SOLUTIONS OF SELECTED PROBLEMS OF EXERCISE 3

ANS 3.6
Given differential equation is,

d 2
  1  0
dx 2

Assume   a  bx  cx 2

Given boundary conditions here all boundary conditions should be used,

  1 at x  0
d
 1 at x  1
dx
Applying boundary conditions gives,

a  1 which gives

  1  bx  cx 2
d
 b  2cx which gives,
dx

b  1  2c

  (1  x)  c( x 2  2 x)
d
 1  c (2 x  2)
dx
d 2
 2c
dx 2

d 2
  1  0
dx 2

R  2c  (1  x)  c ( x 2  2 x )  1

Let W= x 2  2 x

Weighted integral method

 WRdD  0

x  2 x   2c  1  x  c ( x 2  2 x )  1 dx  0
D
1
 2

0
Which gives as,

c = -2.1875
b  1  2c =5.375

  1  5.375 x  2.1875 x 2

ANS 3.8
Variational functional:

 1 
I     AE    pu dx+  Pu  x  l
 du 
l 2

 2  dx  
0  2

x l  x 
approximating function: u  a1
l2

du
 a1
l  2x 
dx l2

essential boundary condition: u(0)=0 and u(l)=0 are satisfied by u.

du
To solve this problem by Ritz’s method, value of u and are put in variational form,
dx

 1  d  x l  x    x l  x     x l  x   
I     AE   a1
2


l

 dx    p  a1  dx+  P  a1  
 2   l 2
  l 2
   l 2
  x l
0
 2

 1   l  2x    x l  x      x l  x  
 I     AE  a1
l 2

  p  1
a  dx+  P  a1  
0
 2  l 2
  l 2
   l 2
  x l
2

 1  l  2x    x l  x      x l  x   
I     AE  a1
l 2

  p  a1  dx+  P  a1 
0
 2  l 2
  l 2
   l 2   l
x
2
 A E a1 2 2 4l 3 p a1 lx 2 x3 a1
l l

I   l x  2 lx 2  2  P
2l 4
3 0
l 2 3 0
4

AE a12 pa1l P a1
 I   
6l 4
6 4
This completes part (a).
Extremizing wih respect to a1 :
I
0
 a1

 A E a1 pl P
   0
3l 6 4

3l  pl P 
 a1    
AE  6 4 
putting a1 in u:

3lx  l  x   pl P 
u   
AEl  6 4
2

This completes part (B).

ANS 3.9

The given differential equation is

d2u
 -cu+x 2 =o
dx 2
Let u  l  mx  nx 2 ………………………………(1)

du
=m +2nx
dx

By putting the boundary conditions,

du du 4
dx
(x = 0 )= 1 and d x ( x = 1 ) = 3
We get m=1, n= 1
6
1
From (1) u  l  ( x  6 x )
2

Residue is given by
1 1
R  c (l  x  x 2 )  x 2
3 6
Here weight W=1

Weighted integral method

 W RdD
D
 0

  3  c(l  x  6 x
 1 1 
1
2
)  x 2 dx  0
0 

5
This gives l  
9

5 x
2
u  x
9 6

ANS 3.10
d du
Given: (u )  f  0
dx dx

Bc’s: u= 2 at x=0

du
=0 at x=1
dx
f= b1+b2x (b1,b2 are constants)
u= 2 +a1x+a2x2
Now applying the boundary conditions

du
Since =0 at x=1
dx
Also
d
( 2  a1 x  a 2 x 2)
dx
 a1+2a2=0
a1=-2a2

Therefore, u= 2 +a2(x2-2x)

d du
since residue, R=
(u )  f
dx dx
d du
R= (u )  (b1  b 2 x)
dx dx
d d
= (u ( 2 +a2(x2-2x))- (b1  b 2 x ) )
dx dx
d
=a2 { 2 +a2(x2-2x)(2x-2)}- (b1  b 2 x )
dx
=a2 (2 2 +6a2x2-12a2x+4a2)-(b1+b2x)

Since the basis function here is that connected with a2

i.e. the weight being  i =(x2-2x)

 [a2 {2 2 +6a2x2-12a2x+4a2}-(b1+b2x)]*(x2-2x)dx
1
Thus,
0

On solving the above integral we get:

8/15 a22-4/3a2+2/3b1+5/12b2=0

4 / 3 2  32 / 9  32 /15  (2 / 3b1  5 /12b 2)

Or , a2=
(16 /15)

Therefore ,
 4 / 3 2  32 / 9  32 /15 * (2 / 3b1  5 /12b 2) 2
U= 2+ (x -2x) (Answer)
(16 /15)

ANS 3.11

Step 1: Convert into variational form. For this let

 d 2 
 I    2    1 dx
1

0 
dx
Integrating the first term by parts
 d  d  
dx      dx    dx
d
dx 0 0  dx  dx
1
1 1 1 1
 I     
2

0
2 0
Applying boundary conditions in first term become zero. Then equations becomes

1  d  d  
 I      dx     2  dx    dx
1
1 1 1

0
2  dx  dx 0
2 0

Or

 I          2    dx
1
 1  d  1
2

 2  dx  2 
0 
Therefore

 
 1  d  2 1 2 
I   
1
    dx (1)
 2  dx  2
0 
Let us solve this problem using Ritz method.
If approximate  by   a  bx , applying the essential boundary conditions i.e.  (0)
=1. So a=1, then
  bx  1
Applying to equation (1)

I      b    bx  1  (bx  1) dx
 1 2 1 
1
2

0 
2 2
Extremize this respect to b

    x 2  1  2  bx   dx    b  1  0
dI
1
 b2 3 b 
db 0  2 2 3 

This gives b= 3/2

 3
Thus,   x 1
2
d 2
Exact solution of the differential equation (    1  0 ) is
dx 2
  A cos x  B sin x  1
Where the constants A and B are to be determined from the boundary conditions. Since
 (0) =1, A=2
d (1) d
Also, 1 ,   A sin x  B cos x
dx dx

Hence 2sin1  B cos1  1 or B= 1.035.

The value of exact  at x=0.5, 0.75 and 1 are

 (0.5) =1.008955  (0.75) =1.0133  (1) =1.0133
The approximate solutions at these points are
 (0.5) =1.75  (0.75) = 2.125  (1) =2.125
Now increase the accuracy of answer, add one more term in the approximation function
and take

  a  bx  cx 2
Essential boundary conditions are  (0) =1. So a=1, thus
  bx  cx 2  1
Applying to equation (1)

I      b  2cx    bx  cx 2  1  (bx  cx 2  1) dx

 1 1 
1
2 2

0 
2 2
Extremize with respect to b and c

  b  2cx  bx 2  cx3  2 x dx   24b  9c  12  0

db 0
dI
1

And

  2bx  4cx 2  bx 3  cx 4  2 x 2 dx   45b  68c  40  0

dc 0
dI
1

By Solving these equation,

b=0.37037, c=0.34568. Thus the approximate solution is

  1  0.37037 x  0.34568 x 2
It gives,  (0.5) =1.2716  (0.75) = 1.472225  (1) =1.71169

ANS 3.12
 d 2u
  cu  x 2  0 0  x 1
dx 2
du du 4
 1, and 
dx x 0 dx x 1 3
d 2u
 cu  x 2  0
dx 2
in varitional form
 d 2u 
I=   2  cu  x 2  udx
1

0  dx 
int egrating by parts

 u   ( u ) dx   cu udx   x 2 udx

du d du
1 1 1 1

dx 0 0
dx dx 0 0

 u      dx    c u 2 dx   x 2 udx
du 1  du  1
1 1 2 1 1


dx 0 0
2  dx  0
2 0

  u   u x 0      dx    cu 2 dx    x 2 udx
4 1  du  1
1 2 1 1

3 x 1 0
2  
dx 0
2 0

4  1  du 2  
   u  u x 0        c u 2  x 2 u  dx 
1
1

 3 x 1 0  2  dx  2  
var itional form
 1  du  2 
u  u x 0       c u 2  x 2 u  dx
4 1
1
I
3 x 1 0  2  dx  2 
Let us take u=a+bx+ex 2
du
 b  2ex
dx
now putting the boundry condition
du
 1, b=1
dx x 0

du 4 4 1
 , =1+2e , e=
dx x 1 3 3 6
x2 du x
u=a+x+ , 1
6 dx 3
1
4 x2   x2  c x2  2  
 a+x+    a+x+     1 
 1 x x2
2 2

I    a+x+   x  a+x+  dx

3 6  x 1  6  x 0 0  2  3 2 6   6  

By Ritz method
I
0
a
  x2  2 
     c  a+x+   x dx  0
1
1

 3 0   6  
1   x2 x3  x3 
1

 c ax      0
3   2 18  3 
0

c  0, on solving we get a=  0.555

x2
u  0.555  x 
6