Bessel’s function

The differential equation

2
d y dy
x 2 2  x  ( x 2  n 2 ) y 0 (1)
dx dx
where n is a non-negative constant, is called Bessel’s differential
equation of order n and its solutions are known as Bessel functions.
Bessel functions have turned out to have very diverse applications in
physics and engineering, in connection with the propagation of waves,
elasticity, fluid motion and especially in many problems of potential
theory and diffusion involving cylindrical symmetry. They also occur in
some interesting problems of pure mathematics.
1
 Bessel’s function
Solution of the Bessel’s Equation in Series
Bessel’s equation is

d2y 1 dy  n 2 
  1  y 0.
dx 2 x dx  x 2 

Let y
r 0
ar x k  r , a0 0


dy

dx
 
r 0
ar ( k  r ) x k  r  1

2
 Bessel’s function

2 
d y
And
dx 2
  r 0
ar (k  r )(k  r  1) x k r  2 .

dy d 2 y
Now putting the values of y, , 2
dx dx
in the Bessel’s equation, we have
 
k r  2 1

r 0
ar (k  r )(k  r  1) x 
x r 0
ar (k  r ) x k  r  1

 n2  

  1  2  ar x k r 0.
 x  r 0
3
 Bessel’s function

On simplification, we have
 


r 0

ar [(k  r )2  (k  r )  (k  r )  n2 ]x k  r  2  ar x k r 0
r 0
 
or 
r 0
ar [(k  r ) 2  n 2 ]x k r  2  
r 0
ar x k r 0. (2)

k 2
Equating the coefficient of x to zero which is lowest
2 2
a
power of x, we have 0 ( k  n ) 0.
Since a0 0 the indicial equation is 4
 Bessel’s function

2 2
 k  n 0 or k n. (3)
Now equating the coefficient of x k  1 to zero by
2 2
a
putting r = 1 in (2), we get 1 [( k  1)  n ] 0.
2 2
But (k  1)  n 0 since k n
hence we have a1 0.
k r
Equating the coefficient of general term, i.e., x to
2 2
zero in (2), we get ar 2 {( k  r  2)  n }  ar 0
5
 Bessel’s function

1
or a r 2  2
a.
2 r (4)
( k  r  2)  n

Since a1 0, therefore from (4) a3 a5 a7  0.

Again for k n , we have the following, from (4).
Case 1. When k = n.
1 ar
ar  2   a 
2 2 r (r  2)( 2n  r  2)
(n  r  2)  n
1 1
Putting r = 0, we get a 2   a 0   a 0 .
2( 2 n  2) 2.2(n  1)
6
 Bessel’s function
Again when r = 2, we have
1 1
a 4  a2  2
a0 ,
4.2(n  2) 2.4.2 (n  1)(n  2)

and so on.With these values of ai (i 0, 1, 2,  , )

and k the solution in the form of the series is
 2 4
x x
y a0 x n 1   2
 2 . 2( n  1) 2 . 4. 2 (n  1)( n  2)

r x 2r 
   ( 1) r r
 . -- (5)
2 .(r!)2 (n  1)( n  2)(n  r ) 
where a0 is an arbitrary constant
7
 Bessel’s function

Case II. When k =  n, then

1 1
a r 2  2 2
a r  ar
( r  2  n)  n (r  2)(r  2  2n)
1 1
When r = 0, we have a2  2.2(1  n) a0  2.2(n  1) a0

1 1
When r = 2, we have a4  4.2(2  n) a2  2.4.2 2 (n  1)(n  2) a0

and so on. Also a1 a3 a5  0

8
 Bessel’s function

Therefore putting the values of k and a0 , a1 , a2 , etc. in

y  x k (a0  a1 x  a2 x 2  )
 x 2
x 4

we have y  a0 x  n  1   2

 2.2(n  1) 2.4.2 (n  1)(n  2)
x 2r 
 r r   . (6)
2 .r!.2 (n  1)(n  2)  (n  r ) 
(5) and (6) are two solutions of Bessel’s equation

9
 Bessel’s function
Bessel’s function
1
If we put
a 0  n in equn. (5) , we get
2 (n  1)
 n  2r
r
( 1)  x
y 
r 0
 
r !(n  r  1)  2 
.
---- (7)

the equation (7) is called Bessel’s function of

first kind of order n and is denoted by J n (x), i.e.
 n  2r
r
( 1)  x
J n ( x) 
r 0
 
r !(n  r  1)  2 
. (8)
10
 Bessel’s function

Show that J  n ( x) ( 1) n J n ( x)

 n  2r
( 1)r  x
Solution J n ( x) 
r 0
 
r !(n  r  1)  2 
.

replace n by - n
 2r  n
r 1  x
J  n ( x)   ( 1)   (9)
r 0
r! ( n  r  1)  2 
 2r  n
r 1  x
J  n ( x)   ( 1)  
r n
r! ( n  r  1)  2 
11
 Bessel’s function

Since n is an integer ( n  r  1) is infinite

(and hence 1 / ( n  r  1) is zero) for r 0, 1,  (n  1)
Changing the variable of summation to m r  n so that
r m  n and m 0 when r n and m  when
 2( m  n)  n
1  x
r  J  n ( x) 
m 0
 ( 1) m n
 
(m  n)!(m  1)  2 

12
 Bessel’s function

 2 m n
1  x
 
m 0
m
( 1) ( 1) n
 
(m  n  1)m!  2 
 2r n
1  x
( 1)
r 0
n

( 1)
( r
r
 n  1) r!
 
 2 
( 1)n J n ( x),

(Changing the variable of summation from m to r while

keeping the limits of summation unchanged

13
 Bessel’s function

Formula I: xJ n nJ n  xJ n 1.

 r n 2 r
( 1)  x
Jn   
Proof We have r!(n  r  1)  2 
r 0
Differentiating above w.r.t. x, we get
 n 2 r  1
( 1) r (n  2r ) 1  x 
J n 
r 0
.  
r!(n  r  1) 2  2 
 n 2r  n 2r  1
r r
n ( 1)  x ( 1) r  x
   
x r 0 r!(n  r  1)  2 
 
r 0
. 
r!(n  r  1)  2 
14
 Bessel’s function
Multiplying both sides by x, we have
 n 2 r
r
( 1)  x
xJ n n 
r 0
. 
r!(n  r  1)  2 
 n 2 r  1
r
( 1)  x
x 
r 0
. 
(r  1)!(n  r  1)  2 

 n 2 r  1
r
( 1)  x
or xJ n nJ n  x 
r 0
. 
(r  1)!(n  r  1)  2 

15
 Bessel’s function
 s 1 n 2 s 1
( 1)  x
nJ n  x  .  {by putting(r  1) = s}
s 0
s!(n  s  2)  2 

 s n  2 s 1
( 1)  x
nJ n  x  .  nJ n  xJ n1 .
s 0
s!(n  1  s  1)  2 

Formula II: xJ n  nJ n  xJ n  1.
 r n 2 r
( 1)  x
Proof: We have Jn  
r 0
 
r!(n  r  1)  2 
Differentiating above w.r.t. x, we get
16
 Bessel’s function
 r n 2r  1
( 1) (n  2r ) 1  x 
J n   .  
r 0
r!(n  r  1) 2  2 

 n 2 r
( 1) r (n  2r )  x 
  xJ n  
r 0
. 
r!(n  r  1)  2 
 n 2 r
r
( 1) (2n  2r  n)  x 
 
r 0
. 
r!(n  r  1)  2 
 n 2 r  1 n 2 r
r 
( 1) 2(n  r ) x  x  r
  r 0
.  
r!(n  r  1) 2  2 
n 
r 0
( 1)  x
. 
r!(n  r  1)  2 
17
 Bessel’s function
 n 2 r  1
( 1) r (n  r )  x
x 
r 0
. 
(r!)(n  r )(n  r )  2 
 nJ n

 ( n  1) 2 r
r
( 1)  x
x 
r 0
. 
(r!)(n  1  r  1)  2 
 nJ n

 xJ n  1  nJ n
or xJ n  nJ n  xJ n  1.

18
 Bessel’s function

Formula III: 2 J n  J n  1  J n 1.

Proof: This is obtained directly by adding the recurrence
relations I and II. We give an alternative proof as follows:
 n 2r
r
( 1)  x
Jn  r 0
. 
(r )!(n  r  1)  2 
Differentiating w.r.t. x, we get
 n 2r  1
r
( 1) (n  2r )  x 
2 J n  r 0
. 
(r )!(n  r  1)  2 
19
 Bessel’s function
 n 2r  1
r
( 1) {( n  r )  r}  x 
 
r 0
. 
r!(n  r  1)  2 
 n  1 2 r  n 2
r r
( 1)  x ( 1)  x
  r 0
. 
(r )!(n  1  r  1)  2 
 
r 1
. 
(r  1)!(n  r  1)  2 
 n 1 2 j
j 1
( 1)  x
J n  1   j 0
. 
( j )!(n  1  j  1)  2 
taking j = r  1.

Or 2 J n  J n  1  J n 1. ,
This proves the result.

20
 Bessel’s function
Formula IV: 2nJ n  x( J n  1  J n 1 ).
Proof: This result is also directly obtained by subtracting
formula II from I. A direct proof is given below. We can
 n 2r
r
( 1)  x
write 2nJ n   r 0
(r )!(n  r  1)
(2n). 
 2
 n 2 r
( 1) r  x
 
r 0
( r )! ( n  r  1)
(2n  2r  2r ) 
 2 

21
 Bessel’s function

 n 2r  1  n
r r
( 1)  x ( 1)  x
x 
r 1
. 
(r )!(n  r )  2 
 x 
r 0
 
(r  1)!(n  r  1)  2 
 n  2 j 1
( 1) j 1  x
 xJ n  1  x 
j 0
 
( j )!(n  1  j  1)  2 

[taking j = r  1 in the second term]

 xJ n 1  xJ n 1.
Thus 2nJ n  x[ J n  1  J n 1 ].

22
 Bessel’s function

Formula V.
d n
dx
 n

x J n x J n 1

 n 2 r
r
( 1)  x
Proof: We have
J n  
r 0
 
r!(n  r  1)  2 
n
Multiplying both sides by x , we have

 2 n 2 r
r n
( 1) 2  x
n
x Jn  
r 0
 
r!(n  r  1)  2 
23
 Bessel’s function
 2 n 2 r  1
r n
or
d n
dx

x Jn    ( 1) 2 (2n  2r ) 1  x 
 
r!(n  r  1) 2  2 
r 0

 n  12 r
r

dx

d n

x J n x n 
( 1)  x
 
r!(n  1  r  1)  2 
r 0
d n

dx
x J n x n J n 1

24
 Bessel’s function

Formula VI. Show that

d n
dx
 n

x J n  x J n 1

 r n 2 r
( 1)  x
Proof: We have
Jn  
r 0
 
r!(n  r  1)  2 
 n
Multiplying both sides by x , we have
r 2r n 2 r

( 1) x  1
x  n
J n   
r 0 r!( n  r  1)  2 

25
 Bessel’s function
r 2r  1 n 2 r
d n ( 1) 2r x  1 


or 
x J n    
dx r 0 r!( n  r  1)  2 

 r n 2 r  1
d n ( 1)  x
 x J n  x 
n
 
dx r 0 ( r  1)! ( n  r  1)  2 

 k 1 n 1 2 k
d n ( 1)  x
 x J n  x 
n
 
dx k 0 k!( n  1  k  1)  2 

d n

dx
 
n
x J n   x J n 1

26
 Bessel’s function
Prove that
2 2
(i) J 1  sin x (ii) J 1  cos x
2
x 
2
x \
Proof: (i) We know that
 r n 2 r
( 1)  x
Jn  
r 0
. 
(r )!(n  r  1)  2 

xn x n 2 x n 4
 n  n
 2 n

2 (n  1) 2.2.2 (n  2) 2.4.2 .2 (n  3)
xn  x2 x4 
 n 1   2
 
2 (n  1)  2.2(n  1) 2.4.2 (n  1)(n  2)  ---- (1)
27
 Bessel’s function
1
In the above, putting n = 2 in (1) , we have
1/ 2 2 4
x  x x 
J1 / 2  1 / 2 3 1    
2 2  2.3 2.3.4.5 
1/ 2 3 5
x 1  x x 
 1/ 2 . x    
2 . 12 .12  x  2.3 2.3.4.5 
2  x3 x5   2 
 x       sin x.
x 3! 5!   x
28
 Bessel’s function

1
(ii) Putting n  in (1), we get
2
x 1/ 2  x2 x4 
J  1 / 2 ( x )   1 / 2 1 1    
2 2  2 2.3.4 
x1/ 2  x2 x4 
 1/ 2  1   
2 .12   2! 4! 
 1/ 2
x  2 
  1/ 2 (cos x)    cos x.
2 .   x
29
 Bessel’s function
Prove that
(i)  2   (3  x 2 ) 3 cos x 
J5/ 2     2
sin x  
 x x x 

(ii)  2   (3  x 2 ) 3 sin x 
J 5/ 2     2
cos x  
 x
  x x 
Proof: We know that 2nJ n  x[ J n  1  J n 1 ].
2n 2n
 J n J n  1  J n 1  J n 1  J n  J n  1 ---(1)
x x
1 1
putting n  in (1) , we get J 3 / 2  J1/ 2  J  1/ 2
2 x
30
 Bessel’s function
1 2 2  2   sin x 
J 3/ 2  sin x  cos x      cos x 
x x x  x x 

3
putting n  in (1) , we get
2
3
J5/ 2  J 3 / 2  J1/ 2
x

3  2   sin x   2 
 J 5 / 2      cos x     sin x
x  x x   x

 2   3 sin x 3 cos x 
 J 5 / 2       sin x 
 x x
2
x

 2   (3  x 2 ) 3 cos x 
 J 5 / 2  
 x
  2
sin x 
x
 proved.
  x 
31
 Bessel’s function
(ii) We know that 2nJ n  x[ J n  1  J n 1 ].
2n 2n
 J n J n  1  J n1  J n 1  J n  J n1 ---(1)
x x
1
putting n   in (1) , we get
2
1 1 2 2
J  3 / 2  J  1 / 2  J 1 / 2  J  3 / 2  cos x  sin x
x x x x

 2   cos x 
     sin x 
 x x 

32
 Bessel’s function
3
putting n   in (1) , we get J  5 / 2   3 J  3 / 2  J  1/ 2
2 x

1  2   3 cos x   2 
 J  5 / 2      3 sin x     cos x
x  x x   x

 2   3 cos x 3 sin x 
 J  5 / 2     2   cos x 
 x x x 

 2   (3  x 2 ) 3 sin x 
 J  5/ 2    
 x 2
cos x 
x
 proved.
  x 

33
 Bessel’s function
Generating function for J n (x).
n
When n is a positive integer J n (x) is the coefficient of t
x
(t  1 / t )
in the expansion of e 2 in ascending powers of t.
1 
x (t  1 / t )
n
i.e
e2   J n ( x)
t
n 
1  r
xt 1  xt 
Proof: We have e2    
(r )! 2  and
r 0

34
 Bessel’s function
x  s
 1   x
e 2t  
s 0
( s )!

 2t
 .

1  r  s
x (t  1 / t ) 1  xt  1  x 
Therefore, e2       
(r )! 2  s 0 ( s )! 2t 
r 0

  s r s
( 1)  x 
    tr s.
(r )!( s )! 2  (1)
r 0s 0
n
We want to collect coefficient of t from the right.
Let n = r  s, i. e., r = n + s. Now the limit
of s tends from 0 to infinity
35
 Bessel’s function
n
Thus the coefficient of t on the right of (9.16) is
 n 2 s  n 2 s
( 1) s  x  ( 1) s  x
 (n  s)!(s)! 2    
( s )! (n  s  1)! 2 
J n ( x)
s 0 0

as (n + s)! = (n + s + 1)
Again if we collect the coefficient of t  n from the right
of (16), the limit of r tends from 0 to infinity.
Thus coefficient of t  n on the right of (16)
 n  2r
( 1) n  r  x 
   ( 1) n J n ( x)  J  n (x )
r 0
( r )! ( n  r )! 2

1 
x (t  1 / t )
n

e2   J n ( x)
t
36
n 
 Bessel’s function

Bessel’s integral for Bessel’s function J n (x)

1
J n ( x)  0 cos(n  x sin  )d , where n is a +ve integer.

1 
x (t  1 / t )
e2   t n J n ( x)
Proof: We know that
n 
1
x (t  1 / t )
e2  J 0  (t  t  1 ) J1  (t 2  t  2 ) J 2  (t 3  t  3 ) J 3  (1)

i n in n  in
Let t e so tht t e and t e .
Then (1) gives
37
 Bessel’s function
1 i  i 
x (e  e )
i  i 2 i  2 i
e2  J 0  (e e ) J 1  (e e )J 2

 (e 3i  e  3i ) J 3   (2)

1 ni  ni 1 ni  ni
Since cos n  ( e  e ), sin n  ( e  e )
2 2i
Equation (2 ) may be re-written as
eix sin   J 0  2i sin  J1  2 cos 2 J 2  2i sin 3 J 3  

or cos( x sin n )  i sin( x sin  ) ( J 0  2 cos 2J 2  )

 2i (sin J1  sin 3J 2  ) (3)

38
 Bessel’s function
Equating real and imaginary parts in (3), we get
cos( x sin  ) J 0  2 cos 2J 2  2 cos 4J 4   (4)
sin( x sin  ) 2 sin J1  2 sin 3J 3  2 sin 5J 5   (5)
Multiplying both sides of (4) and (5) by cosn and sin n
and then integrating w.r.t. ‘’ from 0 to  and using the
following result
 
 / 2, m n

0

cos m cos n d  sin m sin n d 
0
0. m n

and we get,
39
 Bessel’s function
 0 , if n is an odd
0 cos( x sin  ) cos n d 
 J n , if n is an even -----(6)
and
  J n , if n is an odd
0 sin( x sin  ) sin n d 
 0 , if n is an even -----(7)
Adding (6) and (7), we have

0 [cos( x sin  ) cos n  sin( x sin  ) sin n ] d  J n

or 0 cos(n  x sin  )d J n
1
or J n ( x)  0 cos(n  x sin  )d Proved
 40
 Bessel’s function
Trigonometric expansions involving Bessel functions
To show that
(i) cos( x sin  )  J 0 ( x)  2 cos 2.J 2 ( x)  2 cos 4 J 4 ( x)  

(ii) sin( x sin  ) 2 sin .J1 ( x)  2 sin 3 J 2 ( x)  

(iii) cos( x cos  )  J 0 ( x)  2 cos 2 J 2 ( x)  2 cos 4 J 4 ( x)  

(iv) sin( x cos  ) 2 cos  J1 ( x)  2 cos 3 J 3 ( x)  2 cos 5 J 5 ( x)  

(v) cos x  J 0 ( x)  2 J 2 ( x)  2 J 4 ( x)  
(vi) sin x 2 J1 ( x)  2 J 3 ( x)  2 J 5 ( x)   41
 Bessel’s function
1 
x (t  1 / t )
n
Proof: We know that
e 2   t J n ( x)
n 
1
x (t  1 / t )
e2  J 0  (t  t  1 ) J1  (t 2  t  2 ) J 2  (t 3  t  3 ) J 3  (1)

i n in n  in
Let t e so tht t e and t e .
Then (1) gives
1
x ( e i  e  i  )
e2  J 0  ( e i  e  i ) J 1  ( e 2 i  e  2 i ) J 2
3i  3i
 (e e ) J 3   (2)
42
 Bessel’s function

Equation (2 ) may be re-written as

eix sin   J 0  2i sin  J1  2 cos 2 J 2  2i sin 3 J 3  

or cos( x sin n )  i sin( x sin  ) ( J 0  2 cos 2J 2  )

 2i(sin J1  sin 3J 2  ) (3)
Equating real and imaginary parts in (3), we get
(i) cos( x sin  ) J 0  2 cos 2J 2  2 cos 4J 4   (4)
(ii) sin( x sin  ) 2 sin J1  2 sin 3J 3  2 sin 5J 5   (5)

43
 Bessel’s function
(i) Replacing  by  / 2   in (3) and simplifying,
we have
cos( x cos  ) J 0  2 cos 2J 2  2 cos 4J 4   (5)
(ii) Replacing  by  / 2   in (4) and simplifying,
we have

sin( x cos  ) 2 cos J1  2 cos 3J 3  2 cos 5J 5   (6)

Replacing  by 0 in (5) and (6), we get
(v) cos x J 0  2 J 2  2 J 4   ----(7)
and
(vi) sin x 2 J1  2 J 3  2 J 5   ---(8) 44
 Bessel’s function
Orthogonality of Bessel’s function J n (x)

1  0, when  

 0
x J n (x) J n ( x)dx  1
 J
 2 n 1 ( ) 2
, when  

where  and  are roots of J n ( x) .

Proof: We know that the solutions of
x 2u //  xu /  ( 2 x 2  n 2 )u 0     (1)
and
2 // / 2 2 2
x v  xv  ( x  n )v 0     (2) 45
 Bessel’s function
are u J n (x) and v J n ( x) respectively.
v u
Multiplying (1) by and (2) by and then subtracting ,
x x
we get
// // / / 2 2
x(u v  uv )  (u v  uv )  (   ) xuv 0
d
x(u / v  uv / )  ( 2   2 ) xuv 0
dx
Now integrating on both sides

2
(   ) 2
1
0  /
xuvdx  x(u v  
/ 1
uv ) 0
46
 Bessel’s function
1 J n/ ( ) J n (  )  J n/ (  ) J n ( )
0 xJ n (x) J n ( x)dx  2
  2
  (3)

If  and  u J n (x) and v J n ( x are distinct roots of

J n ( x) 0 , then J n ( ) 0 and J n (  ) 0 . Therefore the
equation(3) reduced to
1
0 xJ n (x) J n ( x)dx o, when  

47
 Bessel’s function
0
when   , the right hand side of (3) is of the form
0
Its value can be obtained by considering  as root of
J n ( x) 0 and  as variable tends to  . Thus equation(3)
/
1 J n ( ) J n (  )
reduces to 0 xJ n (x) J n ( x)dx  Lt

2
  2

1 J n/ ( ) J n/ (  )   /
J n ( ) 
2

0
xJ n2 (x)dx  Lt
 2

2
--(4) by L’

Hospital’s rule
48
 Bessel’s function
From recurrence relation we have
xJ n nJ n  xJ n 1.
n
 J n ( )  J n ( )  J n1 ( )  J n ( )  J n1 ( )

Thus from (4) , we get
1   J ( )  2

0
xJ n2 (x)dx  n1
2 ,
Hence

1  0, when  

0
x J n (x) J n ( x)dx  1
 J
 2 n 1 ( ) 2
, when  
proved
49
 Bessel’s function

Prove that (i)

d
dx

xJ n J n1   x J n2  J n21 
(ii )
d 2
dx
 2

 n 2 n 1 2 
J n  J n 1 2 J n 
2 x
J n 1 .

d
Solution (i) L.H.S= dx ( xJ n J n1 ) J n J n1  x( J n J n1  J n J n1 )
 J n J n1  ( xJ n J n1  xJ n J n ) (1)
From recurrence formulae I and II for J n (x) are
xJ n nJ n  xJ n1 (2)
and xJ n  nJ n  xJ n 1. (3)
50
 Bessel’s function

Replacing n by n + 1 in (3) we get

xJ n1  (n  1) J n1  xJ n (4)
/
Putting the values of xJ n and xJ n/ 1 in (1) , we get

d
( xJ n J n1 )  J n J n1  (nJ n  xJ n1 ) J n1
dx
 J n { (n  1) J n1  xJ n } x 2
Jn  2
J n1  Proved

51
 Bessel’s function

(ii) L.H.S =
d 2
dx
 
J n  J n21 2 J n J n/  2 J n 1 J n/ 1.
1 1
2 J n . nJ n  xJ n1   2 J n1. [ (n  1) J n1  xJ n ]
x x
by (3) and (4)
 n 2 n 1 2 
2  J n  J n 1  R.H .S
x x  Proved

52
 Bessel’s function
2
1 a  4 2
Show that 0 x 3 J 0 (ax) dx  3 J1 (a )  2 J 0 (a )
a a
1
Proof: L.H .S  0 x 3 J 0 (ax) dx put ax  y , We know
d n
1
 4
a
 y  yJ 0 ( y ) dy
2

dx
 
x J n x n J n 1
a 0


1
a 4
y J ( y ) 
3
1
a 2
0  4
a
a

0
2
y J1 ( y ) dy n=1, d
dx
xJ1  xJ 0


J1 ( a ) 2  2
a
a

 4   y J 0 ( y) 0 
a 
   2 yJ ( y) dy
a

0
0


xJ 0 dx  xJ1

53
 Bessel’s function
J1 ( a ) 2 J 0 ( a ) 4 d
    yJ 1 ( y ) a
0
n=0, J 0   J  1  J1
a a2 a4 dx
a2  4 2 J (a)
 3 J1 (a)  02 =R.H.S
a a

J1 dx  J 0
Proved

54
 Bessel’s function
2  cos xt
Show that J 0 ( x) 
 0
dt
1 t 2
Proof: we know that
cos( x cos  )  J 0  2 cos 2J 2  2 cos 4J 4  

Integrating w.r.t  from 0 to 2
   
2 2 2 2
cos( x cos )d J 0 d  2 J 2 cos 2 d  2 J 4 cos 4 d  
0 0 0 0

55
 Bessel’s function

2

 cos( x cos )d  J 0  0  0   Put cos  t
0 2
20  dt 2 1 cos( xt )
 J 0  cos( xt )   dt proved
1 sin   0 1  t 2

56
 Bessel’s function
Bessel’s expansion for f (x)
Let f ( x)  A1 J 0 (1 x)  A2 J 0 ( 2 x)  A3 J 0 ( 3 x)     (1)
where 1 , 2 ,  3    are roots of J 0 ( x) 0 and 0  x  1
multiplying (1) by xJ 0 ( n x) and integrating w. r.t x from
0 to 1 , we get
1 1 1

xJ 0 ( n x) f ( x) dx  A1 xJ 0 ( n x) J 0 (1 x) dx  A2 xJ 0 ( n x) J 0 ( 2 x) dx    

0 0 0
1
   An xJ 02 ( n x) dx     (2)
0

57
 Bessel’s function

by orthogonal property
1  0, when  

 0
x J n (x) J n ( x)dx  1
 2  J n 1 ( ) , when  
2

of Bessel’s function , the equation (2) becomes

1
J12 ( n )
xJ 0 ( n x) f ( x) dx 0  0      An 2
0
2 1
 An  2  xJ 0 ( n x) f ( x) dx
J1 ( n ) 0

58
 Bessel’s function

Problem: If 1 , 2 ,  3    are the successive positive

2
 n  4

roots of J 0 ( x) 0 , show that x  3
2
J 0 ( n x)
n 1  n J1 ( n )


Proof: Let x 2  An J 0 ( n x) --- (1) ,where
n 1

2 1 2 1 2
An  2  xJ 0 ( n x ) f ( x ) dx  2  xJ 0 ( n x ) x dx
J1 ( n ) 0 J1 ( n ) 0

59
 Bessel’s function

1
Now , 0 x 3 J 0 ( n x ) dx put  n x  y , We know
d n

1

 n
y 2  yJ 0 ( y ) dy dx
 
x J n x n J n 1
 n4 0
1 3

 4 y J1 ( y )
n

0
n


2
 n4
n
0
2
y J1 ( y ) dy n=1, d
dx
xJ1  xJ 0


J1 ( n )
n

 n4
2
 y J ( y)2
0
n
0 
n
0  2 yJ 0 ( y ) dy  
xJ 0 dx xJ1
J1 ( n ) 2 J 0 ( n ) 4 n=0, d
    yJ1 ( y )0 n J 0   J  1  J1
n  n3  n4 dx

60
 Bessel’s function

 n2  4
 3 J1 ( n )
n

J dx  J
1 0

Since  n is the root of J 0 ( x) 0  J 0 ( n ) 0

Hence 2  n2  4 , putting the value of A in (1)

An  3 n
J1 ( n )  n


 n2  4
 x 2
2
3
J 0 ( n x)
n 1  n J1 ( n ) proved

61
Bessel’s function

62
 Series solution
Solution in series of the deferential equation

P0 ( x) y2  P1 ( x) y1  P2 ( x) y 0 -----------------(1)

Definition: If P0 (a) 0 then x = a is called an ordinary point of (1),

otherwise

Singular point: A singular point x = a of (1) is called regular if, when (1)
Q1 ( x) Q2 ( x)
is put in the form y2  y1  2 y 0 , where Q1 ( x ) and
x a x  a 
Q2 ( x) possess derivatives of all orders in the neighbourhood of a.

63
 Series solution
(i) When x = a is an ordinary point of (1), its every solution can be

expressed in the form y a0  a1 ( x  a )  a2 ( x  a ) 2          (2

(ii) When x = a is a regular singularity of (1), at least one of its solutions

can be expressed as

y ( x  a )c  a0  a1 ( x  a )  a2 ( x  a ) 2          
 (3)
 ar ( x  a ) r c , a0 0
0

The series (2) and (3) are convergent at every point within the circle of

convergence at a.
64
 Series solution

Working rule: P0 ( x) y2  P1 ( x) y1  P2 ( x) y 0 ---------------------(1)

Where P0 ( x), P1 ( x) and P2 ( x) are function of x and x=0 is a

regular singularity.

Step-1 Assume the solution of (1) be y  ar x r c

, a0 0 ---(2)
0

This is the method of Frobenious

2
2
dy d y
Step-2 Find and y2 and then put the values of y,
dy d
and 2 in (1)
dx dx dx dx
65
 Series solution
Step-3 Equate to zero the coefficient of lowest power of x it gives a

quadratic equation in c which is known as the Indicial equation.

Step-4 Find the roots of Indicial equation

Step-5 Equate to zero the coefficient of height power of x it gives a

equation in c which is known as the recurrence relation

Step-6 Find the values of a1, a2, a3 ---------

66
 Series solution

Depending upon the roots of the indicial equation, the following cases
arise:
(1) The roots of the indicial equation are unequal and not differing by
an integer i.e. their difference is not an integer.
(2) The roots of the indicial equation are equal.
(3) The roots of the indicial equation are unequal, differing by an
integer, yielding some coefficient infinite.
(4) The roots of the indicial equation are unequal, differing by an
integer, making some coefficient indeterminate.

67
 Series solution

The solution in series of the above cases are given below:

Case I when the roots of the indicial equation are different and

differing not by an integer.

If the indicial equation has two unequal roots c =  and c =  differing by

a quantity not an integer, we get two independent solutions by putting

these values of c in the series for y. The series solution is

y a( y )c   b( y )  , where a and b are arbitrary constant.

68
 Series solution
Example: Solve in series the differential equation
2x(1 x) y2  (1 x) y1 3 y 0 --------(1)
by the method of Frobenious.

Solution:
 r c
Let y   a r x , a0 0 --------(2) be a
r 0
solution of (1), then

y1   ar (r  c) xr c 1 ,
r 0
Cont 69
 Series solution

y2   ar (r c)(r c  1) xr c 2
r 0
Putting the Value of y, y1 and y2 in (1)

2x (1 x)  ar (r  c)(r  c  1) xr c 2
r 0
 
 (1 x)  ar (r  c) xr c 1 3  ar xr c 0
r 0 r 0
Cont 70
 Series solution

 
   2(r  c)(r  c  1)  (r  c) ar x r  c  1
r 0
 
   2(r  c)(r  c  1)  (r  c)  3 ar x r c 0
r 0
Equating the coefficient of lowest power of x (i,e
xc-1) to zero


 2c(c  1)  c ao 0  2c 2  c 0 as a 0 ,
0
 c 0, 1
2 Cont 71
 Series solution

c
Equating the coefficient of x to zero


 2c(c 1)  (c 1) a1   2c(c  1)  c  3 a0 0

 a1  2c(c  1)  c  3 a0 (c 1)(2c  3) a0

2c(c 1)  (c 1) (c 1)(2c 1)

(2c  3)
 a0
(2c 1)
Cont 72
 Series solution

 2(r 1 c)(r  c)  (r 1 c) ar 1

  2(r  c)(r  c  1)  (r  c)  3 ar 0

  2(r 1 c)(r  c)  (r 1 c) ar 1

 2(r  c)(r  c  1)  (r  c)  3 ar

Cont 73
 Series solution

2(r  c) 2  2(r  c)  (c  r)  3
 ar 1  ar
2(c  r)(r  c 1)  (r  c 1)

(r  c 1)(2r  2c  3) (2r  2c  3)
 ar  ar
(r  c 1)(2r  2c 1) (2r  2c 1)

when
2c  3
r 0, a1  a0
2c 1
2c  1 (2c  1)(2c  3)
r 1, a2  a1  a0
2c  3 (2c  3)(2c 1)
74
Cont
 Series solution
2c 1 (2c 1)(2c  1)(2c  3)
r 2, a3  a2  a0
2c  5 (2c  5)(2c  3)(2c 1)

(2c  1)(2c  3)
 a0
(2c  5)(2c  3)

Similarly
(2c  1)(2c  3) (2c  1)(2c  3)
a4  a0 , a5  a0
(2c  5)(2c  7) (2c  7)(2c  9)
and so on
Cont 75
 Series solution

Putting the values of a1, a2, a3 in (2)

c  2 3 4 
y x  a  a x  a x  a x  a x        
 0 1 2 3 4 

y xc[a  2c  3 a0 x  (2c  1)(2c  3) a0 x2

0 2c 1 (2c  3)(2c 1)

a x3  
(2c  1)(2c  3)
  ]
(2c  5)(2c  3) 0

Cont 76
 Series solution

when c = 0 we get
 3x 3 2 3 3 
y a0 1  x  x    au

 1 1.3 3.5 

Where a a when c 0 and

0
when c = ½ we get y a0 1 x bv

Where a0 b when c 1/ 2 and v x1 / 2 1 x

The general solution is y au bv Ans

77
 Series solution

Case II When the roots of the indicial equation are equal.

If the indicial equation has two equal roots c = , we get two independent

y
solutions by putting this value of c in the series for y and . The series
c

 y 
solution of (1) is y a ( y )c   b   ,where a and b are arbitrary
 c  c 

constant.

78
 Series solution
Example: Solve in series the differential equation
x y  (1 x) y  2 y 0 -------------------------------1)
2 1
by the method of Frobenious.
 rc
Solution: Let y   ar x , a 0 ---(2) be a solution of (1)
r 0 0

y   ar (r  c) xrc 1 ,
1 r 0


y   ar (r  c)(r  c  1)xrc 2
2 r 0

Cont 79
 Series solution

Putting the Value of y, y1 and y2 in (1)

   
x  ar (r  c)(r  c  1) x r c 2  (1 x)  ar (r  c)xr c 1
r 0 r 0

2  ar xrc 0
r 0
 
   (r  c)(r  c  1)  (r  c) ar x r  c  1
r 0
 
   r  c  2 ar x r c 0
r 0
Cont 80
 Series solution

Equating the coefficient of lowest power of x (i,e xc-1 ) to zero


 c(c  1)  c ao 0  c2 0 as a 0 ,  c 0, 0
0
Equating the coefficient of xc to zero


c(c 1)  (c 1) a   c  2 a 0

1 0
c 2
 a  a
1 2 0
(c 1)
Equating the coefficient of xr+c to zero

81
 Series solution


(r  c)(r  c 1)  (r  c 1) a   r  c  2 ar 0

r 1
r c 2
 a  ar
r 1
(r  c 1) 2
Putting r = 0, 1, 2, 3 ---------------------------

(c  2) (c  3) (c  2) (c  3)
a  a , a  a  a etc.
1 0 2 1 0
(c 1) 2 (c  2) 2 (c 1) 2 (c  2) 2

Cont 82
 Series solution

Substituting the values of a1, a2, a3, a4, in equ (2)

c  2 3 4 
y x  a  a x  a x  a x  a x      
 0 1 2 3 4 
 
(c  2) (c  2) (c  3)
a x2   
 
c
x ao   a x 

(c 1) 2 0 (c 1) 2 (c  2) 2 0
 
 
 
 
 32 2
y c 0 a 1 2x 
 x          au , where
0  2 2 
 1 2 

a a for c 0 and u 1 2x  32 x2         

0 12 22
Cont 83
 Series solution

 
y c (c  2)
 (c  2) (c  3) 2 
a x ln x 1  x x   
c 0 2 2 2
 
(c 1) 
 (c 1) (c  2) 



(c  2)  1 2  (c  2) (c  3) 

    x 



(c 1) 2 c  2
 c 1 
 2
(c 1) (c  2) 2 
 
 a xc  
0 



  1 1 2 2  2 
    x          

  c  2 c  3) c 1 c  2  


84
 Series solution

 
 y 
 2 2 3 2 



 a ln x 1  x x         

 c c 0 0



12 12 2 2


 
 2  1 2  2 3  1 1 2 2  2 
 a   x     x          
0  2  2 1  2 2

2 3 1 2

 
 1 1 2   
 y   13 2 



 a u ln x  a   3x  x           bv


 c  c 0
  0 0  4 

13 2  
b a and v u ln x   3x  x          

0 4

 

 The general solution is y au  bv . Ans

85
 Series solution
Case III Root of the indicial equation differing by an
integer (making coefficient of y infinite)
If the indicial equation has two unequal roots ,  (say  > )
differing by an integer and if some of the coefficient of y become
infinite when c = , we modify the form of y by replacing ao by
k(c  ).Then we get two independent solutions by putting c =  in
y
the modified form of y and . The series solution of (1) is
c
 y 
y a ( y )c   b   ,where a and b are arbitrary constant. The
 c  c 
result of putting c =  in y gives a numerical multiple of that
obtained by putting c = . 86
 Series solution
Example 3. Solve in series the differential equation
x2 y2  xy1  ( x2  4) y 0 ------------(1)
by the method of Frobenious.

Solution:

Let y   ar x r c , a 0 -------(2) be a solution of (1)
r 0 0

y1   ar (r  c) x r c 1 ,
r 0

y2   ar (r  c)(r  c  1) xr c 2
r 0
Cont 87
 Series solution

Putting the Value of y, y1 and y2 in (1)

   
2
x  ar (r  c)(r  c  1) x r c 2  x  ar (r  c) x r c  1
r 0 r 0

 ( x 2  4)  ar x r c 0
r 0
 
   (r  c)(r  c  1)  (r  c)  4 ar x r  c
r 0

  ar x r c2 0
r 0
Cont 88
 Series solution

Equating the coefficient of lowest power of x (i,e xc ) to zero

c(c  1)  c  4 ao 0



 c 2  4 0 as a 0 , 0  c 2,  2
Equating the coefficient of xc+1 to zero


c(c 1)  (c 1)  4 a 0  a 0

1 1
c+2
Equating the coefficient of x to zero
a

(c 1)(c  2)  (c  2)  4 a  a 0  a   0

2 0 2 c(c  4)
Cont 89
 Series solution

r+c
Equating the coefficient of x to zero


(r  c  2)(r  c 1)  (r  c  2)  4 a  ar 0

r 2
1 1
 a  ar   ar
r 2 2 (r  c  4)(r  c)
(r  c  2)  4

Putting r = 0, 1, 2, 3 ---------------------------

1 1
a  a a   a 0
2 c(c  4) 0 3 (c  5)(1 c) 1
90
 1 1
a   a  a etc.
4 (c  6)(2  c) 2 c(c  2)(c  4)(c  6) 0

Substituting the values of a1, a2, a3, a4, in equ (2)

y xc  a0  a1x  a2 x 2  a3x3  a4 x 4         

 

a  1 2 1 4 
 o
a x  a x 

 c(c  4) 0 c(c  2)(c  4)(c  6) 0 
xc  

 
1

 a x 6        
 c(c  2)(c  4)(c  6)(c  4)(c  8)
 0 


Cont 91
 Series solution

If we put c = - 2 in above equation , the coefficient becomes

infinite owing to the factor c+2 in denominator , so we shall

replace a0 by k(c+2). Then

k (c  2) 2 k
x4
 

 k (c  2)  x  

 c(c  4) c(c  4)(c  6) 
y xc 



 
k
x6       
 
  

 c(c  4)(c  6)(c  4)(c  8) 


Cont 92
 Series solution
 k 4 
 0  0  x 
  2( 2  4)( 2  6) 
y c  2 x  2 
 
 k 6 
  x  
  2( 2  4)( 2  6)( 2  4)( 2  8) 

 1 4 1 6 
y   2
kx   x  x     au
c  2  2 2 4 23 4 6 

Where a k for c  2 and

 1 4 1 6 
 2
u x   x  x    
 2 2 4 23 4 6 

Cont 93
 Series solution

 (c  2) 2 1 4 
 (c  2)  x  x 
y c  c(c  4) c(c  4)(c  6) 
kx ln x   \
c 1
 x6       
 c(c  4)(c  6)(c  4)(c  8) 

 (c  2)  1 1 1  2 
1       x 
 c(c  4)  c  2 c c  4  
 kxc  

 1  1 1 1  4 
    x   
 c(c  4)(c  6)  c c  4 c  6  
Cont 94
 Series solution

 y   2  1 2 1 4 
  ku ln x  kx 1  x  x     bv
 c  c  2  2 2 2
2 4 2 

 2  1 2 1 4 
b k and v u ln x  x 1  x  x   
 2 2 2
2 4 2 

 The general solution is y au  bv . Ans

95
 Series solution
Case IV Root of the indicial equation differing by an integer
(making coefficient of y indeterminate)
If the indicial equation has two unequal roots ,  (say  > )
differing by an integer and if some of the coefficient of y become
indeterminate when c = , Then we get two independent solutions
by putting c =  in the modified form of y and . The result of
putting c =  in y gives a numerical multiple of that obtained by
putting c = .

Cont 96
 Series solution

Example4 Find the series solution of the equation

2
d y
2 dy
(1  x ) 2  2 x  y 0 -------(1)
dx dx
Solution Let the series solution of (1) be
k 2 3
y x (c0  c1 x  c2 x  c3 x  ) (2)
Differentiating it twice w.r.t. ‘x’ we get
dy
c0 .kx k  1  c1 (k  1) x k  c2 (k  2) x k 1  c3 (k  3) x k 2  
dx
d2y k 2 k1 k
And 2
 c 0 k ( k  1) x  c1 ( k  1) kx  c2 ( k  2)( k  1) x 
dx
Cont 97
 Series solution
Putting these values in (1), we get
(1  x 2 )[c 0 k (k  1) x k  2  c1 (k  1)kx k  1
k
 c 2 (k  2)(k  1) x  ]
k1 k k 1
 2 x[c0 kx  c1 (k  1) x  c2 (k  2) x  ]
k k 1 k 2 k 3
 [c0 x  c1 x  c2 x    c3 x ] 0
k 2 k 1
or c0 k (k  1) x  [c1 (k  1)k ]x
k
 [c 2 (k  2)(k  1)  c 0 k (k  1)  2c 0 k  c 0 ]x
 [c3 (k  3)(k  2)  c1 (k  1).k  2c1 (k  1)  c1 ]x k 1   0

Cont 98
 Series solution
The indicial equation is given by equating the
coefficient of x k  2 to zero
i. e. c0 k (k  1) 0 or k 0, 1 as c0 0
Here the roots are different and differ by the integer 1. Now
equating the coefficient of x k  1 to zero, we obtain
c1 (k  1)k 0
When k = 0, c1 becomes indeterminate (0/0). But in this case we
get the identity c1.0 0, which is satisfied by every value of c1.

Cont 99
 Series solution
Therefore in this case we can also take c1 as arbitrary constant.
Now equating the coefficients of higher order terms of x to zero,
we obtain
c2 (k  2)(k  1)  c0 k (k  1)  2c0 k  c0 0
c3 (k  3)(k  2)  c1 (k  1)k  2c1 (k  1)  c1 0
--- --- --- --- --- --- --- --- --- --- ---
--- --- --- --- --- --- --- --- --- --- ---

Cont 100
 Series solution

k (k  3)  1 (k  1)(k  2)  1
c2  c0 c3  c1
or (k  2)(k  1) (k  3)(k  2)
(k  2)(k  1)  1 {(k  2)(k  1)  1}{k (k  3)  1}
c4  c2  c0
(k  4)(k  3) (k  4)(k  3)(k  2)(k  1)

(k  3)(k )  1 {(k  3)(k )  1}{(k  1)(k  2)  1} c

c5  c3 (k  5)(k  4)(k  3)(k  2)
1
(k  5)(k  4)
--- --- --- --- --- --- --- --- --- --- ---
--- --- --- --- --- --- --- --- --- --- ---

Cont 101
 Series solution

Substituting these values of coefficients in (2, we obtain

k {k (k  3)  1} 2 {(k  1)(k  2)  1} 3
y  x  c0  c1 x  c0 x  c1 x
 (k  2)(k  1) (k  3)(k  2)
{(k  2)(k  1)  1}{k (k  3)  1} 4
 c0 x
(k  4)(k  3)(k  2)(k  1)
{(k  3)k  1}{( k  1)(k  2)  1} 5 
 c1 x  
(k  5)(k  4)(k  3)(k  2)  (3)
Substituting k = 0 in (3), we get
Cont 102
 Series solution

0 1 2 1 3 1 4 1 5 
y  x  c0  c1 x  c0 x  c1 x  c0 x  c1 x 
 2 2 8 40 
 1 2 1 4   1 3 1 5 
c0 1  x  x   c1  x  x  x 
 2 8   2 40 
 1 2 1 4   1 3 1 5 
a 1  x  x   b  x  x  x 
 2 8   2 40  (4)
where a, b being arbitrary constants.
This is the required general series solution

103