Solution of x^2y''+xy'-y=0 and Regular Point of (1-x^2)y''-2xy'+2y=0

  • Thread starter Thread starter latentcorpse
  • Start date Start date
  • Tags Tags
    Point Regular
Join the discussion
Registration is free. Start your own thread to ask a follow-up.
4 replies · 2K views
latentcorpse
Messages
1,411
Reaction score
0
What is the general solution of

[itex]x^2y''+xy'-y=0[/itex]
i tried a series solution [itex]y=\sum_{n=0}^{\infty} a_n x^n[/itex]
and whitteld it down to

[itex]\sum_{n=0}^{\infty} (n^2-1) a_n x^n=0[/itex] but this isn't getting me anywhere






secondly how do i show that x=0 is a regular point of
[itex](1-x^2)y''-2xy'+2y=0[/itex]

the definition of a regular point is that :
x0 is a regular point of y''+a1(x)y'+a0(x)y=0 if the coefficients a1 and a0 can be expressed as convergent power series in (x-x0) for this value of x0.

so if i rearrange our equation into the desired form i get that

[itex]a_1(x)=\frac{-2x}{1-x^2},a_2(x)=\frac{2}{1-x^2}[/itex]
for x0=0, a1=0 and a2=2.

what do i say to round of my argument. something like:

"clearly 0 and 2 can be expressed as a convergent power series in x" or
"clearly a1 and a0 arent singular for this value of x0"
how do i round this off?
 
Physics news on Phys.org
hi latentcorpse - does you series solution imply all [tex]a_n[/tex] are zero except for a certain n?
 
A series solution is not the best way to do this. Are you required to use series?

If so, look closely at lanedance's suggestion. If not, try y= xm for some number m.

For your second question, remember that a geometric series, [itex]\sum r^n[/itex] has sum 1/(1- r).
 
ok if i was just using series solution, i see now that [itex]a_1[/itex] will be non zero - is there any way of finding its value or is the solution just going to be [itex]y=a_1x[/itex]?

i amen't required to use series solutions but it was a past paper exam question and i couldn't think of any other way to do it - why did you decide to use that substitution?



i don't see how the geometric sum fits into this last bit either.
 
You said you had [itex]\sum_{n=0}^{\infty} (n^2-1) a_n x^n=0[/itex].

For Every n, either [itex]a_n= 0[/itex] or [itex]n^2- 1= 0[/itex].

The reason I suggested xm is that (xm)'= mxm-1 and (xm)"= m(m-1)xm-1. so that x(xm)'= mxm and x2(xm)"= m(m-1)xm so every term has the same power of x.

Finally, since the sum of the geometric series [itex]\sum ar^n[/itex] is [itex]a/(1- r)[/itex], [itex]2/(1-x^2)[/itex] can be written as [itex]\sum 2(x^2)^n= \sum 2x^{2n}[/itex]. [itex]2x/(1- x^2)[/itex] is just that multiplied by x: [itex]\sum 2x^{2n+1}[/itex].