Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Partial Di erential Equations

Module 5: Bessel functions - I


Dr.rer.nat. Narni Nageswara Rao

Friedrich Wilhelm Bessel (1784-1846) was a distinguished Ger-


man astronomer and an intimate friend of Gauss. The two cor-
responded for many years. Bessel was the rst man to determine
accurately the distance of a xed star (the star 61 Cygni). In
1844 he discovered the binary (or twin) star Sirius. The compan-
ion star to Sirius has the size of a planet but the mass of a star;
its density is many thousands of times the density of water. It
was the rst dead star to be discovered, and it occupies a special
place in the modern theory of stellar evolution.

1 Bessel functions
1.1 Bessel's equation:

Consider the di erential equation


x2 y 00 (x) + xy 0 (x) + (x2 n2 )y (x) = 0; n  0 (1)
The above equation is known as the Bessel's equation, which occurs in many
physical problems. In the present example we restrict ourselves to the values
of n 6= 0; 21 ; 1; 32 ; 2; ::: i.e, when 2n is not an integer or zero.
Step:1 In the rst step, the power series
X1
y (x) = cm xm+r ; c0 6= 0
m=0

and its derivatives 1


X
0
y (x) = cm (m + r)xm+r 1 ;
m=0
 nnrao maths@yahoo.co.in

1
X1
00
y (x) = cm (m + r)(m + r 1)xm+r 2
m=0

are substituted into equation (1) gives


1
X 1
X X1
cm (m + r)(m + r 1)xm+r + cm (m + r)xm+r +(x2 n2 ) cm xm+r = 0
m=0 m=0 m=0

or 1 1
X X
cm [(m + r)2 2
n ]x +
m
cm xm+2 = 0
m=0 m=0
or
X1
c 0 (r 2 n2 ) + c1 [(r + 1)2 2
n ]x + [cm (m + r + n)(m + r n) + cm 2 ]xm = 0
m=0

Step2: gives

(i) r2 n2 = 0 INDICIAL EQUATION

(ii)
c1 [(r + 1)2 n2 ] = 0 (2)
and
(iii)
cm 2
cm = for m  2 Recurrence Relation (3)
(m + r + n)(m + r n)
Thus the roots of the indicial equation are r1 = n and r2 = n which
are distinct and do not di er by an integer due to our assumption that
n 6= 0; 12 ; 1; 32 ;   

Step:3 For the larger root r = r1 = n equations (2) and (3) give c1 = 0
and cm 2
cm =
m(2n + m)
Thus
c1 = c3 = c5 =    = 0;
and c0 c0
c2 = ; c4 = ;    etc:
2(2n + 2) 2:4(2n + 2)(2n + 4)

2
one of the solutions is, therefore, given by
 
( x2 )2 ( x2 )4
y1 (x) = c0 x n
1 +
1!(n + 1) 2!(n + 1)(n + 2)

Because of the importance of the Bessel's equation, the above solution with
a particular choice of c0 = [2n (n + 1)] 1 is denoted by Jn (x) and is known
as Bessel's function of the rst kind. Using the recurrence relation of the
Gamma function
(x + 1) = x (x)
we can write 1
X ( x2 )n+2m
Jn (x) = ( 1)m (4)
m=0
m! (n + m + 1)
Step:4 Repeating the process of step 3 with r = r2 = n; we will get the
second independent solution y2 (x) given by
 
( x2 )2 ( x2 )4
y2 (x) = c0 x n
1 +
1!(1 n) 2!(1 n)(2 n)

This solution with a particular choice of
c0 = [2 n
( n + 1)] 1
is denoted by J n (x) and is given by (4) with n replaced by n.
Step:5 Thus the general solution is given by
y (x) = AJn (x) + BJ n (x)
where A and B are arbitrary constants.

1.2 Series solution and Bessel Function of the First


kind

Bessel function of rst kind is given by


1  x n+2m
X 1
Jn (x) = ( 1)m
m=0
m! (n + m + 1) 2
1  x n 1  x n+2 1  x n+4
=
(n + 1) 2 1! (n + 2) 2
+
2! (n + 3) 2
(5)


3
In particular 2 4 6
x x x
J0 (x) = 1 2 + 2 2 2 +  (6)
(1!)2 (2!) (3!)2
and
x 2
1 x 2
5 x 2
7
J 12 (x) = 2 2 + + 
2
(3) 1! ( 25 ) 2! ( 27 )
r2  
2 x3 x5
=
x
x
3!
+
5!

r
2
= sin x (7)
x
It follows immediately from (5) and (6) that
Jn (0) = 0 for n > 0
and
J0 (0) = 1
It was shown earlier that if n 6= 0; 1; 2; 3; ::: then
X1 ( 1)m x  n+2m
J n (x) = 2 (8)
m=0
m! ( n + m + 1)
represents the second independent solution of the Bessel's equation.
Example: Determine the value of J 21 (x) from equation (8).

Solution:
1 3 7
(x) 2 ( x2 ) 2 ( x2 ) 2
J 21 (x) = 2 1 +
( 2 ) 1! ( 32 ) 2! ( 52 )

r  
2 x2 x4
=
x
1
2! 4!
+ 
r
2
= cos x
x
which is linearly independent of J 12 (x).
If can be veri ed that J 12 (x) does in fact satisfy Bessel's equation for n = 21 .
Thus r
2
J 12 (x) = sin x (9)
x
4
and r
2
J 12 (x) = cos x (10)
x
Example: Show that J n (x) = ( 1)n Jn (x) for n = 0; 1; 2; 3;   

Solution: we know that


1
X ( x2 ) n+2m
J n (x) = ( 1) m
(11)
m=n
m! ( n + m + 1)

If we put m = k + n, we would obtain


1
X ( x2 )n+2k
J n (x) = ( 1)k+n
k=0
(k + n)! (k + 1)
X1 ( x2 )n+2k
= ( 1)n ( 1)k
k=0
k ! (k + n + 1)

J n (x) = ( 1)n Jn (x) (12)


Thus for n = 0; 1; 2; 3;    ; J n (x) does not represent the second independent
solution of Bessel equation.

1.3 Recurrence relations

1)
xJn0 (x) = nJn (x) xJn+1 (x) (13)
= xJn 1 (x) nJn (x) (14)
Solution: Let us consider the relation
1  x n+2m
X 1
Jn (x) = ( 1)m
m=0
m! (n + m + 1) 2

Di erentiating the above equation w.r.t. 'x' gives


1
X (n + 2m)  x n+2m 1 1
0
xJn (x) = ( 1)m
m! (n + m + 1) 2
 2x (15)
m=0

5
or
1  x n+2m
0 X 1
xJn (x) = n ( 1)m

m=0
m! (n + m + 1) 2
1  x n+2m 1
X 1
+x ( 1)m
m=0
(m 1)! (n + m + 1) 2
1  x n+1+2m
X 1
= nJn (x) x ( 1)m (16)
m=0
m ! (n + 1 + m + 1) 2
xJn0 (x) = nJn (x) xJn+1 (x) (17)
equation (15) can also be written as
1 x n+2m 1 1 ( x2 )n+2m
m 2(n + m)( 2 )
X x X
0
xJn (x) = ( 1) n ( 1)m

m=0
m! (n + m + 1) 2 m=0
m! (n + m + 1)
X1 ( x )n 1+2m
= x ( 1)m 2 nJn (x)
m=0
m ! (n + m )
xJn0 (x) = xJn 1 (x) nJn (x) (18)
which proves equation (14).
2) Show that 2Jn0 (x) = Jn 1 (x) Jn+1 (x)
Proof: We know the recurrence formula
xJn0 (x) = nJn (x) xJn+1 (x)
and
xJn0 (x) = nJn (x) + xJn 1 (x)
by adding, we have
2xJn0 (x) = x[Jn 1 (x) Jn+1 (x)]
Hence
2Jn0 (x) = Jn 1 (x) Jn+1 (x)

3) 2nJn (x) = x[Jn 1 (x) + Jn+1 (x)]

6
Proof: We know that
xJn0 (x) = nJn (x) xJn+1 (x)
xJn0 (x) = nJn (x) + xJn 1 (x)
substacting, we have
0 = 2nJn (x) x[Jn+1 (x) + Jn 1 (x)]
Hence
2nJn (x) = x[Jn 1 (x) + Jn+1 (x)]
4) dxd [x n Jn (x)] = x n Jn+1 (x)
Proof:
d 1 J (x) + x n J 0 (x)
[x n Jn (x)] = nx n
n n
dx
= x n 1 [ nJ (x) + xJ 0 (x)]
n n

we know that
xJn0 (x) = nJn (x) xJn+1 (x)
d
[x n Jn (x)] = x n 1 [ nJn (x) + nJn (x) xJn+1 (x)]
dx
= x n 1 [ xJn+1 (x)]
= x n Jn+1 (x)
Hence
d
[x n Jn (x)] = x n Jn+1 (x)
dx
5) Show that dxd [xn Jn (x)] = xn Jn 1 (x)
Proof:
d n
[x Jn (x)] = nxn 1 Jn (x) + xn Jn0 (x)
dx
= xn 1 [nJn (x) + Jn0 (x)]
we know that
Jn0 (x) = nJn (x) + xJn 1 (x)

7
d n
) [x Jn (x)] = xn 1 [nJn (x) nJn (x) + xJn 1 (x)]
dx
= xn 1 [xJn 1 (x)]
= xn Jn 1 (x)
Hence
d n
[x Jn (x)] = xn Jn 1 (x)
dx

You might also like