . We have that a 1,i = b1,j and also that A(i) = B(j) . Since j = i + 1 , we have
1+j 1+i+1 1+i
(−1 ) = (−1 ) = −(−1 ) (13.2.10)
and therefore a cof (A) = −b cof (B) and a cof (A) = −b cof (B) . Putting this together with [E2] into
1i 1i 1j 1j 1j 1j 1i 1i
[E1] we see that if in the formula for det A we change the sign of each of the summands we obtain the formula for
det B .
n n
det A = ∑ a1l cof (A)1l = − ∑ b1l B1l = det B. (13.2.11)
l=1 l=1
We have therefore proved the case of (1) when j = i + 1 . In order to prove the general case, one needs the following
fact. If i < j , then in order to interchange ith and j th row one can proceed by interchanging two adjacent rows
2(j − i) + 1 times: First swap ith and i + 1 st, then i + 1 st and i + 2 nd, and so on. After one interchanges j − 1 st
and j th row, we have ith row in position of j th and lth row in position of l − 1 st for i + 1 ≤ l ≤ j . Then proceed
backwards swapping adjacent rows until everything is in place.

Since 2(j − i) + 1 is an odd number (−1) 2(j−i)+1
= −1 and we have that det A = − det B .
(2) This is like (1)… but much easier. Assume that (2) is true for all n − 1 × n − 1 matrices. We have that a = kb ji ji
for 1 ≤ j ≤ n . In particular a = kb , and for l ≠ i matrix A(l) is obtained from B(l) by multiplying one of its
1i 1i
rows by k . Therefore cof (A) = kcof (B) for l ≠ i , and for all l we have a cof (A) = kb cof (B) . By [E1],
1l 1l 1l 1l 1l 1l
we have det A = k det B .

(3) This is a consequence of (1). If two rows of A are identical, then A is equal to the matrix obtained by
interchanging those two rows and therefore by (1) det A = − det A . This implies det A = 0 .
(4) Assume (4) is true for all n − 1 × n − 1 matrices and fix A and B such that A is obtained by multiplying ith row
of B by k and adding it to j th row of B (i ≠ j ) then det A = det B . If k = 0 then A = B and there is nothing to
prove, so we may assume k ≠ 0 .
Let C be the matrix obtained by replacing the j th row of B by the ith row of B multiplied by k . By Lemma [lem:L2],
we have that
det A = det B + det C (13.2.12)
and we ‘only’ need to show that det C = 0 . But ith and j th rows of C are proportional. If D is obtained by
multiplying the j th row of C by then by (2) we have det C = det D (recall that k ≠ 0 !). But ith and j th rows of
1
k
1
D are identical, hence by (3) we have det D = 0 and therefore det C = 0 .
Theorem 13.2.8 :
If A is an n × n matrix such that one of its rows consists of zeros, then det A = 0 .
Proof
We will prove this lemma using Mathematical Induction.
If n = 2 this is easy (check!).
Let n ≥ 3 be such that every matrix of size n − 1 × n − 1 with a row consisting of zeros has determinant equal to
zero. Let i be such that the ith row of A consists of zeros. Then we have a = 0 for 1 ≤ j ≤ n .
ij
Fix j ∈ {1, 2, … , n} such that j ≠ i . Then matrix A(j) used in computation of cof (A)1,j has a row consisting of
zeros, and by our inductive assumption cof (A) = 0 . 1,j
On the other hand, if j = i then a 1,j =0 . Therefore a 1,j cof (A)1,j =0 for all j and by [E1] we have
n
det A = ∑ a1,j cof (A)1,j = 0 (13.2.13)
j=1
as each of the summands is equal to 0.
Theorem 13.2.9 :
Let A and B be two n × n matrices. Then
det (AB) = det (A) det (B) (13.2.14)
Proof
If A is an elementary matrix of either type, then multiplying by A on the left has the same effect as performing the
corresponding elementary row operation. Therefore the equality det(AB) = det A det B in this case follows by
Example [exa:EX1] and Theorem [thm:T1].
If C is the reduced row-echelon form of A then we can write A = E ⋅ E ⋅ ⋯ ⋅ E ⋅ C for some elementary
1 2 m
matrices E , … , E .
1 m
Now we consider two cases.

Assume first that C = I . Then A = E ⋅ E ⋅ ⋯ ⋅ E
1 2 m and AB = E1 ⋅ E2 ⋅ ⋯ ⋅ Em B . By applying the above
equality m times, and then m − 1 times, we have that
det AB = det E1 det E2 ⋅ det Em ⋅ det B
= det(E1 ⋅ E2 ⋅ ⋯ ⋅ Em ) det B
= det A det B.
Now assume C ≠ I . Since it is in reduced row-echelon form, its last row consists of zeros and by (4) of Example
[exa:EX1] the last row of C B consists of zeros. By Lemma [lem:L1] we have det C = det(C B) = 0 and therefore
det A = det(E1 ⋅ E2 ⋅ Em ) ⋅ det(C ) = det(E1 ⋅ E2 ⋅ Em ) ⋅ 0 = 0 (13.2.15)
and also
det AB = det(E1 ⋅ E2 ⋅ Em ) ⋅ det(C B) = det(E1 ⋅ E2 ⋅ ⋯ ⋅ Em )0 = 0 (13.2.16)
hence det AB = 0 = det A det B .
The same ‘machine’ used in the previous proof will be used again.
Theorem 13.2.10 :
Let A be a matrix where A is the transpose of A . Then,
T
T
det (A ) = det (A) (13.2.17)
Proof
Note first that the conclusion is true if A is elementary by (5) of Example [exa:EX1].
Let C be the reduced row-echelon form of A. Then we can write A = E 1 ⋅ E2 ⋅ ⋯ ⋅ Em C . Then
A
T
=C
T
⋅E
T
m ⋅⋯⋅E
T
2
⋅E1 . By Theorem [thm:T2] we have
T T T T
det(A ) = det(C ) ⋅ det(Em ) ⋅ ⋯ ⋅ det(E ) ⋅ det(E1 ). (13.2.18)
2
By (5) of Example [exa:EX1] we have that det E = det E for all j . Also, det C is either 0 or 1 (depending on
j
T
j
whether C = I or not) and in either case det C = det C . Therefore det A = det A .
T T
The above discussions allow us to now prove Theorem [thm:welldefineddeterminant]. It is restated below.
Theorem 13.2.11 :
Expanding an n × n matrix along any row or column always gives the same result, which is the determinant.
Proof
We first show that the determinant can be computed along any row. The case n = 1 does not apply and thus let n ≥ 2 .
Let Abe an n × n matrix and fix j > 1 . We need to prove that
n
det A = ∑ aj,i cof (A)j,i . (13.2.19)
i=1
Let us prove the case when j = 2 .

Let B be the matrix obtained from A by interchanging its 1st and 2nd rows. Then by Theorem [thm:T1] we have
det A = − det B. (13.2.20)
Now we have

n
det B = ∑ b1,i cof (B)1,i . (13.2.21)
i=1
Since B is obtained by interchanging the 1st and 2nd rows of A we have that b 1,i = a2,i for all i and one can see that
minor(B) = minor(A)
1,i . 2,i
Further,
1+i 2+i
cof (B)1,i = (−1 ) minorB1,i = −(−1 ) minor(A)2,i = −cof (A)2,i (13.2.22)
n n
hence det B = − ∑ i=1
a2,i cof (A)2,i , and therefore det A = − det B = ∑ i=1
a2,i cof (A)2,i as desired.
The case when j>2 is very similar; we still have minor(B)1,i = minor(A)j,i but checking that
det B = − ∑
n
i=1
aj,i cof (A)j,i is slightly more involved.
Now the cofactor expansion along column j of A is equal to the cofactor expansion along row j of A , which is by T
the above result just proved equal to the cofactor expansion along row 1 of A , which is equal to the cofactor
T
expansion along column 1 of A. Thus the cofactor cofactor along any column yields the same result.
Finally, since det A = det A by Theorem [thm:T.T], we conclude that the cofactor expansion along row 1 of A is
T
equal to the cofactor expansion along row 1 of A , which is equal to the cofactor expansion along column 1 of A.
T
Thus the proof is complete.

13.3: Finding Determinants using Row Operations
Theorems [thm:switchingrows], [thm:multiplyingrowbyscalar] and [thm:addingmultipleofrow] illustrate how row operations
affect the determinant of a matrix. In this section, we look at two examples where row operations are used to find the
determinant of a large matrix. Recall that when working with large matrices, Laplace Expansion is effective but timely, as
there are many steps involved. This section provides useful tools for an alternative method. By first applying row operations,
we can obtain a simpler matrix to which we apply Laplace Expansion.
While working through questions such as these, it is useful to record your row operations as you go along. Keep this in mind
as you read through the next example.
Example 13.3.1 : Finding a Determinant

Find the determinant of the matrix
1 2 3 4
⎡ ⎤
⎢5 1 2 3⎥
A =⎢ ⎥ (13.3.1)
⎢4 5 4 3⎥
⎣ ⎦
2 2 −4 5
Solution
We will use the properties of determinants outlined above to find det (A) . First, add −5 times the first row to the second
row. Then add −4 times the first row to the third row, and −2 times the first row to the fourth row. This yields the matrix
1 2 3 4
⎡ ⎤
⎢0 −9 −13 −17 ⎥
B =⎢ ⎥ (13.3.2)
⎢0 −3 −8 −13 ⎥
⎣ ⎦
0 −2 −10 −3
Notice that the only row operation we have done so far is adding a multiple of a row to another row. Therefore, by
Theorem [thm:addingmultipleofrow], det (B) = det (A) .
At this stage, you could use Laplace Expansion to find det (B) . However, we will continue with row operations to find an
even simpler matrix to work with.
Add −3 times the third row to the second row. By Theorem [thm:addingmultipleofrow] this does not change the value of
the determinant. Then, multiply the fourth row by −3. This results in the matrix
1 2 3 4
⎡ ⎤
⎢0 0 11 22 ⎥
C =⎢ ⎥ (13.3.3)
⎢0 −3 −8 −13 ⎥
⎣ ⎦
0 6 30 9
Here, det (C ) = −3 det (B) , which means that det (B) = (− 1
3
) det (C )
Since det (A) = det (B) , we now have that det (A) = (− ) det (C ) . Again, you could use Laplace Expansion here to
1
find det (C ) . However, we will continue with row operations.

Now replace the add 2 times the third row to the fourth row. This does not change the value of the determinant by
Theorem [thm:addingmultipleofrow]. Finally switch the third and second rows. This causes the determinant to be
multiplied by −1. Thus det (C ) = − det (D) where
1 2 3 4
⎡ ⎤
⎢0 −3 −8 −13 ⎥
D =⎢ ⎥ (13.3.4)
⎢0 0 11 22 ⎥
⎣ ⎦
0 0 14 −17

Hence, det (A) = (− 1
3
) det (C ) = (
1
3
) det (D)
You could do more row operations or you could note that this can be easily expanded along the first column. Then,
expand the resulting 3 × 3 matrix also along the first column. This results in
∣ 11 22
∣
det (D) = 1 (−3) ∣ ∣ = 1485 (13.3.5)
∣ 14 −17 ∣
and so det (A) = ( 1
3
) (1485) = 495.
You can see that by using row operations, we can simplify a matrix to the point where Laplace Expansion involves only a few
steps. In Example [exa:findingdeterminant], we also could have continued until the matrix was in upper triangular form, and
taken the product of the entries on the main diagonal. Whenever computing the determinant, it is useful to consider all the
possible methods and tools.
Consider the next example.
Example 13.3.2 : Find the Determinant

Find the determinant of the matrix
1 2 3 2
⎡ ⎤
⎢1 −3 2 1⎥
A =⎢ ⎥ (13.3.6)
⎢2 1 2 5⎥
⎣ ⎦
3 −4 1 2
Solution
Once again, we will simplify the matrix through row operations. Add −1 times the first row to the second row. Next add
−2 times the first row to the third and finally take −3 times the first row and add to the fourth row. This yields
1 2 3 2
⎡ ⎤
⎢0 −5 −1 −1 ⎥
B =⎢ ⎥ (13.3.7)
⎢0 −3 −4 1⎥
⎣ ⎦
0 −10 −8 −4
By Theorem [thm:addingmultipleofrow], det (A) = det (B) .

Remember you can work with the columns also. Take −5 times the fourth column and add to the second column. This
yields
1 −8 3 2
⎡ ⎤
⎢0 0 −1 −1 ⎥
C =⎢ ⎥ (13.3.8)
⎢0 −8 −4 1⎥
⎣ ⎦
0 10 −8 −4
By Theorem [thm:addingmultipleofrow] det (A) = det (C ) .

Now take −1 times the third row and add to the top row. This gives.
1 0 7 1
⎡ ⎤
⎢0 0 −1 −1 ⎥
D =⎢ ⎥ (13.3.9)
⎢0 −8 −4 1⎥
⎣ ⎦
0 10 −8 −4
which by Theorem [thm:addingmultipleofrow] has the same determinant as A .

Now, we can find det (D) by expanding along the first column as follows. You can see that there will be only one non
zero term.

0 −1 −1
⎡ ⎤
det (D) = 1 det ⎢ −8 −4 1 ⎥+0 +0 +0 (13.3.10)
⎣ ⎦
10 −8 −4
Expanding again along the first column, we have

−1 −1 −1 −1
det (D) = 1 (0 + 8 det [ ] + 10 det [ ]) = −82 (13.3.11)
−8 −4 −4 1
Now since det (A) = det (D) , it follows that det (A) = −82 .
Remember that you can verify these answers by using Laplace Expansion on A . Similarly, if you first compute the determinant
using Laplace Expansion, you can use the row operation method to verify.

13.4: Applications of the Determinant
Learning Objectives
1. Use determinants to determine whether a matrix has an inverse, and evaluate the inverse using cofactors.
2. Apply Cramer’s Rule to solve a 2 × 2 or a 3 × 3 linear system.
3. Given data points, find an appropriate interpolating polynomial and use it to estimate points.
A Formula for the Inverse

The determinant of a matrix also provides a way to find the inverse of a matrix. Recall the definition of the inverse of a matrix
in Definition [def:invertiblematrix]. We say that A , an n × n matrix, is the inverse of A , also n × n , if AA = I and
−1 −1
A =I.
−1
A
We now define a new matrix called the cofactor matrix of A . The cofactor matrix of A is the matrix whose ij th
entry is the
ij cofactor of A . The formal definition is as follows.
th
Definition 13.4.1 : The Cofactor Matrix

Let A = [a ij ] be an n × n matrix. Then the cofactor matrix of A, denoted cof (A), is defined by cof (A) = [cof (A) ij
]
where cof (A) is the ij cofactor of A .

ij
th
Note that cof (A) denotes the ij entry of the cofactor matrix.
ij
th
We will use the cofactor matrix to create a formula for the inverse of A . First, we define the adjugate of A to be the transpose
of the cofactor matrix. We can also call this matrix the classical adjoint of A , and we denote it by adj (A) .
In the specific case where A is a 2 × 2 matrix given by
a b
A =[ ] (13.4.1)
c d
then adj (A) is given by

d −b
adj (A) = [ ] (13.4.2)
−c a
In general, adj (A) can always be found by taking the transpose of the cofactor matrix of A . The following theorem provides
a formula for A using the determinant and adjugate of A .
−1
Theorem 13.4.1 : The Inverse and the Determinant

Let A be an n × n matrix. Then
A adj (A) = adj (A) A = det (A)I (13.4.3)
Moreover A is invertible if and only if det (A) ≠ 0 . In this case we have:
−1
1
A = adj (A) (13.4.4)
det (A)
Notice that the first formula holds for any n × n matrix A , and in the case A is invertible we actually have a formula for A −1
.
Example 13.4.1 : Find Inverse Using the Determinant

Find the inverse of the matrix
1 2 3
⎡ ⎤
A = ⎢3 0 1⎥ (13.4.5)
⎣ ⎦
1 2 1
using the formula in Theorem [thm:inverseanddeterminant].

Solution
According to Theorem [thm:inverseanddeterminant],
1
−1
A = adj (A) (13.4.6)
det (A)
First we will find the determinant of this matrix. Using Theorems [thm:switchingrows], [thm:multiplyingrowbyscalar],
and [thm:addingmultipleofrow], we can first simplify the matrix through row operations. First, add −3 times the first row
to the second row. Then add −1 times the first row to the third row to obtain
1 2 3
⎡ ⎤
B = ⎢0 −6 −8 ⎥ (13.4.7)
⎣ ⎦
0 0 −2
By Theorem [thm:addingmultipleofrow], det (A) = det (B) . By Theorem [thm:determinantoftriangularmatrix],

det (B) = 1 × −6 × −2 = 12 . Hence, det (A) = 12 .
Now, we need to find adj (A) . To do so, first we will find the cofactor matrix of A . This is given by
−2 −2 6
⎡ ⎤
cof (A) = ⎢ 4 −2 0⎥ (13.4.8)
⎣ ⎦
2 8 −6
Here, the ij entry is the ij cofactor of the original matrix

th th
A which you can verify. Therefore, from Theorem
[thm:inverseanddeterminant], the inverse of A is given by
1 1 1
T −
−2 −2 6 ⎡ 6 3 6 ⎤
⎡ ⎤
−1
1 1 1 2
A = ⎢ 4 −2 0⎥ = ⎢ − − ⎥ (13.4.9)
⎢ 6 6 3 ⎥
12
⎣ ⎦
2 8 −6 ⎣ 1 1 ⎦
0 −
2 2
Remember that we can always verify our answer for A

−1
. Compute the product AA
−1
and −1
A A and make sure each
product is equal to I .
Compute A −1
A as follows
1 1 1
⎡− 6 3 6 ⎤ 1 2 3 1 0 0
⎡ ⎤ ⎡ ⎤
−1 1 1 2
A A = ⎢ − − ⎥
1 ⎥ = ⎢0 0⎥ =I (13.4.10)
⎢ 6 6 3 ⎥⎢3 0 1
⎣ ⎦ ⎣ ⎦
⎣ 1 1 ⎦ 1 2 1 0 0 1
0 −
2 2
You can verify that AA −1

=I and hence our answer is correct.
We will look at another example of how to use this formula to find A −1

.
Example 13.4.2 : Find the Inverse From a Formula

Find the inverse of the matrix

1 1
0
⎡ 2 2 ⎤
1 1 1
A =⎢− − ⎥ (13.4.11)
⎢ 6 3 2 ⎥
⎣− 5 2 1 ⎦
−
6 3 2
using the formula given in Theorem [thm:inverseanddeterminant].

Solution
First we need to find det (A) . This step is left as an exercise and you should verify that det (A) =
1
6
. The inverse is
therefore equal to
−1
1
A = adj (A) = 6 adj (A) (13.4.12)
(1/6)
We continue to calculate as follows. Here we show the 2 × 2 determinants needed to find the cofactors.
T
1 1 1 1 1 1
∣ − ∣ ∣− − ∣ ∣− ∣
⎡ 3 2 6 2 6 3 ⎤
∣ ∣ ∣ ∣ ∣ ∣
−
2 ∣ 1 ∣ ∣ 5 1 ∣ ∣ 5 2 ∣ ⎥
⎢ −
⎢ 3
∣ 2
∣ ∣− 6
−
2
∣ ∣− 6 3
∣⎥
⎢ ⎥
⎢ 1 1 1 1 ⎥
⎢ ∣ 0 ∣ ∣ ∣ ∣ 0∣⎥
−1 ∣ 2 ∣ ∣ 2 2 ∣ ∣ 2 ∣⎥
A = 6⎢ − − (13.4.13)
⎢ ∣ 2 1 ∣ ∣ 5 1 ∣ ∣ 5 2 ∣ ⎥
⎢ ∣ 3 − ∣ ∣− − ∣ ∣− ∣⎥
⎢ 2 6 2 6 3 ⎥
⎢ ⎥
1 1 1 1
⎢ ∣ 0 ∣ ∣ ∣ ∣ 0∣⎥
⎢ ∣ 2 ∣ ∣ 2 2 ∣ ∣ 2 ∣⎥
−
⎣ ∣ 1 1 ∣ ∣ 1 1 ∣ ∣ 1 1 ∣ ⎦
∣ − ∣ ∣− − ∣ ∣− ∣
3 2 6 2 6 3
Expanding all the 2 × 2 determinants, this yields

T
1 1 1
⎡ 6 3 6 ⎤ 1 2 −1
⎡ ⎤
−1 1 1 1
A = 6⎢ − ⎥ = ⎢2 1 1⎥ (13.4.14)
⎢ 3 6 3 ⎥
⎣ ⎦
⎣− 1 1 1 ⎦ 1 −2 1
6 6 6
Again, you can always check your work by multiplying A −1

A and AA −1
and ensuring these products equal I .
1 1
⎡ 0 ⎤
1 2 −1 2 2 1 0 0
⎡ ⎤ ⎡ ⎤
−1 1 1 1
A A = ⎢2 ⎢ ⎥ =
1 1⎥⎢− −
⎥ ⎢0 1 0⎥ (13.4.15)
6 3 2
⎣ ⎦ ⎣ ⎦
1 −2 1 ⎣− 5 2 1 ⎦ 0 0 1
−
6 3 2
This tells us that our calculation for A −1

is correct. It is left to the reader to verify that AA −1
=I .
The verification step is very important, as it is a simple way to check your work! If you multiply A A and AA and these −1 −1
products are not both equal to I , be sure to go back and double check each step. One common error is to forget to take the
transpose of the cofactor matrix, so be sure to complete this step.
We will now prove Theorem [thm:inverseanddeterminant].
The Inverse and the Determinant

Proof
(of Theorem [thm:inverseanddeterminant]) Recall that the (i, j)-entry of adj(A) is equal to cof (A) . Thus the ji (i, j) -
entry of B = A ⋅ adj(A) is :
n n
Bij = ∑ aik adj(A)kj = ∑ aik cof (A)jk (13.4.16)
k=1 k=1
By the cofactor expansion theorem, we see that this expression for B is equal to the determinant of the matrix ij
obtained from A by replacing its j th row by a , a , … a — i.e., its ith row.

i1 i2 in

If i = j then this matrix is A itself and therefore B = det A . If on the other hand ii i ≠j , then this matrix has its ith
row equal to its j th row, and therefore B = 0 in his case. Thus we obtain:
ij
A adj (A) = det (A)I (13.4.17)
Similarly we can verify that:

adj (A) A = det (A)I (13.4.18)
And this proves the first part of the theorem.

Further if A is invertible, then by Theorem [thm:determinantofproduct] we have:
−1 −1
1 = det (I ) = det (AA ) = det (A) det (A ) (13.4.19)
and thus det (A) ≠ 0 . Equivalently, if det (A) = 0 , then A is not invertible.
Finally if det (A) ≠ 0 , then the above formula shows that A is invertible and that:
1
−1
A = adj (A) (13.4.20)
det (A)
This completes the proof.
This method for finding the inverse of A is useful in many contexts. In particular, it is useful with complicated matrices where
the entries are functions, rather than numbers.
Example 13.4.1 : Inverse for Non-Constant Matrix

Suppose
t
e 0 0
⎡ ⎤
A (t) = ⎢ 0 cos t sin t ⎥ (13.4.21)

⎣ ⎦
0 − sin t cos t
−1
Show that A(t) exists and then find it.
Solution
−1
First note det (A (t)) = e t 2
(cos t + sin
2
t) = e
t
≠0 so A(t) exists.
The cofactor matrix is
1 0 0
⎡ ⎤
t t
C (t) = ⎢ 0 e cos t e sin t ⎥ (13.4.22)
⎣ t t ⎦
0 −e sin t e cos t
and so the inverse is

T −t
1 0 0 e 0 0
⎡ ⎤ ⎡ ⎤
1
t t
⎢0 e cos t e sin t ⎥ =⎢ 0 cos t − sin t ⎥ (13.4.23)
t
e
⎣ t t ⎦ ⎣ ⎦
0 −e sin t e cos t 0 sin t cos t
Cramer’s Rule
Another context in which the formula given in Theorem [thm:inverseanddeterminant] is important is Cramer’s Rule. Recall
that we can represent a system of linear equations in the form AX = B , where the solutions to this system are given by X.
Cramer’s Rule gives a formula for the solutions X in the special case that A is a square invertible matrix. Note this rule does

not apply if you have a system of equations in which there is a different number of equations than variables (in other words,
when A is not square), or when A is not invertible.
Suppose we have a system of equations given by AX = B , and we want to find solutions X which satisfy this system. Then
recall that if A exists,
−1
AX = B
−1 −1
A (AX) = A B
−1 −1
(A A) X = A B
−1
IX = A B
−1
X = A B
Hence, the solutions X to the system are given by X = A B . Since we assume that A
−1 −1
exists, we can use the formula for
A
−1
given above. Substituting this formula into the equation for X, we have
−1
1
X =A B = adj (A) B (13.4.24)
det (A)
Let x be the i
i
th
entry of X and b be the j
j
th
entry of B . Then this equation becomes
n n
−1
1
xi = ∑ [ aij ] bj = ∑ adj(A) bj (13.4.25)
ij
det (A)
j=1 j=1
where adj(A) is the ij entry of adj (A) .

ij
th
By the formula for the expansion of a determinant along a column,
∗ ⋯ b1 ⋯ ∗
⎡ ⎤
1
xi = det ⎢
⎢
⎥ (13.4.26)
⋮ ⋮ ⋮ ⎥
det (A)
⎣ ⎦
∗ ⋯ bn ⋯ ∗
T
where here the i column of A is replaced with the column vector [b
th
1 ⋯ ⋅, bn ] . The determinant of this modified matrix is
taken and divided by det (A) . This formula is known as Cramer’s rule.
We formally define this method now.
Procedure 13.4.1 : Using Cramer’s Rule
Suppose A is an n × n invertible matrix and we wish to solve the system AX = B for T

X = [ x1 , ⋯ , xn ] . Then
Cramer’s rule says
det (Ai )
xi = (13.4.27)
det (A)
where A is the matrix obtained by replacing the i

i
th
column of A with the column matrix
b1
⎡ ⎤
B =⎢ ⎥
⎢ ⋮ ⎥ (13.4.28)
⎣ ⎦
bn
We illustrate this procedure in the following example.
Example 13.4.1
Find x, y, z if

1 2 1 x 1
⎡ ⎤⎡ ⎤ ⎡ ⎤
⎢3 2 1 ⎥⎢ y ⎥ = ⎢2 ⎥ (13.4.29)
⎣ ⎦⎣ ⎦ ⎣ ⎦
2 −3 2 z 3
Solution
We will use method outlined in Procedure [proc:cramersrule] to find the values for x, y, z which give the solution to this
system. Let
1
⎡ ⎤
B = ⎢2 ⎥ (13.4.30)
⎣ ⎦
3
In order to find x, we calculate

det (A1 )
x = (13.4.31)
det (A)
where A is the matrix obtained from replacing the first column of A with B .
1
Hence, A is given by
1
1 2 1
⎡ ⎤
A1 = ⎢ 2 2 1⎥ (13.4.32)
⎣ ⎦
3 −3 2
Therefore,
∣1 2 1∣
∣ ∣
2 2 1
∣ ∣
det (A1 ) ∣3 −3 2∣ 1
x = = = (13.4.33)
det (A) ∣1 2 1∣ 2
∣ ∣
3 2 1
∣ ∣
∣2 −3 2∣
Similarly, to find y we construct A by replacing the second column of A with B . Hence, A is given by
2 2
1 1 1
⎡ ⎤
A2 = ⎢ 3 2 1⎥ (13.4.34)
⎣ ⎦
2 3 2
Therefore,
∣1 1 1∣
∣ ∣
3 2 1
∣ ∣
det (A2 ) ∣2 3 2∣ 1
y = = =− (13.4.35)
det (A) ∣1 2 1∣ 7
∣ ∣
3 2 1
∣ ∣
∣2 −3 2∣
Similarly, A is constructed by replacing the third column of A with B . Then, A is given by

3 3
1 2 1
⎡ ⎤
A3 = ⎢ 3 2 2⎥ (13.4.36)
⎣ ⎦
2 −3 3
Therefore, z is calculated as follows.

∣1 2 1∣
∣ ∣
3 2 2
∣ ∣
det (A3 ) ∣2 −3 3∣ 11
z = = = (13.4.37)
det (A) ∣1 2 1∣ 14
∣ ∣
3 2 1
∣ ∣
∣2 −3 2∣
Cramer’s Rule gives you another tool to consider when solving a system of linear equations.
We can also use Cramer’s Rule for systems of non linear equations. Consider the following system where the matrix A
has functions rather than numbers for entries.
Using Cramer’s Rulecramersrule
Example 13.4.1
Solve for z if
1 0 0 x 1
⎡ ⎤⎡ ⎤ ⎡ ⎤
t t
⎢0 e cos t e sin t ⎥ ⎢ y ⎥ = ⎢ t ⎥ (13.4.38)
⎣ t t ⎦⎣ ⎦ ⎣ 2 ⎦
0 −e sin t e cos t z t
Solution
We are asked to find the value of z in the solution. We will solve using Cramer’s rule. Thus
∣1 0 1 ∣
∣ t ∣
0 e cos t t
∣ ∣
t 2
∣0 −e sin t t ∣
−t
z = \vspace.05in = t ((cos t) t + sin t) e (13.4.39)
∣1 0 0 ∣
∣ t t
∣
0 e cos t e sin t
∣ ∣
t t
∣0 −e sin t e cos t ∣
Use Cramer’s Rule for Non-Constant Matrixcramersrulenonconstantmatrix
Polynomial Interpolation
In studying a set of data that relates variables x and y , it may be the case that we can use a polynomial to “fit” to the data. If
such a polynomial can be established, it can be used to estimate values of x and y which have not been provided.
Polynomial Interpolationpolynomialinterpolation Given data points (1, 4), (2, 9), (3, 12), find an interpolating polynomial
p(x) of degree at most 2 and then estimate the value corresponding to x = . 1
We want to find a polynomial given by

2
p(x) = r0 + r1 x1 + r2 x (13.4.40)
2
such that p(1) = 4, p(2) = 9 and p(3) = 12 . To find this polynomial, substitute the known values in for x and solve for
r , r , and r .
0 1 2
p(1) = r0 + r1 + r2 = 4
p(2) = r0 + 2 r1 + 4 r2 = 9
p(3) = r0 + 3 r1 + 9 r2 = 12
Writing the augmented matrix, we have

⎡ 1 1 1 4 ⎤
⎢ 1 2 4 9 ⎥ (13.4.41)
⎢ ⎥
⎣ ⎦
1 3 9 12
After row operations, the resulting matrix is
⎡ 1 0 0 −3 ⎤
⎢ 0 1 0 8 ⎥ (13.4.42)
⎢ ⎥
⎣ 0 0 1 −1 ⎦
Therefore the solution to the system is r 0 = −3, r1 = 8, r2 = −1 and the required interpolating polynomial is
2
p(x) = −3 + 8x − x (13.4.43)
To estimate the value for x = 1
2
, we calculate p( 1
2
) :
1 1 1 2
p( ) = −3 + 8( )−( )
2 2 2
1
= −3 + 4 −
4
3
=
4
This procedure can be used for any number of data points, and any degree of polynomial. The steps are outlined below.
Finding an Interpolating Polynomialfindinginterpolynomial Suppose that values of x and corresponding values of y are given,
such that the actual relationship between x and y is unknown. Then, values of y can be estimated using an interpolating
polynomial p(x). If given x , . . . , x and the corresponding y , . . . , y , the procedure to find p(x) is as follows:
1 n 1 n
1. The desired polynomial p(x) is given by

2 n−1
p(x) = r0 + r1 x + r2 x +. . . +rn−1 x (13.4.44)
2. p(xi) = yi for all i = 1, 2, . . . , n so that

2 n−1
r0 + r1 x1 + r2 x +. . . +rn−1 x = y1
1 1
2 n−1
r0 + r1 x2 + r2 x +. . . +rn−1 x = y2
2 2
(13.4.45)
2 n−1
r0 + r1 xn + r2 xn +. . . +rn−1 xn = yn
3. Set up the augmented matrix of this system of equations

2 n−1
⎡ 1 x1 x ⋯ x y1 ⎤
1 1
⎢ 2 n−1 ⎥
⎢ 1 x2 x ⋯ x y2 ⎥
2 2
⎢ ⎥ (13.4.46)
⎢ ⎥
⎢ ⎥
⎢ ⋮ ⋮ ⋮ ⋮ ⋮ ⎥
2 n−1
⎣ 1 xn xn ⋯ xn yn ⎦
4. Solving this system will result in a unique solution r0 , r1 , ⋯ , rn−1 . Use these values to construct p(x) , and estimate the
value of p(a) for any x = a .
This procedure motivates the following theorem.
Polynomial Interpolationpolynomialinterpolation Given n data points (x , y ), (x , y 1 1 2 2 ), ⋯ , (xn , yn ) with the distinct,
xi
there is a unique polynomial p(x) = r + r x + r x + ⋯ + r x

0 1 such that
2
2
n−1
n−1
p(xi ) = yi for i = 1, 2, ⋯ , n. The
resulting polynomial p(x) is called the interpolating polynomial for the data points.
We conclude this section with another example.

Polynomial Interpolationpolynomialinterpolation2 Consider the data points (0, 1), (1, 2), (3, 22), (5, 66). Find an interpolating
polynomial p(x) of degree at most three, and estimate the value of p(2).
The desired polynomial p(x) is given by:
2 3
p(x) = r0 + r1 x + r2 x + r3 x (13.4.47)
Using the given points, the system of equations is
p(0) = r0 = 1
p(1) = r0 + r1 + r2 + r3 = 2
p(3) = r0 + 3 r1 + 9 r2 + 27 r3 = 22
p(5) = r0 + 5 r1 + 25 r2 + 125 r3 = 66
The augmented matrix is given by:
⎡ 1 0 0 0 1 ⎤
⎢ 1 1 1 1 2 ⎥
⎢ ⎥ (13.4.48)
⎢ ⎥
⎢ 1 3 9 27 22 ⎥
⎣ 1 5 25 125 66 ⎦
The resulting matrix is
⎡ 1 0 0 0 1 ⎤
⎢ 0 1 0 0 −2 ⎥
⎢ ⎥ (13.4.49)
⎢ ⎥
⎢ 0 0 1 0 3 ⎥
⎣ ⎦
0 0 0 1 0
Therefore, r0 = 1, r1 = −2, r2 = 3, r3 = 0 and p(x) = 1 − 2x + 3x

2
. To estimate the value of p(2) , we compute
2
p(2) = 1 − 2(2) + 3(2 ) = 1 − 4 + 12 = 9 .

13.E: Exercises
9/11/2020 13.E.1 https://math.libretexts.org/@go/page/44328

CHAPTER OVERVIEW
14: VECTOR SPACES
14.1: ALGEBRAIC CONSIDERATIONS

14.2: SPANNING SETS
In this section we will examine the concept of spanning introduced earlier in terms of Rn . Here,
we will discuss these concepts in terms of abstract vector spaces.
14.3: LINEAR INDEPENDENCE

In this section, we will again explore concepts introduced earlier in terms of Rn and extend them
to apply to abstract vector spaces.
14.4: SUBSPACES AND BASIS

In this section we will examine the concept of subspaces introduced earlier in terms of Rn. Here, we will discuss these concepts in
terms of abstract vector spaces.
14.5: SUMS AND INTERSECTIONS

14.6: LINEAR TRANSFORMATIONS
1 9/16/2020
14.1: Algebraic Considerations
Learning Objectives
Develop the abstract concept of a vector space through axioms.
Deduce basic properties of vector spaces.
Use the vector space axioms to determine if a set and its operations constitute a vector space.
In this section we consider the idea of an abstract vector space. A vector space is something which has two operations
satisfying the following vector space axioms.
Definition 14.1.1 : Vector Space

A vector space V is a set of vectors with two operations defined, addition and scalar multiplication, which satisfy the
axioms of addition and scalar multiplication.
In the following definition we define two operations; vector addition, denoted by + and scalar multiplication denoted by
placing the scalar next to the vector. A vector space need not have usual operations, and for this reason the operations will
always be given in the definition of the vector space. The below axioms for addition (written +) and scalar multiplication must
hold for however addition and scalar multiplication are defined for the vector space.
It is important to note that we have seen much of this content before, in terms of R . We will prove in this section that R is
n n
an example of a vector space and therefore all discussions in this chapter will pertain to R . While it may be useful to consider
n
all concepts of this chapter in terms of R , it is also important to understand that these concepts apply to all vector spaces.
n
In the following definition, we will choose scalars a, b to be real numbers and are thus dealing with real vector spaces.
However, we could also choose scalars which are complex numbers. In this case, we would call the vector space V complex.
Definition 14.1.1 : Axioms of Addition

Let v,⃗ w⃗ , z ⃗ be vectors in a vector space V . Then they satisfy the following axioms of addition:
Closed under Addition
If v,⃗ w⃗ are in V , then v ⃗ + w⃗ is also in V . (14.1.1)
The Commutative Law of Addition
v ⃗ + w⃗ = w⃗ + v ⃗ (14.1.2)
The Associative Law of Addition

(v⃗ + w⃗ ) + z ⃗ = v⃗ + (w⃗ + z )
⃗ (14.1.3)
The Existence of an Additive Identity

⃗
v ⃗ + 0 = v ⃗ (14.1.4)
The Existence of an Additive Inverse
⃗ = 0⃗
v ⃗ + (−v) (14.1.5)
vectorspaceaxiomsaddition
Definition Axioms of Scalar Multiplication

Let a, b ∈ R and let v,⃗ w⃗ , z ⃗ be vectors in a vector space V . Then they satisfy the following axioms of scalar
multiplication:
Closed under Scalar Multiplication
If a is a real number, and v ⃗ is in V , then av ⃗ is in V . (14.1.6)
a (v ⃗ + w⃗ ) = av ⃗ + aw⃗ (14.1.7)
(a + b) v ⃗ = av ⃗ + b v ⃗ (14.1.8)
⃗ = (ab)v ⃗
a (b v) (14.1.9)
1 v ⃗ = v ⃗ (14.1.10)
Consider the following example, in which we prove that R is in fact a vector space.
n
Example 14.1.1 : R n
Rnvectorspace R , under the usual operations of vector addition and scalar multiplication, is a vector space.
n
Solution
To show that R is a vector space, we need to show that the above axioms hold. Let
n
x⃗ , y ,⃗ z ⃗ be vectors in R
n
. We first
prove the axioms for vector addition.
To show that R is closed under addition, we must show that for two vectors in R their sum is also in R . The sum
n n n
x⃗ + y ⃗ is given by:
x1 y1 x1 + y1
⎛ ⎞ ⎛ ⎞ ⎛ ⎞
⎜ x2 ⎟ ⎜ y2 ⎟ ⎜ x2 + y2 ⎟
⎜ ⎟+⎜ ⎟ =⎜ ⎟ (14.1.11)
⎜ ⎟ ⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠ ⎝ ⎠
xn yn xn + yn
The sum is a vector with n entries, showing that it is in R . Hence R is closed under vector addition.
n n
To show that addition is commutative, consider the following:

x1 y1
⎛ ⎞ ⎛ ⎞
⎜ x2 ⎟ ⎜ y2 ⎟
x⃗ + y ⃗ = ⎜ ⎟+⎜ ⎟
⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠
xn yn
x1 + y1
⎛ ⎞
⎜ x2 + y2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn + yn
y1 + x1
⎛ ⎞
⎜ y2 + x2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
yn + xn
y1 x1
⎛ ⎞ ⎛ ⎞
⎜ y2 ⎟ ⎜ x2 ⎟
= ⎜ ⎟+⎜ ⎟
⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠
yn xn
= y ⃗ + x⃗
Hence addition of vectors in R is commutative.

n

We will show that addition of vectors in R is associative in a similar way.
n
x1 y1 z1
⎛⎛ ⎞ ⎛ ⎞⎞ ⎛ ⎞
⎜⎜ x2 ⎟ ⎜ y2 ⎟⎟ ⎜ z2 ⎟
(x⃗ + y )
⃗ + z ⃗ = ⎜⎜ ⎟+⎜ ⎟⎟ + ⎜ ⎟
⎜⎜ ⎟ ⎜ ⎟⎟ ⎜ ⎟
⎜⎜ ⋮ ⎟ ⎜ ⋮ ⎟⎟ ⎜ ⋮ ⎟
⎝⎝ ⎠ ⎝ ⎠⎠ ⎝ ⎠
xn yn zn
x1 + y1 z1
⎛ ⎞ ⎛ ⎞
⎜ x2 + y2 ⎟ ⎜ z2 ⎟
= ⎜ ⎟+⎜ ⎟
⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠
xn + yn zn
(x1 + y1 ) + z1
⎛ ⎞
⎜ (x2 + y2 ) + z2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
(xn + yn ) + zn
x1 + (y1 + z1 )
⎛ ⎞
⎜ x2 + (y2 + z2 ) ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn + (yn + zn )
x1 y1 + z1
⎛ ⎞ ⎛ ⎞
⎜ x2 ⎟ ⎜ y2 + z2 ⎟
= ⎜ ⎟+⎜ ⎟
⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠
xn yn + zn
x1 y1 z1
⎛ ⎞ ⎛⎛ ⎞ ⎛ ⎞⎞
⎜ x2 ⎟ ⎜⎜ y2 ⎟ ⎜ z2 ⎟⎟
= ⎜ ⎟ + ⎜⎜ ⎟+⎜ ⎟⎟
⎜ ⎟ ⎜⎜ ⎟ ⎜ ⎟⎟
⎜ ⋮ ⎟ ⎜⎜ ⋮ ⎟ ⎜ ⋮ ⎟⎟
⎝ ⎠ ⎝⎝ ⎠ ⎝ ⎠⎠
xn yn zn
= x⃗ + (y ⃗ + z )
⃗
Hence addition of vectors is associative.

0
⎛ ⎞
⎜0⎟
Next, we show the existence of an additive identity. Let ⃗
0 =⎜
⎜
⎟.
⎟
⎜ ⋮ ⎟
⎝ ⎠
0

x1 0
⎛ ⎞ ⎛ ⎞
⎜ x2 ⎟ ⎜0⎟
⃗ ⎜ ⎟+⎜ ⎟
x⃗ + 0 =
⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠
xn 0
x1 + 0
⎛ ⎞
⎜ x2 + 0 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn + 0
x1
⎛ ⎞
⎜ x2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn
= x⃗
Hence the zero vector 0⃗ is an additive identity.

−x1
⎛ ⎞
⎜ −x2 ⎟
Next, we prove the existence of an additive inverse. Let −x⃗ = ⎜
⎜
⎟
⎟
.
⎜ ⋮ ⎟
⎝ ⎠
−xn
x1 −x1
⎛ ⎞ ⎛ ⎞
⎜ x2 ⎟ ⎜ −x2 ⎟
x⃗ + (−x⃗ ) = ⎜ ⎟+⎜ ⎟
⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠
xn −xn
x1 − x1
⎛ ⎞
⎜ x2 − x2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn − xn
0
⎛ ⎞
⎜0⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
0
⃗
= 0
Hence −x⃗ is an additive inverse.

We now need to prove the axioms related to scalar multiplication. Let a, b be real numbers and let x⃗ , y ⃗ be vectors in R . n
We first show that R is closed under scalar multiplication. To do so, we show that ax⃗ is also a vector with n entries.
n
x1 ax1
⎛ ⎞ ⎛ ⎞
⎜ x2 ⎟ ⎜ ax2 ⎟
ax⃗ = a ⎜ ⎟ =⎜ ⎟ (14.1.12)
⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠
xn axn
The vector ax⃗ is again a vector with n entries, showing that R is closed under scalar multiplication.
n
We wish to show that a(x⃗ + y )⃗ = ax⃗ + ay ⃗ .

x1 x1
⎛⎛ ⎞ ⎛ ⎞⎞
⎜⎜ x2 ⎟ ⎜ x2 ⎟⎟
a(x⃗ + y )
⃗ = a ⎜⎜ ⎟+⎜ ⎟⎟
⎜⎜ ⎟ ⎜ ⎟⎟
⎜⎜ ⋮ ⎟ ⎜ ⋮ ⎟⎟
⎝⎝ ⎠ ⎝ ⎠⎠
xn xn
x1 + y1
⎛ ⎞
⎜ x2 + y2 ⎟
= a⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn + yn
a(x1 + y1 )
⎛ ⎞
⎜ a(x2 + y2 ) ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
a(xn + yn )
ax1 + ay1
⎛ ⎞
⎜ ax2 + ay2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
axn + ayn
ax1 ay1
⎛ ⎞ ⎛ ⎞
⎜ ax2 ⎟ ⎜ ay2 ⎟
= ⎜ ⎟+⎜ ⎟
⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠
axn ayn
= ax⃗ + ay ⃗
Next, we wish to show that (a + b)x⃗ = ax⃗ + bx⃗ .

x1
⎛ ⎞
⎜ x2 ⎟
(a + b)x⃗ = (a + b) ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn
(a + b)x1
⎛ ⎞
⎜ (a + b)x2 ⎟
⎜ ⎟
=
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
(a + b)xn
ax1 + b x1
⎛ ⎞
⎜ ax2 + b x2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
axn + b xn
ax1 bx1
⎛ ⎞ ⎛ ⎞
⎜ ax2 ⎟ ⎜ bx2 ⎟
= ⎜ ⎟+⎜ ⎟
⎜ ⎟ ⎜ ⎟
⎜ ⋮ ⎟ ⎜ ⋮ ⎟
⎝ ⎠ ⎝ ⎠
axn bxn
= ax⃗ + b x⃗
We wish to show that a(bx⃗ ) = (ab)x⃗ .

x1
⎛ ⎛ ⎞⎞
⎜ ⎜ x2 ⎟⎟
a(b x⃗ ) = a ⎜b ⎜ ⎟⎟
⎜ ⎜ ⎟⎟
⎜ ⎜ ⋮ ⎟⎟
⎝ ⎝ ⎠⎠
xn
bx1
⎛⎛ ⎞⎞
⎜⎜ bx2 ⎟⎟
= a ⎜⎜ ⎟⎟
⎜⎜ ⎟⎟
⎜⎜ ⋮ ⎟⎟
⎝⎝ ⎠⎠
bxn
a(b x1 )
⎛ ⎞
⎜ a(b x2 ) ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
a(b xn )
(ab)x1
⎛ ⎞
⎜ (ab)x2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
(ab)xn
x1
⎛ ⎞
⎜ x2 ⎟
= (ab) ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn
= (ab)x⃗
Finally, we need to show that 1x⃗ = x⃗ .

x1
⎛ ⎞
⎜ x2 ⎟
1x⃗ = 1⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn
1x1
⎛ ⎞
⎜ 1x2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
1xn
x1
⎛ ⎞
⎜ x2 ⎟
= ⎜ ⎟
⎜ ⎟
⎜ ⋮ ⎟
⎝ ⎠
xn
= x⃗
By the above proofs, it is clear that R satisfies the vector space axioms. Hence,
n
R
n
is a vector space under the usual
operations of vector addition and scalar multiplication.
We now consider some examples of vector spaces.
Example 14.1.1 : Vector Space of Polynomials

Let P be the set of all polynomials of at most degree 2 as well as the zero polynomial. Define addition to be the standard
2
addition of polynomials, and scalar multiplication the usual multiplication of a polynomial by a number. Then P is a 2
vector space.

Solution
We can write P explicitly as
2
2
P2 = { a2 x + a1 x + a0 | ai ∈ R for all i} (14.1.13)
To show that P is a vector space, we verify the axioms. Let p(x), q(x), r(x) be polynomials in P and let
2 2 a, b, c be real
numbers. Write p(x) = p x + p x + p , q(x) = q x + q x + q , and r(x) = r x + r x + r .
2
2
1 0 2
2
1 0 2
2
1 0
We first prove that addition of polynomials in P is closed. For two polynomials in P we need to show that their sum
2 2
is also a polynomial in P . From the definition of P , a polynomial is contained in P if it is of degree at most 2 or the
2 2 2
zero polynomial.
2 2
p(x) + q(x) = p2 x + p1 x + p0 + q2 x + q1 x + q0
2
= (p2 + q2 )x + (p1 + q1 )x + (p0 + q0 )
The sum is a polynomial of degree 2 and therefore is in P . It follows that P is closed under addition.
2 2
We need to show that addition is commutative, that is p(x) + q(x) = q(x) + p(x) .
2 2
p(x) + q(x) = p2 x + p1 x + p0 + q2 x + q1 x + q0
2
= (p2 + q2 )x + (p1 + q1 )x + (p0 + q0 )
2
= (q2 + p2 )x + (q1 + p1 )x + (q0 + p0 )
2 2
= q2 x + q1 x + q0 + p2 x + p1 x + p0
= q(x) + p(x)
Next, we need to show that addition is associative. That is, that (p(x) + q(x)) + r(x) = p(x) + (q(x) + r(x)) .
2 2 2
(p(x) + q(x)) + r(x) = (p2 x + p1 x + p0 + q2 x + q1 x + q0 ) + r2 x + r1 x + r0
2 2
= (p2 + q2 )x + (p1 + q1 )x + (p0 + q0 ) + r2 x + r1 x + r0
2
= (p2 + q2 + r2 )x + (p1 + q1 + r1 )x + (p0 + q0 + r0 )
2 2
= p2 x + p1 x + p0 + (q2 + r2 )x + (q1 + r1 )x + (q0 + r0 )
2 2 2
= p2 x + p1 x + p0 + (q2 x + q1 x + q0 + r2 x + r1 x + r0 )
= p(x) + (q(x) + r(x))
Next, we must prove that there exists an additive identity. Let 0(x) = 0x 2
+ 0x + 0 .
2 2
p(x) + 0(x) = p2 x + p1 x + p0 + 0 x + 0x + 0
2
= (p2 + 0)x + (p1 + 0)x + (p0 + 0)
2
= p2 x + p1 x + p0
= p(x)
Hence an additive identity exists, specifically the zero polynomial.

Next we must prove that there exists an additive inverse. Let −p(x) = −p2 x
2
− p1 x − p0 and consider the
following:
2 2
p(x) + (−p(x)) = p2 x + p1 x + p0 + (−p2 x − p1 x − p0 )
2
= (p2 − p2 )x + (p1 − p1 )x + (p0 − p0 )
2
= 0x + 0x + 0
= 0(x)
Hence an additive inverse −p(x) exists such that p(x) + (−p(x)) = 0(x) .
We now need to verify the axioms related to scalar multiplication.
First we prove that P is closed under scalar multiplication. That is, we show that ap(x) is also a polynomial of degree
2
at most 2.
2 2
ap(x) = a (p2 x + p1 x + p0 ) = ap2 x + ap1 x + ap0 (14.1.14)

Therefore P is closed under scalar multiplication.
2
We need to show that a(p(x) + q(x)) = ap(x) + aq(x) .

2 2
a(p(x) + q(x)) = a (p2 x + p1 x + p0 + q2 x + q1 x + q0 )
2
= a ((p2 + q2 )x + (p1 + q1 )x + (p0 + q0 ))
2
= a(p2 + q2 )x + a(p1 + q1 )x + a(p0 + q0 )
2
= (ap2 + aq2 )x + (ap1 + aq1 )x + (ap0 + aq0 )
2 2
= ap2 x + ap1 x + ap0 + aq2 x + aq1 x + aq0
= ap(x) + aq(x)
Next we show that (a + b)p(x) = ap(x) + bp(x) .

2
(a + b)p(x) = (a + b)(p2 x + p1 x + p0 )
2
= (a + b)p2 x + (a + b)p1 x + (a + b)p0
2 2
= ap2 x + ap1 x + ap0 + b p2 x + b p1 x + b p0
= ap(x) + bp(x)
The next axiom which needs to be verified is a(bp(x)) = (ab)p(x).

2
a(bp(x)) = a (b (p2 x + p1 x + p0 ))
2
= a (b p2 x + b p1 x + b p0 )
2
= ab p2 x + ab p1 x + ab p0
2
= (ab) (p2 x + p1 x + p0 )
= (ab)p(x)
Finally, we show that 1p(x) = p(x).

2
1p(x) = 1 (p2 x + p1 x + p0 )
2
= 1 p2 x + 1 p1 x + 1 p0
2
= p2 x + p1 x + p0
= p(x)
Since the above axioms hold, we know that P as described above is a vector space.
2
Another important example of a vector space is the set of all matrices of the same size.
Example 14.1.1 : Vector Space of Matrices

Let M 2,3 be the set of all 2 × 3 matrices. Using the usual operations of matrix addition and scalar multiplication, show
that M 2,3 is a vector space.
Solution
Let A, B be 2 × 3 matrices in M 2,3 . We first prove the axioms for addition.
In order to prove that M is closed under matrix addition, we show that the sum
2,3 A+B is in M2,3 . This means
showing that A + B is a 2 × 3 matrix.
a11 a12 a13 b11 b12 b13
A+B = ( ) +( )
a21 a22 a23 b21 b22 b23
a11 + b11 a12 + b12 a13 + b13

= ( )
a21 + b21 a22 + b22 a23 + b23
You can see that the sum is a 2 × 3 matrix, so it is in M . It follows that M is closed under matrix addition.
2,3 2,3
The remaining axioms regarding matrix addition follow from properties of matrix addition. Therefore M satisfies 2,3
the axioms of matrix addition.

We now turn our attention to the axioms regarding scalar multiplication. Let A, B be matrices in M 2,3 and let c be a real
number.
We first show that M 2,3 is closed under scalar multiplication. That is, we show that cA a 2 × 3 matrix.
a11 a12 a13

cA = c( )
a21 a22 a23
ca11 ca12 ca13

= ( )
ca21 ca22 ca23
This is a 2 × 3 matrix in M 2,3 which proves that the set is closed under scalar multiplication.
The remaining axioms of scalar multiplication follow from properties of scalar multiplication of matrices. Therefore
M 2,3satisfies the axioms of scalar multiplication.
In conclusion, M 2,3 satisfies the required axioms and is a vector space.
While here we proved that the set of all 2 × 3 matrices is a vector space, there is nothing special about this choice of matrix
size. In fact if we instead consider M , the set of all m × n matrices, then M
m,n is a vector space under the operations of
m,n
matrix addition and scalar multiplication.

We now examine an example of a set that does not satisfy all of the above axioms, and is therefore not a vector space.
Example 14.1.1 : Not a Vector Space

Let V denote the set of 2 × 3 matrices. Let addition in V be defined by A + B = A for matrices A, B in V . Let scalar
multiplication in V be the usual scalar multiplication of matrices. Show that V is not a vector space.
Solution
In order to show that V is not a vector space, it suffices to find only one axiom which is not satisfied. We will begin by
examining the axioms for addition until one is found which does not hold. Let A, B be matrices in V .
We first want to check if addition is closed. Consider A + B . By the definition of addition in the example, we have
that A + B = A . Since A is a 2 × 3 matrix, it follows that the sum A + B is in V , and V is closed under addition.
We now wish to check if addition is commutative. That is, we want to check if A + B = B + A for all choices of A
and B in V . From the definition of addition, we have that A + B = A and B + A = B . Therefore, we can find A , B
in V such that these sums are not equal. One example is
1 0 0 0 0 0
A =( ),B =( ) (14.1.15)
0 0 0 1 0 0
Using the operation defined by A + B = A , we have

A+B = A
1 0 0
= ( )
0 0 0
B+A = B
0 0 0
= ( )
1 0 0
It follows that A + B ≠ B + A . Therefore addition as defined for V is not commutative and V fails this axiom.
Hence V is not a vector space.
Consider another example of a vector space.
Example 14.1.1

Let S be a nonempty set and define F to be the set of real functions defined on S . In other words, we write F
S S : S ↦ R .
Letting a, b, c be scalars and f , g, h functions, the vector operations are defined as
(f + g) (x) = f (x) + g (x)
(af ) (x) = a (f (x))
Show that F is a vector space.

S
Solution
To verify that F is a vector space, we must prove the axioms beginning with those for addition. Let f , g, h be functions
S
in F .
S
First we check that addition is closed. For functions f , g defined on the set S , their sum given by
(f + g)(x) = f (x) + g(x) (14.1.16)
is again a function defined on S . Hence this sum is in F and F is closed under addition.
S S
Secondly, we check the commutative law of addition:
(f + g) (x) = f (x) + g (x) = g (x) + f (x) = (g + f ) (x) (14.1.17)
Since x is arbitrary, f + g = g + f .
Next we check the associative law of addition:
((f + g) + h) (x) = (f + g) (x) + h (x) = (f (x) + g (x)) + h (x) (14.1.18)
= f (x) + (g (x) + h (x)) = (f (x) + (g + h) (x)) = (f + (g + h)) (x) (14.1.19)
and so (f + g) + h = f + (g + h) .
Next we check for an additive identity. Let 0 denote the function which is given by 0 (x) = 0. Then this is an additive
identity because
(f + 0) (x) = f (x) + 0 (x) = f (x) (14.1.20)
and so f + 0 = f .
Finally, check for an additive inverse. Let −f be the function which satisfies (−f ) (x) = −f (x) . Then
(f + (−f )) (x) = f (x) + (−f ) (x) = f (x) + −f (x) = 0 (14.1.21)
Hence f + (−f ) = 0 .
Now, check the axioms for scalar multiplication.
We first need to check that F is closed under scalar multiplication. For a function f (x) in F and real number a , the
S S
function (af )(x) = a(f (x)) is again a function defined on the set S . Hence a(f (x)) is in F and F is closed under
S S
scalar multiplication.
((a + b) f ) (x) = (a + b) f (x) = af (x) + bf (x) = (af + bf ) (x) (14.1.22)
and so (a + b) f = af + bf .
(a (f + g)) (x) = a (f + g) (x) = a (f (x) + g (x)) (14.1.23)
= af (x) + bg (x) = (af + bg) (x) (14.1.24)
and so a (f + g) = af + bg .
((ab) f ) (x) = (ab) f (x) = a (bf (x)) = (a (bf )) (x) (14.1.25)
so (abf ) = a (bf ) .
Finally (1f ) (x) = 1f (x) = f (x) so 1f =f .
It follows that V satisfies all the required axioms and is a vector space.
Consider the following important theorem.

Theorem 14.1.1 : Uniqueness
In any vector space, the following are true:
1. 0⃗ , the additive identity, is unique
2. −x⃗ , the additive inverse, is unique
3. 0x⃗ = 0⃗ for all vectors x⃗
4. (−1) x⃗ = −x⃗ for all vectors x⃗
Proof
When we say that the additive identity, 0⃗ , is unique, we mean that if a vector acts like the additive identity, then it is
the additive identity. To prove this uniqueness, we want to show that another vector which acts like the additive
identity is actually equal to 0⃗ .
′
Suppose 0⃗ is also an additive identity. Then,

′
⃗ ⃗ ⃗
0+0 = 0 (14.1.26)
Now, for 0⃗ the additive identity given above in the axioms, we have that
′ ′
⃗ ⃗ ⃗
0 +0 = 0 (14.1.27)
So by the commutative property:

′ ′ ′
0 = 0 +0 = 0 +0 = 0 (14.1.28)
This says that if a vector acts like an additive identity (such as 0⃗ ), it in fact equals 0⃗ . This proves the uniqueness of 0⃗ .
When we say that the additive inverse, −x⃗ , is unique, we mean that if a vector acts like the additive inverse, then it is
the additive inverse. Suppose that y ⃗ acts like an additive inverse:
⃗
x⃗ + y ⃗ = 0 (14.1.29)
Then the following holds:

⃗ ⃗ = −x⃗ + 0⃗ = −x⃗
y ⃗ = 0 + y ⃗ = (−x⃗ + x⃗ ) + y ⃗ = −x⃗ + (x⃗ + y ) (14.1.30)
Thus if y ⃗ acts like the additive inverse, it is equal to the additive inverse −x⃗ . This proves the uniqueness of −x⃗ .
This statement claims that for all vectors x⃗ , scalar multiplication by 0 equals the zero vector 0 . Consider the
⃗
following, using the fact that we can write 0 = 0 + 0 :
0 x⃗ = (0 + 0) x⃗ = 0 x⃗ + 0 x⃗ (14.1.31)
We use a small trick here: add −0x⃗ to both sides. This gives
0 x⃗ + (−0 x⃗ ) = 0 x⃗ + 0 x⃗ + (−x⃗ )
⃗
0+0 = 0 x⃗ + 0
⃗
0 = 0x⃗
This proves that scalar multiplication of any vector by 0 results in the zero vector 0⃗ .
Finally, we wish to show that scalar multiplication of −1 and any vector x⃗ results in the additive inverse of that
vector, −x⃗ . Recall from 2. above that the additive inverse is unique. Consider the following:
(−1) x⃗ + x⃗ = (−1) x⃗ + 1 x⃗
= (−1 + 1) x⃗
= 0x⃗
⃗
= 0

By the uniqueness of the additive inverse shown earlier, any vector which acts like the additive inverse must be equal
to the additive inverse. It follows that (−1) x⃗ = −x⃗ .
An important use of the additive inverse is the following theorem.
Theorem 14.1.1
Let V be a vector space. Then v ⃗ + w⃗ = v ⃗ + z ⃗ implies that w⃗ = z ⃗ for all v,⃗ w⃗ , z ⃗ ∈ V
Proof
The proof follows from the vector space axioms, in particular the existence of an additive inverse (−u⃗ ). The proof is
left as an exercise to the reader.

14.2: Spanning Sets
Learning Objectives
Determine if a vector is within a given span.
In this section we will examine the concept of spanning introduced earlier in terms of R . Here, we will discuss these concepts
n
in terms of abstract vector spaces.

Definition 14.2.1 : Subset

Let X and Y be two sets. If all elements of X are also elements of Y then we say that X is a subset of Y and we write
X ⊆Y (14.2.1)
In particular, we often speak of subsets of a vector space, such as X ⊆ V . By this we mean that every element in the set X is
contained in the vector space V .
Definition 14.2.1 : Linear CombinationLinear Combination

Let V be a vector space and let v ⃗ , v ⃗ , ⋯ , v ⃗
1 2 n ⊆V . A vector v ⃗ ∈ V is called a linear combination of the v ⃗ if there exist
i
scalars c ∈ R such that

i
v⃗ = c1 v⃗1
+ c2 v⃗2
+ ⋯ + cn v⃗n
(14.2.2)
This definition leads to our next concept of span.
Definition 14.2.1 : Span of Vectors

Let {v ⃗ , ⋯ , v ⃗ } ⊆ V . Then
1 n
⃗ , ⋯ , vn
span { v1 ⃗ } = { ∑ ci vi⃗ : ci ∈ R} (14.2.3)
i=1
When we say that a vector w⃗ is in span {v ⃗ , ⋯ , v ⃗ } we mean that w⃗ can be written as a linear combination of the v ⃗ . We say
1 n 1
that a collection of vectors {v ⃗ , ⋯ , v ⃗ } is a spanning set for V if V = span{v ⃗ , ⋯ , v ⃗ } .

1 n 1 n
Example 14.2.1 : Matrix Span

1 0 0 1
Let A = [ ] ,B=[ ] . Determine if A and B are in
0 2 1 0
1 0 0 0
span { M1 , M2 } = span {[ ],[ ]} (14.2.4)
0 0 0 1
Solution
First consider A . We want to see if scalars s, t can be found such that A = sM 1 + tM2 .
1 0 1 0 0 0
[ ] = s[ ]+t [ ] (14.2.5)
0 2 0 0 0 1

The solution to this equation is given by
1 = s
2 = t
and it follows that A is in span {M 1, M2 } .

Now consider B . Again we write B = sM 1 + tM2 and see if a solution can be found for s, t .
0 1 1 0 0 0
[ ] = s[ ]+t [ ] (14.2.6)
1 0 0 0 0 1
Clearly no values of s and t can be found such that this equation holds. Therefore B is not in span {M 1, .
M2 }
Consider another example.
Example 14.2.1 : Polynomial Span

Show that p(x) = 7x 2
+ 4x − 3 is in span {4x 2
+ x, x
2
− 2x + 3} .
Solution
To show that p(x) is in the given span, we need to show that it can be written as a linear combination of polynomials in
the span. Suppose scalars a, b existed such that
2 2 2
7x + 4x − 3 = a(4 x + x) + b(x − 2x + 3) (14.2.7)
If this linear combination were to hold, the following would be true:
4a + b = 7
a − 2b = 4
3b = −3
You can verify that a = 2, b = −1 satisfies this system of equations. This means that we can write p(x) as follows:
2 2 2
7x + 4x − 3 = 2(4 x + x) − (x − 2x + 3) (14.2.8)
Hence p(x) is in the given span.
Example 14.2.1 : Spanning Set

Let S = {x 2
+ 1, x − 2, 2 x
2
− x} . Show that S is a spanning set for P , the set of all polynomials of degree at most 2.
2
Solution
Let p(x) = ax + bx + c be an arbitrary polynomial in P . To show that S is a spanning set, it suffices to show that p(x)
2
2
can be written as a linear combination of the elements of S . In other words, can we find r, s, t such that:
2 2 2
p(x) = ax + bx + c = r(x + 1) + s(x − 2) + t(2 x − x) (14.2.9)
If a solution r, s, t can be found, then this shows that for any such polynomial p(x), it can be written as a linear
combination of the above polynomials and S is a spanning set.
2 2 2
ax + bx + c = r(x + 1) + s(x − 2) + t(2 x − x)
2 2
= rx + r + sx − 2s + 2tx − tx
2
= (r + 2t)x + (s − t)x + (r − 2s)
For this to be true, the following must hold:

a = r + 2t
b = s−t
c = r − 2s
To check that a solution exists, set up the augmented matrix and row reduce:
1 1
⎡ 1 0 0 a + 2b + c ⎤
⎡ 1 0 2 a ⎤ 2 2
⎢ 1 1 ⎥
⎢ 0 1 −1 b ⎥ → ⋯ → ⎢ 0 1 0 a− c ⎥ (14.2.10)
⎢ ⎥ 4 4
⎢ ⎥
⎣ 1 −2 0 c ⎦ 1 1
⎣ 0 0 1 a−b − c ⎦
4 4
Clearly a solution exists for any choice of a, b, c. Hence S is a spanning set for P . 2

14.3: Linear Independence
Learning Objectives
1. Determine if a set is linearly independent.
In this section, we will again explore concepts introduced earlier in terms of R

n
and extend them to apply to abstract vector
spaces.
Definition 14.3.1 : Linear Independence

Let V be a vector space. If {v ⃗ , ⋯ , v ⃗ } ⊆ V , then it is linearly independent if
1 n
⃗
∑ ai vi⃗ = 0 implies a1 = ⋯ = an = 0 (14.3.1)
i=1
where the a are real numbers.

i
The set of vectors is called linearly dependent if it is not linearly independent.
Example 14.3.1 : Linear Independence

Let S ⊆ P be a set of polynomials given by
2
2 2
S = {x + 2x − 1, 2 x − x + 3} (14.3.2)
Determine if S is linearly independent.

Solution
To determine if this set S is linearly independent, we write
2 2 2
a(x + 2x − 1) + b(2 x − x + 3) = 0 x + 0x + 0 (14.3.3)
If it is linearly independent, then a = b = 0 will be the only solution. We proceed as follows.

2 2 2
a(x + 2x − 1) + b(2 x − x + 3) = 0x + 0x + 0
2 2 2
ax + 2ax − a + 2b x − bx + 3b = 0x + 0x + 0
2 2
(a + 2b)x + (2a − b)x − a + 3b = 0x + 0x + 0
It follows that
a + 2b = 0
2a − b = 0
−a + 3b = 0
The augmented matrix and resulting are given by
⎡ 1 2 0 ⎤ ⎡ 1 0 0 ⎤
⎢ 2 −1 0 ⎥ → ⋯ → ⎢ 0 1 0 ⎥ (14.3.4)
⎢ ⎥ ⎢ ⎥
⎣ −1 3 0 ⎦ ⎣ 0 0 0 ⎦
Hence the solution is a = b = 0 and the set is linearly independent.
The next example shows us what it means for a set to be dependent.

Example 14.3.1 : Dependent Set
Determine if the set S given below is independent.
⎧ −1 1 1
⎪⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎫
⎪
S = ⎨⎢ 0 ⎥ , ⎢ 1 ⎥ , ⎢ 3 ⎥⎬ (14.3.5)
⎩
⎪⎣ ⎭
⎦ ⎣ ⎦ ⎣ ⎦⎪
1 1 5
Solution
To determine if S is linearly independent, we look for solutions to
−1 1 1 0
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
a⎢ 0 ⎥+b ⎢ 1 ⎥+c ⎢ 3 ⎥ = ⎢ 0 ⎥ (14.3.6)

⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 5 0
Notice that this equation has nontrivial solutions, for example a = 2 , b = 3 and c = −1 . Therefore S is dependent.
The following is an important result regarding dependent sets.
Lemma 14.3.1 : Dependent Sets

Let V be a vector space and suppose W = {v ⃗ , v ⃗ , ⋯ , v ⃗ } is a subset of V . Then W is dependent if and only if v ⃗ can
1 2 k i
be written as a linear combination of {v ⃗ , v ⃗ , ⋯ , v ⃗ , v ⃗ , ⋯ , v ⃗ } for some i ≤ k .

1 2 i−1 i+1 k
Revisit Example [exa:dependent] with this in mind. Notice that we can write one of the three vectors as a combination of the
others.
1 −1 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢3⎥ =2⎢ 0 ⎥+3 ⎢ 1 ⎥ (14.3.7)
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
5 1 1
By Lemma [lem:dependent] this set is dependent.

If we know that one particular set is linearly independent, we can use this information to determine if a related set is linearly
independent. Consider the following example.
Example 14.3.1 : Related Independent Sets

Let V be a vector space and suppose S ⊆ V is a set of linearly independent vectors given by S = {u⃗ , v,⃗ w⃗ } . Let R ⊆ V
be given by R = {2u⃗ − w⃗ , w⃗ + v,⃗ 3v ⃗ + u⃗ } . Show that R is also linearly independent.
1
Solution
To determine if R is linearly independent, we write
1
a(2 u⃗ − w⃗ ) + b(w⃗ + v⃗)
+ c(3 v⃗ + u⃗ ) = 0⃗ (14.3.8)
2
If the set is linearly independent, the only solution will be a = b = c = 0 . We proceed as follows.
1
⃗
a(2 u⃗ − w⃗ ) + b(w⃗ + v)
⃗ + c(3 v ⃗ + u⃗ ) = 0
2
1
⃗
2au⃗ − aw⃗ + b w⃗ + b v ⃗ + 3c v ⃗ + c u⃗ = 0
2
1
⃗
(2a + c)u⃗ + (b + 3c)v ⃗ + (−a + b)w⃗ = 0
2
We know that the set S = {u⃗ , v,⃗ w⃗ } is linearly independent, which implies that the coefficients in the last line of this
equation must all equal 0. In other words:

1
2a + c = 0
2
b + 3c = 0
−a + b = 0
The augmented matrix and resulting are given by:

1
⎡ 2 0 0 ⎤ ⎡ 1 0 0 0 ⎤
2
⎢ ⎥ → ⋯ → ⎢ ⎥ (14.3.9)
⎢ 0 1 3 0 ⎥ ⎢ 0 1 0 0 ⎥
⎣ ⎦ ⎣ 0 0 1 0 ⎦
−1 1 0 0
Hence the solution is a = b = c = 0 and the set is linearly independent.
The following theorem was discussed in terms in R . We consider it here in the general case.
n
Theorem 14.3.1 : Unique Representation

Let V be a vector space and let U = {v ⃗ , ⋯ , v ⃗ } ⊆ V be an independent set. If
1 k v ⃗ ∈ span U , then v ⃗ can be written
uniquely as a linear combination of the vectors in U .
uniquerepresentation
Consider the span of a linearly independent set of vectors. Suppose we take a vector which is not in this span and add it to the
set. The following lemma claims that the resulting set is still linearly independent.
Lemma 14.3.1 : Adding to a Linearly Independent Set

Suppose v ⃗ ∉ span {u⃗ , ⋯ , u⃗ } and {u⃗ , ⋯ , u⃗ } is linearly independent. Then the set
1 k 1 k
{ u⃗ 1 , ⋯ , u⃗ k , v}
⃗ (14.3.10)
is also linearly independent.
Proof
Suppose ∑ c u⃗ + dv ⃗ = 0⃗ . It is required to verify that each
k
i=1 i i ci = 0 and that d = 0. But if d ≠ 0, then you can

solve for v ⃗ as a linear combination of the vectors, {u⃗ , ⋯ , u⃗ }, 1 k
k
ci
v ⃗ = − ∑ ( ) u⃗ i (14.3.11)
d
i=1
k
contrary to the assumption that v ⃗ is not in the span of the u⃗ . Therefore, d = 0. But then ∑i i=1
ci u⃗ i = 0
⃗
and the linear
independence of {u⃗ , ⋯ , u⃗ } implies each c = 0 also.
1 k i
Example 14.3.1 : Adding to a Linearly Independent Set

Let S ⊆ M 22 be a linearly independent set given by
1 0 0 1
S = {[ ],[ ]} (14.3.12)
0 0 0 0
Show that the set R ⊆ M 22 given by
1 0 0 1 0 0
R = {[ ],[ ],[ ]} (14.3.13)
0 0 0 0 1 0

is also linearly independent.
Solution
Instead of writing a linear combination of the matrices which equals 0 and showing that the coefficients must equal 0, we
can instead use Lemma [lem:addinglinearlyindependent].
To do so, we show that
0 0 1 0 0 1
[ ] ∉ span {[ ],[ ]} (14.3.14)
1 0 0 0 0 0
Write
0 0 1 0 0 1
[ ] = a[ ]+b [ ]
1 0 0 0 0 0
a 0 0 b
= [ ]+[ ]
0 0 0 0
a b
= [ ]
0 0
Clearly there are no possible a, b to make this equation true. Hence the new matrix does not lie in the span of the matrices
in S . By Lemma [lem:addinglinearlyindependent], R is also linearly independent.

14.4: Subspaces and Basis
Learning Objectives
1. Utilize the subspace test to determine if a set is a subspace of a given vector space.
2. Extend a linearly independent set and shrink a spanning set to a basis of a given vector space.
In this section we will examine the concept of subspaces introduced earlier in terms of R . Here, we will discuss these concepts in terms of abstract vector spaces.
n
Consider the definition of a subspace.
Definition 14.4.1 : Subspace

Let V be a vector space. A subset W ⊆V is said to be a subspace of V if ax⃗ + by ⃗ ∈ W whenever a, b ∈ R and x⃗ , y ⃗ ∈ W .
The span of a set of vectors as described in Definition [def:span] is an example of a subspace. The following fundamental result says that subspaces are subsets of a vector space which are
themselves vector spaces.
Theorem 14.4.1 : Subspaces are Vector Spaces

Let W be a nonempty collection of vectors in a vector space V . Then W is a subspace if and only if W satisfies the vector space axioms, using the same operations as those defined on V .
Proof
Suppose first that W is a subspace. It is obvious that all the algebraic laws hold on W because it is a subset of V and they hold on V . Thus u⃗ + v ⃗ = v ⃗ + u⃗ along with the other
axioms. Does W contain 0⃗ ? Yes because it contains 0u⃗ = 0⃗ . See Theorem [thm:axiomuniqueness].
Are the operations of V defined on W ? That is, when you add vectors of W do you get a vector in W ? When you multiply a vector in W by a scalar, do you get a vector in W ? Yes.
This is contained in the definition. Does every vector in W have an additive inverse? Yes by Theorem [thm:axiomuniqueness] because −v ⃗ = (−1) v ⃗ which is given to be in W
provided v ⃗ ∈ W .
Next suppose W is a vector space. Then by definition, it is closed with respect to linear combinations. Hence it is a subspace.
Consider the following useful Corollary.
Theorem 14.4.2 : Span is a Subspace

Let V be a vector space with W ⊆V . If W , ⋯ , v⃗n
= span { v⃗1 } then W is a subspace of V .
When determining spanning sets the following theorem proves useful.
Theorem 14.4.3 : Spanning Set

Let W ⊆V for a vector space V and suppose W , v⃗2
= span { v⃗1 , ⋯ , v⃗n
} .
Let U ⊆V be a subspace such that v⃗ , v⃗ , ⋯ , v⃗
1 2 n ∈ U . Then it follows that W ⊆U .
In other words, this theorem claims that any subspace that contains a set of vectors must also contain the span of these vectors.
The following example will show that two spans, described differently, can in fact be equal.
Example 14.4.1 : Equal Span

Let p(x), q(x) be polynomials and suppose U = span {2p(x) − q(x), p(x) + 3q(x)} and W = span {p(x), q(x)} . Show that U =W .
Solution
We will use Theorem 14.4.3 to show that U ⊆W and W ⊆U . It will then follow that U =W .
1. U ⊆W
Notice that 2p(x) − q(x) and p(x) + 3q(x) are both in W = span {p(x), q(x)} . Then by Theorem 14.4.3 W must contain the span of these polynomials and so U ⊆W .
2. W ⊆U
Notice that
3 2
p(x) = (2p(x) − q(x)) + (p(x) + 3q(x))
7 7
1 2
q(x) = − (2p(x) − q(x)) + (p(x) + 3q(x))
7 7
Hence p(x), q(x) are in span {2p(x) − q(x), p(x) + 3q(x)}. By Theorem 14.4.3 U must contain the span of these polynomials and so W ⊆U .
1.
To prove that a set is a vector space, one must verify each of the axioms given in Definition [def:vectorspaceaxiomsaddition] and [def:vectorspaceaxiomsscalarmult]. This is a cumbersome
task, and therefore a shorter procedure is used to verify a subspace.
Procedure 14.4.1 : Subspace Test

Suppose W is a subset of a vector space V . To determine if W is a subspace of V , it is sufficient to determine if the following three conditions hold, using the operations of V :
1. The additive identity 0⃗ of V is contained in W .

2. For any vectors w⃗ , w⃗ in W , w⃗ + w⃗ is also in W .
1 2 1 2
3. For any vector w⃗ in W and scalar a , the product aw⃗ is also in W .
1 1

Therefore it suffices to prove these three steps to show that a set is a subspace.
Example 14.4.1 : Improper Subspaces

Let V be an arbitrary vector space. Then V is a subspace of itself. Similarly, the set {0⃗ } containing only the zero vector is also a subspace.
Solution
Using the subspace test in Procedure [proc:subspacetest] we can show that V and {0⃗ } are subspaces of V .
Since V satisfies the vector space axioms it also satisfies the three steps of the subspace test. Therefore V is a subspace.
Let’s consider the set {0⃗ }.
1. The vector 0⃗ is clearly contained in {0⃗ }, so the first condition is satisfied.
2. Let w⃗ , w⃗ be in {0⃗ }. Then w⃗
1 2 1 = 0⃗ and w⃗ 2 = 0⃗ and so
⃗ ⃗ ⃗
w⃗ 1 + w⃗ 2 = 0 + 0 = 0 (14.4.1)
It follows that the sum is contained in ⃗

{0} and the second condition is satisfied.
3. Let w⃗ be in {0⃗ } and let a be an arbitrary scalar. Then
1
aw⃗ 1 = a0⃗ = 0⃗ (14.4.2)
Hence the product is contained in {0⃗ } and the third condition is satisfied.
It follows that {0⃗ } is a subspace of V .
The two subspaces described above are called improper subspaces. Any subspace of a vector space V which is not equal to V or {0⃗ } is called a proper subspace.
Consider another example.
Example 14.4.1 : Subspace of Polynomials

Let P be the vector space of polynomials of degree two or less. Let W
2 ⊆ P2 be all polynomials of degree two or less which have 1 as a root. Show that W is a subspace of P .
2
Solution
First, express W as follows:
2
W = {p(x) = ax + bx + c, a, b, c, ∈ R|p(1) = 0} (14.4.3)
We need to show that W satisfies the three conditions of Procedure [proc:subspacetest].

1. The zero polynomial of P is given by 0(x) = 0x + 0x + 0 = 0 . Clearly 0(1) = 0 so 0(x) is contained in W .
2
2
2. Let p(x), q(x) be polynomials in W . It follows that p(1) = 0 and q(1) = 0 . Now consider p(x) + q(x). Let r(x) represent this sum.
r(1) = p(1) + q(1)
= 0 +0
= 0
Therefore the sum is also in W and the second condition is satisfied.

3. Let p(x) be a polynomial in W and let a be a scalar. It follows that p(1) = 0 . Consider the product ap(x).
ap(1) = a(0)
= 0
Therefore the product is in W and the third condition is satisfied.

It follows that W is a subspace of P . 2
Recall the definition of basis, considered now in the context of vector spaces.
Definition 14.4.1 : Basis

Let V be a vector space. Then {v⃗ , ⋯ , v⃗ } is called a basis for V if the following conditions hold.
1 n
1. span {v⃗ , ⋯ , v⃗ } = V

1 n
2. {v⃗ , ⋯ , v⃗ } is linearly independent

1 n
Example 14.4.1 : Polynomials of Degree Two

Let P be the set polynomials of degree no more than 2. We can write P
2 2
2
= span { x , x, 1} . Is {x 2
, x, 1} a basis for P ?
2
Solution
It can be verified that P is a vector space defined under the usual addition and scalar multiplication of polynomials.
2
Now, since P 2
2
= span { x , x, 1} , the set {x 2
, x, 1} is a basis if it is linearly independent. Suppose then that
2 2
ax + bx + c = 0 x + 0x + 0 (14.4.4)
where a, b, c are real numbers. It is clear that this can only occur if a = b = c = 0 . Hence the set is linearly independent and forms a basis of P . 2

The next theorem is an essential result in linear algebra and is called the exchange theorem.
Theorem 14.4.1 : Exchange Theorem

Let {x⃗ , ⋯ , x⃗ } be a linearly independent set of vectors such that each x⃗ is contained in span{y ⃗ , ⋯ , y ⃗ } . Then r ≤ s.
1 r i 1 s
Proof
The proof will proceed as follows. First, we set up the necessary steps for the proof. Next, we will assume that r > s and show that this leads to a contradiction, thus requiring that
r ≤s.
Define span{y ⃗ , ⋯ , y ⃗ } = V . Since each x⃗ is in span{y ⃗ , ⋯ , y ⃗ }, it follows there exist scalars c , ⋯ , c such that
1 s i 1 s 1 s
x⃗ 1 = ∑ ci y i⃗ (14.4.5)
i=1
Note that not all of these scalars c can equal zero. Suppose that all the c
i i =0 . Then it would follow that x⃗ 1 = 0
⃗
and so { x⃗ 1 , ⋯ , x⃗ r } would not be linearly independent. Indeed, if
⃗
,
x⃗ 1 = 0 1 x⃗ 1 + ∑
r
i=2
0 x⃗ i = x⃗ 1 = 0
⃗
and so there would exist a nontrivial linear combination of the vectors {x⃗ , ⋯ , x⃗ } which equals zero. Therefore at least one c is nonzero.
1 r i
Say c k ≠ 0. Then solve [lincomb] for y ⃗ and obtain k
s-1 vectors here

⎧ ⎫
 ⎪
⎪
⃗ ∈ span ⎨x⃗ 1 , y ⃗ , ⋯ , y ⃗
yk ⃗
, y k+1 ⃗ ⎬ .
, ⋯ , ys (14.4.6)
1 k−1
⎩
⎪ ⎭
⎪
Define {z ⃗ , ⋯ , z ⃗
1 s−1 } to be
⃗ , ⋯ , z s−1
{z 1 ⃗ ⃗ , ⋯ , y ⃗
} = {y 1 ⃗
, y k+1 ⃗ }
, ⋯ , ys (14.4.7)
k−1
Now we can write

⃗ ∈ span { x⃗ 1 , z 1
yk ⃗ , ⋯ , z s−1
⃗ } (14.4.8)
Therefore, span { x⃗ 1 , z 1

⃗ , ⋯ , z s−1
⃗ } =V . To see this, suppose v⃗ ∈ V . Then there exist constants c 1, ⋯ , cs such that
s−1
v⃗ = ∑ ci z i⃗ + cs y k
⃗ . (14.4.9)
i=1
Replace this y ⃗ with a linear combination of the vectors {x⃗ , z ⃗ , ⋯ , z ⃗ } to obtain v⃗ ∈ span {x⃗ , z ⃗ , ⋯ , z ⃗ } . The vector
k 1 1 s−1 1 1 s−1
⃗ ,
yk in the list ⃗ , ⋯ , y ⃗ } ,
{y 1 s
has now been replaced
with the vector x⃗ and the resulting modified list of vectors has the same span as the original list of vectors, {y ⃗ , ⋯ , y ⃗ } .
1 1 s
We are now ready to move on to the proof. Suppose that r > s and that
span { x⃗ 1 , ⋯ , x⃗ l , z 1

⃗ , ⋯ , z p
⃗ } = V (14.4.10)
where the process established above has continued. In other words, the vectors z ⃗ , ⋯ , z ⃗ are each taken from the set {y ⃗ , ⋯ , y ⃗ } and
1 p 1 s
l + p = s. This was done for l =1 above.
Then since r > s, it follows that l ≤ s < r and so l + 1 ≤ r. Therefore, x⃗ is a vector not in the list, {x⃗ , ⋯ , x⃗ } and since
l+1 1 l
span { x⃗ 1 , ⋯ , x⃗ l , z 1

⃗ , ⋯ , z p
⃗ } = V (14.4.11)
there exist scalars, c and d such that i j
l p
x⃗ l+1 = ∑ ci x⃗ i + ∑ dj z j
⃗ . (14.4.12)
i=1 j=1
Not all the d can equal zero because if this were so, it would follow that {x⃗ , ⋯ , x⃗ } would be a linearly dependent set because one of the vectors would equal a linear combination
j 1 r
of the others. Therefore, [lincomb2] can be solved for one of the z ⃗ , say z ⃗ , in terms of x⃗ and the other z ⃗ and just as in the above argument, replace that z ⃗ with x⃗ to obtain
i k l+1 i i l+1
p-1 vectors here

⎧
⎪ ⎫
⎪  ⎪
⎪
span ⎨x⃗ 1 , ⋯ x⃗ l , x⃗ l+1 , z 1

⃗ , ⋯ z k−1
⃗ ⃗
, z k+1 ⃗ ⎬ = V
, ⋯ , zp (14.4.13)
⎪
⎩
⎪ ⎪
⎭
⎪
Continue this way, eventually obtaining
span { x⃗ 1 , ⋯ , x⃗ s } = V . (14.4.14)
But then x⃗ r ∈ span { x⃗ 1 , ⋯ , x⃗ s } contrary to the assumption that {x⃗ , ⋯ , x⃗ } is linearly independent. Therefore, r ≤ s as claimed.
1 r
The following corollary follows from the exchange theorem.
Corollary 14.4.1 : Two Bases of the Same Length

Let B , B be two bases of a vector space V . Suppose B contains m vectors and B contains n vectors. Then m = n .
1 2 1 2
Proof
By Theorem [thm:exchangetheorem], m ≤ n and n ≤ m . Therefore m = n .
This corollary is very important so we provide another proof independent of the exchange theorem above.
Proof
Suppose n > m. Then since the vectors {u⃗ , ⋯ , u⃗ 1 m} span V , there exist scalars c such that ij
∑ cij u⃗ i = v⃗j

. (14.4.15)
i=1
Therefore,
n n m
= 0⃗
∑ dj v⃗j if and only if ∑ ∑ cij dj u⃗ i = 0
⃗
(14.4.16)
j=1 j=1 i=1
if and only if

m n
⃗
∑ ( ∑ cij dj ) u⃗ i = 0 (14.4.17)
i=1 j=1
Now since {u⃗ , ⋯ , u⃗ } is independent, this happens if and only if

1 n
∑ cij dj = 0, i = 1, 2, ⋯ , m. (14.4.18)
j=1
However, this is a system of m equations in n variables, d 1, ⋯ , dn and m < n. Therefore, there exists a solution to this system of equations in which not all the d are equal to zero.
j
Recall why this is so. The augmented matrix for the system is of the form [ C
⃗
0 ] where C is a matrix which has more columns than rows. Therefore, there are free variables and
hence nonzero solutions to the system of equations. However, this contradicts the linear independence of {u⃗ , ⋯ , u⃗ }. Similarly it cannot happen that m > n .
1 m
Given the result of the previous corollary, the following definition follows.
Definition 14.4.1 : Dimension

A vector space V is of dimension n if it has a basis consisting of n vectors.
Notice that the dimension is well defined by Corollary [cor:baseslength]. It is assumed here that n < ∞ and therefore such a vector space is said to be finite dimensional.
Example 14.4.1 : Dimension of a Vector Space

Let P be the set of all polynomials of degree at most 2. Find the dimension of P .
2 2
Solution
If we can find a basis of P then the number of vectors in the basis will give the dimension. Recall from Example [exa:polydegreetwo] that a basis of P is given by
2 2
2
S = { x , x, 1} (14.4.19)
There are three polynomials in S and hence the dimension of P is three. 2
It is important to note that a basis for a vector space is not unique. A vector space can have many bases. Consider the following example.
Example 14.4.1 : A Different Basis for Polynomials of Degree Two

Let P be the polynomials of degree no more than 2. Is {x
2
2
+ x + 1, 2x + 1, 3 x
2
+ 1} a basis for P ? 2
Solution
Suppose these vectors are linearly independent but do not form a spanning set for P . Then by Lemma [lem:addinglinearlyindependent], we could find a fourth polynomial in P to create
2 2
a new linearly independent set containing four polynomials. However this would imply that we could find a basis of P of more than three polynomials. This contradicts the result of
2
Example [exa:dimension] in which we determined the dimension of P is three. Therefore if these vectors are linearly independent they must also form a spanning set and thus a basis for
2
P .
2
Suppose then that

2 2
a (x + x + 1) + b (2x + 1) + c (3 x + 1) = 0
2
(a + 3c) x + (a + 2b) x + (a + b + c) = 0
We know that {x 2
, x, 1} is linearly independent, and so it follows that
a + 3c = 0
a + 2b = 0
a+b +c = 0
and there is only one solution to this system of equations, a = b = c = 0 . Therefore, these are linearly independent and form a basis for P . 2
Consider the following theorem.
Theorem 14.4.1 : Every Subspace has a Basis

Let W be a nonzero subspace of a finite dimensional vector space V . Suppose V has dimension n . Then W has a basis with no more than n vectors.
Proof
Let v⃗ ∈ V where v⃗ ≠ 0. If span {v⃗ } = V , then it follows that {v⃗ } is a basis for V . Otherwise, there exists v⃗ ∈ V which is not in span {v⃗ } . By Lemma
1 1 1 1 2 1
[lem:addinglinearlyindependent] {v⃗ , v⃗ } is a linearly independent set of vectors. Then {v⃗ , v⃗ } is a basis for V and we are done. If span {v⃗ , v⃗ } ≠ V , then there exists
1 2 1 2 1 2
v⃗ ∉ span { v⃗ , v⃗ } and { v⃗ , v⃗ , v⃗ } is a larger linearly independent set of vectors. Continuing this way, the process must stop before n + 1 steps because if not, it would be possible to
3 1 2 1 2 3
obtain n + 1 linearly independent vectors contrary to the exchange theorem, Theorem [thm:exchangetheorem].
If in fact W has n vectors, then it follows that W =V .
Theorem 14.4.1 : Subspace of Same Dimension

Let V be a vector space of dimension n and let W be a subspace. Then W =V if and only if the dimension of W is also n .
Proof
First suppose W = V . Then obviously the dimension of W = n.
Now suppose that the dimension of W is n . Let a basis for W be {w⃗ , ⋯ , w⃗ }. If W is not equal to V , then let v⃗ be a vector of V which is not contained in W . Thus v⃗ is not in
1 n
span { w⃗ , ⋯ , w⃗ } and by Lemma [lem:basesisomorphism], { w⃗ , ⋯ , w⃗ , v⃗}

1 n
is linearly independent which contradicts Theorem [thm:exchangetheorem] because it would be an
1 n
independent set of n + 1 vectors even though each of these vectors is in a spanning set of n vectors, a basis of V .

Example 14.4.1 : Basis of a Subspace

∣ 1 0 1 1
Let U = {A ∈ M22 ∣ A [ ] =[ ] A} . Then U is a subspace of M 22 Find a basis of U , and hence dim(U ) .
∣ 1 −1 0 −1
Solution
a b
Let A = [ ] ∈ M22 . Then
c d
1 0 a b 1 0 a+b −b
A[ ] =[ ][ ] =[ ] (14.4.20)
1 −1 c d 1 −1 c +d −d
and
1 1 1 1 a b a+c b +d
[ ]A = [ ][ ] =[ ]. (14.4.21)
0 −1 0 −1 c d −c −d
a+b −b a+c b +d
If A ∈ U , then [ ] =[ ] .
c +d −d −c −d
Equating entries leads to a system of four equations in the four variables a, b, c and d .
\begin{array}{ccc} a+b & = & a + c \\ -b & = & b + d \\ c + d & = & -c \\ -d & = & -d \end{array} \hspace*{.2in}\mbox{ or }\hspace*{.2in} \begin{array}{rcc} b - c & = & 0 \\ -2b - d & = & 0 \\ 2c + d & = & 0 \e
The solution to this system is a = s , b = − 1
2
t ,c=− 1
2
t , d = t for any s, t ∈ R , and thus
t 1
s 1 0 0 −
2 2
A =[ ] = s[ ]+t [ ]. (14.4.22)
t 1
− t 0 0 − 1
2 2
Let
1
1 0 0 −
2
B = {[ ],[ ]} . (14.4.23)
1
0 0 − 1
2
Then span(B) = U , and it is routine to verify that B is an independent subset of M . Therefore B is a basis of U , and dim(U ) = 2 . 22
The following theorem claims that a spanning set of a vector space V can be shrunk down to a basis of V . Similarly, a linearly independent set within V can be enlarged to create a basis of V .
Theorem 14.4.1 : Basis of V

If V = span {u⃗ , ⋯ , u⃗ } is a vector space, then some subset of {u⃗ , ⋯ , u⃗ } is a basis for
1 n 1 n V . Also, if { u⃗ 1 , ⋯ , u⃗ k } ⊆ V is linearly independent and the vector space is finite
dimensional, then the set {u⃗ , ⋯ , u⃗ }, can be enlarged to obtain a basis of V .
1 k
Proof
Let
S = {E ⊆ { u⃗ 1 , ⋯ , u⃗ n } such that (14.4.24)
For E ∈ S, let |E| denote the number of elements of E. Let

m = min{|E| such that E ∈ S}. (14.4.25)
Thus there exist vectors

, ⋯ , v⃗m
{ v⃗1 } ⊆ { u⃗ 1 , ⋯ , (14.4.26)
such that
, ⋯ , v⃗m
span { v⃗1 } =V (14.4.27)
and m is as small as possible for this to happen. If this set is linearly independent, it follows it is a basis for V and the theorem is proved. On the other hand, if the set is not linearly
independent, then there exist scalars, c , ⋯ , c such that 1 m
⃗
0 = ∑ ci v⃗i (14.4.28)
i=1
and not all the c are equal to zero. Suppose c

i k ≠ 0. Then solve for the vector v⃗ in terms of the other vectors. Consequently,
k
, ⋯ , v⃗k−1
V = span { v⃗1
, v⃗k+1 }
, ⋯ , v⃗m (14.4.29)
contradicting the definition of m. This proves the first part of the theorem.
To obtain the second part, begin with {u⃗ , ⋯ , u⃗ } and suppose a basis for V is
1 k
, ⋯ , v⃗n
{ v⃗1 } (14.4.30)
If
span { u⃗ 1 , ⋯ , u⃗ k } = V , (14.4.31)
then k = n . If not, there exists a vector

u⃗ k+1 ∉ span { u⃗ 1 , ⋯ , u⃗ k } (14.4.32)
Then from Lemma [lem:addinglinearlyindependent], { u⃗ 1 , ⋯ , u⃗ k , u⃗ k+1 } is also linearly independent. Continue adding vectors in this way until n linearly independent vectors have
been obtained. Then
span { u⃗ 1 , ⋯ , u⃗ n } = V (14.4.33)
because if it did not do so, there would exist u⃗ as just described and {u⃗ , ⋯ , u⃗ } would be a linearly independent set of vectors having n + 1 elements. This contradicts the fact
n+1 1 n+1
that {v⃗ , ⋯ , v⃗ } is a basis. In turn this would contradict Theorem [thm:exchangetheorem]. Therefore, this list is a basis.
1 n

Recall Example [exa:addinglinind] in which we added a matrix to a linearly independent set to create a larger linearly independent set. By Theorem [thm:basisfromspanninglinind] we can
extend a linearly independent set to a basis.
Example 14.4.1 : Adding to a Linearly Independent Set

Let S ⊆ M 22 be a linearly independent set given by
1 0 0 1
S = {[ ],[ ]} (14.4.34)
0 0 0 0
Enlarge S to a basis of M . 22
Solution
Recall from the solution of Example [exa:addinglinind] that the set R ⊆ M 22 given by
1 0 0 1 0 0
R = {[ ],[ ],[ ]} (14.4.35)
0 0 0 0 1 0
0 0
is also linearly independent. However this set is still not a basis for M 22 as it is not a spanning set. In particular, [ ] is not in spanR. Therefore, this matrix can be added to the set by
0 1
Lemma [lem:addinglinearlyindependent] to obtain a new linearly independent set given by
1 0 0 1 0 0 0 0
T = {[ ],[ ],[ ],[ ]} (14.4.36)
0 0 0 0 1 0 0 1
This set is linearly independent and now spans M . Hence T is a basis.

22
Next we consider the case where you have a spanning set and you want a subset which is a basis. The above discussion involved adding vectors to a set. The next theorem involves removing
vectors.
Theorem 14.4.1 : Basis from a Spanning Set

Let V be a vector space and let W be a subspace. Also suppose that W = span { w⃗ 1 , ⋯ , w⃗ m } . Then there exists a subset of {w⃗ , ⋯ , w⃗ 1 m} which is a basis for W .
Proof
Let S denote the set of positive integers such that for k ∈ S, there exists a subset of {w⃗ , ⋯ , w⃗ } consisting of exactly k vectors which is a spanning set for W . Thus m ∈ S . Pick
1 m
the smallest positive integer in S . Call it k . Then there exists {u⃗ , ⋯ , u⃗ } ⊆ {w⃗ , ⋯ , w⃗ } such that \limfuncspan {u⃗ , ⋯ , u⃗ } = W . If
1 k 1 m 1 k
⃗
∑ ci w⃗ i = 0 (14.4.37)
i=1
and not all of the c i = 0, then you could pick cj ≠0 , divide by it and solve for u⃗ in terms of the others.
j
ci
w⃗ j = ∑ (− ) w⃗ i (14.4.38)
cj
i≠j
Then you could delete w⃗ from the list and have the same span. In any linear combination involving w⃗ , the linear combination would equal one in which w⃗ is replaced with the above
j j j
sum, showing that it could have been obtained as a linear combination of w⃗ for i ≠ j . Thus k − 1 ∈ S contrary to the choice of k . Hence each c = 0 and so {u⃗ , ⋯ , u⃗ } is a basis
i i 1 k
for W consisting of vectors of {w⃗ , ⋯ , w⃗ }.

1 m
Consider the following example of this concept.
Example 14.4.1 : Basis from a Spanning Set

Let V be the vector space of polynomials of degree no more than 3, denoted earlier as P . Consider the following vectors in V .
3
2 3 2 3 2
2x + x + 1, x + 4x + 2x + 2, 2 x + 2x + 2x + 1,
3 2 3 2
x + 4x − 3x + 2, x + 3x + 2x + 1
Then, as mentioned above, V has dimension 4 and so clearly these vectors are not linearly independent. A basis for V is {1, x, x 2 3
,x } . Determine a linearly independent subset of these
which has the same span. Determine whether this subset is a basis for V .
Solution
Consider an isomorphism which maps R to V in the obvious way. Thus
1
⎡ ⎤
⎢1⎥
⎢ ⎥ (14.4.39)
⎢2⎥
⎣ ⎦
0
corresponds to 2x 2
+x +1 through the use of this isomorphism. Then corresponding to the above vectors in V we would have the following vectors in R 4
.
1 2 1 2 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢ 1 ⎥ ⎢ 2 ⎥ ⎢ 2 ⎥ ⎢ −3 ⎥ ⎢ 2 ⎥
⎢ ⎥,⎢ ⎥,⎢ ⎥,⎢ ⎥,⎢ ⎥ (14.4.40)
⎢2⎥ ⎢4⎥ ⎢2⎥ ⎢ 4⎥ ⎢3⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦
0 1 2 1 1
Now if we obtain a subset of these which has the same span but which is linearly independent, then the corresponding vectors from V will also be linearly independent. If there are four in
the list, then the resulting vectors from V must be a basis for V . The for the matrix which has the above vectors as columns is
1 0 0 −15 0
⎡ ⎤
⎢0 1 0 11 0⎥
⎢ ⎥ (14.4.41)
⎢0 0 1 −5 0⎥
⎣ ⎦
0 0 0 0 1
Therefore, a basis for V consists of the vectors

2 3 2 3 2
2x + x + 1, x + 4x + 2x + 2, 2 x + 2x + 2x + 1,
3 2
x + 3x + 2x + 1.
Note how this is a subset of the original set of vectors. If there had been only three pivot columns in this matrix, then we would not have had a basis for V but we would at least have
obtained a linearly independent subset of the original set of vectors in this way.
Note also that, since all linear relations are preserved by an isomorphism,
2 3 2 3 2
−15 (2 x + x + 1) + 11 (x + 4x + 2x + 2) + (−5) (2 x + 2x + 2x + 1)
3 2
= x + 4x − 3x + 2
Example 14.4.1 : Shrinking a Spanning Set

Consider the set S ⊆ P given by
2
2 2
S = {1, x, x , x + 1} (14.4.42)
Show that S spans P , then remove vectors from S until it creates a basis.
2
Solution
First we need to show that S spans P . Let ax
2
2
+ bx + c be an arbitrary polynomial in P . Write 2
2 2 2
ax + bx + c = r(1) + s(x) + t(x ) + u(x + 1) (14.4.43)
Then,
2 2 2
ax + bx + c = r(1) + s(x) + t(x ) + u(x + 1)
2
= (t + u)x + s(x) + (r + u)
It follows that
a = t +u
b = s
c = r+u
Clearly a solution exists for all a, b, c and so S is a spanning set for P . By Theorem [thm:basisfromspanninglinind], some subset of S is a basis for P .
2 2
Recall that a basis must be both a spanning set and a linearly independent set. Therefore we must remove a vector from S keeping this in mind. Suppose we remove x from S . The
resulting set would be {1, x , x + 1} . This set is clearly linearly dependent (and also does not span P ) and so is not a basis.
2 2
2
Suppose we remove x + 1 from S . The resulting set is {1, x, x

2 2
} which is both linearly independent and spans P . Hence this is a basis for P . Note that removing any one of 1, x , or
2 2
2
x + 1 will result in a basis.

2
Now the following is a fundamental result about subspaces.
Theorem 14.4.1 : Basis of a Vector Space

Let V be a finite dimensional vector space and let W be a non-zero subspace. Then W has a basis. That is, there exists a linearly independent set of vectors {w⃗ , ⋯ , w⃗ } such that 1 r
\limfuncspan { w⃗ 1 , ⋯ , w⃗ r } = W (14.4.44)
Also if {w⃗ , ⋯ , w⃗ } is a linearly independent set of vectors, then W has a basis of the form {w⃗ , ⋯ , w⃗ , ⋯ , w⃗ } for r ≥ s .
1 s 1 s r
Proof
Let the dimension of V be n . Pick w⃗ ∈ W where w⃗ ≠ 0⃗ . If w⃗ , ⋯ , w⃗ have been chosen such that {w⃗ , ⋯ , w⃗ } is linearly independent, if \limfuncspan {w⃗ , ⋯ , w⃗ } = W ,
1 1 1 s 1 s 1 r
stop. You have the desired basis. Otherwise, there exists w⃗ ∉ \limfuncspan {w⃗ , ⋯ , w⃗ } and {w⃗ , ⋯ , w⃗ , w⃗ } is linearly independent. Continue this way until the process
s+1 1 s 1 s s+1
stops. It must stop since otherwise, you could obtain a linearly independent set of vectors having more than n vectors which is impossible.
The last claim is proved by following the above procedure starting with {w⃗ , ⋯ , w⃗ } as above.
1 s
This also proves the following corollary. Let V play the role of W in the above theorem and begin with a basis for W , enlarging it to form a basis for V as discussed above.
Corollary 14.4.1 : Basis Extension

Let W be any non-zero subspace of a vector space V . Then every basis of W can be extended to a basis for V .
Example 14.4.1 : Basis Extension

Let V =R
4
and let
⎧ 1 0
⎪⎡ ⎤ ⎡ ⎤⎫
⎪
⎪ ⎪
⎪ ⎪
⎢0⎥ ⎢ 1⎥
W = span ⎨⎢ ⎥,⎢ ⎥⎬ (14.4.45)
⎢1⎥ ⎢ 0 ⎥⎪
⎪
⎪
⎩⎣
⎪ ⎪
⎦ ⎣ ⎦⎪
⎭
1 1
Extend this basis of W to a basis of V .

Solution
An easy way to do this is to take the of the matrix

1 0 1 0 0 0
⎡ ⎤
⎢0 1 0 1 0 0⎥
⎢ ⎥ (14.4.46)
⎢1 0 0 0 1 0⎥
⎣ ⎦
1 1 0 0 0 1
Note how the given vectors were placed as the first two and then the matrix was extended in such a way that it is clear that the span of the columns of this matrix yield all of 4
R . Now
determine the pivot columns. The is
1 0 0 0 1 0
⎡ ⎤
⎢0 1 0 0 −1 1⎥
⎢ ⎥ (14.4.47)
⎢0 0 1 0 −1 0⎥
⎣ ⎦
0 0 0 1 1 −1
These are
1 0 1 0
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
⎢0⎥ ⎢1⎥ ⎢0⎥ ⎢1⎥

⎢ ⎥,⎢ ⎥,⎢ ⎥,⎢ ⎥ (14.4.48)
⎢1⎥ ⎢0⎥ ⎢0⎥ ⎢0⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 0 0
and now this is an extension of the given basis for W to a basis for R .
4
Why does this work? The columns of [vectorspaceeq1] obviously span R the span of the first four is the same as the span of all six.
4

14.5: Sums and Intersections
Learning Objectives
1. Show that the sum of two subspaces is a subspace.
2. Show that the intersection of two subspaces is a subspace.
We begin this section with a definition.
Definition 14.5.1 : Sum and Intersection

Let V be a vector space, and let U and W be subspaces of V . Then
1. U + W = {u⃗ + w⃗ | u⃗ ∈ U and w⃗ ∈ W } and is called the sum of U and W .
2. U ∩ W = {v⃗ |
v⃗ ∈ U and v⃗ ∈ W } and is called the intersection of U and W .
sumintersection
Therefore the intersection of two subspaces is all the vectors shared by both. If there are no vectors shared by both subspaces,
meaning that U ∩ W ⃗
= {0} , the sum U + W takes on a special name.
Definition 14.5.1 : Direct Sum

Let V be a vector space and suppose U and W are subspaces of V such that U ∩ W = {0⃗ } . Then the sum of U and W
is called the direct sum and is denoted U ⊕ W .
directsum
An interesting result is that both the sum U + W and the intersection U ∩ W are subspaces of V .
Example 14.5.1 : Intersection is a Subspace

Let V be a vector space and suppose U and W are subspaces. Then the intersection U ∩ W is a subspace of V .
Solution
By the subspace test, we must show three things:
1. 0⃗ ∈ U ∩ W
2. For vectors v⃗ , v⃗ ∈ U ∩ W , v⃗ + v⃗ ∈ U ∩ W
1 2 1 2
3. For scalar a and vector v⃗ ∈ U ∩ W , av⃗ ∈ U ∩ W

We proceed to show each of these three conditions hold.
1. Since U and W are subspaces of V , they each contain 0⃗ . By definition of the intersection, 0⃗ ∈ U ∩ W .
2. Let v ⃗ , v ⃗ ∈ U ∩ W , . Then in particular, v ⃗ , v ⃗ ∈ U . Since U is a subspace, it follows that v ⃗ + v ⃗ ∈ U . The same
1 2 1 2 1 2
argument holds for W . Therefore v ⃗ + v ⃗ is in both U and W and by definition is also in U ∩ W .

1 2
3. Let a be a scalar and v ⃗ ∈ U ∩ W . Then in particular, v ⃗ ∈ U . Since U is a subspace, it follows that av ⃗ ∈ U . The same
argument holds for W so av ⃗ is in both U and W . By definition, it is in U ∩ W .
intersectionsubspace
Therefore U ∩ W is a subspace of V .
It can also be shown that U + W is a subspace of V .

We conclude this section with an important theorem on dimension.
Theorem 14.5.1 : Dimension of Sum

Let V be a vector space with subspaces U and W . Suppose U and W each have finite dimension. Then U + W also has
finite dimension which is given by
dim(U + W ) = dim(U ) + dim(W ) − dim(U ∩ W ) (14.5.1)
dimensionsum
Notice that when U ∩ W ⃗

= {0} , the sum becomes the direct sum and the above equation becomes
dim(U ⊕ W ) = dim(U ) + dim(W ) (14.5.2)

14.6: Linear Transformations
Learning Objectives
1. Understand the definition of a linear transformation in the context of vector spaces.
Recall that a function is simply a transformation of a vector to result in a new vector. Consider the following definition.
Definition 14.6.1 : Linear Transformation

Let V and W be vector spaces. Suppose T : V ↦ W is a function, where for each x⃗ ∈ V , T (x⃗ ) ∈ W . Then T is a
linear transformation if whenever k, p are scalars and v ⃗ and v ⃗ are vectors in V
1 2
T (kv⃗1
+ p v⃗2
) = kT (v⃗1
) + pT (v⃗2
) (14.6.1)
Several important examples of linear transformations include the zero transformation, the identity transformation, and the
scalar transformation.
Example 14.6.1 : Linear Transformations

Let V and W be vector spaces.
1. 0 : V → W is defined by 0(v)⃗ = 0⃗ for all v ⃗ ∈ V .
2. 1 : V → V is defined by 1 (v)⃗ = v ⃗ for all v ⃗ ∈ V .
V V
3. The scalar transformation Let a ∈ R .

sa : V → V is defined by s
a (v)
⃗ = av ⃗ for all v ⃗ ∈ V .
Solution
We will show that the scalar transformation s is linear, the rest are left as an exercise.
a
By Definition [def:lineartransformation] we must show that for all scalars k, p and vectors ⃗
v1 and ⃗
v2 in V ,
s (kv ⃗ + p v ⃗ ) = ks (v ⃗ ) + p s (v ⃗ ) . Assume that a is also a scalar.
a 1 2 a 1 a 2
⃗ + p v2
sa (kv1 ⃗ ) = ⃗ + p v2
a (kv1 ⃗ )
= ⃗ + ap v2
akv1 ⃗
= ⃗ ) + p (av2
k (av1 ⃗ )
= ⃗ ) + p sa (v2
ksa (v1 ⃗ )
Therefore s is a linear transformation.

a
Consider the following important theorem.
Theorem 14.6.1 : Properties of Linear Transformations

Let V and W be vector spaces, and T : V ↦ W a linear transformation. Then
1. T preserves the zero vector.
⃗ ⃗
T (0) = 0 (14.6.2)
2. T preserves additive inverses. For all v ⃗ ∈ V ,

⃗ = −T (v)
T (−v) ⃗ (14.6.3)
3. T preserves linear combinations. For all v ⃗ , v ⃗ , … , v ⃗

1 2 m ∈ V and all k1, k2 , … , km ∈ R ,

T (k1 v⃗1
+ k2 v⃗2
+ ⋯ + km v⃗m
) = k1 T (v⃗1
) + k2 T (v⃗2
) + ⋯ + km T (v⃗m
). (14.6.4)
Proof
Let 0⃗ denote the zero vector of V and let 0⃗ denote the zero vector of W . We want to prove that T (0⃗
V W V
⃗
) = 0W . Let
v ⃗ ∈ V . Then ⃗
0 v ⃗ = 0V and
⃗ ⃗ = 0⃗ W .
⃗ = 0T (v)
T (0V ) = T (0 v) (14.6.5)
Let v ⃗ ∈ V ; then −v ⃗ ∈ V is the additive inverse of v ,⃗ so v ⃗ + (−v)⃗ = 0⃗ . Thus V
⃗
T (v ⃗ + (−v))
⃗ = T (0V )
⃗ + T (−v))
⃗ ⃗
T (v) = 0W
⃗ = ⃗ ⃗ = −T (v).
⃗
T (−v) 0W − T (v)
This result follows from preservation of addition and preservation of scalar multiplication. A formal proof would be
by induction on m.
Consider the following example using the above theorem.
Example 14.6.1 : Linear Combination

Let T : P2 → R be a linear transformation such that
2 2 2
T (x + x) = −1; T (x − x) = 1; T (x + 1) = 3. (14.6.6)
Find T (4x 2
+ 5x − 3) .
We provide two solutions to this problem.
Solution 1:
Suppose a(x 2
+ x) + b(x
2
− x) + c(x
2
+ 1) = 4 x
2
+ 5x − 3 . Then
2 2
(a + b + c)x + (a − b)x + c = 4 x + 5x − 3. (14.6.7)
Solving for a , b , and c results in the unique solution a = 6 , b = 1 , c = −3 .

Solution 2:
Notice that S = {x + x, x − x, x
2 2 2
+ 1} is a basis of P2 , and thus x
2
, x , and 1 can each be written as a linear
combination of elements of S .
2 1 2 1 2
x = (x + x) + (x − x)
2 2
1 2 1 2
x = (x + x) − (x − x)
2 2
2 1 2 1 2
1 = (x + 1) − (x + x) − (x − x).
2 2
Then
2 1 2 1 2 1 2 1 2
T (x ) = T ( (x + x) + (x − x)) = T (x + x) + T (x − x)
2 2 2 2
1 1
= (−1) + (1) = 0.
2 2
1 2 1 2 1 2 1 2
T (x) = T ( (x + x) − (x − x)) = T (x + x) − T (x − x)
2 2 2 2
1 1
= (−1) − (1) = −1.
2 2
2 1 2 1 2
T (1) = T ((x + 1) − (x + x) − (x − x))
2 2
2 1 2 1 2
= T (x + 1) − T (x + x) − T (x − x)
2 2
1 1
= 3− (−1) − (1) = 3.
2 2
Therefore,

2 2
T (4 x + 5x − 3) = 4T (x ) + 5T (x) − 3T (1)
= 4(0) + 5(−1) − 3(3) = −14.
The advantage of Solution 2 over Solution 1 is that if you were now asked to find T (−6x 2
− 13x + 9) , it is easy to use
T (x ) = 0 , T (x) = −1 and T (1) = 3 :
2
2 2
T (−6 x − 13x + 9) = −6T (x ) − 13T (x) + 9T (1)
= −6(0) − 13(−1) + 9(3) = 13 + 27 = 40.
More generally,
2 2
T (ax + bx + c) = aT (x ) + bT (x) + cT (1)
= a(0) + b(−1) + c(3) = −b + 3c.
Suppose two linear transformations act in the same way on v ⃗ for all vectors. Then we say that these transformations are equal.
Definition 14.6.1 : Equal Transformations

Let S and T be linear transformations from V to W . Then S = T if and only if for every v ⃗ ∈ V ,
S (v⃗)
= T (v⃗)
(14.6.8)
The definition above requires that two transformations have the same action on every vector in order for them to be equal. The
next theorem argues that it is only necessary to check the action of the transformations on basis vectors.
Theorem 14.6.1 : Transformation of a Spanning Set

Let V and W be vector spaces and suppose that S and T are linear transformations from V to W . Then in order for S
and T to be equal, it suffices that S(v ⃗ ) = T (v ⃗ ) where V = \funcspan{v ⃗ , v ⃗ , … , v ⃗ }.

i i 1 2 n
This theorem tells us that a linear transformation is completely determined by its actions on a spanning set. We can also
examine the effect of a linear transformation on a basis.
Theorem 14.6.1 : Transformation of a Basis

Suppose V and W are vector spaces and let {w⃗ , w⃗ , … , w⃗ } be any given vectors in W that may not be distinct. Then
1 2 n
there exists a basis {v ⃗ , v ⃗ , … , v ⃗ } of V and a unique linear transformation T : V ↦ W with T (v ⃗ ) = w⃗ .

1 2 n i i
Furthermore, if
v⃗ = k1 v⃗1
+ k2 v⃗2
+ ⋯ + kn v⃗n
(14.6.9)
is a vector of V , then
T (v⃗)
= k1 w⃗ 1 + k2 w⃗ 2 + ⋯ + kn w⃗ n . (14.6.10)

CHAPTER OVERVIEW
APPENDICES AND ANSWERS TO EXERCISES
A BRIEF TABLE OF INTEGRALS

INDEX
SECTION 1.2 ANSWERS
SECTION 2.1 ANSWERS
SECTION 2.2 ANSWERS
SECTION 2.3 ANSWERS
SECTION 2.4 ANSWERS
SECTION 2.5 ANSWERS
SECTION 2.6 ANSWERS
SECTION 3.1 ANSWERS
SECTION 3.2 ANSWERS
SECTION 3.3 ANSWERS
SECTION 4.1 ANSWERS
SECTION 4.2 ANSWERS
SECTION 4.3 ANSWERS
SECTION 4.4 ANSWERS
SECTION 4.5 ANSWERS
SECTION 5.1 ANSWERS
SECTION 5.2 ANSWERS
SECTION 5.3 ANSWERS
SECTION 5.4 ANSWERS
SECTION 5.5 ANSWERS
SECTION 5.6 ANSWERS
SECTION 5.7 ANSWERS
SECTION 6.1 ANSWERS
SECTION 6.2 ANSWERS
SECTION 6.3 ANSWERS
SECTION 6.4 ANSWERS
SECTION 7.1 ANSWERS
SECTION 7.2 ANSWERS
SECTION 7.3 ANSWERS
SECTION 7.4 ANSWERS
SECTION 7.5 ANSWERS
SECTION 7.6 ANSWERS
SECTION 7.7 ANSWERS
SECTION 8.1 ANSWERS
SECTION 8.2 ANSWERS
SECTION 8.3 ANSWERS
SECTION 8.4 ANSWERS
SECTION 8.5 ANSWERS
SECTION 8.6 ANSWERS
SECTION 8.7 ANSWERS
SECTION 9.1 ANSWERS
SECTION 9.2 ANSWERS
SECTION 9.3 ANSWERS
SECTION 9.4 ANSWERS
1 9/16/2020
A Brief Table of Integrals
α+1
α u
∫ u du = + c, α ≠ −1
α+1
du
∫ = ln |u| + c
u
∫ cos u du = sin u + c
∫ sin u du = − cos u + c
∫ tan u du = − ln | cos u| + c
∫ cot u du = ln | sin u| + c
2
∫ sec u du = tan u + c
2
∫ csc u du = − cot u + c
∫ sec u du = ln | sec u + tan u| + c
2 u 1
∫ cos u du = + sin 2u + c
2 4
2 u 1
∫ sin u du = − sin 2u + c
2 4
du −1
∫ 2
du = tan u +c
1+u
du −1
∫ du = sin u +c
√1−u2
1 1 u−1
∫ du = ln | | +c
u2 −1 2 u+1
∫ cosh u du = sinh u + c
∫ sinh u du = cosh u + c
∫ u dv = uv − ∫ v du
∫ u cos u du = u sin u + cos u + c
∫ u sin u du = −u cos u + sin u + c
u u u
∫ u e du = u e −e +c
λu
e (λ cos ωu+ω sin ωu)
λu
∫ e cos ωu du = 2
+c
λ +ω2
λu
e (λ sin ωu+ω cos ωu)
λu
∫ e sin ωu du = 2
+c
2
λ +ω
∫ ln |u| du = u ln |u| − u + c
2
u ln |u| u
2
∫ u ln |u| du = − +c
2 4
sin( ω1 +ω2 )u sin( ω1 −ω2 )u

∫ cos ω1 u cos ω2 u du = + +c (ω1 ≠ ±ω2 )
2( ω1 +ω2 ) 2( ω1 −ω2 )
sin( ω1 +ω2 )u sin( ω1 −ω2 )u

∫ sin ω1 u sin ω2 u du = − + +c (ω1 ≠ ±ω2 )
2( ω1 +ω2 ) 2( ω1 −ω2 )
cos( ω1 +ω2 )u cos( ω1 −ω2 )u

∫ sin ω1 u cos ω2 u du = − − +c (ω1 ≠ ±ω2 )
2( ω1 +ω2 ) 2( ω1 −ω2 )
9/11/2020 1 https://math.libretexts.org/@go/page/45874
Index
A F method of undetermined coefficients
Abel's theorem forcing function 5.4: The Method of Undetermined Coefficients I
5.1: Homogeneous Linear Equations 5.5: The Method of Undetermined Coefficients II
9.1: Introduction to Linear Higher Order Equations 9.3: Undetermined Coefficients for Higher Order
10.3: Basic Theory of Homogeneous Linear Equations
Systems midpoint method
augmented matrix G 3.2: The Improved Euler Method and Related
11.3: Gaussian Elimination Gaussian Elimination Methods
autonomous differential equation 11.2: Elementary Operations
minor of a matrix
4.4: Autonomous Second Order Equations 13.1: Basic Techniques
gravity
Autonomous Second Order Equations minors or a matrix
H multiplicity
B Heun’s method 9.2: Higher Order Constant Coefficient
Bernoulli Equations Homogeneous Equations
3.2: The Improved Euler Method and Related
2.4: Transformation of Nonlinear Equations into Methods
Separable Equations
homogeneous N
C Homogeneous Differential Equations 4.2: Cooling and Mixing
characteristic polynomial Nonhomgeneous Linear Equations
5.2: Constant Coefficient Homogeneous Equations with constant Coefficients 5.3: Nonhomogeneous Linear Equations
9.2: Higher Order Constant Coefficient 5.2: Constant Coefficient Homogeneous Equations
Chebyshev’s equation O
I open interval of convergence
7.3: Series Solutions Near an Ordinary Point I identity matrix 7.2: Review of Power Series
coefficient matrix 12.6: The Identity and Inverses
open rectangle
10.2: Linear Systems of Differential Equations implicit function Theorem 2.3: Existence and Uniqueness of Solutions of
cofactor 2.2: Separable Equations Nonlinear Equations
13.1: Basic Techniques improved Euler method ordinary point
complementary equation 3.2: The Improved Euler Method and Related 7.3: Series Solutions Near an Ordinary Point I
2.1: Linear First Order Equations Methods 7.4: Series Solutions Near an Ordinary Point II
5.3: Nonhomogeneous Linear Equations initial value problems Overdamped Motion
compound interest 8.3: Solution of Initial Value Problems 6.2: Spring Problems II
4.1: Growth and Decay integrating factor
constant coefficient equation 2.6: Integrating Factors P
9.2: Higher Order Constant Coefficient Inverse Laplace Transform particular solution
8.2: The Inverse Laplace Transform 5.3: Nonhomogeneous Linear Equations
constant matrix phase plane equivalent
K 4.4: Autonomous Second Order Equations
Critically Damped Motion Kronecker delta Poincaré phase plane
6.2: Spring Problems II 12.6: The Identity and Inverses 4.4: Autonomous Second Order Equations
polynomial operator
D L 9.2: Higher Order Constant Coefficient
damping Laplace Expansion Homogeneous Equations
4.4: Autonomous Second Order Equations 13.1: Basic Techniques power series
defective Laplace Transform 7.2: Review of Power Series
10.5: Constant Coefficient Homogeneous Systems 8: Laplace Transforms
II
determinant
8.1: Introduction to the Laplace Transform R
radius of convergence
13.1: Basic Techniques Linear Independence 7.2: Review of Power Series
Determinants 14.3: Linear Independence
reduction of orders
direction field M Riccati equation
1.3: Direction Fields for First Order Equations Malthusian model 2.4E: Transformation of Nonlinear Equations into
1.1: Applications Leading to Differential Separable Equations (Exercises)
E Equations
roundoff errors
Elementary operations matrix 3.1: Euler's Method
11.2: Elementary Operations 12.1: Matrix Arithmetic
Row Operations
escape velocity matrix inverse 13.2: Properties of Determinants
4.3: Elementary Mechanics 12.7: Finding the Inverse of a Matrix
Euler’s Method Matrix Multiplication S
3.1: Euler's Method 12.4: Properties of Matrix Multiplication
Separable Differential Equations
Exact Differentials Mechanics 2.2: Separable Equations
2.5: Exact Equations 4.3: Elementary Mechanics
separatrix
Exact Equations 4.4: Autonomous Second Order Equations
2.5: Exact Equations Simple Harmonic Motion
skew lines T V
11.1: Geometry temperature decay constant van der Pol equation
skew symmetric matrix 4.2: Cooling and Mixing 4.4E: Autonomous Second Order Equations
12.5: The Transpose terminal velocity (Exercises)
span 4.3: Elementary Mechanics Variation of Parameters
14.2: Spanning Sets transpose 5.7: Variation of Parameters
9.4: Variation of Parameters for Higher Order
Spanning Sets 12.5: The Transpose
Equations
14.2: Spanning Sets Triangular Matrix Verhulst model
square matrix 13.1: Basic Techniques
1.1: Applications Leading to Differential
12.1: Matrix Arithmetic truncation errors Equations
subsets 3.1: Euler's Method
14.2: Spanning Sets W
superposition principle U Wronskian
5.3: Nonhomogeneous Linear Equations Underdamped Motion 9.1: Introduction to Linear Higher Order Equations
system of linear equations 6.2: Spring Problems II 10.3: Basic Theory of Homogeneous Linear
11.2: Elementary Operations Systems
Section 1.2 Answers
1. 3
2. 2
3. 1
4. 2
2
13. y = − x
2
+c
14. y = x cos x − sin x + c

2 2
15. y = x
2
ln x −
x
4
+c
16. y = −x cos x + 2 sin x + c 1 + c2 x
17. y = (2x − 4)e x

+ c1 + c2 x
18. y = x
3
− sin x + e
x
+ c1 + c2 x
19. y = sin x + c 1 + c2 x + c3 x
2
20. y = − x
60
+e
x
+ c1 + c2 x + c3 x
2
21. y = 7
64
e
4x
+ c1 + c2 x + c3 x
2
22. y = −(x − 1)e x
23. y = 1 − 1
2
cos x
2
–
24. y = 3 − ln(√2 cos x)
5
25. y = − 47
15
−
37
5
(x − 2) +
x
30
26. y = 1
4
xe
2x
−
1
4
e
2x
+
29
27. y = x sin x + 2 cos x − 3x − 1

2
28. y = (x 2
− 6x + 12)e
x
+
x
2
− 8x − 11
29. y = x
3
+
cos 2x
6
+
7
4
x
2
− 6x +
7
4 3
30. y = x
12
+
x
6
+
1
2
(x − 2 )
2
−
26
3
(x − 2) −
5
40.
a. 576 ft
b. 10 s
41.
b. y = 0
43.
a. (−2c − 2, ∞) (−∞, ∞)
Section 2.1 Answers
1. y = e −ax
2. y = ce
3
−x
3. y = ce −(ln x) /2
4. y = c
x
3
5. y = ce 1/x
−( x−1)
6. y = e
7. y = x ln x
e
8. y = x sin x
π
9. y = 2(1 + x 2
)
10. y = 3x −k
11. y = c(cos kx) 1/k
12. y = 1
3
+ ce
−3x
13. y = 2
x
+
x
c
e
x
14. y = e
2
−x x
( + c)
2
−x
15. y = − e
1+x
+c
2
7 ln |x|
16. x
+
3
2
x+
c
17. y = (x − 1) −4
(ln |x − 1| − cos x + c)
18. y = e
2
−x x c
( + )
4 x
2 ln |x|
19. y = x2
+
1
2
+
c
x2
20. y = (x + c) cos x
21. y = c−cos x
2
(1+x)
3 5
(x−2) (x−2)
22. y = − 1
2 (x−1)
+c
(x−1)
23. y = (x + c)e
2
− sin x
x x
24. y = e
x2
−
e
x
3
+
c
x2
3x −7x
25. y = e −e
10
26. y = 2x+1
2
2
(1+x )
27. y = x2
1
ln(
1+x
2
)
28. y = 1
2
(sin x + csc x)
2 ln |x|
29. y = x
+
x
2
−
1
2x
30. y = (x − 1) −3
[ln(1 − x) − cos x]
31. y = 2x 2
+
x
1
2
(0, ∞)
32. y = x 2
(1 − ln x)
33. y =
2
1 5 −x
+ e
2 2
ln |x−1|+tan x+1
34. y = 3
(x−1)
2
ln |x|+x +1
35. y = 4
(x+2)
36. y = (x 2
− 1) (
1
2
ln | x
2
− 1| − 4)
37. y = −(x 2
− 5)(7 + ln | x
2
− 5|)
x
38. y = e
2 2
−x 2 t
(3 + ∫ t e dt)
0
x
39. y = 1
x
(2 + ∫
1
sin t
t
dt)
x
40. y = e −x
∫
1
tan t
t
dt
t
x
41. y = 1+x
1
2
(1 + ∫
0
e
1+t
2
dt)
2
x
42. y = 1
x
(2 e
−(x−1)
+e
−x
∫
1
t
e e
t
dt)
43. G = r
λ
+ (G0 −
r
λ
)e
−λt
limt→∞ G(t) =
r
x
45. y = y 0e
−a(x−x0 )
+e
−ax
∫
x0
e
at
f (t)dt
48.
a. y = tan
−1
(
1
3
+ ce
3x
)
1/2
b. y = ±[ln( 1
x
+
c
x2
)]
c. y = exp (x
2
+
x
c
2
)
d. y = −1 + c+3 ln |x|
x
Section 2.2 Answers
−−−− −−−− −−−−−−−
1. y = 2 ± √2(x 3 2
+ x + x + c)
2. ln(| sin y|) = cos x + c; y ≡ kπ, k = integer
3. y = x−c
c
y ≡ −1
2
(ln y) 3
4. 2
=−
x
3
+c
5. y 3
+ 3 sin y + ln |y| + ln(1 + x ) + tan
2 −1
x = c; y ≡0
1/2
2
6. y = ±(1 + ( 1+cx
x
) ) ; y ≡ ±1
7. y = tan( x
3
+ c)
8. y = c
√1+x2
2
( x−1) /2
2−ce
9. y = ( x−1)
2
/2
; y ≡1
1−ce
10. y = 1 + (3x 2
+ 9x + c )
1/3
−−−−−−−−−−−−−−−−−
11. y = 2 + √ 2
3
x
3
+ 3x
2
+ 4x −
11
2
−( x −4) /2
12. y = e
2
2−e−( x −4) /2
13. y 3
+ 2y
2
+x
2
+ sin x = 3
14. (y + 1)(y − 1) −3
(y − 2 )
2
= −256(x + 1 )
−6
15. y = −1 + 3e −x
16. y = 1
2
√2 e−2x −1
17. y ≡ −1; (−∞, ∞)
2
−x
18. y = 4−e
−x
2
; (−∞, ∞)
2−e
−1+√4 x2 −15 √15

19. y = 2
; (
2
, ∞)
20. y = 1+e
2
−2x
(−∞, ∞)
−−−−−−
21. y = −√25 − x 2
; (−5, 5)
22. y ≡ 2, (−∞, ∞)
1/3
23. y = 3( x+1
2x−4
) ; (−∞, 2)
24. y = x+c
1−cx
−−−−−
25. y = −x cos c + √1 − x 2
sin c; y ≡ 1; y ≡ −1
26. y = −x + 3π/2
P0
28. P =
αP0 +(1−αP0 ) e
−a t
; limt→∞ P (t) = 1/α
SI0
29. I =
I0 +(S−I0 ) e−rSt
I0 αI0
30. If q = rS then I =
1+rI0 t
and limt→∞ I (t) = 0. If q ≠ Rs, then I =
I0 +(α−I0 ) e
−rαt
. If q < rs, then limt→∞ I (t)
q
=α =S− if q > rS, then limt→∞ I (t) = 0
r
34. f = ap, where a = constant
−− −−−−
35. y = e −x 2
(−1 ± √2 x + c )
−− −−−
36. y = x 2 2
(−1 + √x + c )
37. y = e x
(−1 + (3x e
x
+ c)
1/3
)
−−− −−
38. y =e
2x 2
(1 ± √c − x )
39.
a. y 1 = 1/x; g(x) = h(x)
b. y 1 = x; g(x) = h(x)/ x
2
c. y 1 =e
−x
;
x
g(x) = e h(x)
d. y 1 =x
−r
; g(x) = x
r−1
h(x)
e. y 1 = 1/v(x); g(x) = v(x)h(x)
Section 2.3 Answers
1. (b) x 0 ≠ kπ(l = integer)
2. (b) x 0, y0 ) ≠ (0, 0)
3. (b) x 0 y0 ≠ (2k + 1)
π
2
(k = integer)
4. (b) x 0 y0 > 0 and x0 y0 ≠ 1
5.
a. all (x , y ) 0 0
b. (x , y ) woth y
0 0 0 ≠0
6. (b) all (x 0, y0 )
7. (b) all (x 0, y0 )
8. (b) (x 0, y0 ) such that x 0 ≠ 4 y0
9.
a. all (x 0, y0 )
b. all (x 0, y0 ) ≠ (0, 0)
10.
a. all (x 0, y0 )
b. all (x 0, y0 ) with y 0 ≠ ±1
11. (b) all (x 0, y0 )
12. (b) all (x 0, y0 ) such that (x 0, y0 ) > 0
13. (b) all (x 0, y0 ) with x 0 ≠ 1, y0 ≠ (2k + 1)

π
2
(k = integer)
5/3
16. y = ( 3
5
x + 1) , −∞ < x < ∞ , is a solution.
Also,
5
0, −∞ < x ≤ −
3
y ={
3 5/3 5
( x + 1) , − <x <∞
5 3
is a solution. For every a ≥ 5
3
, the following function is also a solution:
3 5/3
⎧ ( (x + a)) , −∞ < x < −a,
⎪ 5
⎪
5
y =⎨ 0, −a ≤ x ≤ −
3
⎪
⎩
⎪ 3 5/3 5
( x + 1) , − <x <∞
5 3
17.
a. all (x 0, y0 )
b. all (x 0, y0 ) with y 0 ≠1
18. y 1
3
= 1; y2 = 1 + |x | ; y3 = 1 − |x | ; y4 = 1 + x ; y5 = 1 − x
3 3 3
3 3
1 +x , x ≥ 0, 1 −x , x ≥ 0,
y6 = { ; y7 = { ;
1, x <0 1, x <0
1, x ≥ 0, 1, x ≥ 0,
y8 = { ; y9 = {
3 3
1 +x , x <0 1 −x , x <0
19. y = 1 + (x 2
+ 4)
3/2
, −∞ < x < ∞
20.
a. The solution is unique on (0, ∞). It is given by
–
1, 0 < x ≤ √5
y ={ –
2 3/2
1 − (x − 5) , √5 < x < ∞
–
b. 1, −∞ < x ≤ √5,
y ={
2 3/2 –
1 − (x − 5) , √5 < x < ∞
is a solution of (A) on (−∞, ∞). If α ≥ 0 , then

2 2 3/2
⎧ 1 + (x
⎪
−α ) , −∞ < x < −α,
–
y =⎨ 1, −α ≤ x ≤ √5,
⎩
⎪ 2 3/2 –
1 − (x − 5) , √5 < x < ∞,
and
2 2 3/2
⎧ 1 − (x −α ) , −∞ < x < −α,
⎪
–
y =⎨ 1, −α ≤ x ≤ √5,
⎩
⎪ –
2 3/2
1 − (x − 5) , √5 < x < ∞,
are also solutions of (A) on (−∞, ∞).
Section 2.4 Answers
1. y = 1−ce
1
x
2. y = x 2/7
(c − ln |x| )
1/7
3. y = e 2/x
(c − 1/x )
2
√2x+c
4. y = ± 1+x
2
5. y = ±(1 − x
2
2 −x −1/2
+ ce )
1/3
6. y = [ 3(1−x)+ce
x
−x
]
2 √2
7. y = √1−4x
−2
8. y = [1 −
2
3 −( x −1)/4
e ]
2
9. y = 1
1/3
x(11−3x)
10. y = (2e x
− 1)
2
11. y = (2e 12x

− 1 − 12x )
1/3
1/2
12. y = [ 5x
2(1+4 x )
5
]
13. y = (4e x/2

− x − 2)
2
⎧ ∞ if L = 0,
⎪
at
P0 e
14. P = t
; limt→∞ P (t) = ⎨ 0 if L = ∞,
1+aP0 ∫ α(τ) ea τ dτ
0 ⎩
⎪
1/aL if 0 < L < ∞
15. y = x(ln |x| + c)

2
16. y = cx
1−cx
y = −x
17. y = ±x(4 ln |x| + c) 1/4
18. y = x sin −1
(ln |x| + c)
19. y = x tan(ln |x| + c)

−−− −−−
20. y = ±x √cx 2
−1
21. y = ±x ln(ln |x| + c)

22. y = − 2x
2 ln |x|+1
23. y = x(3 ln x + 27) 1/3
4
1/2
9−x
24. y = 1
x
(
2
)
25. y = −x
x(4x−3)
26. y = − (2x−3)
−−−−−−
27. y = x √4x 6
−1
y
28. tan −1
x
−
1
2
ln(x
2
+y ) = c
2
29. (x + y) ln |x| + y(1 − ln |y|) + cx = 0

30. (y + x ) 3 3
= 3 x (ln |x| + c)
31. (y + x) = c(y − x ) 3
; y = x; y = −x
32. y 2
(y − 3x) = c; y ≡ 0; y = 3x
33. (x − y ) 3
(x + y) = c y x ;
2 4
y = 0; y = x; y = −x
3
y y
34. x
+
x
3
= ln |x| + c
40. Choose X and Y so that 0 0
aX0 + b Y0 = α
c X0 + dY0 = β
41. (y + 2x + 1) 4
)2y − 6x − 3) = c; y = 3x + 3/2; y = −2x − 1
42. (y + x − 1)(y − x − 5) 3
= c; y = x + 5; y = −x + 1
2(x+2)
43. ln |y − x − 6| − y−x−6
= c; y = x +6
44. (y 1 =x
1/3
y =x
1/3
(ln |x| + c )
1/3
−− − −−−
45. y 1 =x ;
3 3 6
y = ±x √c x − 1
2 4
x (1+cx )
46. y 1 =x ;
2
y =
1−cx
4
y = −x
2
x x
e (1−2ce
47. y 1 =e ;
x
y =−
1−cex
; y = −2 e
x
48. y 1 = tan x; y = tan x tan(ln | tan x| + c
4
2 ln x(1+c(ln x) )
49. y 1 = ln x; y =
4
; y = −2 ln x
1−c(ln x)
−−−−−− −
50. y 1 =x
1/2
; y =x
1/2
(−2 ± √ln |x| + c )
−− −−−−
51. y
2 2
x x 2
1 =e ; y =e (−1 ± √2 x + c )
−3+√1+60x
52. y = 2x
−5+√1+48x
53. y = 2x2
56. y = 1 + x+1+ce
1
x
57. y = e x
−
1+ce
1
−x
58. y = 1 − x(1−cx)
1
59. y = x − 2x
x2 +c
Section 2.5 Answers
1. 2x 3
y
2
=c
2. 3y sin x + 2x 2
e
x
+ 3y = c
3. Not exact
4. x 2
− 2x y
2
+ 4y
3
=c
5. x + y = c
6. Not exact
7. 2y 2
cos x + 3x y
3
−x
2
=c
8. Not exact
9. x 3
+ x y + 4x y
2 2
+ 9y
2
=c
10. Not exact

11. ln |xy| + x 2
+y
2
=c
12. Not exact

13. x 2
+y
2
=c
14. x 2
y e
2 x
+ 2y + 3 x
2
=c
15. x
2
3 x +y 3 2
e − 4y + 2x =c
16. x 4
e
xy
+ 3xy = c
17. x 3
cos xy + 4 y
2
+ 2x
2
=c
2
x+√2 x +3x−1
18. y = x
2
−−−−−−−
19. y = sin x − √1 − tan x
x
1/3
20. y = ( e −1
ex +1
)
21. y = 1 + 2 tan x
2
22. y = x −x+6
(x+2)(x−3)
2 2
3y
23. 7x
2
+ 4xy +
2
=c
24. (x 4
y
2
+ 1)e
x
+y
2
=c
29.
a. M (x, y) = 2xy + f (x)
b. M (x, y) = 2(sin x + x cos x)(y sin y + cos y) + f (x)
c. M (x, y) = y e − e cos x + f (x) x y
30.
4
x y
a. N (x, y) = + x + 6xy + g(y)
2
2
b. N (x, y) = + 2y sin x + g(y) x
c. N (x, y) = x(sin y + y cos y) + g(y)

33. B = C
34. B = 2D, E = 2C
37.
a. 2x + x y + y = c
2 4 4 2
b. x + 3x y = c
3 2
c. x + y + 2xy = c
3 2
38. y = −1 − x
1
2
2 4 2
−3( x +1)+√9 x +34 x +21
39. y = x 3
(
2
)
2
2x+√9−5x
40. y = −e
2
−x
( )
3
44.
a. G(x, y) = 2xy + c
b. G(x, y) = e sin y + c x
c. G(x, y) = 3x y − y + c 2 3
d. G(x, y) = − sin x sinh y + c

e. G(x, y) = cos x sinh y + c
Section 2.6 Answers
3. μ(x) = 1/x 2
; y = cx and μ(y) = 1/ y ;
2
x = cy
4. μ(x) = x −3/2
; x
3/2
y =c
5. μ(y) = 1/y 3
; y e
3 2x
=c
6. μ(x) = e 5x/2
; e
5x/2
(xy + 1) = c
7. μ(x) = e x
;
x
e (xy + y + x) = c
8. μ(x) = x; 2 2
x y (9x + 4y) = c
9. μ(y) = y 2
;
3 2
y (3 x y + 2x + 1) = c
10. μ(y) = y e y
;
y
e (x y
3
+ 1) = c
11. μ(y) = y 2
; y (3 x
3 4 3
+ 8 x y + y) = c
12.μ(x) = x e x
;
2
x y(x + 1)e
x
=c
13. μ(x) = (x 3
− 1)
−4/3
; xy(x
3
− 1)
−1/3
= c and x ≡ 1
14. μ(y) = e y
;
y
e (sin x cos y + y − 1) = c
15. μ(y) = e
2 2
−y −y
; xy e (x + y) = c
xy
16. sin y
= c and y = kπ(k = integer)
17. μ(x, y) = x 4
y ;
3
x y
5 4
ln x = c
18. μ(x, y) = 1/xy; |x |

α
|y | e
β γx
e
δy
= c and x ≡ 0, y ≡ 0
19. μ(x, y) = x −2
y
−3
; 3x y
2 2 2
+ y = 1 + cx y and x ≡ 0, y ≡ 0
20. μ(x, y) = x −2
y
−1
; −
2
x
+y
3
+ 3 ln |y| = c and x ≡ 0, y ≡ 0
21. μ(x, y) = e ax
e
by
; e
ac
e
by
cos xy = c
22. μ(x, y) = x −4
y
−3
(and others) xy = c
23. μ(x, y) = x e y
; x ye
2 y
sin x = c
3 3
x y y
24. μ(x, y) = 1/x 2
;
3
−
x
=c
25. μ(x) = x + 1; y(x + 1 ) (x + y) = c

2
26. μ(x, y) = x 2
y ;
2 3 3
x y (3x + 2 y ) = c
2
27. μ(x, y) = x −2
y
−2
;
2
3 x y = cxy + 2 and x ≡ 0, y ≡ 0
Section 3.1 Answers
1. y 1 = 1.450000000, y2 = 2.085625000, y3 = 3.079099746
2. y 1 = 1.200000000, y2 = 1.440415946, y3 = 1.729880994
3. y 1 = 1.900000000, y2 = 1.781375000, y3 = 1.646612970
4. y 1 = 2.962500000, y2 = 2.922635828, y3 = 2.880205639
5. y 1 = 2.513274123, y2 = 1.814517822, y3 = 1.216364496
6.
x h = 0.1 h = 0.05 h = 0.025 Exact
1.0 48.298147362 51.492825643 53.076673685 54.647937102
7.
x h = 0.1 h = 0.05 h = 0.025 Exact
2.0 1.390242009 1.370996758 1.361921132 1.353193719
8.
x h = 0.05 h = 0.025 h = 0.0125 Exact
1.50 7.886170437 8.852463793 9.548039907 10.500000000
9.
x h = 0.1 h = 0.05 h = 0.025 h = 0.1 h = 0.05 h = 0.025
3.0 1.469458241 1.462514486 1.459217010 0.3210 0.1537 0.0753
Approximate Solutions Residuals
10.
x h = 0.1 h = 0.05 h = 0.025 h = 0.1 h = 0.05 h = 0.025
2.0 0.473456737 0.483227470 0.487986391 −0.3129 −0.1563 −0.0781
11.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.691066797 0.676269516 0.668327471 0.659957689
12.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
2.0 −0.772381768 −0.761510960 −0.756179726 −0.750912371
13.
Euler's Method
x h = 0.1 h = 0.05 h = 0.025 Exact
1.0 0.538871178 0.593002325 0.620131525 0.647231889
Euler semilinear Method
x h = 0.1 h = 0.05 h = 0.025 Exact
1.0 0.647231889 0.647231889 0.647231889 0.647231889
Applying variation of parameters to the given initial value problem yields y = ue −3x
, where (A) u = 7, u(0) = 6 . Since
′
u = 0 , Euler’s method yields the exact solution of (A). Therefore the Euler semilinear method produces the exact solution of
′′
the given problem

14.
Euler's Method
x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 12.804226135 13.912944662 14.559623055 15.282004826
Euler semilinear method

x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 15.354122287 15.317257705 15.299429421 15.282004826
15.
Euler's method
x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.867565004 0.885719263 0.895024772 0.904276722
Euler's semilinear method

x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.569670789 0.720861858 0.808438261 0.904276722
16.
Euler's method
x h = 0.2 h = 0.1 h = 0.05 "Exact"
3.0 0.922094379 0.945604800 0.956752868 0.967523153

x h = 0.2 h = 0.1 h = 0.05 "Exact"
3.0 0.993954754 0.980751307 0.974140320 0.967523153
17.
Euler's method
x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 0.319892131 0.330797109 0.337020123 0.343780513

x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 0.305596953 0.323340268 0.333204519 0.343780513
18.
Euler's method
x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.754572560 0.743869878 0.738303914 0.732638628

x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.722610454 0.727742966 0.730220211 0.732638628
19.
Euler's method
x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 2.175959970 2.210259554 2.227207500 2.244023982

x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 2.117953342 2.179844585 2.211647904 2.244023982
20.
Euler's method
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.032105117 0.043997045 0.050159310 0.056415515
Euler's semilinear method

x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.056020154 0.056243980 0.056336491 0.056415515
21.
Euler's method
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 28.987816656 38.426957516 45.367269688 54.729594761
Euler's semilinar method

x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 54.709134946 54.724150485 54.728228015 54.729594761
22.
Euler's method
x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 1.361427907 1.361320824 1.361332589 1.361383810
Euler's semilinar method

x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 1.291345518 1.326535737 1.344004102 1.361383810
Section 3.2 Answers
1. y 1 = 1.542812500, y2 = 2.421622101, y3 = 4.208020541
2. y 1 = 1.220207973, y2 = 1.489578775. y3 = 1.819337186
3. y 1 = 1.890687500, y2 = 1.763784003, y3 = 1.622698378
4. y 1 = 2.961317914, y2 = 2.920132727, y3 = 2.876213748
5. y 1 = 2.478055238, y2 = 1.844042564, y3 = 1.313882333
6.
x h = 0.1 h = 0.05 h = 0.025 Exact
1.0 56.134480009 55.003390448 54.734674836 54.647937102
7.
x h = 0.1 h = 0.05 h = 0.025 Exact
2.0 1.353501839 1.353288493 1.353219485 1.353193719
8.
x h = 0.5 h = 0.025 h = 0.0125 Exact
1.50 10.141969585 10.396770409 10.472502111 10.500000000
9.
x h = 0.1 h = 0.05 h = 0.025 h = 0.1 h = 0.05 h = 0.025
3.0 1.455674816 1.455935127 1.456001289 −0.00818 −0.00207 −0.000518
10.
x h = 0.1 h = 0.05 h = 0.025 h = 0.1 h = 0.05 h = 0.025
2.0 0.492862999 0.492709931 0.492674855 0.00335 0.000777 0.000187
11.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.660268159 0.660028505 0.659974464 0.659957689
12.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
2.0 −0.749751364 −0.750637632 −0.750845571 −0.750912371
13. Applying variation of parameters to the given initial value problem y = ue , where (A)u = 1 − 2x, u(0) = 2 . Since
−3x ′
= 0 , the improved Euler method yields the exact solution of (A). Therefore the improved Euler semilinear method
′′′
u
produces the exact solution of the given problem.
Improved Euler method

x h = 0.1 h = 0.05 h = 0.025 Exact
1.0 0.105660401 0.100924399 0.099893685 0.099574137
Improved Euler semilinar method

x h = 0.1 h = 0.05 h = 0.025 Exact
1.0 0.099574137 0.099574137 0.099574137 0.099574137
14.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 15.107600968 15.234856000 15.269755072 15.282004826
Improved Euler semilinar method

x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 15.285231726 15.282812424 15.282206780 15.282004826
15.

x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.924335375 0.907866081 0.905058201 0.904276722
Improved Euler semilinear method

x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.969670789 0.920861858 0.908438261 0.904276722
16.
x h = 0.2 h = 0.1 h = 0.05 "Exact"
3.0 0.967473721 0.967510790 0.967520062 0.967523153

x h = 0.2 h = 0.1 h = 0.05 "Exact"
3.0 0.967473721 0.967510790 0.967520062 0.967523153
17.
x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 0.349176060 0.345171664 0.344131282 0.343780513

x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 0.349350206 0.345216894 0.344142832 0.343780513
18.

x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.732679223 0.732721613 0.732667905 0.732638628

x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.732166678 0.732521078 0.732609267 0.732638628
19.
x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 2.247880315 2.244975181 2.244260143 2.244023982

x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 2.248603585 2.245169707 2.244310465 2.244023982
20.

x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.059071894 0.056999028 0.056553023 0.056415515

x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.056295914 0.056385765 0.056408124 0.056415515
21.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 50.534556346 53.483947013 54.391544440 54.729594761

x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 54.709041434 54.724083572 54.728191366 54.729594761
22.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 1.361395309 1.361379259 1.361382239 1.361383810

x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 1.375699933 1.364730937 1.362193997 1.361383810
23.
x h = 0.1 h = 0.05 h = 0.025 Exact
2.0 1.349489056 1.352345900 1.352990822 1.353193719
24.
x h = 0.1 h = 0.05 h = 0.025 Exact
2.0 1.350890736 1.352667599 1.353067951 1.353193719
25.
x h = 0.05 h = 0.025 h = 0.0125 Exact
1.50 10.133021311 10.391655098 10.470731411 10.500000000
26.
x h = 0.05 h = 0.025 h = 0.0125 Exact
1.50 10.136329642 10.393419681 10.470731411 10.500000000
27.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.660846835 0.660189749 0.660016904 0.659957689
28.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.660658411 0.660136630 0.660002840 0.659957689
29.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
2.0 −0.750626284 −0.750844513 −0.750895864 −0.751331499
30.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
2.0 −0.750335016 −0.750775571 −0.750879100 −0.751331499
Section 3.3 Answers
1. y 1 = 1.550598190, y2 = 2.469649729
2. y 1 = 1.221551366, y2 = 1.492920208
3. y 1 = 1.890339767, y2 = 1.763094323
4. y 1 = 2.961316248, y2 = 2.920128958
5. y 1 = 2.475605264, y2 = 1.825992433
6.
x h = 0.1 h = 0.05 h = 0.025 Exact
1.0 54.654509699 54.648344019 54.647962328 54.647937102
7.
x h = 0.1 h = 0.05 h = 0.025 Exact
2.0 1.353191745 1.353193606 1.353193712 1.353193719
8.
x h = 0.05 h = 0.025 h = 0.0125 Exact
1.50 10.498658198 10.499906266 10.499993820 10.500000000
9.
x h = 0.1 h = 0.05 h = 0.025 h = 0.1 h = 0.05 h = 0.025
3.0 1.456023907 1.456023403 1.456023379 0.0000124 0.000000611 0.0000000333
10.
x h = 0.1 h = 0.05 h = 0.025 h = 0.1 h = 0.05 h = 0.025
2.0 0.492663789 0.492663738 0.492663736 0.000000902 0.0000000508 0.00000000302
11.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.659957046 0.659957646 0.659957686 0.659957689
12.
x h = 0.1 h = 0.05 h = 0.025 "Exact"
2.0 −0.750911103 −0.750912294 −0.750912367 −0.750912371
13. Applying variation of

parameters to the given initial value problem yields y = ue , where −3x
(A)u = 1 − 4x + 3 x − 4 x , u(0) = −3 . Since u = 0 , the Runge-Kutta method yields the exact solution of (A).
′ 2 3 (5)
Therefore the Euler semilinear method produces the exact solution of the given problem.
14.
Runge-Kutta method
x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 15.281660036 15.281981407 15.282003300 15.282004826
Runge-Kutta semilinear method

x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 15.282005990 15.282004899 15.282004831 15.282004826
15.
Runge-Kutta method
x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.904678156 0.904295772 0.904277759 0.904276722

x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.904592215 0.904297062 0.904278004 0.904276722
16.
Runge-Kutta method
x h = 0.2 h = 0.1 h = 0.05 "Exact"
3.0 0.967523147 0.967523152 0.967523153 0.967523153

x h = 0.2 h = 0.1 h = 0.05 "Exact"
3.0 0.967523147 0.967523152 0.967523153 0.967523153
17.
Runge-Kutta method
x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 0.343839158 0.343784814 0.343780796 0.343780513
18.
Runge-Kutta method
x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.732633229 0.732638318 0.732638609 0.732638628

x h = 0.2 h = 0.1 h = 0.05 "Exact"
2.0 0.732639212 0.732638663 0.732638630 0.732638628
19.
Runge-Kutta method
x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 2.244025683 2.244024088 2.244023989 2.244023982

x h = 0.0500 h = 0.0250 h = 0.0125 "Exact"
1.50 2.244025081 2.244024051 2.244023987 2.244023982
20.
Runge-Kutta method
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.056426886 0.056416137 0.056415552 0.056415515

x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 0.056415185 0.056415495 0.056415514 0.056415515
21.
Runge-Kutta method
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 54.695901186 54.727111858 54.729426250 54.729594761
x h = 0.1 h = 0.05 h = 0.025 "Exact"
1.0 54.729099966 54.729561720 54.729592658 54.729594761
22.
Runge-Kutta method
x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 1.361384082 1.361383812 1.361383809 1.361383810

x h = 0.1 h = 0.05 h = 0.025 "Exact"
3.0 1.361456502 1.361388196 1.361384079 1.361383810
24.
x h = .1 h = .05 h = .025 Exact
2.00 −1.000000000 −1.000000000 −1.000000000 −1.000000000
25.
x h = .1 h = .05 h = .025 "Exact"
1.00 1.000000000 1.000000000 1.000000000 1.000000000
26.
x h = .1 h = .05 h = .025 "Exact"
1.50 4.142171279 4.142170553 4.142170508 4.142170505
27.
x h = .1 h = .05 h = .025 "Exact"
3.0 16.666666988 16.666666687 16.666666668 16.666666667
Section 4.1 Answers
1. Q = 20e −(t ln 2)/3200
2. 2 ln 10
ln 2
days
3. τ = 10
ln 2
ln 4/3
minutes
ln( p0 / p1 )
4. τ ln 2
tp ln p
5. tq
=
ln q
Q1
6. k = 1
t2 −t1
ln
Q
2
7. 20 g
8. 50 ln 2
3
yrs
9. 25
2
ln 2
10.
a. = 20 ln 3yr
b. Q = 100000e
0
−5
11.
a. Q(t) = 5000 − 4750e −t/10
b. 5000lbs
12. 25
1
yrs
13. V = V0 e
t ln 10/2
4 hours
4
1500 ln
14. ln 2
3
yrs; 2
−4/3
Q0
15. W (t) = 20 − 19e −t/20

; limt→∞ W (t) = 20 ounces
16. S(t) = 10(1 + e −t/10

; limt→∞ S(t) = 10 g
17. 10 gallons
18. V (t) = 15000 + 10000e t/20
19. W (t) = 4 × 10 6
(t + 1 )
2
dollars t years from now
20. p = 100
−t/2
25−24e
21.
.06t
a. P (t) = 1000 e
.06t
+ 50
e
.06/52
−1
e −1
b. 5.64 × 10 −4
22.
a. P − = rP − 12M
b. P = 12M
(1 − e ) + P
r
rt
0e
rt
rP0
c. M ≈ 12(1−e−rN )
d. For (i) approximate M = $402.25 , exact M = $402.80 for (ii) approximate M = $1206.05 , exact M = $1206.93 .
23.
P0
a. T (α) = −
1
r
ln(1 − (1 − e
−rN
)/α)) years S(α) =
(1−e
−rN
)
[rN + α ln(1 − (1 − e
−rN
)/α)]
b. T (1.05) = 13.69yrs, S(1.05) = $3579.94T (1.10) = 12.61yrs, S(1.10) = $6476.63T (1.15) = 11.70yrs, S(1.15) =
$8874.
( a −r) T
S0 (1−e )
if a ≠ r
24. P0 ={ r−a
S0 T if a = r
Section 4.2 Answers
1. ≈ 15.15 ∘
F
11
2. T = −10 + 110e
−t ln
9
3. ≈ 24.33 ∘
F
4.
a. ≈ 91.30 F ∘
b. 8.99 minutes after being placed outs

c. never
5.
a. 12 : 11 : 32
b. 12 : 47 : 33
6. (85/3) ∘
C
7. 32 ∘
F
8. Q(t) = 40(1 − e −3t/40

)
9. Q(t) = 30 − 20e −t/20
10. K(t) = .3 − .2e −t/20
11. Q(50) = 47.5 (pounds)

12. 50 gallons
13. min q 2 = q1 √c
14. Q = t + 300 − 234×10

2
, 0 ≤ t ≤ 300
(t+300)
15.
a. Q + Q = 6 − 2e
′ 2
25
−t/25
b. Q = 75 − 50e − 25 e
−t/25 −2t/25
c. 75
16.
k( S0 −Tm )
a. T = Tm + (T0 − Tm )e
−kt
+
(k−km )
(e
−kmt
−e
−kt
)
b. T = T m + k(S0 − Tm )te
−kt
+ (T0 − Tm )e
−kt
c. lim t→∞ T (t) = limt→∞ S(t) = Tm
17.
a. T
′
= −k(1 +
a
am
)T + k(Tm0 +
a
am
T0 )
aT0 +am Tm0 am ( T0 −Tm0 ) aT0 +am Tm0 a( Tm0 −T0 )

b. T =
a+am
+
a+am
e
−k(1+a/ am )t
, Tm =
a+am
+
a+am
e
−k(a+a/ am )t
aT0 +am Tm0

c. limt→∞ T (t) = limt→∞ Tm (t) =
a+am
V0
18. V =
a
b V0 −( V0 −a/b) e
−a t
; limt→∞ V (t) = a/b
19. c 1 = c(1 − e
−rt/W
), c2 = c(1 − e
−rt/W
−
r
W
te
−rt/W
)
20.
n−1 j
a. cn = c (1 − e
−rt/W
∑j=0
1
j!
(
W
rt
) )
b. c
c. 0
2 2
c1 W1 +c2 W2 c2 W −c1 W W1 +W2
21. Let c ∞ =
W1 +W2
, α =
2
W1 +W2
1
, andβ =
W1 W2
. Then:
a. c1 (t) = c∞ +
α
W1
e
−rβt
, c2 (t) = c∞ −
α
W2
e
−rβt
b. lim t→∞ c1 (t) = limt→∞ c2 (t) = c∞
Section 4.3 Answers
1. v = − 384
5
(1 − e
−5t/12
); −
384
5
ft/s
2. k = 12; v = −16(1 − e
−2t
)
3. v = 25(1 − e −t
); 25ft/s
4. v = 20 − 27e −t/40
5. ≈ 17.10 ft
−4t/5
40(13+3 e )
6. v = − −4t/5
; −40ft/s
13−3e
7. v = −128(1 − e −t/4
)
v0 k mg v0 k
9. T =
m
k
ln(1 +
mg
); ym = y0 =
m
k
[v0 −
l
ln(1 +
mg
)]
−t
64(1−e )
10. v = − 1+e−t
; 64ft/s
−βt −βt −
−− −−
v0 (1+e )−α(1−e ) mg kg
11. v = α α(1+e−βt )−v0 (1−e−βt )
; −α where α = √ k
and β = 2√ m
ak
−− −
−− −
−− −2√
m
( t−T )
mg
12. T =√
m
kg
tan
−1
(v0 √
mg
k
) v = −√
k
;
1−e
ak
−2√ ( t−T )
m
1+e
13. s ′
= mg −
s+1
as
; a0 = mg
14. (a) ms ′
= mg − f (s)
15.
a. v ′
= −9.8 + v /81
4
b. v T ≈ −5.308m/s
16.
−−
a. v = −32 + 8√|v| ; v = −16 ft/s
′
T
b. From Exercise 4.3.14

−−−
c. v is the negative number such that −32 + 8√|v | = 0 ; thus, v
T T T = −16 ft/s
17. ≈ 6.76miles/s
18. ≈ 1.47miles/s
2
gR
20. α = 2
( ym +R)
Section 4.4 Answers
2
y
1. ȳ = 0 0 is a stable equilibrium; trajectories are v
¯
¯ 2
+
4
=c
3
2y
2. ȳ = 0 0 is an unstable equilibrium; trajectories are v
¯
¯ 2
+
3
=c
3
2|y|
3. ȳ = 0 0 is a stable equilibrium; trajectories are v
¯
¯ 2
+
3
=c
4. y = 0 0 is a stable equilibrium; trajectories are v

¯
¯¯ 2
−e
−y
(y + 1) = c
5. equilibria: 0 (stable) and −2, 2 (unstable); trajectories: 2v 2

−y
4
+ 8y
2
=c ; separatrix: 2v 2
−y
4
+ 8y
2
= 16
6. equilibria: 0 (unstable) and −2, 2 (stable); trajectories: 2v 2

+y
4
− 8y
2
=c ; separatrix: 2v 2
+y
4
− 8y
2
=0
7. equilibria: 0, −2, 2 (stable), −1, 1 (unstable); trajectories: 6v

2 2
+ y (2 y
4
− 15 y
2
+ 24) = c ; separatrix:
2 2 4 2
6v + y (2 y − 15 y + 24) = 11
8. equilibria: 0, 2 (stable) and (unstable); trajectories: 30v + y

−2, 1
2 2
(12 y
3
− 15 y
2
− 80y + 120) = c ; separatrices:
− 80y + 120) = 496 and 30 v + y (12 y − 15 y
2 2 3 2 2 2 3 2
30 v + y (12 y − 15 y − 80y + 120) = 37
−
− −
−
9. No equilibria if a < 0; 0 is unstable if a = 0 ; √a is stable and −√a is unstable if a > 0 .
−
− −
−
10. 0 is a stable equilibrium if a ≤ 0 ; −√a and √a are stable and 0 is unstable if a > 0 .
11. 0 is unstable if a ≤ 0 ; −√−
− −
−
a and √a are unstable and 0 is stable if a > 0 .
12. 0 is stable if a ≤ 0; 0 is stable and −√−

− −
−
a and √a are unstable if a ≤ 0 .
22. An equilibrium solution ȳ of y + p(y) = 0 is unstable if there’s an € > 0 such that, for every δ > 0 , there’s a solution of
¯
¯ ′′
− −−−−−−−−−−−−− − −−−−−−−−−−−−− −
(A) with √(y(0) − y) + v (0) < δ , but √(y(t) − y) + v (t) ≥ € for some t > 0 .
¯
¯¯ 2 2 ¯
¯¯ 2 2
Section 4.5 Answers
2xy
1. y ′
=
x2 +3 y 2
2
y
2. y ′
=−
(xy−1)
2 2 2
y( x +y −2 x ln |xy|)
3. y ′
=−
x( x +y
2 2
−2 y
2
ln |xy|)
1/2
4. x y ′
−y = −
x
5. y
2
′ x
+ 2xy = 4x e
6. x y ′
+ y = 4x
3
7. y ′
− y = cos x − sin x
8. (1 + x 2 ′
)y − 2xy = (1 − x ) e
2 x
10. y ′
g − yg
′
= f g −fg
′ ′
11. (x − x 0 )y
′
= y − y0
12. y ′
(y
2
−x
2
+ 1) + 2xy = 0
13. 2x(y − 1)y ′

−y
2
+x
2
+ 2y = 0
14.
a. y = −81 + 18x, (9, 81) y = −1 + 2x, (1, 1)
b. y = −121 + 22x, (11, 121) y = −1 + 2x, (1, 1)
c. y = −100 − 20x, (−10, 100) y = −4 − 4x, (−2, 4)
d. y = −25 − 10x, (−5, 25) y = −1 − 2x, (−1, 1)
15. (e) y = 5+3x
4
, (−3/5, 4/5) y =−
5−4x
3
, (4/5, −3/5)
17.
a. y = − (1 + x), (1, −1); y = + , (25, 5)
1
2
5
2
x
10
b. y = (4 + x), (4, 2) y = − (4 + x), (4, −2)

1
4
1
c. y = (1 + x), (1, 1) y = + , (49, 7)

1
2
7
2
x
14
d. y = − (1 + x), (1, −1) y = − − , (25, −5)

1
2
5
2
x
10
18. y = 2x 2
19. y = cx
√| x2 =1|
20. y = y 1 + c(x − x1 )
21. y = − x
2
−
x
22. y = −x ln |x| + cx
−−−−−
23. y = √2x + 4
−−−−−
24. y = √x 2
−3
25. y = kx 2
26. (y − x ) 3
(y + x) = k
27. y 2
= −x + k
28. y 2
=−
1
2
ln(1 + 2 x ) + k
2
29. y 2
= −2x − ln(x − 1 )
2
+k
−−−−
2
30. y = 1 + √ 9−x
2
; those with c > 0
y
33. tan −1
x
−
1
2
ln(x
2 2
+y ) = k
y
34. 1
2
ln(x
2 2
+ y ) + (tan α) tan
−1
x
=k
Section 5.1 Answers
1.
(c) y = −2e 2x
+e
5x
2x 5x
(d) y = (5k 0 − k1 )
e
3
+ (k1 − 2 k0 )
e
2.
(c) y = e 3x
(d) y = e x
(k0 cos x + (k1 − k0 ) sin x
3.
(c) y = e x
(7 − 3x)
(d) y = e x
(k0 + (k1 − k0 )x)
4.
c1 c2
a. y = x−1
+
x+1
b. y = 2
x−1
−
3
x+1
; (−1, 1)
5.
a. e x
b. e cos x
2x
c. x + 2x − 2
2
d. − x 5
6
−5/6
e. − x
1
2
f. (x ln |x|) 2
2x
g. e
2 √x
6. 0
7. W (x) = (1 − x 2
)
−1
8. W (x) = 1
10. y 2 =e
−x
11. y 2 = xe
3x
12. y 2 = xe
ax
13. y 2 =
1
14. y 2 = x ln x
15. y 2 =x
a
ln x
16. y 2 =x
1/2
e
−2x
17. y 2 =x
18. y 2 = x sin x
19. y 2 =x
1/2
cos x
20. y 2 = xe
−x
21. y 2 =
1
x2 −4
22. y 2 =e
2x
23. y 2 =x
2
35.
a. y " −2x y + 5y = 0 ′
b. (2x − 1)y " −4x y + 4y = 0 ′
c. x y " −x y + y = 0
2 ′
d. x y " +x y + y = 0
2 ′
e. y " −y = 0
f. xy " −y = 0 ′
37. (c) y = k 0 y1 + k1 y2
38. y 1 = 1, y2 = x − x0 ; y = k0 + k1 (x − x0 )
39. y 1 = cosh(x − x0 ), y2 = sinh(x − x0 ); y = k0 cosh(x − x0 ) + k1 sinh(x − x0 )
k1
40. y 1 = cos ω(x − x0 ), y2 =
1
ω
sin ω(x − x0 )y = k0 cos ω(x − x0 ) +
ω
sin ω(x − x0 )
k0 +k1 x
41. y 1 =
1
1−x
2
y2 =
x
1−x
2
y =
1−x
2
42.
2 3
c1 x + c2 x x ≥0
(c) k 0 = k1 = 0; y = {
2 3
c1 x + c3 x x <0
(d) (0, ∞) if x 0 > 0, (−∞, 0) if x 0 <0
43. (c) k 0 = 0, k1 arbitrary y = k 1x

2
+ c2 x
44.
3 4
a1 x + a2 x x ≥0
(c) k 0 = k1 = 0; y = {
3 4
b1 x + b2 x x <0
(d) (0, ∞) if x 0 > 0, (−∞, 0) if x 0 <0
Section 5.2 Answers
1. y = c 1e
−6x
+ c2 e
x
2. y = e 2x
(c1 cos x + c2 sin x)
3. y = c 1e
−7x
+ c2 e
−x
4. y = e 2x
(c1 + c2 x)
5. y = e −x
(c1 cos 3x + c2 sin 3x)
6. y = e −3x
7. y = e 4x
(c1 + c2 x)
8. y = c 1 + c2 e
−x
– –
9. y = e x
(c1 cos √2x + c2 sin √2x)
10. y = e −3x
11. y = e −x/2
(c1 cos
3x
2
+ c2 sin
3x
2
)
12. y = c 1e
−x/5
+ c2 e
x/2
13. y = e −7x
14. y = 4e x/2
+ 6e
−x/3
15. y = 3e x/3
− 4e
−x/2
−x/2 3x/2
16. y = e
3
+
3e
17. y = e 3x/2
(3 − 2x)
18. y = 3 −4x
− 4e
−3x
19. y = 2xe 3x
20. y = e x/6
(3 + 2x)
– 2 √6 –
21. y = e −2x
(3 cos √6x +
3
sin √6x)
23. y = 2e −(x−1)
− 3e
−2(x−1)
24. y = 1
3
e
−(x−2)
−
2
3
e
7(x−2)
25. y = e 7(x−1)
(2 − 3(x − 1))
26. y = e −(x−2)/3
(2 − 4(x − 2))
27. y = 2 cos 2
3
(x −
π
4
) − 3 sin
2
3
(x −
π
4
)
– –
28. y = 2 cos √3 (x −
π
3
)−
1
√3
sin √3 (x −
π
3
)
k0 k1
30. y = r2 −r1
(r2 e
r1 (x−x0 )
− r1 e
r2 (x−x0 )
)+
r2 −r1
(e
r2 (x−x0 )
−e
r1 (x−x0 )
)
31. y = e r1 (x−x0 )
[ k0 + (k1 − r1 k0 )(x − x0 )]
k1 −λk0
32. y = e λ(x−x0 )
[k0 cos ω(x − x0 ) + (
ω
) sin ω(x − x0 )]
Section 5.3 Answers
1. y
p = −1 + 2x + 3 x ; y = −1 + 2x + 3 x
2 2
+ c1 e
−6x
+ c2 e
x
2. y
p = 1 + x; y = 1 + x + e
2x
3. y
p = −x + x ; y = −x + x
3 3
+ c1 e
−7x
+ c2 e
−x
4. y
p = 1 −x ; y = 1 −x
2 2
+e
2x
(c1 + c2 x)
5. y
p = 2x + x ; y = 2x + x
3 3
+e
−x
(c1 cos 3x + c2 sin 3x); y = 2x + x
3
+e
−x
(2 cos 3x + 3 sin 3x)
6. y
p = 1 + 2x; y = 1 + 2x + e
−3x
(c1 cos x + c2 sin x); y = 1 + 2x + e
−3x
(cos x − sin x)
8. y
p =
2
9. y
p = 4x
1/2
10. y p =
x
11. y p =
1
x3
12. y p = 9x
1/3
13. y p =
2x
13
3x 3x
16. y p =
e
3
; y =
e
3
+ c1 e
−6x
+ c2 e
x
17. y p =e
2x
; y =e
2x
(1 + c1 cos x + c2 sin x)
18. y = −2e −2x

; y = −2 e
−2x
+ c1 e
−7x
+ c2 e
−x
; y = −2 e
−2x
−e
−7x
+e
−x
19. y p
x
=e ; y =e
x
+e
2x
(c1 + c2 x); y = e
x
+e
2x
(1 − 3x)
20. y p =
4
25
e
x/2
; y =
45
4
e
x/2
+e
−x
21. y p =e
−3x
; y =e
−3x
24. y p = cos x − sin x; y = cos x − sin x + e

4x
(c1 + c2 x)
25. y p = cos 2x − 2 sin 2x; y = cos 2x − 2 sin 2x + c1 + c2 e

−x
– –
26. y p
x
= cos 3x; y = cos 3x + e (c1 cos √2x + c2 sin √2x)
27. y p = cos x + sin x; y = cos x + sin x + e

−3x
28. y p = −2 cos 2x + sin 2x; y = −2 cos 2x + sin 2x + c1 e

−4x
+ c2 e
−3x
; y = −2 cos 2x + sin 2x + 2 e
−4x
− 3e
−3x
29. y p = cos 3x − sin 3x; y = cos 3x − sin 3x + e

3x
(c1 + c2 x) y = cos 3x − sin 3x + e
3x
(1 + 2x)
30. y = 2
ω −ω
1
2
(M cos ωx + N sin ωx) + c1 cos ω0 x + c2 sin ω0 x
0
3x 3x
33. y p = −1 + 2x + 3 x
2
+
e
3
; y = −1 + 2x + 3 x
2
+
e
3
+ c1 e
−6x
+ c2 e
x
34. y p = 1 +x +e
2x
; y = 1 +x +e
2x
35. y p = −x + x
3
− 2e
−2x
; y = −x + x
3
− 2e
−2x
+ c1 e
−7x
+ c2 e
−x
36. y p = 1 −x
2 x
+e ; y = 1 −x
2
+e
x
+e
2x
(c1 + c2 x)
37. y p = 2x + x
3
+
45
4
e
x/2
; y = 2x + x
3
+
45
4
e
x/2
+e
−x
38. y p = 1 −x
2 x
+e ; y = 1 −x
2
+e
x
+e
2x
Section 5.4 Answers
1. y
p =e
3x
(−
1
4
+
x
2
)
2. y
p =e
−3x
(1 −
x
4
)
3. y
p =e
x
(2 −
3x
4
)
4. y
p =e
2x
(1 − 3x + x )
2
5. y
p =e
−x
(1 + x )
2
6. y
p
x
= e (−2 + x + 2 x )
2
7. y
p = xe
−x
(
1
6
+
x
2
)
8. y
p
x
= x e (1 + 2x)
9. y
p = xe
3x
(−1 +
x
2
)
10. y p = xe
2x
(−2 + x)
11. y p =x e
2 −x
(1 +
x
2
)
12. y p =x e
2 x
(
1
2
− x)
2 2x
13. y p =
x e
2
(1 − x + x )
2
2 −x/3
14. y p =
x e
27
(3 − 2x + x )
2
3x
15. y = e
4
(−1 + 2x) + c1 e
x
+ c2 e
2x
16. y = e x
(1 − 2x) + c1 e
2x
+ c2 e
4x
2x
17. y = e
5
(1 − x) + e
−3x
(c1 + c2 x)
18. y = x e x
(1 − 2x) + c1 e
x
+ c2 e
−3x
19. y = e x 2
[ x (1 − 2x) + c1 + c2 x]
20. y = −e 2x
(1 + x) + 2 e
−x
−e
5x
21. y = x e 2x
+ 3e
x
−e
−4x
22. y = e −x
(2 + x − 2 x ) − e
2 −3x
23. y = e −2x
(3 − x) − 2 e
5x
24. y p =−
e
3
(1 − x) + e
−x
(3 + 2x)
25. y p
x
= e (3 + 7x) + x e
3x
26. y p =x e
3 4x
+ 1 + 2x + x
2
27. y p = xe
2x
(1 − 2x) + x e
x
28. y p
x
= e (1 + x) + x e
2 −x
29. y p =x e
2 −x
+e
3x
(1 − x )
2
31. y p = 2e
2x
32. y p = 5x e
4x
33. y p =x e
2 4x
3x
34. y p =−
e
4
(1 + 2x − 2 x )
2
35. y p = xe
3x
(4 − x + 2 x )
2
36. y p =x e
2 −x/2
(−1 + 2x + 3 x )
2
37.
a. y =e
−x
(
4
3
x
3/2
+ c1 x + c2 )
2
b. y = e −3x
[
x
4
(2 ln x − 3) + c1 x + c2 ]
c. y =e
2x
[(x + 1) ln |x + 1| + c1 x + c2 ]
3
d. y = e −x/2
(x ln |x| +
x
6
+ c1 x + c2 )
39.
a. e (3 + x) + c
x
b. −e (1 + x ) + c
−x 2
−2x
c. − e
(3 + 6x + 6 x = 4 x
8
2 3
)+c
d. e (1 + x ) + c
x 2
e. e (−6 + 4x + 9x ) + c
3x 2
f. −e (1 − 2x + 3x ) + c
−x 3 4
k αx r
(−1 ) k! e (−αx)
40. α
k+1
∑
k
r=0 r!
+c
Section 5.5 Answers
1. y p = cos x + 2 sin x
2. y p = cos x + (2 − 2x) sin x
3. y p
x
= e (−2 cos x + 3 sin x)
2x
4. y p =
e
2
(cos 2x − sin 2x)
5. y p
x
= −e (x cos x − sin x)
6. y p =e
−2x
(1 − 2x)(cos 3x − sin 3x)
7. y p = x(cos 2x − 3 sin 2x)
8. y p = −x[(2 − x) cos x + (3 − 2x) sin x]
9. y p = x [x cos(
x
2
) − 3 sin(
x
2
)]
10. y p = xe
−x
(3 cos x + 4 sin x)
11. y p
x
= x e [(−1 + x) cos 2x + (1 + x) sin 2x]
12. y p = −(14 − 10x) cos x − (2 + 8x − 4 x ) sin x

2
13. y p
2
= (1 + 2x + x ) cos x + (1 + 3 x ) sin x
2
14. y p =
x
2
(cos 2x − sin 2x)
15. y p = e (x
x 2
cos x + 2 sin x)
16. y p
x
= e (1 − x )(cos x + sin x)
2
17. y p = e (x
x 2 3
− x )(cos x + sin x)
18. y p =e
−x
[(1 + 2x) cos x − (1 − 3x) sin x]
19. y p = x(2 cos 3x − sin 3x)
20. y p = −x
3
cos x + (x + 2 x ) sin x
2
21. y p = −e
−x 2
[(x + x ) cos x − (1 + 2x) sin x]
22. y = e x
(2 cos x + 3 sin x) + 3 e
x
−e
6x
23. y = e x
[(1 + 2x) cos x + (1 − 3x) sin x]
24. y = e x
(cos x − 2 sin x) + e
−3x
(cos x + sin x)
25. y = e 3x
[(2 + 2x) cos x − (1 + 3x) sin x]
26. y = e 3x
[(2 + 3x) cos x + (4 − x) sin x] + 3 e
x
− 5e
2x
27. y p = xe
3x
−
e
5
(cos x − 2 sin x)
x −x
28. y p = x(cos x + 2 sin x) −

e
2
(1 − x) +
e
29. y p =−
xe
2
(2 + x) + 2x e
2x
+
1
10
(3 cos x + sin x)
x 2
30. y p
x
= x e (cos x + x sin x) +
e
25
(4 + 5x) + 1 + x +
x
2 2x
31. y p =
x e
6
(3 + x) − e
2x
(cos x − sin x) + 3 e
3x
+
1
4
(2 + x)
32. y = (1 − 2x + 3x 2
)e
2x
+ 4 cos x + 3 sin x
33. y = x e −2x
cos x + 3 cos 2x
34. y = − 3
8
cos 2x +
1
4
sin 2x + e
−x
−
13
8
e
−2x
−
3
4
xe
−2x
40.
a. 2x cos x − (2 − x ) sin x + c
2
b. − [(1 − x ) cos x − (1 − x ) sin x] + c

e
2
2 2
−x
c. − [(4 + 10x) cos 2x − (3 − 5x) sin 2x] + c

e
25
−x
d. − [(1 + x ) cos x − (1 − x ) sin x] + c

e
2
2 2
e. − [x(3 − 3x + x ) cos x − (3 − 3x + x ) sin x] + c

e
2
2 3
f. −e [(1 − 2x) cos x + (1 + x) sin x] + c

x
g. e [x cos x + x(1 + x) sin x] + c

−x
Section 5.6 Answers
1. y = 1 − 2x + c 1e
−x
+ c2 x e ; { e
x −x
, xe }
x
c2
2. y = 4
3x
2
+ c1 x +
x
; {x, 1/x}
2
x(ln |x|)
3. y = 2
+ c1 x + c2 x ln |x|; {x, x ln |x|}
4. y = (e 2x x
+ e ) ln(1 + e
−x
) + c1 e
2x
+ c2 e ; { e
x 2x x
,e }
5. y = e x
(
4
5
x
7/2
+ c1 + c2 x) ; { e , x e }
x x
6. y = e x
(2 x
3/2
+x
1/2
ln x + c1 x
1/2
+ c2 x
−1/2
); { x
1/2 x
e ,x
−1/2
e
−x
}
7. y = e x
(x sin x + cos x ln | cos x| + c1 cos x + c2 sin x); { e
x
cos x, e
x
sin x}
8. y = e
2 2 2
−x −2x −x −x
(2 e + c1 + c2 x); { e , xe }
c2
9. y = 2x + 1 + c 1x
2
+
x
2
2
; { x , 1/ x }
2
2x
10. y = xe
9
+ xe
−x
(c1 + c2 x); {x e
−x
,x e
2 −x
}
11. y = x e x
(
x
3
+ c1 +
c2
x2
) ; {x e , e /x}
x x
2 x
(2x−1) e
12. y = − 8
+ c1 e
x
+ c2 x e
−x
; {e , xe
x −x
}
13. y = x 4
+ c1 x
2
+ c2 x
2
ln |x|; { x , x
2 2
ln |x|}
14. y = e −x
(x
3/2
+ c1 + c2 x
1/2
); { e
−x
,x
1/2
e
−x
}
15. y = e x
(x + c1 + c2 x ); { e , x e }
2 x 2 x
2x
16. y = x 1/2
(
e
2
+ c1 + c2 e ) ; { x
x 1/2
,x
1/2
e }
x
17. y = −2x 2
ln x + c1 x
2 4
+ c2 x ; { x , x }
2 4
18. {e x x
, e /x}
19. {x 2
,x }
3
20. {ln |x|, x ln |x|}

21. {sin √− −
x , cos √x }
22. {e x
,x e }
3 x
23. {x a
,x
a
ln |x|}
24. {x sin x, x cos x}

25. {e 2x
,x e
2 2x
}
26. {x 1/2
,x
1/2
cos x}
27. {x 1/2
e
2x
,x
1/2
e
−2x
}
28. {1/x, e 2x
}
29. {e x
,x }
2
30. {e 2x
,x e
2 2x
}
31. y = x 4
= 6x
2
− 8x
2
ln |x|
32. y = 2e 2x
− xe
−x
(x+1)
33. y = 4
x
[−e (3 − 2x) + 7 e
−x
]
2
34. y = x
4
+x
2
(x+2)
35. y = 6(x−2)
+
2x
2
x −4
38.
−kc1 sin kx+kc2 cos kx
a. y =
c1 cos kx+c2 sin kx
x
c1 +2 c2 e
b. y = c1 +c2 e
x
7x
−6 c1 +c2 e
c. y =
c1 +c2 e
7x
6x
7 c1 +c2 e
d. y = − c1 +c2 e6x
(7 c1 −c2 ) cos x+( c1 +7 c2 ) sin x

e. y =−
c1 cos x+c2 sin x
5x/6
−2 c1 +3 c2 e
f. y = 5x/6
6( c1 +c2 e )
c1 +c2 (x+6)
g. y = 6( c1 +c2 x)
39.
x
c1 +c2 e (1+x)
a. y =
x( c1 +c2 e )
x
2
−2 c1 x+c2 (1−2 x )
b. y = c1 +c2 x
2x
−c1 +c2 e (x+1)
c. y =
c1 +c2 xe2x
−3x
2 c1 +c2 e (1−x)
d. y = c1 +c2 xe
−3x
(2 c2 x−c1 ) cos x−(2 c1 x+c2 ) sin x

e. y =
2x( c1 cos x+c2 sin x)
6
c1 +7 c2 x
f. y = x( c1 +c2 x6 )
Section 5.7 Answers
− cos 3x ln | sec 3x+tan 3x|
1. y
p =
9
sin 2x ln | cos 2x|

2. y
p =−
4
+
x cos 2x
3. y
p
x
= 4 e (1 + e ) ln(1 + e
x −x
)
4. y
p
x
= 3 e (cos x ln | cos x| + x sin x)
5. y
p =
8
5
x
7/2
e
x
6. y
p =e
x
ln(1 − e
−2x
)−e
−x
ln(e
2x
− 1)
3
2( x −3)
7. y
p =
3
2x
8. y
p =
e
9. y
p =x
1/2
e
x
ln x
10. y p =e
−x(x+2)
11. y p = −4 x
5/2
12. y p = −2 x
2
sin x − 2x cos x
−x
xe (x+1)
13. y p =−
2
√x cos √x
14. y p =−
2
4 x
15. y p =
3x e
16. y p =x
a+1
17. y p =
x sin x
18. y p = −2 x
2
19. y p = −e
−x
sin x
√x
20. y p =−
2
3/2
21. y p =
x
22. y p = −3 x
2
3 x
23. y p =
x e
3/2
24. y p =−
4x
15
25. y p =x e
3 x
26. y p = xe
x
27. y p =x
2
28. y p = x e (x − 2)
x
− x
29. y p = √x e (x − 1)/4
2x 2
e (3 x −2x+6) −x
30. y = 6
+
xe
31. y = (x − 1) 2
ln(1 − x) + 2 x
2
− 5x + 3
32. y = (x 2
− 1)e
x
− 5(x − 1)
2
x( x +6)
33. y = 3( x2 −1)
2
34. y = − x
2
+x +
1
2x
2
2
x (4x+9)
35. y = 6(x+1)
38.
x
a. y = k 0 cosh x + k1 sinh x + ∫
0
sinh(x − t)f (t)dt
x
b. y = k
′
0 sinh x + k1 cosh x + ∫
0
cosh(x − t)f (t)dt
39.
x
a. y(x) = k cos x + k sin x + ∫ sin(x − t)f (t)dt
0 1
0
x
b. y (x) = −k sin x + k cos x + ∫ cos(x − t)f (t)dt
′
0 1 0
Section 6.1 Answers
– –
1. y = 3 cos 4√6t − 1
sin 4 √6t ft
2 √6
– –
2. y = − 1
4
cos 8 √5t −
1
sin 8 √5t ft
4 √5
−−
−
3. y = 1.5 cos 14√10t cm
√17
4. y = 1
4
cos 8t −
16
1
sin 8t ft; R =
16
ft; ω0 = 8 rad/s; T = π/4 s; ϕ ≈ −.245 rad ≈ −14.04
∘
√5 −−−
5. y = 10 cos 14t + 25
14
sin 14t cmt; R =
14
√809 cm; ω0 = 14 rad/s; T = π/7 s; ϕ ≈ .177 rad ≈ 10.12
∘
−−
−− −− −− −−
6. y = − 1
4
cos √70t +
2
√70
sin √70t m; R =
1
4
√
67
35
m; ω0 = √70 rad/s; T = 2π/ √70 s; ϕ ≈ 2.38 rad ≈ 136.28
∘
7. y = 2
3
cos 16t −
1
4
sin 16t ft
8. y = 1
2
cos 8t −
3
8
sin 8t ft
9. .72 m
10. y = 1
3
sin t +
1
2
cos 2t +
5
6
sin 2t ft
11. y = 16
5
(4 sin
t
4
− sin t)
– –
12. y = − 1
16
sin 8t +
1
3
cos 4 √2t −
1
sin 4 √2t
8 √2
13. y = −t cos 8t − 1
6
cos 8t +
1
8
sin 8t ft
–
14. T = 4 √2 s
15. ω = 8 rad/s y = − t
16
(− cos 8t + 2 sin 8t) +
128
1
sin 8t ft
– −
− −
− –
16. ω = 4√6 rad/s; y =−
t
√6
[
8
3
cos 4 √6t + 4 sin 4 √6t] +
1
9
sin 4 √6t ft
17. y = t
2
cos 2t −
t
4
sin 2t + 3 cos 2t + 2 sin 2t m
v0 −−−−−−−−−−− − y0 ω0 v0
18. y = y 0 cos ω0 t +
ω0
sin ω0 t; R =
1
ω0
√(ω0 y0 )2 + (v0 )2 ; cos ϕ =
2 2
; sin ϕ =
2 2
√( ω y ) +( v ) √( ω y ) +( v )
0 0 0 0 0 0
19. The object with the longer period weighs four times as much as the other.
–
20. T 2 = √2T1 , where T is the period of the smaller object
1
21. k 1 = 9 k2 , where k is the spring constant of the system with the shorter period.
1
Section 6.2 Answers
−2t
−
−
1. y = e
2
(3 cos 2t − sin 2t) ft; √
5
2
e
−2t
ft
√82
2. y = −e −t
(3 cos 3t +
1
3
sin 3t) ft
3
e
−t
ft
3. y = e −16t
(
1
4
+ 10t) ft
−3t
4. y = − e
4
(5 cos t + 63 sin t) ft
5. 0 ≤ c < 8 lb-sec/ft
−− −−
6. y = 1
2
e
−3t
(cos √91t +
11
√91
sin √91t) ft
−4t
7. y = − e
3
(2 + 8t) ft
– –
8. y = e −10t
(9 cos 4 √6t +
45
sin 4 √6t) cm
2 √6
√41 √41
9. y = e −3t/2
(
3
2
cos
2
t+
9
sin
2
t) ft
2 √41
3
√119 √119
10. y = e −
2
t
(
1
2
cos
2
t−
2 √119
9
sin
2
t) ft
– –
11. y = e −8t
(
1
4
cos 8 √2t −
1
4 √2
sin 8 √2t) ft
−− −−
12. y = e −t
(−
1
3
cos 3 √11t +
14
sin 3 √11t) ft
9 √11
13. yp =
22
61
cos 2t +
61
2
sin 2t ft
14. y = − 2
3
(e
−8t
− 2e
−4t
)
15. y = e −2t
(
1
10
cos 4t −
1
5
sin 4t) m
16. y = e −3t
(10 cos t − 70 sin t) cm
17. yp =−
2
15
cos 3t
1
15
sin 3t ft
18. yp =
11
100
cos 4t +
100
27
sin 4t cm
19. yp =
42
73
cos t +
39
73
sin t ft
20. y = − 1
2
cos 2t +
1
4
sin 2t m
21. yp =
1
cω0
(−β cos ω0 t + α sin ω0 t)
cy
24. y = e −ct/2m
(y0 cos ω1 t +
1
ω1
(v0 +
2m
0
) sin ω1 t)
r2 y0 −v0 v0 −r1 y0
25. y = r2 −r1
e
r1 t
+
r2 −r1
e
r2 t
26. y = e r1 t
(y0 + (v0 − r1 y0 )t)
Section 6.3 Answers
−−
− −−
1. I =e
−15t
(2 cos 5 √15t −
6
√31
sin 5 √31t)
2. I =e
−20t
(2 cos 40t − 101 sin 40t)
3. I =−
200
3
e
−10t
sin 30t
4. I = −10 e
−30t
(cos 40t + 18 sin 40t)
5. I = −e
−40t
(2 cos 30t − 86 sin 30t)
6. I
p =−
1
3
(cos 10t + 2 sin 10t)
7. I
p =
20
37
(cos 25t − 6 sin 25t)
8. I
p =
3
13
(8 cos 50t − sin 50t)
9. I
p =
20
123
(17 sin 100t − 11 cos 100t)
10. I p =−
45
52
(cos 30t + 8 sin 30t)
−−
− −−−−−−−
12. ω 0 = 1/ √LC
2
maximum amplitude = √U + V
2
/R
Section 6.4 Answers
2 2 2
eρ ρ
1. If e =1 , then Y
2
= ρ(ρ − 2X) ; if e ≠ 1 (X +
1−e
2
) +
Y
1−e
2
= 2
if; e <1 let
(1−e2 )
eρ ρ ρ
X0 = − 2
, a = 2
, b = 2
1−e 1−e 1−e
1/2
2 ′ 2
2
ρ ρr
2. Let h =r θ
2
0
′
0
; then ρ =
h
k
, e = [(
r0
− 1) +(
h
0
) ] . If e =0 , then θ0 is undefined, but also irrelevant if e ≠0
′
ρ pr
then ϕ = θ 0 −α , where −π ≤ α < π cos α = 1
e
(
r0
− 1) and sin α = eh
0
.
3.
γ2 −γ1
a. e =
γ +γ
1 2
1/2
2gγ2
b. r 0 = Rγ1 , r
′
0
=0 , θ arbitrary, θ
0
′
0
=[ 3
]
Rγ ( γ1 +γ2 )
1
4. f (r) = −mh 2
(
6c
r
4
+
1
r
3
)
2 2
mh ( γ +1)
5. f (r) = − r3
6.
2 ′
du( θ0 ) r0
a. d u
2
+ (1 −
k
2
) u = 0, u(θ0 ) =
1
r0
,
dθ
=−
h
dθ h
1/2
b. with γ = ∣∣1 − k
2
∣
∣
h
′ −1
r0 r
i. r = r0 (cosh γ(θ − θ0 ) −
γh
0
sinh γ(θ − θ0 ))
′ −1
r0 r
ii. r = r 0 (1 −
h
0
(θ − θ0 ))
′ −1
r0 r
iii. r = r0 (cos γ(θ − θ0 ) −
γh
0
sin γ(θ − θ0 ))
Section 7.1 Answers
1.
a. R = 2; I = (−1, 3)
b. R = 1/2; I = (3/2, 5/2)
c. R = 0
d. R = 16; I = (−14, 18)
e. R = ∞; I = (−∞, ∞)
f. R = 4/3; I = (−25/3, −17/3)
3.
a. R = 1; I = (0, 2)
– – –
b. R = √2; I = (−2 − √2, −2 + √2)
c. R = ∞; I = (−∞, ∞)
d. R = 0
– – –
e. R = √3; I = (−√3, √3)
f. R = 1
: I = (0, 2)
5.
a. R = 3; I = (0, 6)
b. R = 1; I = (−1, 1)
– – –
c. R = 1/√3; I = (3 − 1/√3, 3 + 1/√3)
d. R = ∞; I = (−∞, ∞)
e. R = 0
f. R = 2; I = (−1, 3)
11. b n = 2(n + 2)(n + 1)an+2 + (n + 1)nan+1 + (n + 3)an
12. b 0 = 2 a2 − 2 a0 bn = (n + 2)(n + 1)an+2 + [3n(n − 1) − 2] an + 3(n − 1)an−1 , n ≥ 1
13. b n = (n + 2)(n + 1)an+2 + 2(n + 1)an+1 + (2 n

2
− 5n + 4)an
14. b n = (n + 2)(n + 1)an+2 + 2(n + 1)an+1 + (n

2
− 2n + 3)an
15. b n = (n + 2)(n + 1)an+2 + (3 n

2
− 5n + 4)an
16. b 0 = −2 a2 + 2 a1 + a0 , bn = −(n + 2)(n + 1)an+2 + (n + 1)(n + 2)an+1 + (2n + 1)an + an−1 , n ≥ 2
17. b 0
2
= 8 a2 + 4 a1 − 6 a0 , bn = 4(n + 2)(n + 1)an+2 + 4(n + 1 ) an+1 + (n
2
+ n − 6)an − 3 an−1 , n ≥ 1
21. b 0
2
= (r + 1)(r + 2)a0 , bn = (n + r + 1)(n + r + 2)an − (n + r − 2 ) an−1 , n ≥ 1.
22. b 0 = (r − 2)(r + 2)a0 , bn = (n + r − 2)(n + r + 2)an + (n + r + 2)(n + r − 3)an−1 , n ≥ 14
23. b 0
2 2 2
= (r − 1 ) a0 , b1 = r a1 + (r + 2)(r + 3)a0 , bn = (n + r − 1 ) an + (n + r + 1)(n + r + 2)an−1
+ (n + r − 1)an−2 , n ≥ 2
24. b 0 = r(r + 1)a0 , b1 = (r + 1)(r + 2)a1 + 3(r + 1)(r + 2)a0 , bn = (n + r)(n + r + 1)an
+ 3(n + r)(n + r + 1)an−1 + (n + r)an−2 , n ≥ 2
25. b 0 = (r + 2)(r + 1)a0 , b1 = (r + 3)(r + 2)a1 , bn = (n + r + 2)(n + r + 1)an + 2(n + r − 1)(n + r − 3)an−2 , n
≥2
26. b 0 = 2(r + 1)(r + 3)a0 , b1 = 2(r + 2)(r + 4)a1 , bn = 2(n + r + 1)(n + r + 3)an + (n + r − 3)(n + r)an−2 , n ≥ 2
Section 7.2 Answers
1. y = a0 ∑
∞
m=0
(−1 )
m
(2m + 1)x
2m
+ a1 ∑
∞
m=0
(−1 )
m
(m + 1)x
2m+1
2m
∞
2. y = a0 ∑m=0 (−1 )
m+1
2m−1
x
+ a1 x
3. y = a0 (1 − 10 x
2 4
+ 5 x ) + a1 (x − 2 x
3
+
1
5
x )
5
4. y = a0 ∑
∞
m=0
(m + 1)(2m + 1)x
2m
+
a1
3
∑
∞
m=0
(m + 1)(2m + 3)x
2m+1
4j+1 2m+1
5. y = a0 ∑
∞
m=0
(−1 )
m
[∏
m−1
j=0 2j+1
]x
2m
+ a1 ∑
∞
m=0
(−1 )
m
[∏
m−1
j=0
(4j + 3)]
x
2
m
m!
2 2
∞ m−1 (4j+1) 2m
∞ m−1 (4j+3) 2m+1
6. y = a0 ∑m=0 (−1 )
m
[∏j=0
2j+1
]
8
x
m
m!
+ a1 ∑m=0 (−1 )
m
[∏j=0
2j+3
]
x
m
8 m!
m−1
m ∏ (2j+3)
∞ ∞
7. y = a 2 m! j=0
2m 2m+1
0 ∑m=0 m−1
x + a1 ∑m=0 m x
∏ (2j+1) 2 m!
j=0
8. y = a0 (1 − 14 x
2
+
35
3
4
x ) + a1 (x − 3 x
3
+
3
5
x
5
+
1
35
x )
7
2m 2m+1
∞ ∞
9. (a) y = a 0 ∑
m=0
(−1 )
m
m−1
x
+ a1 ∑
m=0
(−1 )
m x
2
m
m!
∏j=0 (2j+1)
∞ m−1 4j+3 2m
∞ m−1 4j+5 2m+1
10. (a) y = a 0 ∑
m=0
(−1 )
m
[∏
j=0 2j+1
]
2
x
m
m!
+ a1 ∑
m=0
(−1 )
m
[∏
j=0 2j+3
]
x
2
m
m!
11. y = 2 − x − x 2
+
1
3
x
3
+
5
12
x
4
−
1
6
x
5
−
17
72
x
6
+
126
13
x
7
+…
12. y = 1 − x + 3x 2
−
5
2
x
3
+ 5x
4
−
21
8
x
5
+ 3x
6
−
11
16
x
7
+…
13. y = 2 − x − 2x 2
+
1
3
x
3
+ 3x
4
−
5
6
x
5
−
49
5
x
6
+
45
14
x
7
+…
2m 2m+1
∞ (x−3) ∞ (x−3)
16. y = a 0 ∑m=0
(2m)!
+ a1 ∑m=0
(2m+1)!
2m 2m+1
(x−3) (x−3)
17. y = a 0 ∑
∞
m=0 2
m
m!
+ a1 ∑
∞
m=0 m−1
∏ (2j+3)
j=0
2m m
(x−1) 4 (m+1)!
18. y = a 0 ∑
∞
m=0
[∏
m−1
j=0
(2j + 3)]
m!
+ a1 ∑
∞
m=0 m−1
(x − 1 )
2m+1
∏j=0 (2j+3)
19. y = a 0 (1 − 6(x − 2 )
2
+
4
3
(x − 2 )
4
+
135
8 6
(x − 2 ) ) + a1 ((x − 2) −
10
9
(x − 2 ) )
3
m m
20. y = a 0 ∑
∞
m=0
(−1 )
m
[∏
m−1
j=0
(2j + 1)]
4
m
3
m!
(x + 1 )
2m
+ a1 ∑
∞
m=0
(−1 )
m
m−1
3 m!
(x + 1 )
2m+1
∏j=0 (2j+3)
21. y = −1 + 2x + 3
8
x
2
−
1
3
x
3
−
128
3
x
4
−
1024
1
x
6
+…
22. y = −2 + 3(x − 3) + 3(x − 3) 2

− 2(x − 3 )
3
−
5
4
(x − 3 )
4
+
3
5
(x − 3 )
5
+
7
24
(x − 3 )
6
−
35
4
(x − 3 )
7
+…
23. y = −1 + (x − 1) + 3(x − 1) 2
−
5
2
(x − 1 )
3
−
27
4
(x − 1 )
4
+
21
4
(x − 1 )
5
+
27
2
(x − 1 )
6
−
81
8
(x − 1 )
7
+…
24. y = 4 − 6(x − 3) − 2(x − 3) 2

+ (x − 3 )
3
+
3
2
(x − 3 )
4
−
5
4
(x − 3 )
5
−
49
20
(x − 3 )
6
+
135
56
(x − 3 )
7
+…
25. y = 3 − 4(x − 4) + 15(x − 4) 2

− 4(x − 4 )
3
+
15
4
(x − 4 )
4
−
1
5
(x − 4 )
5
26. y = 3 − 3(x + 1) − 30(x + 1) 2

+
20
3
(x + 1 )
3
+ 20(x + 1 )
4
−
4
3
(x + 1 )
5
−
8
9
(x + 1 )
6
27.
a. y = a 0 ∑
∞
m=0
(−1 )
m
x
2m
+ a1 ∑
∞
m=0
(−1 )
m
x
2m+1
a0 +a1 x
b. y = 1+x
2
3m 3m+1
∞ ∞
33. y = a 0 ∑m=0
m
x
m−1
+ a1 ∑m=0
m
x
m−1
3 m! ∏ (3j+2) 3 m! ∏ (3j+4)
j=0 j=0
∞ m m−1 3m
∞
m
34. y = a 0 ∑m=0 (
2
3
) [∏j=0 (3j + 2)]
x
m!
+ a1 ∑m=0 m−1
6 m!
x
3m+1
∏ (3j+4)
j=0
m 3m+1
35. y = a 0 ∑
∞
m=0
(−1 )
m
m−1
3 m!
x
3m
+ a1 ∑
∞
m=0
(−1 )
m
[∏
m−1
j=0
(3j + 4)]
x
3
m
m!
∏j=0 (3j+2)
3j−5
36. y = a 0 (1 − 4x
3 6
+ 4 x ) + a1 ∑
∞
m=0
m
2 [∏
m−1
j=0 3j+4
]x
3m+1
37. y = a 0 (1 +
21
2
x
3
+
42
5
x
6
+
7
20
9
x ) + a1 (x + 4 x
4
+
10
7
x )
7
5j+1 m 5m+1
39. y = a 0 ∑
∞
m=0
(−2 )
m
[∏
m−1
j=0 5j+4
]x
5m
+ a1 ∑
∞
m=0
(−
2
5
) [∏
m−1
j=0
(5j + 2)]
x
m!
4m 4m+1
40. y = a 0 ∑
∞
m=0
(−1 )
m
m
x
m−1
+ a1 ∑
∞
m=0
(−1 )
m
m
x
m−1
4 m! ∏ (4j+3) 4 m! ∏ (4j+5)
j=0 j=0
7m 7m+1
41. y = a 0 ∑
∞
m=0
(−1 )
m
∏
∞
x
(7j+6)
+ a1 ∑
∞
m=0
(−1 )
m x
7
m
m!
j=0
42. y = a 0 (1 −
9
7
8
x ) + a1 (x −
7
9
9
x )
43. y = a 0 ∑
∞
m=0
x
6m
+ a1 ∑
∞
m=0
x
6m+1
6m 6m+1
∞ ∞
44. y = a 0 ∑m=0 (−1 )
m
m−1
x
+ a1 ∑m=0 (−1 )
m x
6
m
m!
∏ (6j+5)
j=0
Section 7.3 Answers
1. y = 2 − 3x − 2x 2
+
7
2
x
3
−
55
12
x
4
+
59
8
x
5
−
83
6
x
6
+
9547
336
x
7
+…
2. y = −1 + 2x − 4x 3
+ 4x
4
+ 4x
5
− 12 x
6
+ 4x
7
+…
3. y = 1 + x 2
−
2
3
x
3
+
11
6
x
4
−
9
5
x
5
+
329
90
x
6
−
1301
315
x
7
+…
4. y = x − x 2
−
7
2
x
3
+
15
2
x
4
+
45
8
x
5
−
261
8
x
6
+
207
16
x
7
+…
5. y = 4 + 3x − 15
4
x
2
+
1
4
x
3
+
11
16
x
4
−
5
16
x
5
+
1
20
x
6
+
120
1
x
7
+…
6. y = 7 + 3x − 16
3
x
2
+
13
3
x
3
−
23
9
x
4
+
10
9
x
5
−
27
7
x
6
−
1
9
x7 + …
7. y = 2 + 5x − 7
4
x
2
−
3
16
x
3
+
37
192
x
4
−
192
7
x
5
−
1920
1
x
6
+
19
11520
x
7
+…
8. y = 1 − (x − 1) + 4
3
(x − 1 )
3
−
4
3
(x − 1 )
4
−
4
5
(x − 1 )
5
+
136
45
(x − 1 )
6
−
104
63
(x − 1 )
7
+…
9. y = 1 − (x + 1) + 4(x + 1) 2
−
13
3
(x + 1 )
3
+
77
6
(x + 1 )
4
−
278
15
(x + 1 )
5
+
1942
45
(x + 1 )
6
−
23332
315
(x + 1 )
7
+…
10. y = 2 − (x − 1) − 1
2
(x − 1 )
2
+
5
3
(x − 1 )
3
−
19
12
(x − 1 )
4
+
7
30
(x − 1 )
5
+
59
45
(x − 1 )
6
−
1091
630
(x − 1 )
7
+…
11. y = −2 + 3(x + 1) − 1
2
(x + 1 )
2
−
2
3
(x + 1 )
3
+
5
8
(x + 1 )
4
−
11
30
(x + 1 )
5
+
29
144
(x + 1 )
6
−
101
840
(x + 1 )
7
+…
12. y = 1 − 2(x − 1) − 3(x − 1) 2

+ 8(x − 1 )
3
− 4(x − 1 )
4
−
42
5
(x − 1 )
5
+ 19(x − 1 )
6
−
604
35
(x − 1 )
7
+…
19. y = 2 − 7x − 4x 2
−
17
6
x
3
−
3
4
x
4
−
40
9
x
5
+…
20. y = 1 − 2(x − 1) + 1
2
(x − 1 )
2
−
1
6
(x − 1 )
3
+
5
36
(x − 1 )
4
−
1080
73
(x − 1 )
5
+…
21. y = 2 − (x + 2) − 7
2
(x + 2 )
2
+
4
3
(x + 2 )
3
−
1
24
(x + 2 )
4
+
1
60
(x + 2 )
5
+…
22. y = 2 − 2(x + 3) − (x + 3) 2
+ (x + 3 )
3
−
11
12
(x + 3 )
4
+
67
60
(x + 3 )
5
+…
23. y = −1 + 2x + 1
3
x
3
−
12
5
x
4
+
2
5
x
5
+…
24. y = 2 − 3(x + 1) + 7
2
(x + 1 )
2
− 5(x + 1 )
3
+
197
24
(x + 1 )
4
−
287
20
(x + 1 )
5
+…
25. y = −2 + 3(x + 2) − 9
2
(x + 2 )
2
+
11
6
(x + 2 )
3
+
5
24
(x + 2 )
4
+
7
20
(x + 2 )
5
+…
26. y = 2 − 4(x − 2) − 1
2
(x − 2 )
2
+
2
9
(x − 2 )
3
+
49
432
(x − 2 )
4
+
23
1080
(x − 2 )
5
+…
27. y = 1 + 2(x + 4) − 1
6
(x + 4 )
2
−
10
27
(x + 4 )
3
+
19
648
(x + 4 )
4
+
13
324
(x + 4 )
5
+…
28. y = −1 + 2(x + 1) − 1
4
(x + 1 )
2
+
1
2
(x + 1 )
3
−
65
96
(x + 1 )
4
+
67
80
(x + 1 )
5
+…
31.
c1 c2
a. y = 1+x
+
1+2x
c1 c2
b. y = 1−2x
+
1−3x
c1 c2 x
c. y = 1−2x
+ 2
(1−2x)
c1 c2 x
d. y = 2+x
+ 2
(2+x)
e. y =
c1
2+x
+
2+3x
c2
32. y = 1 − 2x − 3
2
x
2
+
5
3
x
3
+
17
24
x
4
−
11
20
x
5
+…
33. y = 1 − 2x − 5
2
x
2
+
2
3
x
3
−
3
8
x
4
+
1
3
x
5
+…
34. y = 6 − 2x + 9x 2
+
2
3
x
3
−
23
4
x
4
−
10
3
x
5
+…
35. y = 2 − 5x + 2x 2
−
10
3
x
3
+
3
2
x
4
−
25
12
x
5
+…
36. y = 3 + 6x − 3x 2
+x
3
− 2x
4
−
17
20
x
5
+…
37. y = 3 − 2x − 3x 2
+
3
2
x
3
+
3
2
x
4
−
49
80
x
5
+…
38. y = −2 + 3x + 4
3
x
2
−x
3
−
19
54
x
4
+
13
60
x
5
+…
m 2m m 2m+1
∞ (−1) x 2
∞ (−1) x 2
39. y1 = ∑m=0
m!
=e
−x
, y2 = ∑m=0
m!
= xe
−x
40. y = −2 + 3x + x 2
−
1
6
x
3
−
3
4
x
4
+
31
120
x
5
+…
41. y = 2 + 3x − 7
2
x
2
−
5
6
x
3
+
41
24
x
4
+
41
120
x
5
+…
42. y = −3 + 5x − 5x 2
+
23
6
x
3
−
23
12
x
4
+
11
30
x
5
+…
43. y = −2 + 3(x − 1) + 3
2
(x − 1 )
2
−
17
12
(x − 1 )
3
−
12
1
(x − 1 )
4
+
1
8
(x − 1 )
5
+…
44. y = 2 − 3(x + 2) + 1
2
(x + 2 )
2
−
1
3
(x + 2 )
3
+
31
24
(x + 2 )
4
−
53
120
(x + 2 )
5
+…
45. y = 1 − 2x + 3
2
x
2
−
11
6
x
3
+
15
8
x
4
−
71
60
x
5
+…
46. y = 2 − (x + 2) − 7
2
(x + 2 )
2
−
43
6
(x + 2 )
3
−
203
24
(x + 2 )
4
−
167
30
(x + 2 )
5
+…
47. y = 2 − x − x 2
+
7
6
x
3
−x
4
+
120
89
x
5
+…
48. y = 1 + 3
2
(x − 1 )
2
+
1
6
(x − 1 )
3
−
1
8
(x − 1 )
5
+…
49. y = 1 − 2(x − 3) + 1
2
(x − 3 )
2
−
1
6
(x − 3 )
3
+
1
4
(x − 3 )
4
−
1
6
(x − 3 )
5
+…
Section 7.4 Answers
1. y = x 1x
−4
+ c2 x
−2
2. y = c 1x + c2 x
7
3. y = x(c 1 + c2 ln x)
4. y = x −2
(c1 + c2 ln x)
5. y = c 1 cos(ln x) + c2 sin(ln x)
6. y = x 2
[ c1 cos(3 ln x) + c2 sin(3 ln x)]
c2
7. y = c 1x +
x3
8. y = c 1x
2/3
+ c2 x
3/4
9. y = x −1/2
(c1 + c2 ln x)
10. y = c 1x + c2 x
1/3
11. y = c 1x
2
+ c2 x
1/2
12. y = 1
x
[ c1 cos(2 ln x) + c2 sin(2 ln x]
13. y = x −1/3
(c1 + c2 ln x)
14. y = x[c 1 cos(3 ln x) + c2 sin(3 ln x)]
x2
15. y = c 1x
3
+
x
2
c1
16. y = x
+ c2 x
1/2
17. y = x 2
(c1 + c2 ln x)
18. y = x2
1
[c1 cos(
1
ln x) + c2 sin(
1
ln x)]
√2 √2
Section 7.5 Answers
1. y
1 =x
1/2
(1 −
1
5
x−
2
35
x
2
+
315
31
x
3
+ …) ; y2 = x
−1
(1 + x +
1
2
x
2
−
1
6
x
3
+ …)
2. y
1 =x
1/3
(1 −
2
3
x+
8
9
x
2
−
40
81
x
3
+ …) ; y2 = 1 − x +
6
5
x
2
−
4
5
x
3
+…
3. y
1 =x
1/3
(1 −
4
7
x−
7
45
x
2
+
970
2457
x
3
+ …) ; y2 = x
−1
(1 − x
2
+
2
3
x
3
+ …)
4. y
1 =x
1/4
(1 −
1
2
x−
19
104
x
2
+
1571
10608
x
3
+ …) ; y2 = x
−1
(1 + 2x −
11
6
x
2
−
1
7
x
3
+ …)
5. y
1 =x
1/3
(1 − x +
28
31
x
2
−
1111
1333
x
3
+ …) ; y2 = x
−1/4
(1 − x +
7
8
x
2
−
19
24
x
3
+ …)
6. y
1 =x
1/5
(1 −
6
25
x−
1217
625
x
2
+
41972
46875
x
3
+ …) ; y2 = x −
1
4
x
2
−
35
18
x
3
+
11
12
x
4
+…
7. y
1 =x
3/2
(1 − x +
11
26
x
2
−
109
1326
x
3
+ …) ; y2 = x
1/4
(1 + 4x −
131
24
x
2
+
39
14
x
3
+ …)
8. y
1 =x
1/3
(1 −
1
3
x+
2
15
x
2
−
63
5
x
3
+ …) ; y2 = x
−1/6
(1 −
12
1
x
2
+
1
18
x
3
+ …)
9. y
1 =1−
14
1
x
2
+
105
1
x
3
+… ; y2 = x
−1/3
(1 −
1
18
x−
405
71
x
2
+
719
34992
x
3
+ …)
10. y 1 =x
1/5
(1 +
3
17
x−
153
7
x
2
−
547
5661
x
3
+ …) ; y2 = x
−1/2
(1 + x +
14
13
x
2
−
556
897
3
x + …)
n n
(−2) (−1)
14. y 1 =x
1/2
∑
∞
n=0 n
x ;
n
y2 = x
−1
∑
∞
n=0 n!
x
n
∏j=1 (2j+3)
n n
(−1 ) ∏ (3j+1)
15. y 1 =x
1/3
∑
∞
n=0 9 n!
j=1
n
x ;
n
x
−1
n
(−1) n
16. y 1 =x
1/2
∑
∞
n=0 n
x ;
n
y2 =
1
x2
∑
∞
n=0 n
2
x
n
∏j=1 (3j+4) n! ∏j=1 (2j−5)
n n
∞ (−1) ∞ (−1)
17. y 1 =x∑
n=0 ∏
n
(3j+4)
x ;
n
y2 = x
−1/3
∑
n=0 3 n!
n
x
n
j=1
n n
18. y 1 =x∑
∞
n=0 n! ∏
n
2
(2j+1)
x ;
n
y2 = x
1/2
∑
∞
n=0 n! ∏
n
2
(2j−1)
x
n
j=1 j=1
19. y 1 =x
1/3
∑
∞
n=0 n! ∏
n
1
(3j+2)
x ;
n
y2 = x
−1/3
∑
∞
n=0 n! ∏
n
1
(3j−2)
x
n
j=1 j=1
n
∞ (−1) n 3j−13
20. y 1 = x (1 +
2
7
x+
70
1
x );
2
y2 = x
−1/3
∑n=0 n
3 n!
(∏j=1
3j−4
)x
n
n
2j+1 (−1)
21. y 1 =x
1/2
∑
∞
n=0
(−1 )
n
(∏
n
j=1 6j+1
)x ;
n
y2 = x
1/3
∑
∞
n=0 n
9 n!
(∏
n
j=1
(3j + 1)) x
n
n n
(−1 ) (n+2)! (−1)
22. y 1 =x∑
∞
n=0 2 ∏
n
(4j+3)
x ;
n
y2 = x
1/4
∑
∞
n=0 16 n!
n
∏
n
j=1
(4j + 5)x
n
j=1
n n
∞ (1) ∞ (1)
23. y 1 =x
−1/2
∑n=0 n
n! ∏j=1 (2j+1)
x ;
n
y2 = x
−1
∑n=0 n
n! ∏j=1 (2j−1)
x
n
n
(−1) n 2j−1
24. y 1 =x
1/3
∑
∞
n=0 n!
(
2
9
) (∏
n
j=1
(6j + 5)) x ;
n
y2 = x
−1
∑
∞
n=0
(−1 ) 2
n n
(∏
n
j=1 3j−4
)x
n
25. y 1 = 4x
1/3
∑
∞
n=0 n
6 n!(3n+4)
1
x ;
n
x
−1
28. y 1 =x
1/2
(1 −
9
40
x+
128
5
x
2
−
245
39936
x
3
+ …) ; y2 = x
1/4
(1 −
25
96
x+
675
14336
x
2
−
38025
5046272
x
3
+ …)
29. y 1 =x
1/3
(1 +
32
117
x−
1053
28
x
2
+
4480
540189
x
3
+ …) ; y2 = x
−3
(1 +
32
7
x+
48
7
x )
2
30.y 1 =x
1/2
(1 −
5
8
x+
55
96
x
2
−
935
1536
x
3
+ …) ; y2 = x
−1/2
(1 +
1
4
x−
5
32
x
2
−
55
384
x
3
+ …)
31. y 1 =x
1/2
(1 −
3
4
x+
5
96
x
2
+
4224
5
x
3
+ …) ; y2 = x
−2
(1 + 8x + 60 x
2
− 160 x
3
+ …)
32. y 1 =x
−1/3
(1 −
10
63
x+
7371
200
x
2
−
17600
3781323
x
3
+ …) ; y2 = x
−1/2
(1 −
20
3
x+
352
9
x
2
−
105
23936
x
3
+ …)
m m
(−1) 4j−3 (−1) 8j−7
33. y 1 =x
1/2
∑
∞
m=0 8
m
m!
(∏
m
j=1 8j+1
)x
2m
; y2 = x
1/4
∑
∞
m=0 16
m
m!
(∏
m
j=1 8j−1
)x
2m
∞ m 8j−3 ∞ m
34. y 1 =x
1/2
∑
m=0
(∏
j=1 8j+1
)x
2m
; y2 = x
1/4
∑
m=0 2
m
1
m!
(∏
j=1
(2j − 1)) x
2m
35. y1 =x
4
∑
∞
m=0
(−1 )
m
(m + 1)x
2m
; y2 = −x ∑
∞
m=0
(−1 )
m
(2m − 1)x
2m
m
(−1)
36. y1 =x
1/3
∑
∞
m=0 18
m
m!
(∏
m
j=1
(6j − 17)) x
2m
; y2 = 1 +
4
5
x
2
+
55
8
x
4
m
8j+1 ∏ (2j−1)
37. y ∞ m ∞ j=1
1/4 2m −1 2m
1 =x ∑ (∏ )x ; y2 = x ∑ m
x
m=0 j=1 8j+5 m=0 2 m!
∞ m ∞ m 3j−1
38. y1 =x
1/2
∑m=0
8
m
1
m!
(∏j=1 (4j − 1)) x
2m
; y2 = x
1/3
∑m=0 2
m
(∏j=1
12j−1
)x
2m
m
m
∏j=1 (4j+5) (−1) 4j−1
39. y1 =x
7/2
∑
∞
m=0
(−1 )
m
8
m
m!
x
2m
; y2 = x
1/2
∑
∞
m=0 4
m
(∏
m
j=1 2j−3
)x
2m
m m
(−1) 4j−1 (−1)
40. y1 =x
1/2
∑
∞
m=0 4
m
(∏
m
j=1 2j+1
)x
2m
; y2 = x
−1/2
∑
∞
m=0 8
m
m!
(∏
m
j=1
(4j − 3)) x
2m
m
∞ (−1) m ∞ m 4j−3
41. y1 =x
1/2
∑
m=0 m!
(∏
j=1
(2j + 1)) x
2m
; y2 =
1
x
2
∑
m=0
(−2 )
m
(∏
j=1 4j−5
)x
2m
∞ m 3j−4
42. y1 =x
1/3
∑m=0 (−1 )
m
(∏j=1
3j+2
)x
2m
; y2 = x
−1
(1 + x )
2
m
m
2 (m+1)! ∏ (2j−1)
43. y ∞ 1 ∞ j=1
m 2m m 2m
1 =∑ (−1 ) m
x ; y2 = ∑ (−1 ) m
x
m=0 ∏ (2j+3) x
3 m=0 2 m!
j=1
m 2 m 2
(−1) (4j−3) (−1) (2j−3)
44. y1 =x
1/2
∑
∞
m=0 8
m
m!
(∏
m
j=1 4j+3
)x
2m
; y2 = x
−1
∑
∞
m=0 2
m
m!
(∏
m
j=1 4j−3
)x
2m
m
2j+1 (−1)
45. y1 =x∑
∞
m=0
(−2 )
m
(∏
m
j=1 4j+5
)x
2m
; y2 = x
−3/2
∑
∞
m=0 4
m
m!
(∏
m
j=1
(4j − 3)) x
2m
m m
(−1) (−1)
46. y1 =x
1/3
∑
∞
m=0 2
m
∏
m
(3j+1)
x
2m
; y2 = x
−1/3
∑
∞
m=0 m
6 m!
x
2m
j=1
47. y1 =x
1/2
(1 −
6
13
x
2
+
325
36
x
4
−
216
12025
x
6
+ …) ; y2 = x
1/3
(1 −
1
2
x
2
+
1
8
x
4
−
1
48
x
6
+ …)
48. y1 =x
1/4
(1 −
13
64
x
2
+
8192
273
x
4
−
2639
524288
x
6
+ …) ; y2 = x
−1
(1 −
1
3
x
2
+
2
33
x
4
−
209
2
x
6
+ …)
49. y1 =x
1/3
(1 −
3
4
x
2
+
14
9
x
4
−
81
140
x
6
+ …) ; y2 = x
−1/3
(1 −
2
3
x
2
+
5
9
x
4
−
40
81
x
6
+ …)
50. y1 =x
1/2
(1 −
3
2
x
2
+
15
8
x
4
−
35
16
x
6
+ …) ; y2 = x
−1/2
(1 − 2 x
2
+
8
3
x
4
−
16
5
x
6
+ …)
51. y1 =x
1/4
(1 − x
2
+
3
2
x
4
−
5
2
x
6
+ …) ; y2 = x
−1/2
(1 −
2
5
x
2
+
36
65
x
4
−
408
455
x
6
+ …)
m m
∞ (−1) ∞ (−1)
53. (a) y1 =x
v
∑
m=0 4
m
m! ∏
m
(j+v)
x
2m
; y2 = x
−v
∑
m=0 4
m
m! ∏
m
(j−v)
x
2m
y1 =
sin x
√x
; y2 =
cos x
√x
j=1 j=1
1/2
61. y1 =
x
1+x
; y2 =
1+x
x
1/3 1/2
62. y1 =
x
1+2x
2
; y2 =
1+2x
x
2
1/4 2
63. y1 =
x
1−3x
; y2 =
1−3x
x
1/3 −1/3
64. y1 =
x
5+x
; y2 =
x
5+x
1/4 −1/2
65. y1 =
x
2−x
2
; y2 =
x
2−x
2
1/2 3/2
66. y1 =
x
1+3x+x2
; y2 =
x
1+3x+x2
1/3
67. y1 =
x
2
; y2 =
x
2
(1+x) (1+x)
1/4
68. y1 =
3+2x+x
x
2
; y2 =
3+2x+x
x
2
Section 7.6 Answers
1. y
1 = x (1 − x +
3
4
x
2
−
13
36
x
3
+ …) ; y2 = y1 ln x + x
2
(1 − x +
65
108
x
2
+ …)
2. y
1 =x
−1
(1 − 2x +
9
2
x
2
−
20
3
x
3
+ …) ; y2 = y1 ln x + 1 −
15
4
x+
133
18
x
2
+…
3. y
1 = 1 +x −x
2
+
1
3
x
3
+… ; y2 = y1 ln x − x (3 −
1
2
x−
31
18
x
2
+ …)
4. y
1 =x
1/2
(1 − 2x +
5
2
x
2
− 2x
3
+ …) ; y2 = y1 ln x + x
3/2
(1 −
9
4
x+
17
6
x
2
+ …)
5. y
1 = x (1 − 4x +
19
2
x
2
−
49
3
x
3
+ …) ; y2 = y1 ln x + x
2
(3 −
43
4
x+
208
9
x
2
+ …)
6. y
1 =x
−1/3
(1 − x +
5
6
x
2
−
1
2
x
3
+ …) ; y2 = y1 ln x + x
2/3
(1 −
11
12
x+
25
36
x
2
+ …)
7. y
1 = 1 − 2x +
7
4
x
2
−
7
9
x
3
+… ; y2 = y1 ln x + x (3 −
15
4
x+
239
108
x
2
+ …)
8. y
1 =x
−2
(1 − 2x +
5
2
x
2
− 3x
3
+ …) ; y2 = y1 ln x +
3
4
−
13
6
x +…
9. y
1 =x
−1/2
(1 − x +
1
4
x
2
+
18
1
x
3
+ …) ; y2 = y1 ln x + x
1/2
(
3
2
−
13
16
x+
1
54
x
2
+ …)
10. y 1 =x
−1/4
(1 −
1
4
x−
32
7
x
2
+
384
23
x
3
+ …) ; y2 = y1 ln x + x
3/4
(
1
4
+
5
64
x−
2304
157
x
2
+ …)
11. y 1 =x
−1/3
(1 − x +
7
6
x
2
−
23
18
x
3
+ …) ; y2 = y1 ln x − x
5/3
(
1
12
−
108
13
x …)
n n
(−1) (−1)
12. y 1 =x
1/2
∑
∞
n=0 2
x ;
n
y2 = y1 ln x − 2 x
1/2
∑
∞
n=1 2
(∑
n
j=1
1
j
)x
n
(n!) (n!)
n n
n ∏j=1 (3j+1) n ∏j=1 (3j+1)
13. y 1 =x
1/6
∑
∞
n=0
(
2
3
)
n!
n
x ; y2 = y1 ln x − x
1/6
∑
∞
n=1
(
2
3
)
n!
(∑
n
j=1
1
j(3j+1)
)x
n
∞ ∞
14. y 1 =x
2
∑n=0 (−1 ) (n + 1 ) x ;
n 2 n
y2 = y1 ln x − 2 x
2
∑n=1 (−1 ) n(n + 1)x
n n
15. y 1 =x
3
∑
∞
n=0
n
2 (n + 1)x ;
n
y2 = y1 ln x − x
3
∑
∞
n=1
n
2 nx
n
n n n n
(−1 ) ∏j=1 (5j+1) (−1 ) ∏j=1 (5j+1) 5j+2
16. y 1 =x
1/5
∑
∞
n=0 n 2
x ;
n
y2 = y1 ln x − x
1/5
∑
∞
n=1 n 2
(∑
n
j=1 j(5j+1)
)x
n
125 (n! ) 125 (n! )
n n n n
(−1 ) ∏ (2j−3) (−1 ) ∏ (2j−3)
17. y ∞ ∞ n
j=1 j=1
1/2 n 1/2 1 n
1 =x ∑ n
x ; y2 = y1 ln x + 3 x ∑ n
(∑ )x
n=0 4 n! n=1 4 n! j=1 j(2j−3)
n n 2 n n 2
(−1 ) ∏ (6j−7 ) (−1 ) ∏ (6j−7 )
18. y ∞ ∞ n 1
j=1 j=1
1/3 n 1/3 n
1 =x ∑ x ; y2 = y1 ln x + 14 x ∑ (∑ )x
n=0 n 2 n=1 n 2 j=1 j(6j−7)
81 (n! ) 81 (n! )
n n n n
(−1 ) ∏ (2j+5) (−1 ) ∏ (2j+5) j+5
19. y ∞ ∞ n
j=1 j=1
2 n 2 n
1 =x ∑ x ; y2 = y1 ln x − 2 x ∑ (∑ )x
n=0 2 n=1 2 j=1 j(2j+5)
(n!) (n!)
n n n n
(2 ) ∏ (2j−1) (2 ) ∏ (2j−1)
∞ ∞ n
20. y 1 1 1
j=1 j=1
n n
1 = ∑ x ; y2 = y1 ln x + ∑ (∑ )x
x n=0 n! x n=1 n! j=1 j(2j−1)
n n n n
(−1 ) ∏j=1 (2j−5) (−1 ) ∏j=1 (2j−5)
21. y 1 =
1
x
∑
∞
n=0 n!
x ;
n
y2 = y1 ln x +
5
x
∑
∞
n=1 n!
(∑
n
j=1 j(2j−5)
1
)x
n
n n n n
(−1 ) ∏ (2j+3) (−1 ) ∏ (2j+3)
∞ ∞ n
22. y 1
j=1 j=1
2 n 2 n
1 =x ∑n=0 n
x ; y2 = y1 ln x − 3 x ∑n=1 n
(∑j=1 )x
2 n! 2 n! j(2j+3)
2
23. y 1 =x
−2
(1 + 3x +
3
2
−
1
2
x
3
+ …) ; y2 = y1 ln x − 5 x
−1
(1 +
5
4
x−
1
4
x
2
+ …)
24. y 1
3
= x (1 + 20x + 180 x
2
+ 1120 x
3
+ …); y2 = y1 ln x − x
4
(26 + 324x + 69683 x
2
+ …)
25. y 1 = x (1 − 5x +
85
4
x
2
−
3145
36
x
3
+ …) ; y2 = y1 ln x + x
2
(2 −
39
4
x+
4499
108
x
2
+ …)
26. y 1 = 1 −x +
3
4
x
2
−
12
7
x
3
+… ; y2 = y1 ln x + x (1 −
3
4
x+
5
9
x
2
+ …)
27. y 1 =x
−3
(1 + 16x + 36 x
2
+ 16 x
3
+ …); y2 = y1 ln x − x
−2
(40 + 150x +
280
3
x
2
+ …)
m m
∞ (−1) ∞ (−1) ∞
28. y 1 = x ∑m=0
2
m
m!
x
2m
; y2 = y1 ln x −
x
2
∑m=1
2
m
m!
(∑j=1
1
j
)x
2m
2
∞ ∞
29. y 1 =x
2
∑m=0 (−1 )
m
(m + 1)x
2m
; y2 = y1 ln x −
x
2
∑m=1 (−1 )
m
mx
2m
m m
(−1) 1/2 (−1)
30. y 1 =x
1/2
∑
∞
m=0 m
4 m!
x
2m
; y2 = y1 ln x −
x
2
∑
∞
m=1 4
m
m!
(∑
m
j=1
1
j
)x
2m
m m m m
(−1 ) ∏ (2j−1) (−1 ) ∏ (2j−1)
∞ ∞ m
31. y x 1
j=1 j=1
2m 2m
1 = x ∑m=0 m x ; y2 = y1 ln x + ∑m=1 m (∑j=1 )x
2 m! 2 2 m! j(2j−1)
m m m m
(−1 ) ∏ (4j−1) 1/2 (−1 ) ∏ (4j−1)
32. y ∞ x ∞ m 1
j=1 j=1
1/2 2m 2m
1 =x ∑ m
x ; y2 = y1 ln x + ∑ m
(∑ )x
m=0 8 m! 2 m=1 8 m! j=1 j(4j−1)
m m m m
(−1 ) ∏ (2j+1) (−1 ) ∏ (2j+1)
33. y 1 =x∑
∞
m=0 2
m
j=1
m!
x
2m
; y2 = y1 ln x −
x
2
∑
∞
m=1 2
m
j=1
m!
(∑
m
j=1
1
j(2j+1)
)x
2m
m m m m
(−1 ) ∏ (8j−13) (−1 ) ∏ (8j−13)
34. y ∞ 13 ∞ m 1
j=1 j=1
−1/4 2m −1/4 2m
1 =x ∑ m
x ; y2 = y1 ln x + x ∑ m
(∑ )x
m=0 (32 ) m! 2 m=1 (32 ) m! j=1 j(8j−13)
m m m m
(−1 ) ∏j=1 (3j−1) 1/3 (−1 ) ∏j=1 (3j−1)
35. y 1 =x
1/3
∑
∞
m=0 9
m
m!
x
2m
; y2 = y1 ln x +
x
2
∑
∞
m=1 9
m
m!
(∑
m
j=1
1
j(3j−1)
)x
2m
36.
m m m m
(−1 ) ∏ (4j−3)(4j−1) (−1 ) ∏ (4j−3)(4j−1) 8j−3
1/2 ∞ j=1
2m 1/2 ∞ j=1 m 2m
y1 = x ∑ x ; y2 = y1 ln x + x ∑ (∑ )x
m=0 m 2 m=1 m 2 j=1 j(4j−3)(4j−1)
4 (m! ) 4 (m! )
m m
(−1) 5/3 (−1)
37. y 1 =x
5/3
∑
∞
m=0 m
3 m!
x
2m
; y2 = y1 ln x −
x
2
∑
∞
m=1 3
m
m!
(∑
m
j=1
1
j
)x
2m
m m m m
(−1 ) ∏ (4j−7) (−1 ) ∏ (4j−7)
38. y 1 =
1
x
∑
∞
m=0 2
m
j=1
m!
x
2m
; y2 = y1 ln x +
7
2x
∑
∞
m=1 m
2
j=1
m!
(∑
m
j=1 j(4j−7)
1
)x
2m
39. y 1 =x
−1
(1 −
3
2
x
2
+
15
8
x
4
−
35
16
x
6
+ …) ; y2 = y1 ln x + x (
1
4
−
13
32
x
2
+
101
192
x
4
+ …)
40. y 1 = x (1 −
1
2
x
2
+
1
8
x
4
−
1
48
x
6
+ …) ; y2 = y1 ln x + x
3
(
1
4
−
32
3
x
2
+
11
576
x
4
+ …)
41. y 1 =x
−2
(1 −
3
4
x
2
−
9
64
x
4
−
25
256
x
6
+ …) ; y2 = y1 ln x +
1
2
−
128
21
x
2
−
215
1536
x
4
+…
42. y 1 =x
−3
(1 −
17
8
x
2
+
85
256
x
4
−
85
18432
x
6
+ …) ; y2 = y1 ln x + x
−1
(
25
8
−
471
512
x
2
+
1583
110592
x
4
+ …)
43. y 1 =x
−1
(1 −
3
4
x
2
+
45
64
x
4
−
175
256
x
6
+ …) ; y2 = y1 ln x − x (
1
4
−
128
33
x
2
+
395
1536
x
4
+ …)
44. y 1 =
1
x
; y2 = y1 ln x − 6 + 6x −
8
3
x
2
45. y 1 = 1 − x; y2 = y1 ln x + 4x
2
(x−1)
46. y 1 =
x
; y2 = y1 ln x + 3 − 3x + 2 ∑
∞
n=2
1
n( n2 −1)
x
n
n
(−1)
47. y 1 =x
1/2
(x + 1 ) ;
2
y2 = y1 ln x − x
3/2
(3 + 3x + 2 ∑
∞
n=2 n( n −1)
2
x )
n
∞
48. y 1 = x (1 − x ) ;
2 3
y2 = y1 ln x + x
3
(4 − 7x +
11
3
x
2
−6 ∑
n=3 n(n−2)( n −1)
1
2
x )
n
49. y 1 = x − 4x
3
+x ;
5
y2 = y1 ln x + 6 x
3
− 3x
5
50. y 1 =x
1/3
(1 −
1
6
x );
2
y2 = y1 ln x + x
7/3
(
1
4
−
1
12
∑
∞
m=1 6
m
m(m+1)(m+1)!
1
x
2m
)
m
∞ (−1)
51. y 1 = (1 + x ) ;
2 2
y2 = y1 ln x −
3
2
x
2
−
3
2
x
4
+ ∑m=3
m(m−1)(m−2)
x
2m
52. y 1 =x
−1/2
(1 −
1
2
x
2
+
32
1
x );
4
y2 = y1 ln x + x
3/2
(
5
8
−
9
128
x
2
+∑
∞
m=2 m+1
1
x
2m
)
4 (m−1)m(m+1)(m+1)!
m m
∞ (−1) ∞ (−1) m
56. y 1 = ∑m=0
m 2
x
2m
; y2 = y1 ln x − ∑m=1
m 2
(∑j=1
1
j
)x
2m
4 (m! ) 4 (m! )
1/2 1/2
58. x
1+x
;
x
1+x
ln x
1/3 1/3
59. x
3+x
;
x
3+x
ln x
60. 2−x
x
2
;
x ln x
2−x
2
1/3 1/4
61. x
1+x
2
;
x
1+x
ln x
2
62. 4+3x
x
;
x ln x
4+3x
1/2 1/2
63. x
1+3x+x
2
;
x
1+3x+x
ln x
2
64. x
2
;
x ln x
2
(1−x) (1−x)
1/3 1/3
65. x
1+x+x2
;
x
1+x+x2
ln x
Section 7.7 Answers
n n
(−4) (−4) j+1
1. y
1 = 2x
3
∑
∞
n=0 n!(n+2)!
x ;
n
y2 = x + 4 x
2
− 8 (y1 ln x − 4 ∑
∞
n=1 n!(n+2)!
(∑
n
j=1 j(j+2)
)x )
n
n n
∞ (−1) ∞ (−1) n 2j+1
2. y
1 =x∑
n=0 n!(n+1)!
x ;
n
y2 = 1 − y1 ln x + x ∑
n=1 n!(n+1)!
(∑
j=1 j(j+1)
)x
n
n
∞ (−1)
3. y
1 =x
1/2
; y2 = x
−1/2
+ y1 ln x + x
1/2
∑n=1
n
x
n
n n
(−1) (−1)
4. y
1 =x∑
∞
n=0 n!
x
n
= xe
−x
; y2 = 1 − y1 ln x + x ∑
∞
n=1 n!
(∑ j = 1
n 1
j
)x
n
n n
n ∏j=1 (2j+1) n ∏j=1 (2j+1)
5. y
1 =x
1/2
∑
∞
n=0
(−
3
4
)
n!
x ;
n
y2 = x
−1/2
−
3
4
( y1 ln x − x
1/2
∑
∞
n=1
(−
3
4
)
n!
(∑
n
j=1
1
j(2j+1)
n
)x )
n n
(−1) (−1)
6. y
1 =x∑
∞
n=0 n!
x
n
= xe
−x
; y2 = x
−2
(1 +
1
2
x+
1
2
x )−
2 1
2
(y1 ln x − x ∑
∞
n=1 n!
(∑
n
j=1
1
j
)x )
n
n
∞ (−1)
7. y
1 = 6x
3/2
∑
n=0 n
4 n!(n+3)!
x ;
n
n
∞ (−1) n 2j+3
−3/2 1 1 2 1 3/2 n
y2 = x (1 + x+ x )− (y1 ln x − 6 x ∑ n
(∑ )x )
8 64 768 n=1 4 n!(n+3)! j=1 j(j+3)
n
∞ (−1)
8. y
1 =
120
x
2
∑
n=0 n!(n+5)!
x ;
n
y2 = x
−7
(1 +
1
4
x+
1
24
x
2
+
1
144
x
3
+
576
1
x )
4
n
120 ∞ (−1) n 2j+5
1 n
− (y1 ln x − ∑ (∑ )x )
2880 x
2 n=1 n!(n+f )! j=1 j(j+5)
1/2
∞
9. y
1 =
x
6
n
∑n=0 (−1 ) (n + 1)(n + 2)(n + 3)x ;
n
y2 = x
−5/2
(1 +
1
2
2
x + x ) − 3 y1 ln x +
3
2
x
1/2
∞ n n 1 n
∑ (−1 ) (n + 1)(n + 2)(n + 3) (∑ )x
n=1 j=1 j(j+3)
10. y 1 =x
4
(1 −
2
5
x) ; y2 = 1 + 10x + 50 x
2
+ 200 x
3
− 300 (y1 ln x +
27
25
x
5
−
1
30
x )
6
n
(−1 ) 6!
11. y 1 =x ;
3
y2 = x
−3
(1 −
6
5
x+
3
4
x
2
−
1
3
x
3
+
1
8
x
4
−
1
20
x )−
5
120
1
(y1 ln x + x
3
∑
∞
n=1 n(n+6)!
x )
n
∞ n 2j+3 2
12. y 1 =x
2
∑n=0
1
n!
(∏j=1
j+4
)x ;
n
y2 = x
−2
(1 + x +
1
4
x
2
−
12
1
x )−
3
16
1
y1 ln x +
x
2
∞ 1 n 2j+3 n ( j +3j+6)
n
∑ (∏ ) (∑ )x
n=1 n! j=1 j+4 j=1 j(j+4)(2j+3)
13. y 1 =x
5
∑
∞
n=0
n
(−1 ) (n + 1)(n + 2)x ;
n
y2 = 1 −
x
2
+
x
n
(−1) (j+3)(2j−3)
14. y 1 =
1
x
∑
∞
n=0 n!
(∏
n
j=1 j+6
)x ;
n
y2 = x
−7
(1 +
26
5
x+
143
20
x )
2
n
(−1)
15. y 1 =x
7/2
∑
∞
n=0 n
2 (n+4)!
x ;
n
y2 = x
−1/2
(1 −
1
2
x+
1
8
x
2
−
48
1
x )
3
n
∞ (−1 ) (n+1) n 3j+7
16. y 1 =x
10/3
∑n=0
9
n (∏j=1
j+4
)x ;
n
y2 = x
−2/3
(1 +
27
4
x−
1
243
x )
2
j−8
17. y 1 =x
3
∑
7
n=0
(−1 ) (n + 1) (∏
n n
j=1 j+6
)x ;
n
y2 = x
−3
(1 +
52
5
x+
234
5
x
2
+
572
5
x
3
+ 143 x )
4
n 2
(−1) (j+3)
18. y 1 =x
3
∑
∞
n=0 n!
(∏
n
j=1 j+5
)x ;
n
y2 = x
−2
(1 +
1
4
x)
j−5
19. y 1 =x
6
∑
4
n=0
(−1 ) 2
n n
(∏
n
j=1 j+5
)x ;
n
y2 = x(1 + 18x + 144 x
2
+ 672 x
3
+ 2016 x )
4
20. y 1 =x
6
(1 +
2
3
x+
1
7
x );
2
y2 = x (1 +
21
4
x+
21
2
x
2
+
35
4
x )
3
∞
21. y 1 =x
7/2
∑n=0 (−1 ) (n + 1)x ;
n n
y2 = x
−7/2
(1 −
5
6
x+
2
3
x
2
−
1
2
x
3
+
1
3
x
4
−
1
6
x )
5
10
22. y 1 =
x
6
∑
∞
n=0
n n
(−1 ) 2 (n + 1)(n + 2)(n + 3)x ;
n
y2 = (1 −
4
3
x+
5
3
x
2
−
40
21
x
3
+
40
21
x
4
−
32
21
x
5
+
16
21
6
x )
m m
(−1 ) ∏ (2j+5)
∞
23. y 3 15 75
j=1
6 2m 2 2 6
1 =x ∑m=0 m
x ; y2 = x (1 + x )− y1 ln x + x
2 m! 2 2 2
m m
(−1 ) ∏j=1 (2j+5)
∞ m 1 2m
∑ (∑ )x
m=1 m+1 j=1 j(2j+5)
2 m!
m m
(−1) 6 (−1)
24. y
2
6 ∞ 2m 6 −x 2 1 2 1 x ∞ m 1 2m
1 =x ∑ m
x =x e ; y2 = x (1 + x )− y1 ln x + ∑ m
(∑ )x
m=0 2 m! 2 2 4 m=1 2 m! j=1 j
m m
∞ (−1) ∞ (−1( m 2j+3
25. y1 = 6x
6
∑m=0
4
m
m!(m+3)!
x
2m
; y2 = 1 +
1
8
x
2
+
1
64
x
4
−
1
384
(y1 ln x − 3 x
6
∑m=1 m
4 m!(m+3)!
(∑j=1
j(j+3)
)x
2m
)
m m
2
∞ (−1 ) (m+2) ∞ (−1 ) (m+2) m j +4j+2
26. y1 =
x
2
∑m=0
m!
x
2m
; y2 = x
−1
− 4 y1 ln x ∑m=1
m!
(∑j=1
j(j+1)(j+2)
)x
2m
m m
(−1) (−1) j+1
27. y1 = 2x
3
∑
∞
m=0 4
m
m!(m+2)!
x
2m
; y2 = x
−1
(1 +
1
4
2
x )−
1
16
(y1 ln x − 2 x
3
∑
∞
m=1 4
m
m!(m+2)!
(∑
m
j=1 j(j+2)
)x
2m
)
m m
(−1 ) ∏ (2j−1)
28. y ∞ j=1
−1/2 2m
1 =x ∑ m
x ;
m=0 8 m!(m+1)!
m m
(−1 ) ∏ (2j−1) 2
2 j −2j−1
−5/2 1 −1/2 ∞ j=1 m 2m
y2 = x + y1 ln x − x ∑ (∑ )x
4 m=1 m+1 j=1 j(j+1)(2j−1)
8 m!(m+1)!
m m
(−1) (−1)
29. y
2
∞ 2m −x /2 −1 x ∞ m 1 2m
1 =x∑ m
x = xe ; y2 = x − y1 ln x + ∑ m
(∑ )x
m=0 2 m! 2 m=1 2 m! j=1 j
30. y
2
2 ∞ 1 2m 2 x −2 2 2 ∞ 1 m 1 2m
1 =x ∑ x =x e ; y2 = x (1 − x ) − 2 y1 ln x + x ∑ (∑ )x
m=0 m! m=1 m! j=1 j
m
(−1)
31. y1 = 6x
5/2
∑
∞
m=0 16
m
m!(m+3)!
x
2m
; y2 = x
−7/2
(1 +
1
32
x
2
+
1024
1
x )
4
m
∞ (−1) m 2j+3
1 5/2 2m
− (y1 ln x − 3 x ∑ m
(∑ )x )
24576 m=1 16 m!(m+3)! j=1 j(j+3)
m
∏ (3j+1)
∞
32. y
j=1
13/3 2m
1 = 2x ∑m=0 m
x ;
9 m!(m+2)!
m
∏ (3j+1) 2
3 j +2j+2
1/3 2 2 2 13/3 ∞ j=1 m 2m
y2 = x (1 + x )+ ( y1 ln x − x ∑ m
(∑ )x )
9 81 m=0 9 m!(m+2)! j=1 j(j+2)(3j+1)
∞
33. y1 =x ;
2
y2 = x
−2 2
(1 + 2 x ) − 2 (y1 ln x + x
2
∑m=1
1
m(m+2)!
x
2m
)
m
3
( )
34. y ∞ 2
2 1 2 −2 9 2 27 7 4 2 2m
1 =x (1 − x ); y2 = x (1 + x )− ( y1 ln x + x −x ∑ x )
2 2 2 12 m=2 m(m−1)(m+2)!
35. y1 =∑
∞
m=0
(−1 )
m
(m + 1)x
2m
; y2 = x
−4
m
∞ (−1)
36. y1 =x
5/2
∑m=0
(m+1)(m+2)(m+3)
x
2m
; y2 = x
−7/2
(1 + x )
2 2
37. y1 =
x
5
∑
∞
m=0
(−1 )
m
(m + 5)x
2m
; y2 = x
−1
(1 − 2 x
2
+ 3x
4
− 4x )
6
∞ 2j+1
m m+1 m
38. y1 =x
3
∑m=0 (−1 )
2
m (∏j=1
j+5
)x
2m
; y2 = x
−7
(1 +
21
8
x
2
+
35
16
x
4
+
35
64
x )
6
m
∏j=1 (4j+5)
39. y1 = 2x
4
∑
∞
m=0
(−1 )
m
2
m
(m+2)!
x
2m
; y2 = 1 −
1
2
x
2
m m
(−1 ) ∏j=1 (2j−1)
40. y1 =x
3/2
∑
∞
m=0 m−1
x
2m
; y2 = x
−5/2
(1 +
3
2
x )
2
2 (m+2)!
m m
∞ (−1) v−1 (−1)
42. y1 =x
v
∑m=0
4
m
m! ∏
m
(j+v)
x
2m
; y2 = x
−v
∑m=0
4
m
m! ∏
m
(j−v)
x
2m
j=1 j=1
m
2 x
v
∞ (−1) m 2j+v
2m
− v
( y1 ln x − ∑ m m
(∑ )x )
4 v!(v−1)! 2 m=1 4 m! ∏ (j+v) j=1 j(j+v)
j=1
Section 8.1 Answers
1.
a. 1
s
2
b. 1
2
(s+1)
c. 2
b
2
s −b
d. −2s+5
(s−1)(s−2)
e. 2
s3
2.
2
s +2
a. 2 2
[(s−1 ) +1][(s+1 ) +1]
b. s( s +4)
2
2
2
s +8
c. s( s +16)
2
d. s −2
s( s2 −4)
e. 4s
2
2
( s −4 )
f. s2 +4
1
g. 1 s+1
s2 +1
√2
h. 2
( s +4)( s +9)
5s
2
3 2
i. s +2 s +4s+32
2
( s +4)( s +16)
2
4.
a. f (3−) = −1, f (3) = f (3+) = 1
b. f (1−) = 3, f (1) = 4, f (1+) = 1
c. f ( −) = 1, f ( ) = f ( +) = 2, f (π−) = 0, f (π) = f (π+) = −1
π
2
π
2
π
d. f (1−) = 1, f (1) = 2, f (1+) = 1, f (2−) = 0, f (2) = 3, f (2+) = 6

5.
−( s+1) −( s+2)
a. 1−e
s+1
+
e
s+2
b. 1
s
+e
−4s
(
1
s2
+
3
s
)
−s
c. 1−e
s
2
−( s−1)
d. 1−e
2
(s−1)
2 2
(s−λ) −ω 2ω(s−λ)
7. L(e λt
cos ωt) = 2
2
2
L(e
λt
sin ωt) = 2
2
2
((s−λ) +ω ) ((s−λ) +ω )
15.
a. tan −1 ω
s
, s >0
2
b. ln1
2 2
s +ω
s
2
, s >0
c. ln
s−b
s−a
, s > max(a, b)
2
d. 1
2
ln
s2 −1
s
, s >1
2
e. 1
4
ln 2
s −4
s
, s >2
18.
a. 1
s
2
tanh
s
b. 1
s
tanh
s
c. 1
s2 +1
coth
πs
d. 2
1
( s +1)(1−e −πs
)
Section 8.2 Answers
1.
3 7t
a. t e
b. 2e 2t
cos 3t
−2t
c. e
sin 4t
4
d. sin 3t
2
e. t cos t
2t
f. sinh 2t
e
2
2t
g. 2te
3
sin 9t
3t
h. 2e
3
sinh 3t
i. e 2t
t cos t
2.
a. t e + t e
2 7t 17
6
3 7t
b. e ( t + t + t )
2t 1
6
3 1
6
4 1
40
5
c. e (cos 3t + sin 3t)

−3t 2
d. 2 cos 3t + sin 3t 1
e. (1 − t)e −t
f. cosh 3t + sinh 3t 1
g. (1 − t − t − t ) e 2 1
6
3 −t
h. e (2 cos 2t + sin 2t)

t 5
i. 1 − cos t
j. 3 cosh t + 4 sinh t
k. 3e + 4 cos 3t + sin 3t
t 1
l. 3te − 2 cos 2t − 3 sin 2t

−2t
3.
a. e − e − e
1
4
2t 1
4
−2t −t
b. e − e + 5e
1
5
−4t 41
5
t 3t
c. − e − e − e1
2
2t 13
10
−2t 1
5
3t
d. − e − e 2
5
−4t 3
5
t
e. e − e + e + e
3
20
2t 37
12
−2t 1
3
t 8
5
−3t
f. e + e +
39
10
e − e
t
14
3 3t 23
105
−4t 7
3
2t
4.
a. e − e − cos t + sin t
4
5
−2t 1
2
−t 3
10
11
10
b. sin t + cos t + e sin t − e cos t

2
5
6
5
7
5
−t 6
5
−t
c. e − e cos 2t + e sin 2t
8
13
2t
13
8 −t 15
26
−t
d. te + e + e − e
1
2
t 3
8
t −2t 11
8
−3t
e. te + e + te − e
2
3
t 1
9
t −2t 1
9
−2t
f. −e + te + cos t − sin t
t 5
2
t 3
5.
a. cos 2t + sin 2t − cos 3t − sin 3t
3
5
1
5
3
5 15
2
b. − cos t + sin t + cos 4t − sin 4t

4
15
1
15
4
15
1
60
c. cos t + sin t − cos 2t − sin 2t

5
3
5
3
1
d. − cos + sin + cos t − sin t

1
3
t
2
2
3
t
2
1
3
1
e. 1
15
cos
t
4
−
15
8
sin
t
4
−
15
1
cos 4t +
30
1
sin 4t
f. 2
5
cos
t
3
−
3
5
sin
t
3
−
2
5
cos
t
2
+
2
5
sin
t
6.
a. e t
(cos 2t + sin 2t) − e
−t
(cos 3t +
4
3
sin 3t)
b. e 3t
(− cos 2t +
3
2
sin 2t) + e
−t
(cos 2t +
1
2
sin 2t)
c. e −2t
(
1
8
cos t +
1
4
sin t) − e
2t
(
1
8
cos 3t −
12
1
sin 3t)
d. e 2t
(cos t +
1
2
sin t) − e
3t
(cos 2t −
1
4
sin 2t)
e. e t
(
1
5
cos t +
2
5
sin t) − e
−t
(
1
5
cos 2t +
2
5
sin 2t)
f. e t/2
(− cos t +
9
8
sin t) + e
−t/2
(cos t −
1
8
sin t)
7.
a. 1 − cos t
t
b. (1 − cos 4t)
e
16
c. e + e sin 3t − e cos 3t
4
9
2t 5
9
−t 4
9
−t
d. 3e − e sin 2t − 3e cos 2t
t/2 7
2
t t
e. e − e cos 2t
1
4
3t 1
4
−t
f. e − e cos 3t + e sin 3t
1
9
2t 1
9
−t 5
9
−t
8.
a. − 3
10
sin t +
2
5
cos t −
3
4
e +
t
20
7
e
3t
b. − 3
5
e
−t
sin t +
1
5
e
−t
cos t −
1
2
e
−t
+
3
10
e
t
c. − 1
10
t
e sin t −
7
10
t
e cos t +
1
5
e
−t
+
1
2
e
2t
d. − 1
2
e +
t 7
10
e
−t
−
1
5
cos 2t +
3
5
sin 2t
e. 3
10
+
1
10
e
2t
+
10
1 t
e sin 2t −
2
5
t
e cos 2t
f. − 4
9
e
2t
cos 3t +
1
3
e
2t
sin 3t −
5
9
e
2t
+e
t
9. 1
a
e a
t
f (
t
a
)
Section 8.3 Answers
1. y = 1
6
e −
t 9
2
e
−t
+
16
3
e
−2t
2. y = − 1
3
+
15
8
e
3t
+
4
5
e
−2t
3. y = − 23
15
e
−2t
+
1
3
e +
t 1
5
e
3t
4. y = − 1
4
e
2t
+
17
20
e
−2t
+
2
5
e
3t
5. y = 11
15
e
−2t
+
1
6
e +
t 1
10
e
3t
6. y = e t
+ 2e
−2t
− 2e
−t
7. y = 5
3
sin t −
1
3
sin 2t
8. y = 4e t
− 4e
2t
+e
3t
9. y = − 7
2
e
2t
+
13
3
e +
t 1
6
e
4t
10. y = 5
2
t
e − 4e
2t
+
1
2
e
3t
11. y = 1
3
t
e − 2e
−t
+
5
3
e
−2t
12. y = 2 − e −2t
+e
t
13. y = 1 − cos 2t + 1
2
sin 2t
14. y = − 1
3
+
15
8
e
3t
+
4
5
e
−2t
15. y = 1
6
e −
t 2
3
e
−2t
+
1
2
e
−t
16. y = −1 + e t
+e
−t
17. y = cos 2t − sin 2t + sin t

18. y = 7
3
−
7
2
e
−t
+
1
6
e
3t
19. y = 1 + cos t
20. y = t + sin t
21. y = t − 6 sin t + cos t + sin 2t
22. y = e −t
+ 4e
−2t
− 4e
−3t
23. y = −3 cos t − 2 sin t + e −t

(2 + 5t)
24. y = − sin t − 2 cos t + 3e 3t

+e
−t
25. y = (3t + 4) sin t − (2t + 6) cos t

26. y = −(2t + 2) cos 2t + sin 2t + 3 cos t
27. y = e t
(cos t − 3 sin t) + e
3t
28. y = −1 + t + e −t
(3 cos t − 5 sin t)
29. y = 4 cos t − 3 sin t − e t

(3 cos t − 8 sin t)
30. y = e −t
− 2e + e
t −2t
(cos 3t − 11/3 sin 3t)
31. y = e −t
(sin t − cos t) + e
−2t
(cos t + 4 sin t)
32. y = 1
5
e
2t
−
4
3
e +
t 32
15
e
−t/2
33. y = 1
7
e
2t
−
2
5
e
t/2
+
9
35
e
−t/3
34. y = e −t/2
(5 cos(t/2) − sin(t/2)) + 2t − 4
35. y = 1
17
(12 cos t + 20 sin t − 3 e
t/2
(4 cos t + sin t))
−t/2
36. y = e
10
(5t + 26) −
1
5
(3 cos t + sin t)
37. y = 100
1
(3 e
3t
−e
t/3
(3 + 310t))
Section 8.4 Answers
1. 1 + u(t − 4)(t − 1); 1
s
+e
−4s
(
s
1
2
+
3
s
)
−s
2. t + u(t − 1)(1 − t); 1−e
s
2
3. 2t − 1 − u(t − 2)(t − 1); (

2
s2
−
1
s
)−e
−2s
(
1
s2
+
1
s
)
4. 1 + u(t − 1)(t + 1); 1
s
+e
−s
(
s
1
2
+
2
s
)
5. t − 1 + u(t − 2)(5 − t); s

1
2
−
1
s
−e
−2s
(
1
s
2
−
3
s
)
6. t
2
(1 − u(t − 1));
s
2
3
−e
−s
(
s
2
3
+
2
2
s
+
1
s
)
7. u(t − 2)(t 2
+ 3t); e
−2s
(
2
s3
+
s2
7
+
10
s
)
8. t
2
+ 2 + u(t − 1)(t − t
2
− 2);
s3
2
+
2
s
−e
−s
(
s3
2
+
1
s2
+
2
s
)
−( s−1)
9. te t
+ u(t − 1)(e − te );
t t 1−e
2
(s−1)
−( s+1) −( s+2)
10. e −t
+ u(t − 1)(e
−2t
−e
−t
);
1−e
s+1
+
e
s+2
−2s
11. −t + 2u(t − 2)(t − 2) − u(t − 3)(t − 5); −

1
2
s
+
2e
s
2
+e
−3s
(
2
s
−
1
2
s
)
12. [u(t − 1) − u(t − 2)] t; e

−s
(
s
1
2
+
1
s
)−e
−2s
(
1
s
2
+
2
s
)
13. t + u(t − 1)(t 2

− t) − u(t − 2)t ;
2 1
s
2
+e
−s
(
s
2
3
+
s
1
2
)−e
−2s
(
s
2
3
+
4
s
2
+
4
s
)
−s
14. t + u(t − 1)(2 − 2t) + u(t − 2)(4 + t); 1
s2
−2
e
s2
+e
−2s
(
s2
1
+
6
s
)
π
− s
−πs
1+e 2 s−e (s−2)
15. sin t + u(t − π/2) sin t + u(t − π)(cos t − 2 sin t); 2
s +1
16. 2 − 2u(t − 1)t + u(t − 3)(5t − 2); 2
s
−e
−s
(
2
s2
+
2
s
)+e
−3s
(
5
s2
+
13
s
)
17. 3 + u(t − 2)(3t − 1) + u(t − 4)(t − 2); 3
s
+e
−2s
(
s2
3
+
5
s
)+e
−4s
(
1
s2
+
2
s
)
18. (t + 1) 2
+ u(t − 1)(2t + 3);
2
s
3
+
2
s
2
+
1
s
+e
−s
(
s
2
2
+
5
s
)
0, 0 ≤ t < 2,
19. u(t − 2)e 2(t−2)
={
2(t−2)
e , t ≥2
0, 0 ≤ t < 1,
20. u(t − 1) (1 − e −(t−1)
) ={
−(t−1)
1 −e , t ≥1
⎧ 0, 0 ≤ t < 1,
⎪
⎪
2
⎪ 2
(t−1) (t−1)
21. u(t − 1) 2
+ u(t − 2)(t − 2) = ⎨
2
, 1 ≤ t < 2,
⎪
⎪
⎩
⎪ 2
t −3
, t ≥2
2
⎧ 2 + t, 0 ≤ t < 1,
⎪
22. 2 + t + u(t − 1)(4 − t) + u(t − 3)(t − 2) = ⎨ 6, 1 ≤ t < 3,

⎩
⎪
t + 4, t ≥3
⎧ 5 − t, 0 ≤ t < 3,
⎪
23. 5 − t + u(t − 3)(7t − 15) + 3
2
u(t − 6)(t − 6 )
2
=⎨ 6t − 10, 3 ≤ t < 6,
⎩
⎪ 3 2
44 − 12t + t , t ≥6
2
0, 0 ≤ t < π,
24. u(t − π)e −2(t−π)
(2 cos t − 5 sin t) = {
−2(t−π)
e (2 cos t − 5 sin t) t ≥π
π
1 − cos t, 0 ≤t < ,
25. 1 − cos t + u(t − π/2)(3 sin t + cos t) = { π
2
1 + 3 sin t, t ≥
2
0, 0 ≤ t < 2,
26. u(t − 2)(4e −(t−2)
− 4e
2(t−2)
+ 2e
(t−2)
={
−(t−2) 2(t−2) (t−2)
4e − 4e + 2e , t ≥2
⎧ t + 1,
⎪
0 ≤ t < 1,
27. 1 + t + u(t − 1)(2t + 1) + u(t − 3)(3t − 5) = ⎨ 3t + 2, 1 ≤ t < 3,

⎩
⎪
6t − 3, t ≥3
2
⎧ 1 −t , 0 ≤ t < 2,
⎪
⎪
2
2
3t
28. 1 − t 2
+ u(t − 2) (−
t
2
+ 2t + 1) + u(t − 4)(t − 4) = ⎨ −
2
+ 2t + 2, 2 ≤ t < 4,
⎪
⎩
⎪ 2
3t
− + 3t − 2, t ≥4
2
−τ s
29. e
30. For each t only finitely many terms are nonzero.

33. 1 + ∑ ∞
m=1
u(t − m);
1
s(1−e
−s
)
−s
∞ 1−e
34. 1 + 2 ∑ m=1
(−1 )
m
u(t − m);
1
s
;
1+e
−s
−s −s
∞ e (1+e )
35. 1 + ∑ m=1
(2m + 1)u(t − m);
−s
2
s(1−e )
s
∞ (1−e )
36. ∑ m=1
(−1 )
m
(2m − 1)u(t − m);
1
s s
2
(1+e )
Section 8.5 Answers
1. y = 3(1 − cos t) − 3u(t − π)(1 + cos t)
2. y = 3 − 2 cos t + 2u(t − 4)(t − 4 − sin(t − 4))
u(t−1)
3. y = − 15
2
+
3
2
e
2t
− 2t +
2
(e
2(t−1)
− 2t + 1)
4. y = 1
2
e +
t 13
6
e
−t
+
1
3
e
2t
+ u(t − 2) (−1 +
1
2
e
t−2
+
1
2
e
−(t−2)
+
1
2
e
t+2
−
1
6
e
−(t−6)
−
1
3
e
2t
)
5. y = −7e t
+ 4e
2t
+ u(t − 1) (
1
2
−e
t−1
+
1
2
e
2(t−1)
) − 2u(t − 2) (
1
2
−e
t−2
+
1
2
e
2(t−2)
)
6. y = 1
3
sin 2t − 3 cos 2t +
1
3
sin t − 2u(t − π) (
1
3
sin t +
1
6
sin 2t) + u(t − 2π) (
1
3
sin t −
1
6
sin 2t)
7. y = 1
4
−
31
12
e
4t
+
16
3
t
e + u(t − 1) (
2
3
e
t−1
−
1
6
e
4(t−1)
−
1
2
) + u(t − 2) (
1
4
+
12
1
e
4(t−2)
−
1
3
e
t−2
)
8. y = 1
8
(cos t − cos 3t) −
1
8
u (y −
3π
2
) (sin t − cos t + sin 3t −
1
3
cos 3t)
9. y = 4
t
−
1
8
sin 2t +
1
8
u (t −
π
2
) (π cos 2t − sin 2t + 2π − 2t)
10. y = t − sin t − 2u(t − π)(t + sin t + π cos t)

2( t−2)
11. y = u(t − 2) (t − 1
2
+
e
2
− 2e
t−2
)
12. y = t + sin t + cos t − u(t − 2π)(3t − 3 sin t − 6π cos t)

13. y = 1
2
+
1
2
e
−2t
−e
−t
+ u(t − 2) (2 e
−(t−2)
−e
−2(t−2)
− 1)
14. y = − 1
3
−
1
6
e
3t
+
1
2
t
e + u(t − 1) (
2
3
+
1
3
e
3(t−1)
−e
t−1
)
15. y = 1
4
(e + e
t −t
(11 + 6t)) + u(t − 1)(te
−(t−1)
− 1)
16. y = e t
−e
−t
− 2te
−t
− u(t − 1) (e − e
t −(t−2)
− 2(t − 1)e
−(t−2)
)
17. y = te −t
+e
−2t
+ u(t − 1) (e
−t
(2 − t) − e
−(2t−1)
)
2 2t
18. y = t e
2
− te
2t
− u(t − 2)(t − 2 ) e
2 2t
19. y = t
12
+1 −
1
12
u(t − 1)(t
4
+ 2t
3
− 10t + 7) +
1
6
u(t − 2)(2 t
3
+ 3t
2
− 36t + 44)
20. y = 1
2
e
−t
(3 cos t + sin t) +
1
2
− u(t − 2π) (e
−(t−2π)
((π − 1) cos t +
2π−1
2
sin t) + 1 −
t
2
)
1 −(t−3π)
− u(t − 3π)(e (3π cos t + (3π + 1) sin t) + t)
2
2
2
∞ (t−m)
21. y = t
2
+∑
m=1
u(t − m)
2
22.
2m + 1 − cos t, 2mπ ≤ t < (2m + 1)π (m = 0, 1, …)
a. y ={
2m, (2m − 1)π ≤ t < 2mπ (m = 1, 2, …)
b. y = (m + 1)(t − sin t − mπ cos t), 2mπ ≤ t < (2m + 2)π (m = 0, 1, …)

c. y = (−1) − (2m + 1) cos t, mπ ≤ t < (m + 1)π (m = 0, 1, …)
m
m+1
d. y = e
2(e−1)
−1
(e
t−m
+e
−t
) − m − 1, m ≤ t < m +1 (m = 0, 1, …)
2( m+1) π
e. y = (m + 1 − (
e
e
2π
−1
−1
)e
−t
) sin t 2mπ ≤ t < 2(m + 1)π (m = 0, 1, …)
m+1 2m+2
f. y = m+1
2
−e
t−m e
e−1
−1
+
1
2
e
2(t−m) e
e −1
2
−1
, m ≤ t < m +1 (m = 0, 1, …)
Section 8.6 Answers
1.
t
a. 1
2
∫
0
τ sin 2(t − τ )dτ
t
b. ∫ 0
e
−2τ
cos 3(t − τ )dτ
t t
c. 1
2
∫
0
sin 2τ cos 3(t − τ )dτ or 1
3
∫
0
sin 3τ cos 2(t − τ )dτ
t
d. ∫ 0
cos τ sin(t − τ )dτ
t
e. ∫ 0
e
aτ
dτ
t
f. e −t
∫
0
sin(t − τ )dτ
t
g. e −2t
∫
0
τe
τ
sin(t − τ )dτ
−2t
t
h. e
2
∫
0
τ
2
(t − τ )e
3τ
dτ
t
i. ∫ 0
(t − τ )e
τ
cos τ dτ
t
j. ∫ 0
e
−3τ
cos τ cos 2(t − τ )dτ
t
k. 4!5!
1
∫
0
τ
4
(t − τ ) e
5 3τ
dτ
t
l. 1
4
∫
0
τ
2
e
τ
sin 2(t − τ )dτ
t
m. 1
2
∫
0
τ (t − τ ) e
2 2(t−τ)
dτ
t
n. 5!6!
1
∫
0
(t − τ ) e
5 2(t−τ)
τ
6
dτ
2.
a. as
2
( s2 +a2 )( s2 +b )
b. (s−1)( s2 +a2 )
a
c. as
2
( s2 −a2 )
2 2
2ωs( s −ω )
d. 2 2
4
( s +ω )
(s−1)ω
e. 2 2 2
((s−1 ) +ω )
f. 3
2
2
(s−2 ) (s−1 )
s+1
g. 2 2
(s+2 ) [(s+1 ) +ω2 ]
h. 1
2
(s−3)((s−1 ) −1)
i. 2
2
2
(s−2 ) ( s +4)
j. 4
s (s−1)
6
k. 3⋅6!
2
7
s [(s+1 ) +9]
l. 12
s
7
m. 2⋅7!
2
8
s [(s+1 ) +4]
n. 5
48
s ( s +4)
2
3.
t √5τ
a. y =
√5
2
∫
0
f (t − τ )e
−3τ/2
sinh
2
dτ
t
b. y = 1
2
∫
0
f (t − τ ) sin 2τ dτ
t
c. y = ∫ τ e 0
−τ
f (t − τ )dτ
t
d. y(t) = − 1
k
sin kt + cos kt +
1
k
∫
0
f (t − τ ) sin kτ dτ
t
e. y = −2te + ∫ −3t
0
τe
−3τ
f (t − τ )dτ
t
f. y = sinh 2t + 3
2
1
2
∫
0
f (t − τ ) sinh 2τ dτ
t
g. y = e 3t
+∫
0
(e
3τ
−e
2τ
)f (t − τ )dτ
k1 t
h. y = ω
sin ωt + k0 cos ωt +
1
ω
∫
0
f (t − τ ) sin ωτ dτ
4.
a. y = sin t
b. y = te −t
c. y = 1 + 2te t
d. y = t + t
e. y = 4 + t + 5
2
2 1
24
4
t
f. y = 1 − t
5.
a. 7!8!
16!
t
16
b. 13!7!
21!
t
21
c. 6!7!
14!
t
14
d. 1
2
(e
−t
+ sin t − cos t)
e. 1
3
(cos t − cos 2t)
Section 8.7 Answers
1. y = 1
2
e
2t
− 4e
−t
+
11
2
e
−2t
+ 2u(t − 1)(e
−(t−1)
−e
−2(t−1)
)
2. y = 2e −2t
+ 5e
−t
+
5
3
u(t − 1)(e
(t−1)
−e
−2(t−1)
)
3. y = 1
6
e
2t
−
2
3
e
−t
−
1
2
e
−2t
+
5
2
u(t − 1) sinh 2(t − 1)
4. y = 1
8
(8 cos t − 5 sin t − sin 3t) − 2u(t − π/2) cos t
5. y = 1 − cos 2t + 1
2
sin 2t +
1
2
u(t − 3π) sin 2t
6. y = 4e t
+ 3e
−t
− 8 + 2u(t − 2) sinh(t − 2)
7. y = 1
2
e −
t 7
2
e
−t
+ 2 + 3u(t − 6)(1 − e
−(t−6)
)
8. y = e 2t
+ 7 cos 2t − sin 2t −
1
2
u(t − π/2) sin 2t
9. y = 1
2
(1 + e
−2t
) + u(t − 1)(e
−(t−1)
−e
−2(t−1)
)
10. y = 1
4
e +
t 1
4
e
−t
(2t − 5) + 2u(t − 2)(t − 2)e
−(t−2)
11. y = 1
6
(2 sin t + 5 sin 2t) −
1
2
u(t − π/2) sin 2t
12. y = e −t
(sin t − cos t) − e
−(t−π)
sin t − 3u(t − 2π)e
−(t−2π)
sin t
13. y = e −2t
(cos 3t +
4
3
sin 3t) −
1
3
u(t − π/6)e
−2(t−π/6)
cos 3t −
2
3
u(t − π/3)e
−2(t−π/3)
sin 3t
14. y = 7
10
e
2t
−
6
5
e
−t/2
−
1
2
+
1
5
u(t − 2)(e
2(t−2)
−e
−(t−2)/2
)
15. y = 1
17
(12 cos t + 20 sin t) +
1
34
e
t/2
(10 cos t − 11 sin t) − u(t − π/2)e
(2t−π)/4
cos t + u(t − π)e
(t−π)/2
sin t
16. y = 1
3
(cos t − cos 2t − 3 sin t) − 2u(t − π/2) cos t + 3u(t − π) sin t
17. y = e t
−e
−t
(1 + 2t) − 5u(t − 1) sinh(t − 1) + 3u(t − 2) sinh(t − 2)
18. y = 1
4
(e − e
t −t
(1 + 6t)) − u(t − 1)e
−(t−1)
+ 2u(t − 2)e
−(t−2)
)
19. y = 5
3
sin t −
1
3
sin 2t +
1
3
u(t − π)(sin 2t + 2 sin t) + u(t − 2π) sin t
20. y = 3
4
cos 2t −
1
2
sin 2t +
1
4
+
1
4
u(t − π/2)(1 + cos2t) +
1
2
u(t − π) sin 2t +
3
2
u(t − 3π/2) sin 2t
21. y = cos t − sin t

22. y = 1
4
(8 e
3t
− 12 e
−2t
)
23. y = 5(e −2t

−e
−t
)
24. y = e −2t
(1 + 6t)
25. y = 1
4
e
−t/2
(4 − 19t)
29. y = (−1) k
m ω1 Re
−cτ/2m
δ(t − τ ) if ω 1τ − ϕ = (2k + 1)π/2(k = integer )
30.
m+1 t−m −t
(e −1)( e −e )
a. y =
2(e−1)
, m ≤ t < m + 1, (m = 0, 1, …)
b. y = (m + 1) sin t, 2mπ ≤ t < 2(m + 1)π, (m = 0, 1, …)

2m+2 m+1
2(t−m) e −1 (t−m) e −1
c. y =e 2
−e
e−1
, m ≤ t < m + 1, (m = 0, 1, …)
2 −1
0, 2mπ ≤ t < (2m + 1)π,

d. y = { (m = 0, 1, …)
− sin t, (2m + 1)π ≤ t < (2m + 2)π,
Section 9.1 Answers
2. y = 2x 2
− 3x
3
+
x
1
3. y = 2e x
+ 3e
−x
−e
2x
+e
−3x
i−1
(x−x0 )
4. y i =
(i−1)!
, 1 ≤i ≤n
5.
b. y 1 =−
1
2
x
3
+x
2
+
2x
1
, y2 =
1
3
2
x −
1
3x
, y3 =
1
4
3
x −
1
3
2
x +
1
12x
c. y = k 0 y1 + k1 y2 + k2 y3
7. 2e −x
–
8. √2K cos x
9.
a. W (x) = 2e 3x
d. y = e x
(c1 + c2 x + c3 x )
2
10.
a. 2
b. −e 3x
c. 4
d. 4/x 2
e. 1
f. 2x
g. 2/x 2
h. e (x − 2x + 2)
x 2
i. −240/x 5
j. 6e (2x − 1)
2x
k. −128x
24.
a. y = 0 ′′′
b. x y − y − x y + y = 0
′′′ ′′ ′
c. (2x − 3)y − 2y − (2x − 5)y = 0 ′′′ ′′ ′
d. (x − 2x + 2)y − x y + 2x y − 2y = 0
2 ′′′ 2 ′′ ′
e. x y + x y − 2x y + 2y = 9
3 ′′′ 2 ′′ ′
f. (3x − 1)y − (12x − 1)y + 9(x + 1)y − 9y = 0

′′′ ′′ ′
g. x y + 5x y − 3x y − 6x y + 6y = 0
4 (4) 3 ′′′ 2 ′′ ′
h. x y + 3x y − x y + 2x y − 2y = 0
4 (4) 2 ′′′ 2 ′′ ′
i. (2x − 1)y − 4x y + (5 − 2x)y + 4x y − 4y = 0

(4) ′′′ ′′ ′
j. x y − y − 4x y + 4y = 0
(4) ′′′ ′′ ′
Section 9.2 Answers
1. y = e x
(c1 + c2 x + c3 x )
2
2. y = c 1e
x
+ c2 e
−x
+ c3 cos 3x + c4 sin 3x
3. y = c 1e
x
4. y = c 1e
x
+ c2 e
−x
+ c3 e
−3x/2
5. y = c 1e
−x
+e
−2x
6. y = c 1e
x
+e
x/2
(c2 + c3 x)
7. y = e −x/3
(c1 + c2 x + c3 x )
2
8. y = c 1 + c2 x + c3 cos x + c4 sin x
9. y = c 1e
2x
+ c2 e
−2x
– –
10. y = (c 1 + c2 x) cos √6x + (c3 + c4 x) sin √6x
11. y = e 3x/2
(c1 + c2 x) + e
−3x/2
(c3 + c4 x)
12. y = c 1e
−x/2
+ c2 e
−x/3
+ c3 cos x + c4 sin x
13. y = c 1e
x
+ c2 e
−2x
+ c3 e
−x/2
+ c4 e
−3x/2
14. y = e x
(c1 + c2 x + c3 cos x + c4 sin x)
15. y = cos 2x − 2 sin 2x + e 2x
16. y = 2e x
+ 3e
−x
− 5e
−3x
17. y = 2e x
+ 3x e
x
− 4e
−x
18. y = 2e −x
cos x − 3 e
−x
sin x + 4 e
2x
19. y = 9
5
e
−5x/3 x
+ e (1 + 2x)
20. y = e 2x
(1 − 3x + 2 x )
2
21. y = e 3x
(2 − x) + 4 e
−x/2
22. y = e x/2
(1 − 2x) + 3 e
−x/2
23. y = 1
8
(5 e
2x
+e
−2x
+ 10 cos 2x + 4 sin 2x)
24. y = −4e x
+e
2x
−e
4x
+ 2e
−x
25. y = 2e x
−e
−x
26. y = e 2x
+e
−2x
+e
−x
(3 cos x + sin x)
27. y = 2e −x/2
+ cos 2x − sin 2x
28.
a. {e , x e , e } : 1
x x 2x
b. {cos 2x, sin 2x, e } : 26 3x
c. {e cos x, e sin x, e } : 5
−x −x x
d. {1, x, x , e } 2e 2 x x
e. {e , e , cos x, sin x} 8
x −x
f. {cos x, sin x, e cos x, e sin x} x x

: 5
29. {e −3x
cos 2x, e
−3x
sin 2x, e
2x
, xe
2x 2
, 1, x, x }
30. {e x
, xe , e
x x/2
, xe
x/2
,x e
2 x/2
, cos x, sin x}
31. {cos 3x, x cos 3x, x 2
cos 3x, sin 3x, x sin 3x, x
2
sin 3x, 1, x}
32. {e 2x
, xe
2x
,x e
2 2x
,e
−x
, xe
−x
, 1}
33. {cos x, sin x, cos 3x, x cos 3x, sin 3x, x sin 3x, e 2x
}
34. {e 2x
, xe
2x
,e
−2x
, xe
−2x
, cos 2x, x cos 2x, sin 2x, x sin 2x}
35. {e −x/2
cos 2x, x e
−x/2
cos 2x, x e
2 −x/2
cos 2x, e
−x/2
sin 2x, x e
−x/2
sin 2x, x e
2 −x/2
sin 2x}
36. {1, x, x 2
,e
2x
, xe
2x
, cos 2x, x cos 2x, sin 2x, x sin 2x}
37. {cos(x/2), x cos(x/2), sin(x/2), x sin(x/2), cos 2x/3x cos(2x/3), x 2

cos(2x/3), sin(2x/3), x sin(2x/3), x
2
sin
(2x/3)}
38. {e −x
,e
3x
,e
x
cos 2x, e
x
sin 2x}
39. b. e ( a1 +a2 +…+an )x

∏
1≤i<j≤n
(aj − ai )
43.
√3 √3
a. x
{e , e
−x/2
cos(
2
x), e
−x/2
sin(
2
x)}
√3 √3
b. {e −x
,e
x/2
cos(
2
x), e
x/2
sin(
2
x)}
c. {e
2x
cos 2x, e
2x
sin 2x, e
−2x
cos 2x, e
−2x
sin 2x}
√3 √3 √3 √3
d. {e x
,e
−x
,e
x/2
cos(
2
x), e
x/2
sin(
2
x), e
−x/2
cos(
2
x), e
−x/2
sin(
2
x)}
e. {cos 2x, sin 2x, e

−√3x
cos x, e
−√3x
sin x, e
√3x
cos x, e
√3x
sin x}
√3 √3 √3 √3
f. {1, e 2x
,e
3x/2
cos(
2
x), e
3x/2
sin(
2
x), e
x/2
cos(
2
x), e
x/2
sin(
2
x)}
√3 √3 √3 √3
g. {e −x
.e
x/2
cos(
2
x), e
x/2
sin(
2
x), e
−x/2
cos(
2
x), e
−x/2
sin(
2
x)}
45. y = c 1x
r1
+ c2 x
r2
+ c3 x
r3
(r1 , r2 , r3 distinct); y = c1 x
r1
+ (c2 + c3 ln x)x
r2
2 r1 r1 λ
(r1 , r2 distinct); y = [ c1 + c2 ln x + c3 (ln x ) ] x ; y = c1 x + x [ c2 cos(ω ln x) + c3 sin(ω ln x)]
Section 9.3 Answers
1. y
p =e
−x
(2 + x − x )
2
−3x
2. y
p =−
e
4
(3 − x + x )
2
3. y
p
x
= e (1 + x − x )
2
4. y
p =e
−2x
(1 − 5x + x )
2
5. y
p =−
xe
2
(1 − x + x
2
−x )
3
6. y
p
2
= x e (1 + x)
x
−2x
7. y
p =
xe
2
(2 + x)
2 x
8. y
p =
x e
2
(2 + x)
2 2x
9. y
p =
x e
2
(1 + 2x)
10. y p =x e
2 3x
(2 + x − x )
2
11. y p =x e
2 4x
(2 + x)
3 x/2
12. y p =
x e
48
(1 + x)
13. y p =e
−x
(1 − 2x + x )
2
14. y p =e
2x
(1 − x)
15. y p =e
−2x
(1 + x + x
2
−x )
3
16. y p =
e
3
(1 − x)
17. y p
x
= e (1 + x )
2
18. y p = x e (1 + x )
x 3
19. y p = x e (2 + x)
x
2x
20. y p =
xe
6
(1 − x )
2
21. y p = 4x e
−x/2
(1 + x)
22. y p =
xe
6
(1 + x )
2
2 2x
23. y p =
x e
6
(1 + x + x )
2
2 2x
24. y p =
x e
6
(3 + x + x )
2
3 x
25. y p =
x e
48
(2 + x)
3 x
26. y p =
x e
6
(1 + x)
3 −x
27. y p =−
x e
6
(1 − x + x )
2
3 2x
28. y p =
x e
12
(2 + x − x )
2
29. y p =e
−x
[(1 + x) cos x + (2 − x) sin x]
30. y p =e
−x
[(1 − x) cos 2x + (1 + x) sin 2x]
31. y p =e
2x 2
[(1 + x − x ) cos x + (1 + 2x) sin x]
32. y p =
e
2
[(1 + x) cos 2x + (1 − x + x ) sin 2x]
2
33. y p =
x
13
(8 cos 2x + 14 sin 2x)
34. y p = xe
x
[(1 + x) cos x + (3 + x) sin x]
2x
35. yp =
xe
2
[(3 − x) cos 2x + sin 2x]
3x
36. yp =−
xe
12
(x cos 3x + sin 3x)
37. yp =−
e
10
(cos x + 7 sin x)
38. yp =
e
12
(cos 2x − sin 2x)
39. yp = xe
2x
cos 2x
−x
40. yp =−
e
2
[(1 + x) cos x + (2 − x) sin x]
−x
41. yp =
xe
10
(cos x + 2 sin x)
42. yp =
xe
4−
(3 cos 2x − sin 2x)
−2x
43. yp =
xe
8
[(1 − x) cos 3x + (1 + x) sin 3x]
44. yp =−
xe
4
(1 + x) sin 2x
2 −x
45. yp =
x e
4
(cos x − 2 sin x)
2 2x
46. yp =−
x e
32
(cos 2x − sin 2x)
2 2x
47. yp =
x e
8
(1 + x) sin x
48. yp = 2x e
2 x
+ xe
2x
− cos x
49. yp =e
2x
+ xe
x
+ 2x cos x
50. yp = 2x + x
2
+ 2x e
x
− 3x e
−x
+ 4e
3x
51. yp
x
= x e (cos 2x − 2 sin 2x) + 2x e
2x
+1
52. yp =x e
2 −2x
(1 + 2x) − cos 2x + sin 2x
53. yp
2
= 2 x (1 + x)e
−x
+ x cos x − 2 sin x
54. yp = 2x e
x
+ xe
2x
+ cos x
55. yp =
xe
6
(cos x + sin 2x)
56. yp =
x
54
[(2 + 2x)e
x
+ 3e
−2x
]
57. yp =
x
8
sinh x sin x
58. yp
3
= x (1 + x)e
−x
+ xe
−2x
59. yp = x e (2 x
x 2
+ cos x + sin x)
60. y = e 2x
(1 + x) + c1 e
−x x
+ e (c2 + c3 x)
61. y = e 3x
(1 − x −
x
2
) + c1 e
x
+e
−x
62. y = x e 2x
(1 + x )
2
+ c1 e
x
+ c2 e
2x
+ c3 e
3x
63. y = x 2
e
−x
(1 − x )
2
+ c1 + e
−x
(c2 + c3 x)
3 x
64. y = x e
24
(4 + x) + e (c1 + c2 x + c3 x )
x 2
2 −x
65. y = x e
16
(1 + 2x − x ) + e (c1 + c2 x) + e
2 x −x
(c3 + c4 x)
66. y = e −2x
[(1 +
x
2
) cos x + (
3
2
− 2x) sin x] + c1 e
x
+ c2 e
−x
+ c3 e
−2x
67. y = −x e x
sin 2x + c1 + c2 e
x x
+ e (c3 cos x + c4 sin x)
2 x
68. y = − x e
16
(1 + x) cos 2x + e
x
[(c1 + c2 x) cos 2x + (c3 + c4 x) sin 2x]
69. y = (x 2
+ 2)e
x
−e
−2x
+e
3x
70. y = e −x 2
(1 + x + x ) + (1 − x)e
x
71. y = ( x
12
+ 16) x e
−x/2
−e
x
72. y = (2 − x)(x 2
+ 1)e
−x
+ cos x − sin x
73. y = (2 − x) cos x − (1 − 7x) sin x + e −2x
74. 2 + e x
[(1 + x) cos x − sin x − 1]
Section 9.4 Answers
1. y p = 2x
3
2. y
2
8 7/2 −x
p = x e
105
3. y p = x ln |x|
2
2( x +2)
4. y p =−
x
−3x
5. y p =−
xe
64
6. y p =−
2x
−x
e (x+1)
7. y p =−
x
8. y p = 2x
2
ln |x|
9. y p =x
2
+1
10. y p =
2 x +6
2
x ln |x|
11. y p =
3
12. y p = −x
2
−2
13. 1
4
x
3
ln |x| −
25
48
x
3
5/2
14. y p =
x
2
x(12−x )
15. y p =
6
4
x ln |x|
16. y p =
6
3 x
17. y p =
x e
18. y p =x
2
ln |x|
19. y p =
xe
20. y p =
3xe
21. y p = −x
3
22. y = −x(ln x ) 2
+ 3x + x
3
− 2x ln x
23. y = x
2
(ln |x| )
2
+x
2
−x
3
+ 2x
3
ln |x|
24. y = − 1
2
(3x + 1)x e
x
− 3e
x
−e
2x
+ 4x e
−x
25. y = 3
2
4
x (ln x )
2
+ 3x − x
4
+ 2x
4
ln x
26. y = − x +12
6
+ 3x − x
2
+ 2e
x
27. y = ( x
3
−
x
2
) ln |x| + 4x − 2 x
2
x
xe (1+3x) x 3x
28. y = − 2
+
x+1
2
−
e
4
+
e
29. y = −8x + 2x 2
− 2x
3
+ 2e
x
−e
−x
30. y = 3x 2
ln x − 7 x
2
2
3(4 x +9) x −x 2x
31. y = 2
+
x
2
−
e
2
+
e
2
+
e
32. y = x ln x + x − √−
x+
1
x
+
1
√x
33. y = x 3
ln |x| + x − 2 x
3
+
1
x
−
1
x2
( x−t) −( x−t) −2( x−t)

x
35. yp =∫
x0
e −3 e
6
+2 e
F (t)dt
2
x (x−t) (2x+t)
36. yp =∫
x0 6xt
3
F (t)dt
( x−t) 2
x xe −x +x(t−1)
37. yp =∫
x0 4
t
F (t)dt
2 ( x−t)
x x −t(t−2)−2te
38. yp =∫
x0 2
F (t)dt
2x(t−1)
2( x−t) ( x−t) −( x−t) −2( x−t)

x
39. yp =∫
x0
e −2 e +2 e
12
−e
F (t)dt
3
x (x−t)
40. yp =∫
x0 6x
F (t)dt
3
x (x+t)(x−t)
41. yp =∫
x0 12x t
2 3
F (t)dt
2( x−t) −2( x−t) 2 2

x e (1+2x)+e (1−2t)−4 x +4 t −2
42. yp =∫
x0 2
F (t)dt
32t
Section 10.1 Answers
′ 1 1
Q =2− Q1 + Q2
1
1. 10
3
25
1
′
Q =6+ Q1 − Q2
2 50 20
′ 5 1
Q = 12 − Q1 + Q2
1 100+2t 100+3t
2. 1 4
′
Q =5+ Q1 − Q2
2 50+t 100+3t
′′ ′ ′
m1 y = −(c1 + c2 )y + c2 y − (k1 + k2 )y1 + k2 y2 + F1
1 1 2
3. m2 y
2
′′
= (c2 − c3 )y
′
1
− (c2 + c3 )y
′
2
+ c3 y
3
′
+ (k2 − k3 )y1 − (k2 + k3 )y2 + k3 y3 + F2
′′ ′ ′ ′
m3 y = c3 y + c3 y − c3 y + k3 y1 + k3 y2 − k3 y3 + F3
3 1 2 3
2 2 2
gR x gR y gR z
4. x ′′
=−
α
m
′
x +
2 2 2 3/2
y
′′
=−
α
m
′
y +
2 2 2 3/2
z
′′
=−
α
m
′
z +
2 2 2 3/2
( x +y +x ) ( x +y +x ) ( x +y +z )
5.
′
x = x2
1
′
x = x3
2
a. x
′
3
= f (t, x1 , y1 , y2 )
′
y = y2
1
′
y = g(t, y1 , y2 )
2
′
u = f (t, u1 , v1 , v2 , w2 )
1
′
v = v2
1
b. v
′
2
= g(t, u1 , v1 , v2 , w1 )
′
w = w2
1
′
w = h(t, u1 , v1 , v2 , w1 , w2 )
2
′
y = y2
1
c. y
′
2
= y3
′
y = f (t, y1 , y2 , y3 )
3
′
y = y2
1
′
y = y3
d. 2
′
y = y4
3
′
y = f (t, y1 )
4
′
x = x2
1
′
x = f (t, x1 , y1 )
e. 2
′
y = y2
1
′
y = g(t, x1 , y1 )
2
2
′ gR x
′
x = x1 x =−
1 3/2
( x2 +y 2 +x2 )
2
gR y
6. y
′
= y1 y
′
1
=−
2 2 2 3/2
( x +y +x )
2
gR z
′ ′
z = z1 z =−
1 2 2 2
3/2
( x +y +z )
1.
2 4
a. ′
y = [ ]y
4 2
−2 −2
b. y ′
= [ ]y
−5 1
−4 −10
c. ′
y = [ ]y
3 7
2 1
d. y ′
= [ ]y
1 2
2.
−1 2 3
⎡ ⎤
a. ′
y = ⎢ 0 1 6 ⎥y
⎣ ⎦
0 0 −2
0 2 2
⎡ ⎤
b. y ′
= ⎢2 0 2⎥y
⎣ ⎦
2 2 0
−1 2 2
⎡ ⎤
c. ′
y = ⎢ 2 −1 2 ⎥y
⎣ ⎦
2 2 −1
3 −1 −1
⎡ ⎤
d. y ′
= ⎢ −2 3 2 ⎥y
⎣ ⎦
4 −1 −2
3.
1 1 1
a. ′
y = [ ] y, y(0) = [ ]
−2 4 0
5 3 9
b. y ′
= [ ] y, y(0) = [ ]
−1 1 −5
4.
6 4 4 3
⎡ ⎤ ⎡ ⎤
a. ′
y = ⎢ −7 −2 −1 ⎥ y, y(0) = ⎢ −6 ⎥
⎣ ⎦ ⎣ ⎦
7 4 3 4
8 7 7 2
⎡ ⎤ ⎡ ⎤
b. y ′
= ⎢ −5 −6 −9 ⎥ y, y(0) = ⎢ −4 ⎥
⎣ ⎦ ⎣ ⎦
5 7 10 3
5.
−3 2 3 − 2t
a. ′
y =[ ]+[ ]
−5 3 6 − 3t
t
3 1 −5e
b. y ′
=[ ]y+[
t
]
−1 1 e
10.
a. d
dt
Y
2
=Y
′
Y +Y Y
′
b. d
dt
Y
n
=Y
′
Y
n−1
+Y Y
′
Y
n−2
+Y
2
Y
′
Y
n−3
+… +Y
n−1
Y
′ n−1
=∑
r=0
Y
r
Y
′
Y
n−r−1
13. B = (P ′
+ P A)P
−1
0 1
2. y ′
=[ P2 (x) P1 (x) ]y
− −
P0 (x) P0 (x)
0 1 … 0
⎡ ⎤
⎢ ⎥
⎢ ⋮ ⋮ ⋱ ⋮ ⎥
3. ′
y =⎢
⎢
⎥y
⎥
⎢ 0 0 … 1 ⎥
⎢ ⎥
Pn (x) Pn−1 (x) P1 (x)
⎣− − … − ⎦
P0 (x) P0 (x) P0 (x)
7.
6t −2t
3e − 6e
b. y = [ 6t −2t
]
3e + 6e
6t −2t 6t −2t
e +e e −e
c. y = 1
2
[
6t −2t 6t −2t
]k
e −e e +e
8.
−4t 3t
6e + 4e
b. y = [ −4t 3t
]
6e − 10 e
−4t 3t −4t 3t
5e + 2e 2e − 2e
c. y = 1
7
[
−4t 3t −4t 3t
]k
5e − 5e 2e + 5e
9.
2t t
−15 e − 4e
b. y = [ 2t t
]
9e + 2e
2t t 2t t
−5 e + 6e −10 e + 10 e
c. y = [ 2t t 2t t
]k
3e − 3e 6e − 5e
10.
3t t
5e − 3e
b. y = [ 3t t
]
5e + 3e
3t t 3t t
e +e e −e
c. y = 1
2
[
3t t 3t t
]k
e −e e +e
11.
2t 3t −t
e − 2e + 3e
⎡ ⎤
b. y = ⎢ 2e
3t
− 9e
−t
⎥
⎣ 2t 3t −t ⎦
e − 2e + 21 e
2t 3t −t 2t 3t 2t 3t −t
4e + 3e −e 6e − 6e 2e − 3e +e
⎡ ⎤
c. y = 1
6
⎢ −3 e
3t
+ 3e
−t
6e
3t
3e
3t
− 3e
−t
⎥k
⎣ 2t 3t −t 2t 3t 2t 3t −t ⎦
4e + 3e − 7e 6e − 6e 2e − 3e + 7e
12.
−2t 4t
−e +e
⎡ ⎤
b. y = 1
3
⎢ −10 e
−2t
+e
4t
⎥
⎣ −2t 4t ⎦
11 e +e
−2t 4t −2t 4t −2t 4t
2e +e −e +e −e +e
⎡ ⎤
c. y = 1
3
⎢ −e
−2t
+e
4t
2e
−2t
+e
4t
−e
−2t
+e
4t
⎥k
⎣ −2t 4t −2t 4t −2t 4t ⎦
−e +e −e +e 2e +e
13.
t −t −2t
3e + 3e −e
⎡ ⎤
b. y = ⎢ t
3e + 2e
−2t
⎥
⎣ −2t ⎦
−e
−t t −t t −t −2t
e e −e 2e − 3e +e
⎡ ⎤
c. y = ⎢ 0 e
t t
2e − 2e
−2t
⎥k
⎣ −2 ⎦
0 0 e
14. Y Z −1
and ZY −1
1 1
1. y = c1 [ ]e
3t
+ c2 [ ]e
−t
1 −1
1 −1
2. y = c1 [ ]e
−t/2
+ c2 [ ]e
−2t
1 1
−3 −1
3. y = c1 [ ]e
−t
+ c2 [ ]e
−2t
1 2
2 −2
4. y = c1 [ ]e
−3t
+ c2 [ ]e
t
1 1
1 −4
5. y = c1 [ ]e
−2t
+ c2 [ ]e
3t
1 1
3 1
6. y = c1 [ ]e
2t
+ c2 [ ]e
t
2 1
−3 −1
7. y = c1 [ ]e
−5t
+ c2 [ ]e
−3t
1 1
1 −1 −1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
8. y = c1 ⎢2⎥e
−3t
+ c2 ⎢ −4 ⎥ e
−t
+ c3 ⎢ −1 ⎥ e
2t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 1
2 −1 −1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
9. y = c1 ⎢1⎥e
−16t
+ c2 ⎢ 2 ⎥e
2t
+ c3 ⎢ 0 ⎥e
2t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 0 1
−2 −1 −7
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
10. y = c 1 ⎢ −4 ⎥ e + c2 ⎢
t
1 ⎥e
−2t
+ c3 ⎢ −5 ⎥ e
2t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
3 0 4
−1 −1 −2
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
11. y = c 1 ⎢ −1 ⎥ e
−2t
+ c2 ⎢ −2 ⎥ e
−3t
+ c3 ⎢ −6 ⎥ e
−5t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 3
11 1 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
12. y = c 1 ⎢ 7 ⎥e
3t
+ c2 ⎢ 2 ⎥ e
−2t
+ c3 ⎢ 1 ⎥ e
−t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 1
4 −1 −1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
13. y = c 1 ⎢ −1 ⎥ e
−4t
+ c2 ⎢ −1 ⎥ e
6t
+ c3 ⎢ 0 ⎥e
4t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 1 1
1 −1 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
14. y = c 1 ⎢1⎥e
−5t
+ c2 ⎢ 0 ⎥e
5t
+ c3 ⎢ 1 ⎥ e
5t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
5 1 0
1 −1 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
15. y = c 1 ⎢ −1 ⎥ + c2 ⎢ 0 ⎥e
6t
+ c3 ⎢ 3 ⎥ e
6t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 3 0
2 4
16. y = − [ ]e
5t
+[ ]e
−5t
6 2
2 −2
17. y = [ ]e
t/2
+[ ]e
t
−4 1
7 2
18. y = [ ]e
9t
−[ ]e
−3t
7 4
3 4
19. y = [ ]e
5t
−[ ]e
−5t
9 2
5 0 −1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
20. y = ⎢ 5 ⎥ e t/2
+⎢ 0 ⎥e
t/2
+⎢ 2 ⎥e
−t/2
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
0 1 0
3 −2
⎡ ⎤ ⎡ ⎤
21. y = ⎢ 3 ⎥ e t
+ ⎢ −2 ⎥ e
−t
⎣ ⎦ ⎣ ⎦
3 2
2 3 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
22. y = ⎢ −2 ⎥ e t
−⎢ 0 ⎥e
−2t
+⎢ 1 ⎥e
3t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 3 0
1 4 1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
23. y = − ⎢ 2 ⎥ e t
+⎢ 2 ⎥e
−t
+⎢ 1 ⎥e
2t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 4 0
−2 0 4
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
24. y = ⎢ −2 ⎥ e 2t
−⎢ 3 ⎥e
−2t
+ ⎢ 12 ⎥ e
4t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 0 4
−1 2 7
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
25. y = ⎢ −1 ⎥ e −6t
+ ⎢ −2 ⎥ e
2t
+ ⎢ −7 ⎥ e
4t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 2 −7
1 6
⎡ ⎤ ⎡ ⎤
26. y = ⎢ 4 ⎥ e −t
+⎢ 6 ⎥e
2t
⎣ ⎦ ⎣ ⎦
4 −2
4 3 −1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
27. y = ⎢ −2 ⎥ + ⎢ −9 ⎥ e 4t
+⎢ 1 ⎥e
2t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 6 −1
29. Half lines of L : y = y and L : y = −y are trajectories other trajectories are asymptotically tangent to
1 2 1 2 2 1 L1 as
t → −∞ and asymptotically tangent to L as t → ∞ . 2
30. Half lines of L : y = −2y and L : y = −y /3 are trajectories other trajectories are asymptotically parallel to L as
1 2 1 2 2 1 1
31. Half lines of L : y = y /3 and L : y = −y are trajectories other trajectories are asymptotically tangent to
1 2 1 2 2 1 L1 as
t → −∞ and asymptotically parallel to L as t → ∞ . 2
32. Half lines of L : y = y /2 and L : y = −y are trajectories other trajectories are asymptotically tangent to
1 2 1 2 2 1 L1 as
33. Half lines of L : y = −y /4 and L : y = −y are trajectories other trajectories are asymptotically tangent to
1 2 1 2 2 1 L1 as
t → −∞ and asymptotically parallel to L as t → ∞ . 2
34. Half lines of L : y = −y and L : y = 3y are trajectories other trajectories are asymptotically parallel to
1 2 1 2 2 1 L1 as
36. Points on L2 : y2 = y1 are trajectories of constant solutions. The trajectories of nonconstant solutions are half-lines on
1
either side of L , parallel to [
1 ] , traversed toward L1.
−1
37. Points on L 1 : y2 = −y1 /3 are trajectories of constant solutions. The trajectories of nonconstant solutions are half-lines on
−1
1 ] , traversed away from L . 1
2
38. Points on L1 : y2 = y1 /3 are trajectories of constant solutions. The trajectories of nonconstant solutions are half-lines on
1 −1
1 ] [ , ] , traversed away from L .
1
−1 1
39. Points on L1 : y2 = y1 /2 are trajectories of constant solutions. The trajectories of nonconstant solutions are half-lines on
1
1 ] L1, .
−1
40. Points on L2 : y2 = −y1 are trajectories of constant solutions. The trajectories of nonconstant solutions are half-lines on
−4
2 ] , traversed toward L . 1
1
41. Points on L1 : y2 = 3 y1 are trajectories of constant solutions. The trajectories of nonconstant solutions are half-lines on
1
1 ] , traversed away from L . 1
−1
2 −1 2
1. y = c1 [ ]e
5t
+ c2 ([ ]e
5t
+[ ] te
5t
)
1 0 1
1 1 1
2. y = c1 [ ]e
−t
+ c2 ([ ]e
−t
+[ ] te
−t
)
1 0 1
−2 −1 −2
3. y = c1 [ ]e
−9t
+ c2 ([ ]e
−9t
+[ ] te
−9t
)
1 0 1
−1 −1 −1
4. y = c1 [ ]e
2t
+ c2 ([ ]e
2t
+[ ] te
2t
)
1 0 1
−1 −1 −2t −2
5. y = c1 [ ] + c2 ([ ]
e
3
+[ ] te
−2t
)
1 0 1
3 −1 −4t 3
6. y = c1 [ ]e
−4t
+ c2 ([ ]
e
2
+[ ] te
−4t
)
2 0 2
4 −1 −t 4
7. y = c1 [ ]e
−t
+ c2 ([ ]
e
3
+[ ] te
−t
)
3 0 3
−1 1 0 1
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
4t
8. y = c1 ⎢ −1 ⎥ + c2 ⎢ 1 ⎥ e
4t
+ c3 ⎜⎢ 1 ⎥
e
2
+ ⎢ 1 ⎥ te
4t
⎟
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
2 2 0 2
−1 1 0 1
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
9. y = c1 ⎢ 1
t
⎥ e + c2 ⎢ −1 ⎥ e
−t
+ c3 ⎜⎢ 3 ⎥ e
−t
+ ⎢ −1 ⎥ te
−t
⎟
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
1 1 0 1
0 1 1 1
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
−2t
10. y = c 1 ⎢1⎥e
2t
+ c2 ⎢ 0 ⎥ e
−2t
+ c3 ⎜⎢ 1 ⎥
e
2
+ ⎢ 0 ⎥ te
−2t
⎟
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
1 1 0 1
−2 0 1 0
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
4t
11. y = c 1 ⎢ −3 ⎥ e
2t
+ c2 ⎢ −1 ⎥ e
4t
+ c3 ⎜⎢ 0 ⎥
e
2
+ ⎢ −1 ⎥ te
4t
⎟
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
1 1 0 1
−1 1 1 1
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
4t
12. y = c 1 ⎢ −1 ⎥ e
−2t
+ c2 ⎢ 1 ⎥ e
4t
+ c3 ⎜⎢ 0 ⎥
e
2
+ ⎢ 1 ⎥ te
4t
⎟
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
1 1 0 1
6 8
13. y = [ ]e
−7t
−[ ] te
−7t
2 4
5 12
14. y = [ ]e
3t
−[ ] te
3t
8 16
2 8
15. y = [ ]e
−5t
−[ ] te
−5t
3 4
3 12
16. y = [ ]e
5t
−[ ] te
5t
1 6
0 6
17. y = [ ]e
−4t
+[ ] te
−4t
2 6
4 2 −1
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
18. y = ⎢ 8
t
⎥ e + ⎢ −3 ⎥ e
−2t
+⎢ 1 ⎥ te
−2t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
−6 −1 0
3 9 2
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
19. y = ⎢ 3 ⎥ e 2t
−⎢ 5 ⎥+⎢ 2 ⎥t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
6 6 0
2 −4 0
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
20. y = − ⎢ 0 ⎥ e −3t
+⎢ 9
t
⎥ e − ⎢ 4 ⎥ te
t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 1 4
−2 0 3
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
21. y = ⎢ 2 ⎥e
4t
+ ⎢ −1 ⎥ e
2t
+ ⎢ −3 ⎥ te
2t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 1 3
1 −3 8
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
22. y = − ⎢ 1 ⎥ e −4t
+⎢ 2 ⎥e
8t
+⎢ 0 ⎥ te
8t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
0 −3 −8
3 3 8
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
23. y = ⎢ 6 ⎥ e 4t
−⎢ 4 ⎥+⎢ 4 ⎥t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
3 1 4
0 −1 0 1 −1 0
⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞ ⎛⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎞
6t 6t 6t 2 6t
24. y = c 1 ⎢1⎥e
6t
+ c2 ⎜⎢ 1 ⎥
e
4
+ ⎢ 1 ⎥ te
6t
⎟ + c3 ⎜⎢ 1 ⎥
e
8
+⎢ 1 ⎥
te
4
+⎢ 1 ⎥
t e
2
⎟
⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠ ⎝⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎠
1 0 1 0 0 1
−1 1 −1 1 1 −1
⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞ ⎛⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎞
3t 3t 3t 2 3t
25. y = c 1 ⎢ 1 ⎥e
3t
+ c2 ⎜⎢ 0 ⎥
e
2
+⎢ 1 ⎥ te
3t
⎟ + c3 ⎜⎢ 2 ⎥
e
36
+⎢ 0 ⎥
te
2
+⎢ 1 ⎥
t e
2
⎟
⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠ ⎝⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎠
1 0 1 0 0 1
0 −1 0 3 −1 0
⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞ ⎛⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎞
−2t 2 −2t
26. y = c 1 ⎢ −1 ⎥ e
−2t
+ c2 ⎜⎢ 1 ⎥e
−2t
+ ⎢ −1 ⎥ te
−2t
⎟ + c3 ⎜⎢ −2 ⎥
e
4
+⎢ 1 ⎥ te
−2t
+ ⎢ −1 ⎥
t e
2
⎟
⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠ ⎝⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎠
1 0 1 0 0 1
0 1 0 −1 1 0
⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞ ⎛⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎞
2t 2t 2t 2 2t
27. y = c 1 ⎢1⎥e
2t
+ c2 ⎜⎢ 1 ⎥
e
2
+ ⎢ 1 ⎥ te
2t
⎟ + c3 ⎜⎢ 1 ⎥
e
8
+⎢ 1 ⎥
te
2
+⎢ 1 ⎥
t e
2
⎟
⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠ ⎝⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎠
1 0 1 0 0 1
−2 6 −2 12 6 −2
⎡ ⎤ ⎛ ⎡ ⎤ ⎡ ⎤ ⎞ ⎛ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎞
−6t −6t −6t 2 −6t
28. y = c 1 ⎢ 1 ⎥e
−6t
+ c2 ⎜− ⎢ 1 ⎥
e
6
+⎢ 1 ⎥ te
−6t
⎟ + c3 ⎜− ⎢ 1 ⎥
e
36
−⎢ 1 ⎥
te
6
+⎢ 1 ⎥
t e
2
⎟
⎣ ⎦ ⎝ ⎣ ⎦ ⎣ ⎦ ⎠ ⎝ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎠
2 0 2 0 0 2
−4 6 1 2
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
29. y = c 1 ⎢ 0 ⎥e
−3t
+ c2 ⎢ 1 ⎥ e
−3t
+ c3 ⎜⎢ 1 ⎥ e
−3t
+ ⎢ 1 ⎥ te
−3t
⎟
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
1 0 0 1
−1 0 1 −1
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
30. y = c 1 ⎢ 0 ⎥e
−3t
+ c2 ⎢ 1 ⎥ e
−3t
+ c3 ⎜⎢ 1 ⎥ e
−3t
+ ⎢ −1 ⎥ te
−3t
⎟
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
1 0 0 1
2 −3 1 −1
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
−t
31. y = c 1 ⎢0⎥e
−t
+ c2 ⎢ 2 ⎥e
−t
+ c3 ⎜⎢ 0 ⎥
e
2
+⎢ 2 ⎥ te
−t
⎟
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
1 0 0 1
−1 0 −1 1
⎡ ⎤ ⎡ ⎤ ⎛⎡ ⎤ ⎡ ⎤ ⎞
32. y = c 1 ⎢ 1 ⎥e
−2t
+ c2 ⎢ 0 ⎥ e
−2t
+ c3 ⎜⎢ 0 ⎥e
−2t
+ ⎢ −1 ⎥ te
−2t
⎟
⎣ ⎦ ⎣ ⎦ ⎝⎣ ⎦ ⎣ ⎦ ⎠
0 1 0 1
3 cos t + sin t 3 sin t − cos t
1. y = c1e
2t
[ ] + c2 e
2t
[ ]
5 cos t 5 sin t
5 cos 2t + sin 2t 5 sin 2t − cos 2t

2. y = c1e
−t
[ ] + c2 e
−t
[ ]
13 cos 2t 13 sin 2t
cos 2t + sin 2t sin 2t − cos 2t

3. y = c1e
3t
[ ] + c2 e
3t
[ ]
2 cos 2t 2 sin 2t
cos 3t − sin 3t sin 3t + cos 3t

4. y = c1e
2t
[ ] + c2 e
2t
[ ]
cos 3t sin 3t
−1 cos 2t − sin 2t sin 2t + cos 2t

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
5. y = c1 ⎢ −1 ⎥ e
−2t
+ c2 e
4t
⎢ cos 2t + sin 2t ⎥ + c3 e
4t
⎢ sin 2t − cos 2t ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
2 2 cos 2t 2 sin 2t
−1 cos 2t − sin 2t sin 2t + cos 2t

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
6. y = c1 ⎢ −1 ⎥ e
−t
+ c2 e
−2t
⎢ − cos 2t − sin 2t ⎥ + c3 e
−2t
⎢ − sin 2t + cos 2t ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 2 cos 2t 2 sin 2t
1 − sin t cos t
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
7. y = c1 ⎢1⎥e
2t
+ c2 e ⎢
t
sin t
t
⎥ + c3 e ⎢ − cos t ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 cos t sin t
−1 − sin 2t − cos 2t cos 2t − sin 2t

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
8. y = c1 ⎢ 1
t
⎥ e + c2 e
−t
⎢ 2 cos 2t ⎥ + c3 e
−t
⎢ 2 sin 2t ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 2 cos 2t 2 sin 2t
cos 6t − 3 sin 6t sin 6t + 3 cos 6t

9. y = c1e
3t
[ ] + c2 e
3t
[ ]
5 cos 6t 5 sin 6t
cos t − 3 sin t sin t + 3 cos t

10. y = c 1e
2t
[ ] + c2 e
2t
[ ]
2 cos t 2 sin t
3 sin 3t − cos 3t −3 cos 3t − sin 3t

11. y = c 1e
2t
[ ] + c2 e
2t
[ ]
5 cos 3t 5 sin 3t
sin 4t − 8 cos 4t − cos 4t − 8 sin 4t

12. y = c 1e
2t
[ ] + c2 e
2t
[ ]
5 cos 4t 5 sin 4t
−1 sin t − cos t
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
13. y = c 1 ⎢ 1 ⎥e
−2t t
+ c2 e ⎢ − cos t ⎥ + c3 e ⎢ − sin t ⎥
t
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 cos t sin t
2 − cos 3t − sin 3t − sin 3t + cos 3t

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
14. y = c 1 ⎢2⎥e
−2t
+ c2 e
2t
⎢ − sin 3t ⎥ + c3 e
2t
⎢ cos 3t ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 cos 3t sin 3t
1 − sin 3t cos 3t
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
15. y = c 1 ⎢2⎥e
3t
+ c2 e
6t
⎢ sin 3t ⎥ + c3 e
6t
⎢ − cos 3t ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 cos 3t sin 3t
1 2 cos t − 2 sin t 2 sin t + 2 cos t

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
16. y = c 1 ⎢ 1 ⎥ e + c2 e ⎢
t t
cos t − sin t ⎥ + c3 e ⎢
t
cos t + sin t ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
1 2 cos t 2 sin t
5 cos 3t + sin 3t
17. y = e t
[ ]
2 cos 3t + 3 sin 3t
5 cos 6t + 5 sin 6t
18. y = e 4t
[ ]
cos 6t − 3 sin 6t
17 cos 3t − sin 3t
19. y = e t
[ ]
7 cos 3t + 3 sin 3t
cos(t/2) + sin(t/2)
20. y = e t/2
[ ]
− cos(t/2) + 2 sin(t/2)
1 3 cos t + sin t
⎡ ⎤ ⎡ ⎤
21. y = ⎢ −1 ⎥ e t
+e
4t
⎢ cos t − 3 sin t ⎥
⎣ ⎦ ⎣ ⎦
2 4 cos t − 2 sin t
4 4 cos 2t + 8 sin 2t
⎡ ⎤ ⎡ ⎤
22. y = ⎢ 4 ⎥ e 8t
+e
2t
⎢ −6 sin 2t + 2 cos 2t ⎥
⎣ ⎦ ⎣ ⎦
2 3 cos 2t + sin 3t
0 15 cos 6t + 10 sin 6t
⎡ ⎤ ⎡ ⎤
23. y = ⎢ 3 ⎥ e −4t
+e
4t
⎢ 14 cos 6t − 8 sin 6t ⎥
⎣ ⎦ ⎣ ⎦
3 7 cos 6t − 4 sin 6t
6 10 cos 4t − 4 sin 4t
⎡ ⎤ ⎡ ⎤
24. y = ⎢ −3 ⎥ e 8t
+ ⎢ 17 cos 4t − sin 4t ⎥
⎣ ⎦ ⎣ ⎦
3 3 cos 4t − 7 sin 4t
−1 1
29. U = 1
[ ], V =
1
[ ]
√2 √2
1 1
.5257 −.8507
30. U ≈ [ ], V ≈[ ]
.8507 .5257
.8507 −.5257
31. U ≈ [ ], V ≈[ ]
.5257 .8507
−.9732 .2298
32. U ≈ [ ], V ≈[ ]
.2298 .9732
.5257 −.8507
33. U ≈ [ ], V ≈[ ]
.8507 .5257
−.5257 .8507
34. U ≈ [ ], V ≈[ ]
.8507 .5257
−.8817 .4719
35. U ≈ [ ], V ≈[ ]
.4719 .8817
.8817 −.4719
36. U ≈ [ ], V ≈[ ]
.4719 .8817
0 −1
37. U = [ ], V =[ ]
1 0
0 1
38. U = [ ], V =[ ]
1 0
1 −1
39. U = 1
[ ], V =
1
[ ]
√2 √2
1 1
.5257 −.8507
40. U ≈ [ ], V ≈[ ]
.8507 .5257
Index
A F method of undetermined coefficients
Abel's theorem forcing function 5.4: The Method of Undetermined Coefficients I
5.1: Homogeneous Linear Equations 5.5: The Method of Undetermined Coefficients II
9.1: Introduction to Linear Higher Order Equations 9.3: Undetermined Coefficients for Higher Order
10.3: Basic Theory of Homogeneous Linear Equations
Systems midpoint method
augmented matrix G 3.2: The Improved Euler Method and Related
11.3: Gaussian Elimination Gaussian Elimination Methods
autonomous differential equation 11.2: Elementary Operations
minor of a matrix
gravity
Autonomous Second Order Equations minors or a matrix
H multiplicity
B Heun’s method 9.2: Higher Order Constant Coefficient
Bernoulli Equations Homogeneous Equations
3.2: The Improved Euler Method and Related
2.4: Transformation of Nonlinear Equations into Methods
Separable Equations
homogeneous N
C Homogeneous Differential Equations 4.2: Cooling and Mixing
characteristic polynomial Nonhomgeneous Linear Equations
5.2: Constant Coefficient Homogeneous Equations with constant Coefficients 5.3: Nonhomogeneous Linear Equations
9.2: Higher Order Constant Coefficient 5.2: Constant Coefficient Homogeneous Equations
Chebyshev’s equation O
I open interval of convergence
7.3: Series Solutions Near an Ordinary Point I identity matrix 7.2: Review of Power Series
coefficient matrix 12.6: The Identity and Inverses
open rectangle
10.2: Linear Systems of Differential Equations implicit function Theorem 2.3: Existence and Uniqueness of Solutions of
cofactor 2.2: Separable Equations Nonlinear Equations
13.1: Basic Techniques improved Euler method ordinary point
complementary equation 3.2: The Improved Euler Method and Related 7.3: Series Solutions Near an Ordinary Point I
2.1: Linear First Order Equations Methods 7.4: Series Solutions Near an Ordinary Point II
5.3: Nonhomogeneous Linear Equations initial value problems Overdamped Motion
compound interest 8.3: Solution of Initial Value Problems 6.2: Spring Problems II
4.1: Growth and Decay integrating factor
constant coefficient equation 2.6: Integrating Factors P
9.2: Higher Order Constant Coefficient Inverse Laplace Transform particular solution
8.2: The Inverse Laplace Transform 5.3: Nonhomogeneous Linear Equations
constant matrix phase plane equivalent
K 4.4: Autonomous Second Order Equations
Critically Damped Motion Kronecker delta Poincaré phase plane
6.2: Spring Problems II 12.6: The Identity and Inverses 4.4: Autonomous Second Order Equations
polynomial operator
D L 9.2: Higher Order Constant Coefficient
damping Laplace Expansion Homogeneous Equations
4.4: Autonomous Second Order Equations 13.1: Basic Techniques power series
defective Laplace Transform 7.2: Review of Power Series
10.5: Constant Coefficient Homogeneous Systems 8: Laplace Transforms
II
determinant
8.1: Introduction to the Laplace Transform R
radius of convergence
13.1: Basic Techniques Linear Independence 7.2: Review of Power Series
Determinants 14.3: Linear Independence
reduction of orders
direction field M Riccati equation
1.3: Direction Fields for First Order Equations Malthusian model 2.4E: Transformation of Nonlinear Equations into
1.1: Applications Leading to Differential Separable Equations (Exercises)
E Equations
roundoff errors
Elementary operations matrix 3.1: Euler's Method
11.2: Elementary Operations 12.1: Matrix Arithmetic
Row Operations
escape velocity matrix inverse 13.2: Properties of Determinants
4.3: Elementary Mechanics 12.7: Finding the Inverse of a Matrix
Euler’s Method Matrix Multiplication S
3.1: Euler's Method 12.4: Properties of Matrix Multiplication
Separable Differential Equations
Exact Differentials Mechanics 2.2: Separable Equations
2.5: Exact Equations 4.3: Elementary Mechanics
separatrix
Exact Equations 4.4: Autonomous Second Order Equations
2.5: Exact Equations Simple Harmonic Motion
skew lines T V
11.1: Geometry temperature decay constant van der Pol equation
skew symmetric matrix 4.2: Cooling and Mixing 4.4E: Autonomous Second Order Equations
12.5: The Transpose terminal velocity (Exercises)
span 4.3: Elementary Mechanics Variation of Parameters
14.2: Spanning Sets transpose 5.7: Variation of Parameters
9.4: Variation of Parameters for Higher Order
Spanning Sets 12.5: The Transpose
Equations
14.2: Spanning Sets Triangular Matrix Verhulst model
square matrix 13.1: Basic Techniques
1.1: Applications Leading to Differential
12.1: Matrix Arithmetic truncation errors Equations
subsets 3.1: Euler's Method
14.2: Spanning Sets W
superposition principle U Wronskian
5.3: Nonhomogeneous Linear Equations Underdamped Motion 9.1: Introduction to Linear Higher Order Equations
system of linear equations 6.2: Spring Problems II 10.3: Basic Theory of Homogeneous Linear
11.2: Elementary Operations Systems

Euler

Uploaded by

Copyright:

Available Formats

Euler

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Euler

Uploaded by

Copyright:

Available Formats

ELEMENTARY

This text was compiled on 09/16/2020

1.2E: BASIC CONCEPTS (EXERCISES)

1.3E: DIRECTION FIELDS FOR FIRST ORDER EQUATIONS (EXERCISES)

2: FIRST ORDER EQUATIONS

2.1: LINEAR FIRST ORDER EQUATIONS

2.3E: EXISTENCE AND UNIQUENESS OF SOLUTIONS OF NONLINEAR EQUATIONS (EXERCISES)

2.4E: TRANSFORMATION OF NONLINEAR EQUATIONS INTO SEPARABLE EQUATIONS (EXERCISES)

2.5E: EXACT EQUATIONS (EXERCISES)

3.1: EULER'S METHOD

4: APPLICATIONS OF FIRST ORDER EQUATIONS

4.1: GROWTH AND DECAY

4.2E: COOLING AND MIXING (EXERCISES)

4.4E: AUTONOMOUS SECOND ORDER EQUATIONS (EXERCISES)

5: LINEAR SECOND ORDER EQUATIONS

5.1: HOMOGENEOUS LINEAR EQUATIONS

5.2E: CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS (EXERCISES)

5.7E: VARIATION OF PARAMETERS (EXERCISES)

6: APPLICATIONS OF LINEAR SECOND ORDER EQUATIONS

6.1: SPRING PROBLEMS I

7: SERIES SOLUTIONS OF LINEAR SECOND ORDER EQUATIONS

7.1: PRELUDE TO SERIES SOLUTIONS OF LINEAR SECOND ORDER EQUATIONS

8.3E: SOLUTION OF INITIAL VALUE PROBLEMS (EXERCISES)

9: LINEAR HIGHER ORDER DIFFERENTIAL EQUATIONS

9.3E: UNDETERMINED COEFFICIENTS FOR HIGHER ORDER EQUATIONS (EXERCISES)

10: LINEAR SYSTEMS OF DIFFERENTIAL EQUATIONS

10.4E: CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS I (EXERCISES)

11: SYSTEMS OF EQUATIONS

14: VECTOR SPACES

APPENDICES AND ANSWERS TO EXERCISES

1.1: APPLICATIONS LEADING TO DIFFERENTIAL EQUATIONS

1.2: BASIC CONCEPTS

1.2E: BASIC CONCEPTS (EXERCISES)

1.3E: DIRECTION FIELDS FOR FIRST ORDER EQUATIONS (EXERCISES)

y with respect to t ; thus,

Population Growth and Decay

c = P (0) = P , so the applicable solution is

William F. Trench 9/11/2020 1.1.1 CC-BY-NC-SA https://math.libretexts.org/@go/page/44215

This implies that

where P 0 = P (0) > 0 . Therefore lim P (t) = 1/α , independent of P . 0

Figure 1.1.1 shows typical graphs of P versus t for various values of P . 0

Figure 1.1.1 : Solutions of the logistic equation

Newton’s Law of Cooling

T = T (t) is the temperature of the body at time t , then

William F. Trench 9/11/2020 1.1.2 CC-BY-NC-SA https://math.libretexts.org/@go/page/44215

where T is the temperature of the body when t = 0 . Therefore

independent of T (Common sense suggests this. Why?).

Figure 1.1.2 shows typical graphs of T versus t for various values of T . 0

Figure 1.1.2 : Temperature according to Newton’s Law of Cooling

Glucose Absorption by the Body

William F. Trench 9/11/2020 1.1.3 CC-BY-NC-SA https://math.libretexts.org/@go/page/44215

From calculus you know that if c is any constant then

Graphs of this function are similar to those in Figure 1.1.2. (Why?)

that all the susceptible people eventually become infected.

Newton’s Second Law of Motion

the direction opposite to the velocity.

which is usually rewritten as

William F. Trench 9/11/2020 1.1.4 CC-BY-NC-SA https://math.libretexts.org/@go/page/44215

Interacting Species: Competition

Figure 1.1.3 : Populations of competing species