Partial Differential Equation

(1) Suppose
∂2 y 2∂ y
2
= c for 0 < x < a (1)
∂t 2 ∂x 2

(i) Find the separable solutions which satisfy the boundary

conditions
=
y(0, t ) y(=a , t ) 0, t > 0. (2)
(ii) By linearly combining two such solutions find y( x , t ) when
the initial conditions are y( x , 0) = 0, (3)
∂y( x ,0) πx 2π x
= A sin + B sin (4)
∂t a a
(iii) Find
y( x , t ) subject to
 π x  ∂y( x ,0) (5)
= y( x ,0) C= sin 3  , 0
 a  ∂t
solution:
(i) Step 1: Separation of variables.
Assume that y( x , t ) = X ( x )T ( t ) is a solution of (1). Then it becomes
1
X ′′ ( x )T ( t ) = 2 X ( x ) T ′′( t ) ( 6)
c
Divide (4) by y( x , y ) = X ( x )T ( t )
Therefore
X ′′( x ) T ′′( t )
= 2 = − λ 2 (constant ) (7)
X ( x ) c T (t )
Then we have
X ′′( x )
− λ2 (8)
X ( x)
T ′′( t )
2
= −λ 2 (9)
c T (t )
 Partial differential Equations

for all ( x , y ) in D and

solve X equation
=X ( x ) A cos λ x + B sin λ x
From the boundary condition (2) = y(0, t ) X=
(0)T ( t ) 0 Then
X (0) = 0
X (0) = A cos 0 + B sin 0 =A=0 the solution becomes
X ( x ) = B sin λ x (10)
From the boundary conditions in (2) we have y(a , t ) = 0 Then
X (a ) = 0
Equation (1) X (a ) =B sin λ a =0 ⇒ λ a =nπ , n =1,2,3,...
nπ nπ x
Then λn = and substitute in (10) we have X n ( x ) = Bn sin
a a
nπ
Step 3: Solve T equation. (with λn =
U U )
a
From (5)
T ′′( t )
2
= −λ 2
c T (t )
Tn′′( t ) = − λn 2c 2Tn ( t )
nπ ct nπ ct
Tn ( t ) = C n sin λnct + Dn cos λnct = C n sin + Dn cos (11)
a a
From (10), (11)
nπ x  nπ ct nπ ct 
y( x , t ) Bn sin C sin + D cos
a  n
a n
a 
For simplicity the general solution take the following form
nπ x  nπ ct nπ ct 
y( x , t ) sin  Dn sin + En cos (12)
a  a a 
**************************************************
U

(ii) use the initial condition (2) y( x ,0) = 0 in equation (12) we have

2
 Partial differential Equations

nπ x nπ x
yn (= 0 + =
En  0 since ≠ 0 then En = 0 and
a 
x , 0) sin sin
a
nπ x nπ ct
the solution (12) becomes yn ( x , t ) = Dn sin sin .
a a
according the super position rule
let
=y( x , t ) y1 ( x , t ) + y2 ( x , t )
πx π ct2π x 2π ct
= D1 sin sin + D2 sin
sin
a a a a
∂y ( x , t ) π c πx π ct 2π c 2π x 2π ct
D1 sin cos + D2 sin cos
∂t a a a a a a
∂y( x ,0) π c π x 2π c 2π x πx 2π x
= D1 sin + D2 sin = A sin + B sin
∂t a a a a a a
Then
Aa Ba
=D1 = , D2
πc πc
and the general solution in this case
Aa π x π ct Ba 2π x 2π ct
= y( x , t ) sin sin + sin sin
πc a a πc a a

**************************************************
U

(iii) let the solution is

nπ x  nπ ct nπ ct 
y( x , t ) sin D sin + E cos
a  n
a n
a 
∂y ( x , t ) nπ x  nπ c nπ ct nπ c nπ ct 
sin  Dn cos − En sin
∂t a  a a a a 
∂y ( x , 0 ) nπ x  nπ c 
= sin = Dn  0
∂t a  a 
then Dn = 0

3
 Partial differential Equations

nπ x nπ ct
∴ yn ( x , t ) = En sin cos
a a
and the general solution
∞ ∞ nπ x nπ ct
= y ∑ = yn ( x , t ) ∑ En sin cos
= n 1= n 1 a a
apply the condition
πx
y( x ,0) = C sin 3  
 a 
∞ nπ x πx
= y ( x , 0) ∑ = En sin C sin3  
n =1 a  a 

which show that

πx
En is the coefficient of Fourier series for the function C sin 3  
 a 
then
πx nπ x
a
2
En = ∫ C sin 3   sin dx
a0  a  a
and the general solution
2C ∞  a 3 π x  nπ x  nπ x nπ ct
y( x , t ) = ∑ ∫ sin   sin dx  sin cos
a n =1  0  a  a  a a

********************************************************
U

∂T ∂ 2T
(2) Derive the one-dimensional heat equation =k 2,
∂t ∂x
T0  − x2 
and verify that T = exp  2  is a solution.
(1 + 4kt / a )
2
 a + 4 kt 
********************************************************
U

4
 Partial differential Equations

(3) A thermal conductor occupies the region 0 ≤ x ≤ a . The end

x = 0 is maintained at T = 0 , while the end x = a is insulated. Show
that separable solutions of the heat equation satisfying these
conditions have the form
 −(2n + 1)2 π 2 kt  (2n + 1)π x
T ( x , t ) = T0 exp  2  sin .
 4a  2a
Answer:
U

Let T ( x , t ) = F ( x )G ( t ) substitute in the differential equation

∂T ∂ 2T
= k 2 , in 0 < x < a
∂t ∂x
F ( x )G ′( t ) = k F′′( x )G ( t )
G ′( t ) F ′′( x )
= = −α 2
kG ( t ) F ( x )
separate the variables
G ′( t )
= −α 2 ⇒ G ( t ) = Exp( −α 2 kt )
kG ( t )
F ′′( x )
= −α 2 ⇒ F ( x ) = A cos α x + B sin α x
F ( x)
exp( −α 2 kt ) [ A cos α x + B sin α x ]
T ( x, t ) =
T (0, t ) =exp( −α 2 kt ) [ A] =0 ⇒ A =0
so
( x , t ) B exp( −α 2 kt )sin α x
T=
∂T ( x , t )
= α B exp( −α 2 kt )cos α x
∂x
∂T (a ,0)
From the boundary condition =0
∂x
∂T (a ,0) π (2n + 1)π
= α B cos α a ⇒ cos α a = 0 ∴ α a = (2n + 1) ⇒ α =
∂x 2 2a
then

5
 Partial differential Equations

(2n + 1)2 π 2kt (2n + 1)π

T=( x , t ) B exp( − )sin x
4a 2 2a
********************************************************
U

(4) Find the separable solutions of Laplace’s equation in polar

coordinates
∂ 2T 1 ∂T 1 ∂ 2T
Ans.  Ar n +
B
+ + 2
r ∂r r ∂θ
= 0  ( C cos nθ + D sin nθ )
∂r 2 2
 rn 
and solve the equation in r ≤ a when
( i ) T T= 0 cos 2θ on r a
= ( ii )T T= 0 cos θ
2
on r a
Where are the maximum and minimum temperatures T attained in
each case?
(iii)A thermally conducting solid lies in the annular region a ≤ r ≤ b .
If T = 1 on r = a and T = cos 2 θ on r = b , find T ( r ,θ ).
Answer: U

∂ 2T 1 ∂T 1 ∂ 2T
+ + = 0 (1)
∂r 2 r ∂r r 2 ∂θ 2
let T ( r ,θ ) = F ( r )G (θ )
substitute in (1)
1 1
F ′′( r )G (θ ) + F ′( r )G (θ ) + 2 F ( r )G ′′(θ ) =
0
r r
F ′′( r ) F ′( r ) G′′(θ )
r2 +r + = 0
F (r ) F ( r ) G (θ )
F ′′( r ) F ′( r ) G′′(θ )
r2 +r = −
F (r ) F (r ) G (θ )
F ′′( r ) F ′( r ) G′′(θ )
r2 +r = − = λ2
F (r ) F (r ) G (θ )

6
 Partial differential Equations

F ′′( r ) F ′( r )
r2 +r λ2
= (2)
F (r ) F (r )
G ′′(θ )
= −λ 2 (3)
G (θ )
 B
=
F ( r )  Ar n + n 
 r 
G (θ ) ( C cos nθ + D sin nθ )
=
 B
∴ T ( r ,θ ) = Ar n + n  ( C cos nθ + D sin nθ )
 r 

********************************************************
U

(5) Use the method of separation of variables, and Fourier series, to

solve the heat equation

∂u ∂ 2u
= k 2 , in 0 < x < a
∂t ∂x
subject to the boundary conditions
u= (0, t ) u= (a , t ) 0 for t > 0,
and the initial condition
x 1
 a if 0 < x ≤
a
u( x ,0) = 
1 − x if
1
≤ x<a
 a a
Answer:
U

Let u( x , t ) = X ( x )T ( t ) substitute in the differential equation

X ( x )T ′( t ) = k X′′( x )T ( t )
T ′( t ) X ′′( x )
= = −α 2
kT ( t ) X ( x )
separate the variables

7
 Partial differential Equations

T ′( t )
=−α 2 ⇒ T ( t ) =Exp( −α 2 kt )
kT ( t )
X ′′( x )
=−α 2 ⇒ X ( x ) =A sin α x + B cos α x
X ( x)
exp( −α 2 kt ) [ A sin α x + B cos α x ]
u( x , t ) =
From the boundary condition u(0, t ) = 0
u(0, t )= exp( −α 2 kt ) [ A sin 0 + B cos 0]= exp( −α 2 kt ) [ B ]= 0 ⇒ B= 0
so
, t ) exp( −α 2 kt ) [ A sin α x ]
u( x=
From the boundary condition u(a , t ) = 0
∴ u(a , t ) =exp( −α 2 kt ) [ A sin ax ] =0 since exp( −α 2 kt ) ≠ 0
∴ A sin α a =0 ⇒ α a =nπ
nπ
=α = , n 0,1,2,...
a
− n2π 2 kt  nπ  − n2π 2 kt nπ
=
un ( x , t ) exp( )  An sin x  An exp( )sin x
a 2
 a  a 2
a
according the super position theory
∞ ∞ − n2π 2 kt nπ
=u( x , t ) ∑ = un ( x , t ) ∑ An exp( 2
)sin x
= n 1= n 1 a a
from the initial condition
x a
∞  a if 0 < x ≤
nπ 2
=u( x ,0) ∑ = An sin x 
n =1 a 1 − x a
if ≤ x<a
 a 2
which show that An is the coefficient of odd sine Fourier series for the
a
2
4 x (2n − 1)π x
given function An = ∫
a0a
sin
a
dx

8
 Partial differential Equations

a
4 −x a (2n − 1)π x a (2n − 1)π x2
 cos + sin 
a  a (2n − 1)π a (2n − 1)2 π 2 a 0
1 (2n − 1)π ( −1)n −1
= sin
4(2n − 1)2 π 2 2 4(2n − 1)2 π 2
1 ∞ ( −1)n −1 − n2π 2 kt (2n − 1)π
u( x , t ) = ∑ exp( )sin x
4π 2 n =1 (2n − 1) 2
a 2
a
********************************************************
U

(6) Thermally conducting material occupies 0 < x < a , and the

boundaries
at x = 0 and x = a are insulated.
T for 0 < x < ε a ,
If T ( x ,0) =  0
0 for ε a < x < a ,
where 0 < ε < 1 , show that
2T ∞ 1  − n2π 2 kt   nπ x 
T ( x= , t ) ε T0 + 0 ∑ sin( nπε )exp  
 cos  
π n =1 n  a
2
  a 
Answer:
U

Let T ( x , t ) = F ( x )G ( t ) substitute in the differential equation

∂T ∂ 2T
= k 2 , in 0 < x < a
∂t ∂x
F ( x )G ′( t ) = k F′′( x )G ( t )
G ′( t ) F ′′( x )
= = −α 2
kG ( t ) F ( x )
separate the variables
G ′( t )
= −α 2 ⇒ G ( t ) = Exp( −α 2 kt )
kG ( t )
F ′′( x )
= −α 2 ⇒ F ( x ) = A cos α x + B sin α x
F ( x)

9
 Partial differential Equations

T ( x, t ) = exp( −α 2 kt ) [ A cos α x + B sin α x ]

∂T (0, t )
From the boundary condition =0
∂x
∂T ( x , t )
= α exp( −α 2kt )[ A sin α x − B cos α x ]
∂x
∂T (0, t )
= α exp( −α 2 kt ) [ − B ]= 0 ⇒∴ B= 0
∂x
so
T ( x=, t ) exp( −α 2 kt ) [ A cos α x ]
∂T ( x , t )
and −α exp( −α 2 kt ) [ A sin α x ]
=
∂x
∂u(a , t )
From the boundary condition =0
∂x
∂T ( a , t )
= −α exp( −α 2 kt ) [ A sin α a ] = 0 ⇒ sin α a ⇒ α a =
nπ
∂x
nπ
αn =
= , n 0,1,2,...
a
∞ n2π 2 kt
T ( x , t ) ∑ exp( −
= ) [ A cos nπ x ]
n= 0 a2
∞ − n2π 2 kt
= A0 + ∑ exp( )  An cos nπ x 
n =1 a2
apply the initial condition
∞ T0 for 0 < x < ε a ,
A0 + ∑  An cos nπ x  =
T ( x ,0) = 
n =1 0 for ε a < x < a ,
1 εa T0
=
A0 ∫ =
T0dx = ε a ε T0
a 0 a
εa
2 εa nπ x 2  a nπ x  2T0
An = ∫ T0 cos dx T0= sin sin nπε
a 0 a a  nπ a  0 nπ

10
 Partial differential Equations

2T0 ∞ sin nπε − n2π 2 kt nπ

ε T0 +
∴T ( x, t ) =
π
∑ exp( ) cos x
n =1 n a2 a
********************************************************
U

(7) A uniform conductor of thermal diffusivity k occupies the

region 0 ≤ x ≤ a .The initial temperature is T = 0 for 0 < x < a , and
subsequently T = 0 at x = 0 while T = 1 at x = a . Show that, for
t > 0:
x ∞ 2  − n2π 2 kt  nπ x
T ( x, t ) = +∑ ( −1) exp 
n

 sin .
a n =1 nπ  a
2
 a
Answer:
U

Let T ( x , t ) = F ( x )G ( t ) substitute in the differential equation

∂T ∂ 2T
= k 2 , in 0 < x < a
∂t ∂x
F ( x )G ( t ) = k F′′( x )G ( t )
′
G ′( t ) F ′′( x )
= = −α 2
kG ( t ) F ( x )
separate the variables
G ′( t )
= −α 2 ⇒ G ′( t ) = kG ( t ) (1)
kG ( t )
∴ G ( t ) = Exp( −α 2 kt )
F ′′( x )
=−α 2 ⇒ F ′′( x ) =
−α 2 F ( x ) (2)
F ( x)
∴ F ( x= ) A cos α x + B sin α x
exp( −α 2 kt ) [ A cos α x + B sin α x ]
T ( x, t ) =
From the boundary condition T (0, t ) = 0
T (0, t ) =exp( −α 2 kt ) [ A] =0 ⇒ A =0
so

11
Partial differential Equations

, t ) exp( −α 2 kt ) [ B sin α x ]
T ( x=
From the boundary condition T (a , t ) = 0
T (a , t ) =exp( −α 2 kt ) [ B sin α a ] =0 ⇒ α a =nπ
nπ
αn
= = , n 0,1,2,...
a
Since The initial temperature is = T ( x ,0) F= ( x )G (0) 0 [but
G (0) = 1] for 0 < x < a , then F ( x ) = 0 and From equation (2)
F ′′( x ) = 0
Which has a general solution F (= x ) C1 x + C 2 where
C1 , C 2are constants
At x = 0 F (0) =C1 .0 + C 2 =0 ⇒ C 2 =0
1
∴ F ( x) = C1 x at x = a we find F (a ) =C1a =1 ⇒ C1 = then
a
x
∴ F ( x) =
a
so
x n2π 2 k t  nπ 
Tn ( x , t ) = + Bn exp( − ) sin x ,n = 1,2,3,...
a a2  a 
and
x ∞ n2π 2 k t  nπ 
T ( x, t ) =+ ∑ B exp( − ) sin x (3)
a n =1 n a2  a 

at t = 0 we have T ( x ,0) = 0
x ∞  nπ 
T ( x ,0) = + ∑ Bn sin x =0
a n =1  a 
∞
 nπ  − x
∴ ∑ Bn sin = x , a< x<a
n =1  a  a
which show that

12
Partial differential Equations

a
2 a −x nπ x 2  − x  − a nπ x  1 nπ x 
Bn = ∫
a0 a
sin
a
dx  
a  a  nπ
cos  −
a  n2π 2
sin
a  0
 2 nπ x  2( −1)n
= = cos 
 nπ a  nπ

substitute in (3) we get the general solution in the form

x 2 ∞ ( −1)n n2π 2 k t  nπ

T ( x, t ) =+ ∑ exp( − x
) sin
a π n =1 n a2  a
********************************************************
U

(8) T ( x , y ) satisfies Laplace’s equation in the square region

0 < x < π , 0 < y < π . T = 1 on the side y= π , 0 < x < π , and
T = 0 on the other three sides. Show that
4 ∞ sinh(2n + 1) y sin(2n + 1) x
T= ∑ .
π n = 0 sinh(2n + 1)π (2n + 1)
What is the value of T at the centre of the square?
Answer
U

∂ 2T ∂ 2T
+ = 0 (1)
∂x 2 ∂y 2
Assume that T ( x , t ) = X ( x )Y ( y ) is a solution of (1). Then it
becomes
X ′′( x )Y ( y ) + X ( x )Y ′′( y ) =
0
X ′′( x )Y ( y ) X ( x )Y ′′( y )
+ =0
X ( x )Y ( y ) X ( x )Y ( y )
X ′′( x ) Y ′′( y )
+ = 0
X ( x) Y ( y)
X ′′( x ) Y ′′( y )
= − = −α 2
X ( x) Y ( y)

13
 Partial differential Equations

X ′′( x )
= −α 2 (2)
X ( x)
Y ′′( y )
=α2 (3)
Y ( y)

Solution X Equation
U

=X ( x ) A cos α x + B sin α x (4)

Solution Y Equation
U

Y=( y ) Ceα y + De −α y (5)

since
since T ( x ,0) =0 ⇒ X ( x )Y (0) =
0 ∴ Y (0) = 0 (6)
since T ( x , π ) =1 ⇒ X ( x )Y (π ) = 0 ∴ Y (π ) =1 (7)
since T (0, y ) = 0 ⇒ X (0)T ( y ) = 0 ∴ X (0) = 0 (8)
since T (π , y ) = 0 ⇒ X (π )T ( y ) = 0 ∴ X (π ) = 0 (9)
from (8) and (4)
X (0)= A= 0
∴ X ( x) = B sin α x
from (9)
X (π ) =B sin απ =⇒ 0 απ = nπ
=∴ α n n and
= X n (π ) Bn sin nx
from (5) , (6)
C + D =⇒ 0 C= −D (10)
substitute in (5)
Y ( y ) =Ceα y − Ce −α y =C  eα y − e −α y  =2C sinh α y

Yn ( y ) = 2C n sinh ny
the general solution in the form
= n ( x )Yn ( y ) 2 ( C n Bn ) sinh ny=
Tn ( x , y ) X= sin nπ , n 0,1,2,3,

14
Partial differential Equations

∞
T ( x, y) = ∑ 2 Dn sinh ny sin nπ (11)
n =1
∞
T ( x ,π ) ∑
= 2 Dn sinh nπ sin nπ 1
n =1
which show that 2 Dn sinh nπ is the coefficient of Fourier sine series
as follows
π
2π 2  −1  −2
n sinh nπ
2 D= ∫
= sin nxdx  = cos nx  [ cos nπ − 1]
π0 πn  0 nπ
 4
−2  =  nπ
n odd
= ( − 1) n
− 1
nπ  
 0 n even
4
2 Dn sinh nπ =
(2n − 1)π
4
2 Dn =
π (2n − 1)sinh(2n − 1)π
in final form (11) becomes
4 ∞ 1 sinh(2n − 1) y sin(2n − 1)π
T ( x, y) = ∑
π n =1 (2n − 1) sinh(2n − 1)π

15