1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Series solution of an ODE

  1. Mar 9, 2013 #1
    1. The problem statement, all variables and given/known data

    Solve for [tex] y' = x^2y [/tex]

    3. The attempt at a solution

    There's something that's been really bothering me about this question and similar ones.

    We assume that the solution to the ODE will take the form

    [tex] y = \sum_{n=0}{a_nx^n} [/tex]

    After finding y', plugging in the expressions into the ODE, and distributing x2,

    [tex] \sum_{n=1}na_nx^{n-1} - \sum_{n=0}a_nx^{n+2} = 0 [/tex]

    I want to eventually combine the summations, so first I make the exponents match.

    [tex] \sum_{n=0}(n+1)a_{n+1}x^n - \sum_{n=2}a_{n-2}x^n = 0 [/tex]

    I need to make the indices match without manipulating exponents, so I just peel off the first two terms (n = 0 and n = 1) from the first summation. After doing so and combining the summations:

    [tex] \sum_{n=2}((n+1)(a_{n+1} - a_{n-2}))x^n + a_1 + 2a_3x = 0 [/tex]

    Now here's where I start becoming uncertain...

    The summation begins at n = 2, so the smallest power of x that can come out is x2. Since the RHS is 0,

    [tex] a_1 = 0 [/tex]
    [tex] 2a_3x = 0 [/tex]

    This means that the coefficients a_1 and a_3 must be equal to 0. Also, since the coefficients for all powers of x on the RHS are zero,

    [tex] (n+1)a_{n+1} - a_{n-2} = 0 [/tex]

    Rearranging gives me the recursion formula for n ≥ 2:

    [tex] a_{n+1} = \frac{c_{n-2}}{n+1} [/tex]

    My end goal in here is to write the sum in a way that doesn't involve any past terms so that there's no recursion. I start plugging in numbers and hope that I see some pattern...

    n = 2, [itex] a_3 = \frac{a_0}{3} [/itex]

    n = 3, [itex] a_4 = \frac{a_1}{5} = 0 [/itex]

    n = 4, [itex] a_5 = \frac{a_2}{6} [/itex]

    n = 5, [itex] a_6 = \frac{a_3}{7} = 0 [/itex]

    n = 6, [itex] a_7 = \frac{a_4}{8} = \frac{a_1}{5*8} [/itex]

    At this point it's clear that the odd n will give a zero and that I only have to worry about even n.

    What bothers me is the term I get when n = 4. I'm not given any information as far as I know about a2 and since the ODE is first-order, there's only one arbitrary constant which I assume is a0 (if I don't assume it's arbitrary then I have no idea what to say about a0 either). So I'm stumped. How do I write a non-recursive formula for the sum with the mystery number a2? What am I missing in here?

    Apologies for the lengthy post

    EDIT: Disregard the thing I said about odd and even n. n = 7 doesn't equal 0.
    Last edited: Mar 9, 2013
  2. jcsd
  3. Mar 10, 2013 #2
    I guess there's a typo there. You should have a1 = a2 = 0.
  4. Mar 10, 2013 #3


    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper
    Education Advisor

    Because of the error clamtrox pointed out, you had a3=0, and the n=2 equation then tells you a0=0. So a2 would be the arbitrary constant.

    Since you're really supposed to have ##a_1=a_2=0##, the non-vanishing coefficients are ##a_{3k}## (k=0, 1, 2, …), and a0 plays the role of the arbitrary constant.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted