Engineering Mathematics-I

Lecture 38 : Variation of Parameters, Green’s Function

Panchatcharam Mariappan1

1 Assistant Professor

Department of Mathematics and Statistics

IIT Tirupati, Tirupati
Variation of
Parameters and
Green’s Function

1
Second-Order Linear ODE
• A second-order ODE is called linear if it can be written in the following
form
y ′′ + p(x)y ′ + q(x)y = r(x) (1)
• Nonlinear if it cannot be written in this form.
• If the coefficient of y ′′ is not unity, then one can divide the equation by the
coefficient and bring that to the standard form.
• If r(x) ≡ 0, then (1) is called homogeneous. That is,

y ′′ + p(x)y ′ + q(x)y = 0 (2)

• If r(x) ̸= 0 for some x, then (1) is called nonhomogeneous. That is,

y ′′ + p(x)y ′ + q(x)y = r(x) (3)

2
Variation of Parameters
Suppose solution of the second-order homogeneous linear ODE (3) is given by

yh = c1 y1 (x) + c2 y2 (x)

where y1 , y2 are two independent solutions and c1 , c2 are constants.

• Change c1 and c2 as parameters and obtain the solution to the
nonhomogeneous linear ODE.
• Since we have variable parameters, this method is called variation of
parameters.
• The particular solution of (3) can be written as

yp = c1 (x)y1 (x) + c2 (x)y2 (x)

where c1 and c2 are functions of x.
3
Variation of Parameters

yp = c1 y1 + c2 y2
yp′ = c′1 y1 + c′2 y2 + c1 y1′ + c2 y2′
yp′′ = c′′1 y1 + c′′2 y2 + c′1 y1′ + c′2 y2′ + c′1 y1′ + c2 y2′ + c1 y1′′ + c2 y2′′

0 0
z }| { z }| {
yp′′ + p(x)yp′ + q(x)yp = c1 [y1′′ + p(x)y1′ + q(x)y1 ] +c2 [y2′′ + p(x)y2′ + q(x)y2 ]
+ c′′1 y1 + c′1 y1′ + c2 y2′′ + c′2 y2′ + p(x)[c′1 y1 + c′2 y2 ] + c′1 y1′ + c′2 y2′
d ′
= [c y1 + c′2 y2 ] + p(x)[c′1 y1 + c′2 y2 ] + c′1 y1′ + c′2 y2′
dx 1

4
Variation of Parameters
If we assume c′1 y1 + c′2 y2 = 0, then first two terms of the above equation
vanishes and we left with c′1 y1′ + c′2 y2′ = r(x).
Therefore, we have

c′1 y1 + c′2 y2 = 0
c′1 y1′ + c′2 y2′ = r(x)

Then the solution of c1 and c2 are obtained using Cramer’s rule.

−y2 r(x) y1 r(x)

c′1 = , c′2 = (4)
W W

where W denotes the Wronskian of y1 and y2 . Here W ̸= 0 (Why?). We obtain

the particular solution by solving (4). 5
Variation of Parameters
Example 1
Solve x2 y ′′ + xy ′ − 9y = 48x5 .
This equation is similar to Euler-Cauchy equation, here

a = 1, b = −9, r(x) = 48x3 =⇒ m2 − 9 = 0 =⇒ m = 3, −3

=⇒ yh = c1 x3 + c2 x−3
=⇒ y1 = x3 , y2 = x−3 =⇒ y1′ = 3x2 , y2′ = −3x−4
3 3 6
=⇒ W = − − = −
x x x
′ −y2 r(x)
c1 = = 8x =⇒ c1 = 4x2
W
y1 r(x)
c′2 = = −8x7 =⇒ c2 = −x8
W
=⇒ yp = 4x5 − x5 = 3x5 =⇒ y = c1 x3 + c2 x−3 + 3x5
6
Variation of Parameters

Exercise 1: Variation of Parameters

Solve the following ODEs using variation of parameters
1. y ′′ + 9y = sec 3x
2. x2 y ′′ − 2xy ′ + 2y = x3 sin x
3. y ′′ + y = cos x − sin x
4. y ′′ − 4y ′ + 5y = e2x csc x
5. x2 y ′′ − 4xy ′ − 6y = 21x−4

7
Green’s Function
From method of variation of parameters, we observed that the particular
solution of (3) is given by

yp = c1 (x)y1 (x) + c2 (x)y2 (x)

 x x
−y2 (t)r(t) y1 (t)r(t)
yp (x) = −y1 (x) dt + y2 (x) dt
x0 W (t) W (t)
x0
x x
−y2 (t)y1 (x) y1 (t)y2 (x)
yp (x) = r(t)dt + r(t)dt
W (t) W (t)
x0 x0

8
Green’s Function

x
yp (x) = G(x, t)r(t)dt (5)
x0

where the function G(x, t) is called the Green’s function.

The Green’s function for an initial value problem is given by

y1 (t)y2 (x) − y2 (t)y1 (x)

G(x, t) = (6)
W (t)

9
Green’s Function
The following Green’s function works for the boundary value problem.

b
yp (x) = G(x, t)r(t)dt (7)
a


y1 (t)y2 (x)
a≤t≤x


W (t)

G(x, t) = (8)
y 2 (t)y1 (x)
x≤t≤b


W (t)


10
Green’s Function

Example 2
Solve the IVP y ′′ + 4y = x, y(0) = 2, y ′ (0) = 1 using Green’s function

m2 + 4 = 0 =⇒ m = 2i, −2i
=⇒ y1 = cos 2x, y2 = sin 2x
=⇒ y2′ = 2 cos 2x, y1′ = −2 cos 2x
=⇒ W = 2(cos2 2x + sin2 2x) = 2
cos 2t sin 2x − cos 2x sin 2t 1
G(x, t) = = sin 2(x − t)
2 2
x
 
1
=⇒ yp =  t sin 2(x − t)dt
2
0

11
Green’s Function

x sin 2x
=⇒ yp = −
4 8
x sin 2x
=⇒ y = A cos 2x + B sin 2x + −
4 8
y(0) = 0 =⇒ A = 2
1 1 1
y ′ (0) = 0 =⇒ 2B + − = 1 =⇒ B =
4 4 2
3 x
y(x) = 2 cos 2x + sin 2x +
8 4

12
Green’s Function

Example 3
Solve the BVP y ′′ + 4y = 3, y(0) = 1, y(π/4) = 0 using Green’s function

m2 + 4 = 0 =⇒ m = 2i, −2i
=⇒ y1 = cos 2x, y2 = sin 2x
=⇒ y1′ = −2 sin 2x, y1′ = 2 cos 2x
=⇒ W = 2(cos2 2x + sin2 2x) = 2

 cos 2t sin 2x

0≤t≤x
G(x, t) = cos 2x2sin 2t

 x ≤ t ≤ π/4
2

13
Green’s Function

π/4
=⇒ yp = G(x, t)r(t)dt
0
x π/4
 
3
=⇒ yp = sin 2x cos 2tdt + cos 2x sin 2tdt

2
0 x
3 3 cos 2x 3 3 cos 2x
=⇒ yp = − =⇒ y = A cos 2x + B sin 2x + −
4 4 4 4
π  3
y(0) = 0 =⇒ A = 1, y = 0 =⇒ B = −
4 4
1
=⇒ y(x) = (cos 2x − 3 sin 2x + 3)
4
14
Green’s Function

Example 4
Solve the IVP y ′′ − y = e2x , y(0) = 1, y ′ (0) = 0 using Green’s function

m2 − 1 = 0 =⇒ y1 = ex , y2 = e−x =⇒ W = −2
1 1
G(x, t) = − (et e−x − e−t ex ) = (ex−t − e−(x−t) )
2 2
x x

1
=⇒ yp =  ex−t e2t dt − e−(x−t) e2t dt
2
0 0
x x
 
1 x
=⇒ yp = e et dt − e−x e3t dt
2
0 0

15
Green’s Function
 
1 2 2x x −x
=⇒ yp = e −e +e
2 3
 
x −x 1 2 2x x −x
=⇒ y = Ae + Be + e −e +e
2 3
   
1 x 1 −x 1 2x
=⇒ y = A − e + B+ e + e
2 2 3
Using IVP, we obtain
1
A = 0, B = −
3
1 1 1
=⇒ y = − ex + e−x + e2x
2 6 3
16
 Thanks
Doubts and Suggestions
panch.m@iittp.ac.in

17
 Engineering Mathematics-I
Lecture 38 : Variation of Parameters, Green’s Function

Panchatcharam Mariappan1

1 Assistant Professor

Department of Mathematics and Statistics

IIT Tirupati, Tirupati

17