3 Topic: Applications of Partial Differential Equations

3rd Topic
Applications of
Partial Differential Equations

Heat Flow Equation
One-dimensional Heat Flow,
Formulation and Solution of Heat Equation
Prepared by:
Prof. Sunil
Department of Mathematics & Scientific Computing
NIT Hamirpur (HP)
One-dimensional wave equation:
 2y  2y
 c2
t 2 x 2
One-dimensional heat flow equation:

The one-dimensional heat flow equation, also known as the heat equation,
describes how the temperature distribution in a one-dimensional object change
over time due to heat conduction.
u  2u
 c2
t x 2
where:
u (x, t) is the temperature distribution at position a and time t,
c is the thermal diffusivity (a material property),
u
represents the rate of change of temperature with respect to time and
t
Applications of Partial Differential Equations: One-Dimensional Heat flow equation 2
Prepared by: Dr. Sunil, NIT Hamirpur (HP)
 2u
represents the second derivative of temperature with respect to position.
x 2
This equation essentially states that the rate of change of temperature at any
point is proportional to the curvature of the temperature distribution.
The heat equation can be solved with appropriate initial and boundary conditions,
such as specifying the initial temperature distribution u (x, 0) and boundary
conditions at the ends of the object. Various numerical and analytical techniques
can be employed to solve this equation for specific cases.
FORMULATION of One-dimensional heat flow equation:

Consider the flow of heat by conduction (transfer) in a uniform bar/rod of length
ℓ , placed along the x-axis.
Assumptions:
1. Rod is sufficiently thin.
2. Rod has constant density  .
3. Rod has uniform cross section A.
4. The sides of the rod are insulated and the loss of heat from the sides by
the conduction or radiation is negligible.
Take one end of the bar as origin and the direction of flow as the positive
x-axis. The temperature u(x, t) at any point of the bar depends on the distance x
of the point from one end and the time t. Since the rod is sufficiently thin, so that
the temperature at all points of any cross-sectional area of the rod is the same.
y-axis
Q1 Q2
A x-axis
x x+δx ℓ
Figure
The amount of heat crossing any section of the bar per second depends on the
area A of the cross-section, the conductivity K of the material of the bar and the
u
temperature gradient i.e., rate of change of temperature w.r.t. distance
x
normal to the area.
We have following results:
1. The amount of heat Q(t) in a small segment of rod between the cross
x  x
sections at x and x  x is Q( t )   SA us, t ds ,
x
where S is specific heat of the rod, and  , the density of the material.
2. Q1 , the quantity of heat flowing into the section at a distance
 u 
x   KA  per sec.
 x  x
(A negative sign appears because heat flows in the direction of decreasing
temperature i.e. as x increases, u decreases).
3. Q 2 , the quantity of heat flowing out of the section at a distance
 u 
x  x   KA  per sec.
 x  x  x
Hence, the amount of heat retained (preserve) by the slab with thickness x
i.e. net heat flow into this segment of the rod is
 u   u  
Q1  Q 2  KA       per sec. (i)
 x  x  x  x  x 
Now the rate of change of heat content in the segment of the rod between x
and x  x must be equal to net heat flow into this segment of the rod. Thus
Q  u   u  
 KA     .
t  x  x  x  x  x 
Q u
But the rate of increase of heat in the slab i.e.  SAx . (ii)
t t
Calculations:
 Q( t )  x  x x  x
u s, t  u , t  
        
t  
  S A u s , t ds  S  A ds  S A dx 
 t x x
t t 
where  lies between x and x  x .

Thus, from (i) and (ii), we obtain
  u   u  
     
u  u   u   u  x  x  x  x  x 
SA x  KA        S  K .
t  x  x  x  x  x  t  x 
 
 
Taking the limit as x  0 , we have
u  2u u K  2 u
S K  
t x 2 t S x 2
u  2u
  c2 , (iii)
t x 2
K
where c 2  , known as diffusivity of the material of the bar (block).
S
Equation (iii) is one-dimensional heat equation, which is second order,
homogeneous and periodic type.
Boundary conditions:
u  2u
The boundary conditions, satisfied by the equation  c2 are:
t x 2
(i). u  0 when x  0  u 0, t   0
 . This means that .
(ii). u  0 when x  ℓ  u ℓ, t   0 
i.e. the end points are assumed to be zero temperature.

These should be satisfied for every value of t.
Initial condition:
The initial temperature distribution in the bar is given by:
u  f x  , when t = 0. i.e. u x ,0  f x  ,
where f(x) is a given (prescribed) function of x.
SOLUTION of one-dimensional heat flow equation by separation of

variables:
2
u 2  u
The heat flow equation is c . (i)
t x 2
Assume that the solution u (x, t) is separable i.e. u = X(x)T(t). (ii)
u  2u
Then  XT  and  X T .
t x 2
X  1 T 
Substituting in (i), we have XT   c 2 X T   . (iii)
X c2 T
Now the LHS of (iii) is a function of x only and the RHS is a function of t only.
Since x and t are independent variables, this equation can hold only when both
X  1 T 
sides reduce to a constant, say k. i.e.   k.
X c2 T
Then, the above equation leads to the two ordinary differential equations
d 2X dT
 kX  0 and  kc 2 T  0 . (iv)
dx 2 dt
X 0   0
Using (ii), the boundary conditions reduce to .
X ℓ   0 
Calculations:
u 0, t   0 u 0, t   X 0 Tt   0 X 0   0
Boundary conditions  reduces to  
u ℓ, t   0  u ℓ, t   X ℓ T t   0  X ℓ   0 
Solving equations (iv), we get
Case (i): When k is positive and  p 2 (say)
d 2X dT
Then  p 2 X  0 and  p 2c 2T  0 .
dx 2 dt
2 2
Solution is X  c1e px  c 2 e  px and T  c 3 e c p t .
X 0   0
Using boundary conditions  , we get c1  0 and c 2  0 .
X ℓ   0 
Thus Xx   0 for all x or u = 0 for all x, which is a trivial solution.

Case (ii): When k = 0
d2X dT
Then  0 and  0.
2 dt
dx
Solution is X  c1x  c 2 and T  c 3 .
X 0   0
Using boundary conditions  , we get c1  0 and c 2  0 .
X ℓ   0 
Thus Xx   0 for all x or u = 0 for all x, which is a trivial solution.
Case (iii): When k is negative and   p 2 (say)
d 2X dT
Then  p 2 X  0 and  p 2c 2T  0 .
2 dt
dx
2 2
Solution is X  c1 cos px  c 2 sin px and T  c 3 e  c p t .
Find the values of c1 and c 2 using boundary conditions:
X 0   0
Using boundary conditions  , we get c1  0 and 0  c 2 sin pℓ .
X ℓ   0 
We must take c 2  0 , since otherwise Xx   0 .

n
Hence, sin pℓ  0  pℓ  n  p  , n is integer.
ℓ
Assuming c 2  1 , we get infinitely many solutions
nx
X n x   sin , n  1, 2, 3, 4,..........
ℓ
Find the value of c 3 using boundary conditions:
dT dT
Since solution of  kc 2 T  0 i.e.  p 2 c 2 T  0 is
dt dt
c2n 22 t

T  c 3e  c p t  Tn t   b n e
2 2
ℓ2 .
Hence, the ‘eigen function’

n 22t
nx 
u n x , t   X n ( x )Tn ( t )  b n sin
2
e ℓ , n  1, 2, 3, .....
ℓ
u  2u u 0, t   0
satisfy equation  c2 and boundary conditions .
t x 2 u ℓ, t   0 
In order to satisfy the initial condition u x ,0  f x  , we bring into play the linearity
principle, which states that
c2n 22 t
 
nx 
u x , t    u n x , t    b n sin e ℓ2 ,
n 1 n 1
ℓ
u  2u u 0, t   0
satisfy equation  c2 and boundary conditions .
t x 2 u ℓ, t   0 
To determine the unknown constants b n , use initial condition u x ,0  f x  in
c2n 22t
 
nx 
u x , t    u n x , t    b n sin e ℓ2 , we get
n 1 n 1
ℓ

nx
f ( x )  u  x ,0    b n sin .
n 1
ℓ
Since RHS is a half range Fourier sine series expansion of f(x) in the interval
0, ℓ  .
The Fourier coefficients b n ’s are given by
ℓ
2 nx
b n   f ( x ) sin dx , n  1, 2, 3, .......
ℓ ℓ
0
Thus, the temperature distribution in the rod is given by the solution

c2n 22t
 
nx 
u x , t    u n x , t    b n sin e ℓ2
n 1 n 1
ℓ
ℓ
2 nx
with b n ’s calculated from b n 
ℓ  f ( x ) sin
ℓ
dx , n  1, 2, 3, .......
0
Remarks:
2
u 2  u
The various solutions of the heat equation c are
t x 2
   2 2
u  c1e px  c 2 e  px .c 3 e c p t .
 u  c1x  c 2 .c 3 .
2 2
 u  c1 cos px  c 2 sin px .c 3 e  c p t .
Of these solutions, we have to choose that solution which is consistent with the
physical nature of the problem.
Since u decreases as time t increases, the only suitable solution of the heat
equation is
u  C1 cos px  C 2 sin px .e  c p t .

2 2
Steady state condition:

A condition (phenomena) is said to be steady state if the dependent
variables are independent (free) of time t.
Non-homogeneous boundary conditions:
Non-homogeneous boundary conditions are boundary conditions, which
are all not zero.
Rod with both ends insulated:
When both ends of the rod are insulated, i.e. no heat flows through them
and then the corresponding boundary conditions are
u u
 u x 0, t   0,  u x ℓ, t   0.
x x  0 x x  ℓ
FINAL CONCLUSIONS
u  2u
The solution of one-dimensional heat flow equation  c2 (i)
t x 2
u 0, t   0
satisfying the boundary conditions  and the initial condition u x ,0  f x 
u ℓ, t   0 
is given by
u x , t   c1 cos px  c 2 sin px .e  c p t ,

2 2
where u(x, t) is called temperature function.

The most general solution of (i) satisfying the boundary condition is
c2n 22t

nx
u (x, t )   b n sin
ℓ
.e ℓ2 ,
n 1
ℓ
2 nx
with b n ’s calculated from b n 
ℓ  f ( x ) sin
ℓ
dx , n  1, 2, 3, .......
0
*****************************************
Now let us solve some problems related to one-dimensional heat flow equation:
Q.No.1.: A rod of length ℓ with insulated sides is initially at a uniform temperature u0.
Its ends are suddenly cooled to 0oC and are kept at that temperature. Find the
temperature function u(x, t).
Sol.: This is one-dimensional heat flow problem. This is also known as initial boundary
value problem (IBVP).
2
u 2 u
First we consider one-dimensional heat flow equation: c .
t x 2
To find: Temperature function u(x, t).
We know that the suitable solution of the one-dimensional heat flow equation is
u ( x , t )  c1 cos px  c 2 sin px .e  c p t .

2 2
(i)
Boundary conditions:
Since, here the ends x = 0 and x  ℓ are cooled at 0oC and kept at that temperature
throughout.
i.e. u (0, t )  u (ℓ, t )  0 for all t.
Initial condition: Here u x ,0   u 0 .
To find: c1 and c 2 using boundary conditions

2 2
Since u(0, t) = 0, from (i), we have 0  c1e  c p t  c1  0 .
2 2
Thus (i) becomes u ( x , t )  c 2 sin px.e  c p t . (ii)
Since u ℓ, t   0 , from (ii), we have 0  c 2 sin pℓ.e  c p t

2 2
n
 sin pℓ  0  pℓ  n  p  , n being an integer.
ℓ
c2n 22 t
nx
Solution (ii) reduces to u x , t   b n sin .e ℓ2 . (on replacing c2 by bn)
ℓ
The most general solution is obtained by adding all such solutions for n = 1, 2, 3,….
c2n 22 t

nx
 u ( x , t )   b n sin .e ℓ2 . (iii)
n 1
ℓ

nx
Since u(x, 0) = u0, then from (iii), we have u 0   b n sin
ℓ
,
n 1
which is half range sine series for u0.

0, when n is even
2 ℓ nx 
b n   u 0 sin dx   4u 0
ℓ 0 ℓ  n , when n is odd

Hence, the temperature function is given by
c2n 22 t

4u 0 1 nx
u (x, t )  
 n 1,3,5,.... n
sin
ℓ
.e ℓ2
 c 2 2n 12  2 t
4u 0 
1 2n  1x .e
 
 n 1 2n  1
sin
ℓ
ℓ2 . Ans.
Q.No.2.: (a). An insulated rod of length ℓ has its ends A and B maintained at 0oC and
100oC, respectively, until steady state conditions prevail. If B is suddenly
reduced to 0oC and maintained at 0oC, find the temperature at a distance x
from A at time t.
(b). Find also the temperature, if the change consists of raising the temperature
of A to 20oC and reducing that of B to 80oC.
Sol.: This is one-dimensional heat flow problem. This is also known as initial boundary
value problem (IBVP).
u  2u
(a): Consider one-dimensional heat flow equation  c2 . (i)
t x 2
To find: Temperature function u(x, t).

u ( x , t )  c1 cos px  c 2 sin px .e  c p t .

2 2
Initial boundary conditions:

An insulated rod of length ℓ has its ends A and B maintained at 0oC and 100oC,
respectively
i.e. u (0, t )  0, u (ℓ, t )  100 for all values of t.
Prior to the temperature change at the end B, when t = 0, the heat flow was independent
of time (steady state condition).
2u
When the temperature u depends only upon x and not on t, (i) reduces to  0.
x 2
Its general solution is u  ax  b , (ii)
where a, b are arbitrary constants.
To find: a and b using boundary conditions
100
Since u = 0 for x = 0 and u = 100 for x  ℓ , we get from (ii), b = 0 and a  .
ℓ
100
 The initial condition is expressed by u ( x ,0)  x.
ℓ
100
Initial condition: Here u ( x ,0)  x.
ℓ
Boundary conditions: The boundary conditions for the subsequent flow are
u(0, t) = 0 and u ℓ, t   0 for all values of t.
The most general solution of (i) satisfying the boundary condition is
c2n 22t

nx
u (x, t )   b n sin
ℓ
.e ℓ2 . (iii)
n 1

100 100 nx
Since u ( x ,0)  x , we have x   b n sin ,
ℓ ℓ n 1
ℓ
100
which is half-range sine series for x.
ℓ
ℓ
  
  cos nx  
sin
nx 
2 ℓ 100 x
Thus b n  
nx
dx 
200  
 ℓ   1  ℓ 
sin  x
n   2 
ℓ 0 ℓ ℓ ℓ 2      n   
  ℓ    ℓ  
    0
200  ℓ 2  200
   cos n    1n  200  1n 1 .
ℓ  n
2
 n n
Hence, the temperature function

c2n 22 t
200  1
 n 1
nx 
u (x, t )  
 n 1 n
sin
ℓ
.e ℓ2 .
This is the temperature at a distance x from A at time t.

100
(b): Here, the initial condition remains the same as u ( x ,0)  x
ℓ
and the boundary conditions are u(0, t) = 20 for all values of t.
u ℓ, t   80 for all values of t.
Since the boundary values are non-zero, we modify the procedure.
We break up the temperature function u(x, t) into two parts as
u x , t   u s ( x )  u t ( x , t ) , (iv)
where us(x) is a solution of (i) involving x only and satisfying the boundary condition;
ut(x, t) is then a function defined by (iv).
Thus, us(x) is a steady state solution of the form (ii) and ut(x, t) may be regarded as a
transient part of the solution which decreases with increase of time.
Since us(0) = 20 and u s ℓ   80 , we have form (ii)
20 = b and 80  aℓ  b
60
b = 20 and a 
ℓ
60
and u s x   x  20 . (v)
ℓ
Putting x = 0 in (iv), we have u t 0, t   u 0, t   u s 0   20  20  0 . (vi)
Putting x  ℓ in (iv), we have u t ℓ, t   u ℓ, t   u s ℓ   80  80  0 . (vii)

100  60  40
Also u t ( x ,0)  u ( x ,0)  u s ( x )  x   x  20   x  20 . (viii)
ℓ  ℓ  ℓ
Hence, (vi) and (vii) give the boundary conditions and (viii) gives the initial condition
relative to the transient solution u t x , t  . Since the boundary values given by (vi) and
(vii) are both zero, therefore, we have

c2n 22 t

nx
u t ( x, t )   b n sin
ℓ
.e ℓ2 , (ix)
n 1
2 ℓ  40  nx  40
where bn is given by b n    x  20  sin dx  1  cos n
ℓ n 2  ℓ  ℓ n
0, when n is odd


  80
 n when n is even
c2n 22 t

 80 1 nx
Hence (ix) becomes u t ( x , t )  
 n  2,4,6,.... n
sin
ℓ
.e ℓ2 .
 4c 2 m 2  2 t

40 1 2mx
 
 m 1 m
sin
ℓ
.e ℓ2 [taking n = 2m] (x)
Combining (v) and (x), the required solution is

 4c 2 m 2  2 t

60 40 1 2mx
u ( x , t )  20 
ℓ
x 
 m 1 m
sin
ℓ
.e ℓ2 .
This is the required temperature if the change consists of raising the temperature of A to
20oC and reducing that of B to 80oC.
u  2u
Q.No.3.: Solve the differential equation  2 for the conduction of heat along a
t x 2
rod without radiation, subject to the following conditions:
(i) u is not infinite for t   ,
du
(ii)  0 for x = 0 and x  ℓ ,
dx
(iii). u  ℓx  x 2 for t = 0 between x = 0 and x  ℓ .

Sol.: This is one-dimensional heat flow problem.
u  2u
Here one-dimensional heat flow equation is  2 .
t x 2
Substituting u = X(x) T(t) in the given equation, we get
X  T
XT    2 X   T     k 2 (say).
X  T 2
d2X dT
  k 2 X  0 and  k 2 2T  0 . (1)
dx 2 dt
Their solutions are
2 2
X  c1 cos kx  c 2 sin kx and T  c 3e  k  t . (2)
If k2 is changed to  k 2 , the solutions are

2 2
X  c 4 e kx  c 5 e  kx and T  c 6 e k  t . (3)
If k2 = 0, the solutions are

X  c 7 x  c 8 and T = c9. (4)
In (3) T   for t   therefore, u also   i.e. the given condition (i) is not satisfied.
So, we reject the solutions (3) while (2) and (4), satisfy this condition.
Applying the condition (ii) to equation (4), we get c7 = 0
 u  XT  c 8 c 9  a 0 (say). (5)
u 2 2
From (2),   c1 sin kx  c 2 cos kx kc 3e  k  t
x
Applying the condition (ii), we get c2 = 0 and  c1 sin kℓ  c 2 cos kℓ  0
 c2 = 0 and kℓ  n (n is an integer).
n 22 2t

2 2  nx 
 u  c1 cos kx.c 3e  k  t  a n cos e ℓ2 . (6)
 ℓ 
Hence, the general solution being the sum of (5) and (6), is
2 2  nx   n 2  2  2 t / ℓ 2
u  c1 cos kx.c 3 e  k  t  a 0   a n cos e . (7)
 ℓ 
Find: The values of a 0 and a n
 nx 
Now using the condition (3), we get ℓx  x 2  a 0   a n cos .
 ℓ 
This being the expansion of ℓx  x 2 as a half-range cosine series in 0, ℓ  , we get

ℓ
1 ℓ
a 0   ℓx  x dx 
ℓ 0
2

ℓ 2

3

1 ℓx 2 x 3

ℓ2
6
,
0
and
ℓ
an 
2 ℓ
ℓ 0

ℓx  x 2 cosnx
ℓ
2
  ℓ
dx  ℓx  x 2  sin
ℓ  n
nx 
ℓ 

  ℓ  2 x  
ℓ2
 n 
2 2
cos
nx 
ℓ 
 ℓ3
  2 
 n 
3 3
sin
nx 
ℓ 
  0
2  ℓ3  4ℓ 2
 0  2 2 cos n  1  0   2 2 , when n is even, otherwise 0.
ℓ  n   n 
Hence, taking n = 2m, the required solution (7) becomes
ℓ2 ℓ2  1  2mx   4m 2  2  2 t / ℓ 2
u x , t     cos e . Ans.
6  2 m 1 m 2  ℓ 
Q.No.4.: Bar with insulated ends. A bar 100 cm long, with insulated ends, has its ends
kept 0oC and 100oC until steady state conditions prevail. The two ends are then
suddenly insulated and kept so. Find the temperature distribution.
The temperature u(x, t) along the bar satisfies the equation
2
u 2  u
c . (i)
t x 2
By law of heat conduction, the rate of heat flow is proportional to the gradient of the
temperature.
Thus, if the ends x = 0 and x  ℓ 100 cm  of the bar are insulated so that no heat can
flow through the ends, the boundary conditions are
u 0, t  u ℓ, t 
 0,  0 for all t. (ii)
x x
2u
Initially, under steady state conditions,  0 . Its solution is u  ax  b .
x 2
Since u = 0 for x = 0 and u = 100 for x  ℓ .  b  0 and a  1 .
Thus, the initial condition is u ( x ,0)  x , 0  x  ℓ . (iii)

Now the solution of (i) is of the form
u x , t   c1 cos px  c 2 sin px e  c p t .

2 2
(iv)
Differentiating partially w.r.t. x, we get
u 2 2
  c1p sin px  c 2 p cos px e  c p t . (v)
x
 u  2 2
Putting x = 0,    c 2 pe  c p t  0 for all t . [by (ii)]
 x  0
c2  0 .
 u  2 2
Putting x  ℓ in (v),    c1p sin pℓe  c p t for all t . [by (ii)]
 x  ℓ
c1p sin pℓ  0 i. e. p being  0 either c1 = 0  sin pℓ  0 .

When c1 = 0, (iv) gives u(x, t) = 0which is trivial solution, therefore sin pℓ  0
n
 pℓ  n  p  , n = 0, 1, 2, ………
ℓ
nx  c 2 n 2  2 t / ℓ 2
Hence, (iv) becomes u ( x , t )  c1 cos e
ℓ
 The most general solution of (i) satisfying the boundary conditions (ii) is

nx  c 2 n 2  2 t / ℓ 2
u x , t    A n cos e
n 0
ℓ

nx  c 2 n 2  2 t / ℓ 2
 A0   A n cos
ℓ
e (where An = c1) (vi)
n 1

nx
Putting t = 0, u x ,0   A 0   A n cos  x. [by (iii)]
n 1
ℓ
This requires the expansion of x into a half-range cosine series 0, ℓ  .


a nx 2 ℓ
Thus x  0   a n cos , where a 0   xdx  ℓ
2 n 1 ℓ ℓ 0
2 ℓ nx 2ℓ
and a n   x cos dx  cos n  1 = 0, where n is even;
ℓ 0 ℓ n 2
2
4ℓ
 , where n is odd.
n 22
Hence, (vi) take the form

ℓ  4ℓ nx  c 2 n 2  2 t / ℓ 2
u x , t     cos e
2 n 1,3,5,.... n 2  2 ℓ
ℓ 4ℓ  1 2n  1x e  c 2 2n 12  2 t / ℓ 2 . Ans.

  
2  2 n 1 2n  12
cos
ℓ
This is the required temperature distribution.

2
u 2  u
Q.No.5.: Determine the solution of one-dimensional heat equation c
t x 2
where, the boundary conditions are
u (0, t )  0 ,
u ℓ, t   0t  0
and the initial conditions are u(x, 0) = x, ℓ being the length of the bar
u  2u
Here one-dimensional heat flow equation is  c2 . (i)
t x 2
u ( x , t )  c1 cos px  c 2 sin px e c

2 2
p t
. (ii)
Boundary conditions: u (0, t )  0 ,
u ℓ, t   0t  0 .
Initial conditions: u(x, 0) = x.
Now since u 0, t   c1e c

2 2
p t
 0  c1  0 . (iii)
2 2
(ii) reduces to u ( x , t )  c 2 sin pxe c p t
.
2 2
Also u (ℓ, t )  c 2 sin pℓe c p t
0
 sin pℓ  0 [ c 2  0 otherwise from(ii) and (i), u(x, t)=0, which is meaningless]
 pℓ  n , n  Z sin   0    n, n  Z
n
p .
ℓ
c 2 n 2  2 c 2 n 2  2
nx t nx t
Hence, (ii) reduces to u x , t   c 2 sin
2
e ℓ  a n sin e ℓ2
ℓ ℓ
[Replacing c 2 by a n ]
Giving n the values 1, 2, 3, ……… and adding all the solutions, we get the general
solution of (i)
c2 n 22

nx t
u x , t    a n sin e ℓ2 . (iv)
n 1 ℓ
Further u(x, 0) = x.

nx
 (iv) reduces to x   a n sin , which is Fourier half-range sine series in 0, ℓ  .
n 1 ℓ
ℓ
2 nx
Hence, a n   x sin dx
ℓ 0 I II ℓ
Integrating by parts, we get

ℓ
  
   cos nx   sin nx 
a n  x 
2 ℓ   1  ℓ 
 n   .
ℓ    n 
2 2

  
  ℓ   
 ℓ2  0
2   ℓ 2  2ℓ


cos n  0  0   

, n0 sin n  0n  Z
ℓ  n  n
[As n varies from 1 to   n  0 ]

Therefore, from (iv), the required solution is given by
c2 n 22
 2ℓ  cos n nx t
u x , t    sin e ℓ2 . Ans.
 n 1 n ℓ
Q.No.6.: A bar with insulated sides is initially at temperature 0oC, throughout. The end
x  0 is kept at 0oC and heat is suddenly applied at the end x  ℓ so that
u
 A for x  ℓ , where A is a constant. Find the temperature function.
x

u  2u
Consider the one-dimensional heat equation  c2 2 (i)
t x
Here given
u (0, t )  0, u x,0   0

Boundary conditions:  u  . (ii)
  A0 
 x  x ℓ 
As one of the boundary condition is non-zero, we consider
u x , t   u s ( x , t )  u t x , t  , (*)
where us(x, t) is steady state solution satisfying the given boundary conditions (ii) and
u t ( x , t ) is the transient solution given by
u t x , t   u x , t   u s x , t  . (iii)
Now for the steady state solution

2
u 2  u 2u u
0c 0 2 0  a  u  ax  b .
t x 2
x t
To find a, b: Using (iii), we have
u x , t   ax  b  0, t   b  0
u  u 
and a  a A
x  x  x ℓ
Therefore, u s x , t   Ax . (iv)
Again from (iii),

u t 0, t   u 0, t   u 2 0, t   0  0 [From (iv) u s (0, t )  0 from (ii) u 0, t   0 ]

Also, from (iii), u t   u   u s
x x x
    u    
  ut       us  AA 0 [From (iv) and (ii)]
 x   x  x ℓ  x  x ℓ
 u 
 i  0
 x  x ℓ
Again from (iii), we have
u t x ,0   u x ,0   u 2 x ,0   0  Ax   Ax [From (ii) and (iv)]
 u t x ,0   Ax .
u  2u
Hence, u t x ,0  is the solution of  c2 2
t x
satisfying
u t 0, t   0 

 u t  
  0 (**)
 x  x ℓ 

u t x ,0    Ax 
which is given by
u t x , t   c1 cos px  c 2 sin px e p

2 2
c t
. (v)
Here u t 0, t   c1e p

2 2
c t
0 [using (*)]
 c1  0 .
u t 2 2
Also, from (v), we get  c 2 p cos pxe c p t
x
  
  ut   c 2 p cos pℓ  0 [using (*)]
 x  x ℓ
 cos pℓ  0 [ c 2  0 , otherwise from (v), ut(x, t) = 0]
 
 pℓ  2n  1  p  2n  1 .
2 2ℓ
 2c 2
n 2 n 12 t
From (v), u t x , t   c 2 sin 2n  1 xe 4ℓ 2 .
2ℓ
 2c 2
n 2 n 12 t
Replacing c2 by an, we get u t x , t   a n sin 2n  1 xe 4ℓ 2 .
2ℓ
Giving n the values 1, 2, 3, ………. And adding, the general solution is given by
 2c 2

n 2 n 12 t
u t x , t    a n sin 2n  1 xe 4ℓ 2 . (vi)
n 1 2ℓ

n
Now u t x ,0    a n sin 2n  1 xe , which is a Fourier half range sine series in 0, ℓ 
n 1 2ℓ
and hence
ℓ
2 x
a n   u 1 x ,0  sin 2n  1 dx [using (*)]
ℓ0 2ℓ
ℓ
2A x
  x sin 2n  1 dx
ℓ 0 I II 2ℓ

ℓ
  x   x 
   cos2n  1    sin 2n  1 
2A  2ℓ   1 2ℓ  
an    x.
ℓ        
 2n  1   2n  1 
  2ℓ   2ℓ  0
2A  4ℓ 2  8Aℓ 1n
 0  sin 2n  1  
ℓ  2n  12  2 2  2n  12  2
      
sin 2n  1 2  sin  n  2    1 sin   2    1 sin n     1 sin 
n n n
     
Therefore, from (vi), we have
2 n 1z 2c 2
8Aℓ 
 1 n
sin 2n  1
x t
u t (x, t) 
2
 2n  12 2ℓ
e 4ℓ 2 (vii)
n 1
and hence from (*), (iv) and (vii), we have

2 n 1z 2c 2
8Aℓ 
 1n sin 2n  1 x e t
u ( x , t )  Ax   4ℓ 2 . Ans.
2 n 1 2n  1
2 2ℓ
Q.No.7.: A homogeneous rod of conducting material of length 100 cm has its ends kept
at zero temperature and the temperature initially is
x , 0  x  50
u x ,0   
100  x, 50  x  100
Find the temperature u(x, t) at any time.
u  2u
Consider the one-dimensional heat equation  c2 2 . (i)
t x
We know the solution of (i) is given by

2 2
p t
(ii)
Now given
u 0, t   0 
Boundary conditions:  (iii)
u 100, t   0
x , 0  x  50
Initial conditions: u x ,0    (iv)
100  x, 50  x  100
Using (ii) and (iii), we get u 0, t   c1e c

2 2
p t
 0  c1  0 .
Also u 100, t   c 2 sin 100pe c

2 2
p t
0
 sin 100p  0 [ c 2  0 , otherwise from (v), ut(x, t) = 0, which is meaningless]
n
100p  n  p  , nZ [ sin   0,   n, n  Z ]
100
c 2n 22 c2n 22
 t  t
n n
xe 100   a n sin xe 100 
2 2
Therefore (ii) reduces to u ( x , t )  c 2 sin (v)
100 100
Giving n the values 1, 2, 3, ……. in (v) and adding all the solutions, we have
c 2 n 22
  t
n
u x , t    a n sin xe 100  .
2
(vi)
n 1 100

n x , 0  x  50
Using (iv), we get u x ,0    a n sin x
n 1 100 100  x, 50  x  100
which is a Fourier half-range sine series in (0, 100) and hence an is given by
1  nx 
100 50 50
2 n n
an   u ( x ,0) sin xdx   x sin xdx   I II 100 dx 
100  x sin
100 0
100 50  0 I II 100 0 
50
 n    50
 x   cos  n    
x     n  
  sin 100 x     cos n
1   100  x   sin x
an   1  100  x   100   1  100 
50  n 2        
   n    50  n  50   n   
2 2
 100   100    100   

2  
    0    100   0
1  5000 n 100 2 n   100 n 1 100 2 n 

  cos  2 2 sin   0  . cos  . 2 2 sin 
50  n 2 n  2   n 2 50 n  2 
1  100 2 n  400 n
 .2. 2 2 sin   2 2 sin , n0
50  n  2  n  2


  400 n [As n varies from 1 to  , other terms get cancelled]
 2 2 sin , when is odd
n  2


 400
sin
2m  1 , when is odd [As n is odd, take n = 2m + 1, m = 0, 1, 2,…]

 2m  1 
2 2 2


 400
  1m , m  0,1,2,....
 2m  1 
2 2
   
[Here sin 2m  1  sin  m     1m sin   1m ]
2  2 2
Therefore, from (vi) the required solution is given by

400 1m 2m  1 xe  2m 1c / 1002 t . Ans.
u x , t    sin
m  0 2m  1 
2 2 100
Q.No.8.: Find the temperature u(x, t) in a slab whose ends x = 0 and x = L are kept at
zero temperature and whose initial temperature f(x) is given by
 1
k, 0x L

f x   
2
0, 1
 LxL
2
u  2u
Consider the one-dimensional heat equation  c2 2 . (i)
t x
We know the solution of (i) is given by

2 2
p t
(ii)
Now given
Boundary conditions: u(0, t) = 0, u(L, t) = 0 . (iii)
 1
k , 0x L
 2
Initial conditions: u x ,0     f (x) . (iv)
 1
0, L  x  L:
2
Using (ii) and (iii), we get u 0, t   c1e c

2 2
p t
 0  c1  0 .
Also u L, t   c 2 sin Lpe  c p t  0

2 2
 sin Lp  0 [ c 2  0 , otherwise from (v), ut(x, t) = 0, which is meaningless]

n
Lp  n  p  , nZ [ sin   0,   n, n  Z ]
L
c2n 22 c2n 22
n  t n  t
Therefore (ii) reduces to u ( x , t )  c 2 sin xe L2  a n sin xe L2 (v)
L L
Giving n the values 1, 2, 3, ……. in (v) and adding all the solutions, we have
c2n 22
  t
n
u x , t    a n sin xe L2 . (vi)
n 1
L

n
Now f x   u x ,0    a n sin x,
n 1 L
which is a Fourier Half-range sine series in (0, L)

L/2
 nx 
 cos
2 nx  2k 
L L/2
2 nx
dx    k sin
nx L 
L 0
an  f ( x ) sin dx  0 sin dx    
L L  0 L L  L  n 
 L  0
2k  n  4 k n
   cos  1  . sin 2 , n  0 , as n varies from 1 to  .
n  2  n 4
 2 2 1  cos A n 
cos 2A  1  2 sin A  sin A  . Here A  
 2 4
Therefore, from (vi), the required solution is given by
c 2n 2 2 c 2n 2 2

4k n nx  t 4k  1 2 n nx  t
u x , t    
2
sin 2 sin e L  sin sin e L2 . Ans.
n 1 n 4 L  n 1 n 4 L
u  2 u
Q.No.9.: Solve the equation 
t x 2
with boundary conditions u ( x ,0)  3 sin nx , u(0, t) = 0, where 0 < x <1, t < 0.
u  2u
Consider the one-dimensional heat equation  c2 2 .
t x
We know the solution of above is given by

2 2
p t
Now given
Boundary conditions: u(0, t) = 0.
Initial conditions: u x ,0  3 sin nx .
The given is one dimensional heat equation with c 2  1 , ℓ  1 .

 
u x , t    a n sin nxe n
2 2
a m sin mxe  m  t .
2 2
 t
  (as n is given) (i)
n 1 m 1

Now u x ,0    a m sin mx  3 sin nx , ∵ u x,0  3 sin nx 
m 1
which is a Fourier half-range sine series in (0, 1) and hence

1
2
am   3 sin nx sin mxdx ∵ ℓ  1 (ii)
10
1
 3 cosm  1x  cosm  n x dx
0
 sin m  n x sin m  n x 

1
 3   0m , n except m  n , m   n
 m  n  m  n   0
But as m is positive integer ∵ a m  0m except m  n

For m = n,
1 1 1
a n  2  3 sin nx sin nxdx  6. sin nxdx  3 1  cos 2nx dx
2
0 0 0
1
 sin 2nx 
 3 x  3 ∵ sin n  0n  Z
 2n  0
Hence, the required solution is given by, from (i)

u x , t   3  sin mxe m
2 2
 t
. Ans.
m 1
V  2V
Q.No.10.: Find the solution of k 2 (i)
t x
having given that V  V0 sin nt when x = 0, for all values of t, and V = 0,
when x is very large.

V  2V
Here one-dimensional heat equation k 2 .
t x
Given
Boundary conditions: V(0, t )  V0 sin nt and V  , t   0 .
Assume the solution of the (i) to be of the form
Vx, t   Ae ax  bt . (ii)
V ax  bt V ax  bt  2V
Now  Ae .b ,  Ae .a ,  Ae ax  bt .a 2 .
t x x 2
Putting all these derivatives in (i), we have
Ae ax  bt b  kAa 2 e ax  bt [Dividing by e ax bt ]
Ab  kAa 2  b  ka 2 .
 From (ii), Vx , t   Ae ax  ka

2
t
(iii)
Also for given V0, t   V0 sin nt
= Imaginary part of V0 e int

2
 Ae ka t  I.P. of V0 e int  A  V0 ,
in
ka 2  in  a 2  (iv)
k
n 1/ 2 n 1 i 
a (i)  a   . .
k k 2
 1/ 2
 1/ 2
         
 Since (i) 1/2   cos  i sin  i sin    cos  2 r   2   i sin  2 r   2  
  2 2 2      
 
 
  cos 4r  1   i sin 4 r  1 , r  0,1 
 4 4 
 [Using De Movier' s theorem] 
 

 Put r  0, i 1 / 2  cos   i sin   1  i 

4 4 2 2
 
 5  5  1 1 1  i  
 r  1, i 1 / 2  cos  i sin  i   
 4 4 2 2  2  
n  1  i 
Also V, t   0 , a   . (v)
k  

Hence, from (iii), (iv) and (v), we have
 n 
x i nt  2 k x 
n n n n
 1i x int   ix int 
Vx , t   Ae 2k  Ae 2k 2k  Ae 2k e  
n
 x  n   n 
 Ae 2 k cos nt  x   i sin  nt  x  .
 
  2k   2 k 
The two possible solutions are
n
 x  n 
Vx , t   Ae 2k cos nt  x  , (vi)
 2 k 
n
 x  n 
Vx , t   Ae 2k sin  nt  x  . (vii)
 2 k 
But from given V  V0 sin nt , therefore, (vi) is the required solution with A = V0
n
 x  n 
Hence, Vx , t   V0 e 2k sin  nt 
 x  . Ans.
 2k 
  2
Q.No.11.: Solve,  k 2 , such that
t x
(i).  is finite when t   .

(ii).  0 , when x = 0 and   0 , when x  ℓ for all t.
x
(iii).    0 , when t = 0 for all values of x between 0 and ℓ .

  2
Consider k 2. (i)
t x
Then, the three possible solutions are (Here c2 = k)
 
(a). x , t   c1e px  c 2 e px .e kp
2
t
(b). x , t   c1 cos px  c 2 sin px .e  kp

2
t
(c). x , t   c 3 x  c 4
1. Now given when t   ,  is finite.
But the solution (a) does give the finite value of  , when t   .  e 
 
The solution (a) is rejected.
  
2. Also from given condition (ii)    0 and ℓ, t   0 (ii)
 x  x 0
And from the solution (c) given and (ii), c 3  0 , c 4  0 , which gives   0 ,
The solution (c) is rejected.

3. Hence the only possible solution is given by
x , t   c1 cos px  c 2 sin px .e kp t .

2
(iii)

  c1p cos px  c 2 p sin px .e kp t
2
Now
x
   2
   0  c 2 p.e kp t  0  c2  0
 x  x 0
Hence (iii) reduces to
x , t   c1 cos px.e kp t .

2
(iv)
Again from given condition (ii), ℓ, t   0 . (iv) gives
2 n
c1 cos pℓ.e kp t  0  cos pℓ  0  cos pℓ  cos , n is odd.
2
c1  0 , otherwise from (iii), we would have x, t   0 , which is meaningless]
n   
p , n is odd cos   0  2n  1 2 , n  Z
2ℓ  
kn 2 2
n  t
(iv) reduces to x , t   c1 cos xe 4ℓ 2
2ℓ
Replacing c1 by an, we have 1, 2, 3,….. and adding, we have the general solution,
kn 2 2
nx  t
x , t   a n cos e 4ℓ 2 . (v)
2ℓ

n
Now from given condition (iii), x ,0    0   a n cos x,
n 1 2ℓ
which is a Fourier half-range cosine series in 0, ℓ  and hence

ℓ
nx
ℓ sin
2 nx 2 0 2ℓ 4 0 n
ℓ 0
an   0 cos dx  .  sin .
2ℓ ℓ n n 2
2ℓ 0
Therefore, from (iv), we have

kn 22

4 0 n nx  t
x , t    sin cos e 4ℓ 2
n 1 n 2 2ℓ
  
2
 3 
2

4 0      
kt
x 1 kt
3x    
  e   cos
2 ℓ
 e   cos
2 ℓ
 ...... ∵ sin 2  1, sin   0 .
 2ℓ 3 2ℓ  
 
 
Q.No.12.: A bar 10 cm long, with insulated sides, has its ends A and B maintained at
temperature 50oC and 100oC respectively, until steady state conditions prevail.
The temperature A is suddenly raised to 90oC and at the same time that at B is
lowered to 60oC. Find the temperature distribution in the bar at time t.
The one dimensional heat equation is
u  2u
 c2 2 . (i)
t x
For steady state conditions, the temperature u depends only upon x and not on t. therefore
u  2u  2u u
 0 and (i)  c 2 2  0  2  0   B  u x ,0   Bx  A . (ii)
t x x x
Now given, until steady state conditions prevail, we have
u 0,0   50 
 (iii)
u (10,0)  100
Using (iii) in (ii), we have
u 0,0  A  50
and u (10,0)  10B  A  100 B5
Hence u x ,0  5x  50 . (iv)
Further the boundary conditions are
u 0, t   90 
, (v)
u (10, t )  60
which are non-zero. Therefore, assume the required, as
u x , t   u s x , t   u t x , t  , (*)
where us(x, t) is the steady state solution satisfying the boundary conditions (v)
and ut(x, t) is the transient solution given by
u t x , t   u x , t   u s x , t  . (vi)
For steady state solution us(x, t), consider
 2u  2u  u 
c2 2
0 2
0 As t  0
x x  
 u  c1 x  c 2 u x , t   c1 x  c 2 (vii)
From (v) and (vii), we have
u 0, t   c 2  90
u 10, t   10c1  90  60  c1  3
 u s x , t   3x  90 . (viii)
u  2u
Further, u(x, t) is the solution of  c2 2
t x
satisfying u t 0, t   0 [To be proved in eqn. (ix)]
u t 10, t   0 [To be proved in eqn. (x)]
u t x ,0   8x  40 [To be proved in eqn. (xi)]
Now from (vi), we have

u t 0, t   u 0, t   u s 0, t   90  90  0 [From (v) and (viii)]
 u t 0, t   0 . (ix)
Also from (vi), we have

 u t 10, t   0 (x)
Again from (vi), we have

u t x ,0   u x ,0   u s x ,0   5x  50   3x  90  [From (iv) and (viii)]
 u t x ,0   8x  40 . (xi)
u  2u
Hence the solution ut(x, t) of  c 2 2 , satisfying (ix) and (x) is given by
t x
c 2n 2 2
 
nx t
u t x , t    b n sin e 102 .
n 1 10

nx
Using equation (xi), we have u t x , t    b n sin , (xii)
n 1 10

10 10
2 nx 1 nx
bn   u t x ,0  sin dx   8x I 40sin dx
10 0
10 5 0
II 10

10
 nx   


  cos    sin nx 
1 10   8 10  1 10 10 
b n  8x  40      40. cos n  40. 
5  n  2 2  5 n n 
    n  
  10   100  0

 80
n
 
1   1n , n  0 as n varies 1 to 
0, n is odd

  160
 , n is even, say n  2m, m  1, 2,.......
 n
160 80
Hence, b n    , m = 1, 2, 3,……..
2mx m
Hence, from (xii), we have
m 2c 2  2
80  1 mx  t
u t x , t     sin e 25 [n = 2m] (xiii)
 n 1 m 5
Therefore, the required solution is given by

m 2c 2  2
 80   1 mx  t
u t x , t   3x  90      sin e 25 [using (*), (viii) and (xiii)]
  n 1 m 5
 m 2c 2  2
80 1 mx  t
 3x  90  
 n 1 n
sin
5
e 25 . Ans. (Replacing m by n)
Q.No.13.: The ends A and B of a rod 20 cm long have the temperature at 30oC and 80oC
until steady state prevails. The temperature of the ends are changed to 40oC
and 60oC respectively. Find the temperature distribution in the rod at time t.
u  2u
The one dimensional heat equation is  c2 2 . (i)
t x
For steady state conditions, the temperature u depends only upon x and not on t. therefore
u  2u 2u u
 0 and (i)  c 2 2  0  2  0  A.
t x x x
u ( x , t )  Ax  B , at t = 0
 u x,0  Ax  B (ii)
u 0,0   30 
Now given, until steady state conditions prevail, we have . (iii)
u (20,0)  80
Using (iii) in (ii), we have u 0,0  B  30

5
and u (20,0)  20A  30  80  A  .
2
5
Hence u x ,0   x  30 (iv)
2
u 0, t   40 
Further the boundary conditions are . (v)
u (20, t )  60
which are non-zero. Therefore, assume the required, as
u x , t   u s x , t   u s x , t  , (*)
where us(x, t) is the steady state solution satisfying the boundary conditions (v)
and ut(x, t) is the transient solution given by
u t x , t   u x , t   u s x , t  (vi)
u  2u
For steady state solution us(x, t), consider  0  c2 2  0 .
t x
 u  c1 x  c 2  u x, t   c1x  c 2 . (vii)
From (v) and (vii),
u 0, t   c 2  40
u 20, t   20c1  40  60  c1  3
 u s x , t   x  40 . (viii)
u  2u
Further, ut(x, t) is the solution of  c2 2 ,
t x
satisfying u t 0, t   0 [To be proved in eqn. (ix)]
u t 20, t   0 [To be proved in eqn. (x)]
3
u t x ,0   x  40 [To be proved in eqn. (xi)]
2
Now from (vi), we have
 u t 0, t   0 . (ix)
Also from (vi), we have

 u t 20, t   0 . (x)
Again from (vi), we have

5
u t x ,0   u x ,0   u s x ,0   x  30  x  40 [From (iv) and (viii)]
2
3
 u t  x ,0   x  10 . (xi)
2
u  2u
Hence, the solution ut(x, t) of  c 2 2 , satisfying (ix) and (x) is given by
t x
c 2n 2 2
 
nx t
u t x , t    b n sin e 202 .
n 1 20

3 nx
Using equation (xi), we have u t x , t   x  10   b n sin , (xii)
2 n 1 20

20
  
  nx   nx 
20   cos   sin
2 3  nx 1  3  20    3  20 
bn 
20  
2
x  10 

sin
10
dx  
10  2
x  10 
 n
 
  2   n  2 

0    
  20    20  
    0
1   20 2   200
bn  
 20
cos n  0   .10   1  2 cos n   20 1  2 cos n , n  0
10  n  n  10n n
Hence, from (xii), we get

n 2c2 2
20  1 
1  2 cos nsin mx e
t
u t x , t     400 . (xiii)
 n 1 n 20
and hence, from (viii), (*) and (xiii), we get

n 2c 2  2
20  1 nx  t
u x , t   3x  40   1  2 cos n   sin e 400 . Ans.
 n 1 n 20
*** *** *** *** ***

*** *** ***
***
Home Assignments
u  2u
Q. No. 1.: Find a solution of the heat conduction equation  such that
t x 2
(i) u is finite when t   , (ii) u =100 when x = 0 or  , for all t,
(iii) u = 0 when t = 0 for all values of x between 0 and  .
(Here the initially ice-cold rod has its ends in boiling water)
400  sin 2m  1x  2m 12 t

Ans.: u  100  
 m 1 2m  1
e .
Q. No. 2.: Find the temperature u(x, t) in a bar which is perfectly insulated laterally,
whose ends are kept at temperature 0oC and whose initial temperature in (oC)
is f x   x 10  x  given that its length is 10 cm, constant cross-section of
area 1 cm2 , density 10.6 gm/cm3, thermal conductivity 1.04 cal/cm deg./sec
and specific heat 0.056 cal/gm degree.

800 nx  .0175n 2  2 t 
Ans.: u x , t    sin e .
2 2
n 1 n 
10
Q. No. 3.: Solve the boundary value problem:
u  2u u 0, t  u ℓ, t 
 , 0 x ℓ,  0,  0 , u(x, 0) = x.
t x 2 x x
Ans.: u x , t  
5 4ℓ  1 2n  1x e  2n 12 c 2  2 t / 25 .
  cos
2 2 1 2n  12 ℓ
Q. No. 4.: The temperature at one end of a bar, 50 cm long with insulated sides, is kept
at 00C and that the other end is kept at 100oC until steady state conditions
prevail. The two ends are then suddenly insulated, so that the temperature
gradient is zero at each thereafter. Find the temperature distribution.
Ans.: u x , t   50 
400  1 2n  1x e  c 2  2 2n 12 t / 2500 .
 cos
2 1 2n  12 50
************************************************
****************************************
****************************

3 Topic: Applications of Partial Differential Equations

Uploaded by

Copyright:

Available Formats

3 Topic: Applications of Partial Differential Equations

Uploaded by

Document Information

Original Title

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

3 Topic: Applications of Partial Differential Equations

Uploaded by

Copyright:

Available Formats

3rd Topic

Partial Differential Equations

One-dimensional wave equation:

One-dimensional heat flow equation:

FORMULATION of One-dimensional heat flow equation:

where  lies between x and x  x .

i.e. the end points are assumed to be zero temperature.

SOLUTION of one-dimensional heat flow equation by separation of

Solving equations (iv), we get

Case (i): When k is positive and  p 2 (say)

Thus Xx   0 for all x or u = 0 for all x, which is a trivial solution.

Case (ii): When k = 0

Thus Xx   0 for all x or u = 0 for all x, which is a trivial solution.

Case (iii): When k is negative and   p 2 (say)

Find the values of c1 and c 2 using boundary conditions:

We must take c 2  0 , since otherwise Xx   0 .

Hence, the ‘eigen function’

To determine the unknown constants b n , use initial condition u x ,0  f x  in

Thus, the temperature distribution in the rod is given by the solution

u  C1 cos px  C 2 sin px .e  c p t .

Steady state condition:

u x , t   c1 cos px  c 2 sin px .e  c p t ,

where u(x, t) is called temperature function.

u ( x , t )  c1 cos px  c 2 sin px .e  c p t .

To find: c1 and c 2 using boundary conditions

Since u ℓ, t   0 , from (ii), we have 0  c 2 sin pℓ.e  c p t

which is half range sine series for u0.

To find: Temperature function u(x, t).

u ( x , t )  c1 cos px  c 2 sin px .e  c p t .

Initial boundary conditions:

Hence, the temperature function

This is the temperature at a distance x from A at time t.

Putting x  ℓ in (iv), we have u t ℓ, t   u ℓ, t   u s ℓ   80  80  0 . (vii)

(vii) are both zero, therefore, we have

0, when n is odd

Combining (v) and (x), the required solution is

(iii). u  ℓx  x 2 for t = 0 between x = 0 and x  ℓ .

Sol.: This is one-dimensional heat flow problem.

If k2 is changed to  k 2 , the solutions are

If k2 = 0, the solutions are

This being the expansion of ℓx  x 2 as a half-range cosine series in 0, ℓ  , we get

Hence, taking n = 2m, the required solution (7) becomes

Thus, the initial condition is u ( x ,0)  x , 0  x  ℓ . (iii)

u x , t   c1 cos px  c 2 sin px e  c p t .

c1p sin pℓ  0 i. e. p being  0 either c1 = 0  sin pℓ  0 .

This requires the expansion of x into a half-range cosine series 0, ℓ  .

ℓ 4ℓ  1 2n  1x e  c 2 2n 12  2 t / ℓ 2 . Ans.

This is the required temperature distribution.

u ( x , t )  c1 cos px  c 2 sin px e c

Now since u 0, t   c1e c

Integrating by parts, we get

[As n varies from 1 to   n  0 ]

Sol.: This is one-dimensional heat flow problem.

Now for the steady state solution

Again from (iii),

u t x ,0   u x ,0   u 2 x ,0   0  Ax   Ax [From (ii) and (iv)]

u t x , t   c1 cos px  c 2 sin px e p

Here u t 0, t   c1e p

Integrating by parts, we get

* * * * ***