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!

Help with the coefficients of a fourier series

  1. Oct 22, 2011 #1
    1. The problem statement, all variables and given/known data
    This isn't really homework, just research I'm doing for myself...

    we start by defining:
    [tex]\sin_\ell(x) = (1 - \frac{1}{1+\tan^\ell(x)})^{\frac{1}{\ell}}\,\,\,for\, \,\,x \in [0, \pi][/tex]

    for [itex]\ell[/itex] divisible by two, or [itex]\frac{1}{\ell}[/itex] divisible by two.

    [tex]\sin_\ell(x + \pi) = -\sin_\ell(x) = \sin_\ell(-x)[/tex]

    and then for:

    [tex]a_n = \frac{1}{\pi} \int_{-\pi}^{\pi} \sin_\ell(t) \cdot \cos(n\cdot t) dt[/tex]

    [tex]b_n = \frac{1}{\pi} \int_{-\pi}^{\pi} \sin_\ell(t) \cdot \sin(n\cdot t) dt[/tex]

    [tex]\sin_\ell(x) = \sum_{n=1}^{\infty} [a_n \cdot \cos(n\cdot x) + b_n \cdot \sin(n \cdot x)][/tex]

    2. Relevant equations

    what is [tex] b_n = \frac{1}{\pi} \int_{-\pi}^{\pi} \sin_\ell(t) \cdot \sin(n\cdot t) dt[/tex]

    3. The attempt at a solution

    Well, all I've really managed to deduce is that [tex]a_n[/tex] is always zero, this was found through the odd property of [tex]\sin_\ell(x)[/tex]

    so we get
    [tex]\sin_\ell(x) = \sum_{n=1}^{\infty}b_n \cdot \sin(n \cdot x)[/tex]

    plugging in [itex]x = \frac{\pi}{2}[/itex] we get:

    [tex]\sum_{n=0}^{\infty} b_{2n + 1} \cdot (-1)^{n} = 1[/tex]

    otherwise, I have no clue how to tackle this integral. To me it seems insurpassably complicated. Perhaps someone can simply point me towards the correct integration technique. Or perhaps there is an alternative way to find the fourier coefficients that don't involve evaluating this clunky integral.

    This is only a small part of the much bigger problem I'm analysing, so please don't think I'm just trying to find the easy way out. This part of the problem just goes beyond my schooling. Thanks, any help or suggestions would be greatly appreciated.

    edit:
    well I managed to reformulate:
    [tex]\int_{-\pi}^{\pi} \sin_\ell(t) \cdot \sin(n\cdot t) dt = \int_{-\pi}^{0} \sin_\ell(t) \cdot \sin(n\cdot t) dt + \int_{0}^{\pi} \sin_\ell(t) \cdot \sin(n\cdot t) dt[/tex]

    [tex]\int_{0}^{\pi} \sin_\ell(t-\pi) \cdot \sin(n\cdot t - n \cdot \pi) dt + \int_{0}^{\pi} \sin_\ell(t) \cdot \sin(n\cdot t) dt = ((-1)^{n+1} + 1) \cdot \int_{0}^{\pi} (1 - \frac{1}{1+\tan^\ell(t)})^{\frac{1}{\ell}} \cdot \sin(n\cdot t) dt[/tex]

    which means we can rewrite the equation as:
    [tex]\sin_\ell(x) = \sum_{n=0}^{\infty} b_{2n+1} \cdot \sin((2n+1)\cdot x)[/tex]

    or
    [tex]\sin_\ell(x) = \sum_{n=0}^{\infty} \sin((2n + 1)\cdot x) \frac{2}{\pi} \int_{0}^{\pi} \sin_\ell(t) \cdot \sin((2n + 1) \cdot t) dt[/tex]


    Actually, after plugging this into mathematica it does not return a algebraic result, which leaves me to believe this integral is NOT elementary. I'm a little inadequate at how to give the proof of this, but it seems sufficiently complex to not be elementary (save when [itex]\ell = 2[/itex])
     
    Last edited: Oct 22, 2011
  2. jcsd
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Can you offer guidance or do you also need help?
Draft saved Draft deleted



Similar Discussions: Help with the coefficients of a fourier series
Loading...