Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

More Fourier Series

  1. Nov 10, 2005 #1
    Find the Fourier series for f(x)=1 on the interval 0 <=x <= pi in temrs of phi = sin nx. By integrating thi series find a convergent series for hte function g(x) =x oin this interval assuming that the set {sin nx} is complete

    i can find the Fourier series for f(x) =1. But i would like to know why n is supposed to be odd becuase the answer i get is
    [tex] \frac{2}{\pi} \sum_{n=1}^{\infty} \frac{1 - \cos(n \pi)}{n} \sin(nx) [/tex]
    the text ssumes n is odd. Is it becuase if n was even then the summand would be zero?
    the answer in the text is [tex] \frac{4}{\pi} \sum_{n=1}^{\infty} \frac{\sin(2k-1) x}{2k-1} [/tex]

    How would i integrate the series?
    f i integrate the answer for the text i get [tex] \frac{-4}{\pi} \sum_{n=1}^{\infty} \frac{\cos(2k-1) x}{(2k-1)^2} [/tex]
    but hte answer in the book i
    [tex] \frac{-4}{\pi} \sum_{n=1}^{\infty} \frac{1-\cos(2k-1) x}{(2k-1)^2} [/tex]

    why or where is that 1 / (2k-1)^2 term coming from? Can someone please help?

    Thank you
  2. jcsd
  3. Nov 11, 2005 #2


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Precisely. Try plugging different values of n in cos(n pi). You will see that cos(n pi) = (-1)^n. Hence, for n even, cos(n pi)=1 and the summand vanishes. Now your sum is over all odd positive integers. But you can restore the sum to be over all positive integer by noting that the change of variable n-->2k-1 (with k running over the naturals) spans exactly every odd positive integers. This is what the book did.

    If you've seen how to prove the uniform convergence of Fourier series, do it. If not, assume it does converge uniformly such that the integral of the sum is the serie of the integrals.

    I think it comes from the constants of integration and then evaluating the serie at x=0, but I don't understand the mechanism exactly. It is true that setting one constant 0 and the other 1 / (2k-1)^2 makes the serie work when evaluating at x = 0, but how can we be sure that these are the right constant?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook