Lesson 7

R. I.
Badran The Legendre’s Equation Mathematical Physics
The Legendre’s Equation

The Legendre’s DE is written as:
d2y dy
(1  x ) 2  2 x  (  1) y  0 ,
2
dx dx
2x (  1)
where  is a real constant, p( x)   and q ( x )  .
1 x2 1  x2
The point x = 1 represents a singularity (a regular singular point),
since (1-x2) p (x) is finite and (1-x2) q (x) is also finite. However, it
must be noted that x = 0 is ordinary point.
Now Frobenius method can be applied:
Put (  1)   and substitute the assumed solution


y   an x m  n into the DE to obtain:
n 0
 
 a (m  n)(m  n  1) x
n 0
n
m n  2
  an (m  n)(m  n  1) x mn
n 0
 
 2 a n (m  n) x m n
   a n x m n  0
n 0 n 0
which gives

a0 m(m  1) x m2
 a1m(m  1) x m 1
  an (m  n)(m  n  1)x m n2
n2
 ,
 [an (m  n)(m  n  1)  2an (m  n)  an ]x m n
0
n 0
Writing n+2 for n in the 1st summation to get:

a0 m(m  1) x m2
 a1m(m  1) x m 1
 [an 2 (m  n  2)(m  n  1)
n 0 .
 an (m  n)(m  n  1)  2an (m  n)  an ]x m n
0
R. I. Badran The Legendre’s Equation Mathematical Physics
Equating the coefficient of all powers of x to zero to obtain:
a m(m  1)  0 , [This is called the indicial equation that

determines the values of the index m. This equation gives the
roots m = 0 and m = 1].
a1m(m  1)  0 ,
(m  n)(m  n  1)  2(m  n)  
an2  an
(m  n  2)(m  n  1)
Case (1): m = 0
ao and a1 are arbitrary constants.
n(n  1)  2n  
 an2  an
(n  2)(n  1)
put (  1)   and take the values of n in order to get:
(  1) 2  (  1) (  2)(  1)

a2   a , a3  a1   a1 ,
2 3 2 3!
2  4  (  1) (  1)(  2)(  3)

a4  a2  a ,
43 4!
6  6  (  1) (  1)(  2)(  3)(  4)

a5  a3  a1
5 4 5!
 The general solution is:
(  1) 2 (  1)(  2)(  3) 4

y  a [1  x  x    ]
2! 4!
(  1)(  2) 3 (  1)(  2)(  3)(  4) 5
 a1[ x  x  x    ]
3! 5!
These two independent series solutions are called Legendre

functions.
Case (2): m = 1
Again as mentioned before, this gives a dependent solution.
Hence the general solution is same as mentioned above.
Now, for integral  , we are seeking a solution which converges at

x = 1. In addition we need a solution for  x  1.
The obtained general solution, at x = 1, will become a set of

Legendre polynomials when a set of values of integral  is given,
as follows:
a) For  =0 with x =1 we can show that the odd series diverges

1 1
(by using the ratio test) like (1      ) .
3 5
a n  2 n(n  1)  2n  (  1)
[Note: Try the ratio test  and start with
an (n  2)(n  1)
n=1, and then other odd n’s.].
b) For the same  but with even n, use the ratio test e.g. for n=2
we get the ratio a4/a2 not defined. All other even n, give non-
defined ratios as well. Only ao survives such that y = ao at x = 1.
Here, when y = 1  ao = 1. Thus the solution y =1 can be named
as a polynomial P ( x)  1 . This is Legendre polynomial P (x) for
  0.
For  =1 with x = 1, the ao series (even series) can be shown to be
divergent. But a1 series (odd series) will give y = a1x. Again when
y=1 and for x = 1 a1=1.
Thus P1 ( x)  x .
For  =2 with x = 1, the odd series diverges while the even series
23 2
becomes y  a [1  x    ] . Hence at x = 1 and y = 1  ao=
2!
-1/2.
1
 P2 ( x)  (3x 2  1) .
2
5
For  =3, only the odd series survives and we get y  a1 ( x  x 3 ) ,
3
3 5 3
for x = 1 and y = 1  a1= - 3/2.  P3 ( x)  ( x  x) .
2 3
Note: Other Legendre Polynomials can be left as exercise for the

1
students. Try to find P4 ( x)  (35 x 4  30 x 2  3)
8
1
and P5 ( x)  (63x 5  70 x 3  15 x) .
8
Remarks:
a. The second solution for each  which is infinite series at x= 1,
is convergent for  x  1. This latter solution is called a Legendre
function of second kind Q (x) . The functions Q (x) are not used
as frequently as P (x) .
b. For fraction (non-integral)  both solutions are infinite series
and again these occur less frequently in applications.
c. By solving the Legendre DE, we actually have solved what is
called the eigenvalue problem. That is, the values of  , namely,
0, 1, 2, and 3 are called eigenvalues and the corresponding
solutions P (x) are called eigenfunctions.

Lesson 7

Uploaded by

Copyright:

Available Formats

Lesson 7

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

Lesson 7

Uploaded by

Copyright:

Available Formats

R. I.

Badran The Legendre’s Equation Mathematical Physics

The Legendre’s Equation

Now Frobenius method can be applied:

Put (  1)   and substitute the assumed solution

Writing n+2 for n in the 1st summation to get:

Equating the coefficient of all powers of x to zero to obtain:

a m(m  1)  0 , [This is called the indicial equation that

ao and a1 are arbitrary constants.

put (  1)   and take the values of n in order to get:

(  1) 2  (  1) (  2)(  1)

2  4  (  1) (  1)(  2)(  3)

6  6  (  1) (  1)(  2)(  3)(  4)

 The general solution is:

(  1) 2 (  1)(  2)(  3) 4

These two independent series solutions are called Legendre

Again as mentioned before, this gives a dependent solution.

Hence the general solution is same as mentioned above.

Now, for integral  , we are seeking a solution which converges at

The obtained general solution, at x = 1, will become a set of

a) For  =0 with x =1 we can show that the odd series diverges

Note: Other Legendre Polynomials can be left as exercise for the

You might also like