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

Almost Reimann Integral

  1. Mar 25, 2009 #1
    I am trying to show that a certain sum rigorously tends to a Reimann integral, but I have a certain sticking point that you will find below.

    [tex]S = \sum_{n = 1}^{\infty} c_n \phi_n[/tex]

    [tex]
    \phi_n = \sqrt{\frac{2}{L}} \sin(\frac{n \pi x}{L})
    [/tex]

    In this case the c_n can be assumed to be such that the series converges. Basically I want to show that the limit of S as L goes to infinity is a Reimann integral by definition. We have:

    [tex]\lim_{L\to\infty}\sum_{n = 1}^{\infty} \sqrt{\frac{2}{L}} c_n \sin(\frac{n \pi x}{L})[/tex]

    consider the function sin(k), and define a partition of the positive half of the real line by the points n pi / L. Then the mesh of the partition is pi / L. To deal with the constants define c_n = c(n pi x / L) = c(k_n). Then we have:

    [tex]\lim_{L\to\infty}\sum_{n = 1}^{\infty} c(k_n)\sin(k_n) \sqrt{\frac{2}{L}}[/tex]

    This is almost a Reimann integral by definition, we have an infinite sum of values of a function multiplied by something small, and the mesh of the partition of points is going to zero, the problem is that the mesh of the partition is [itex]\Delta k_n = \pi / L [/itex], not the [itex]\sqrt{\frac{2}{L}}[/itex] that I have. The constant factor is obviously not a problem, but I cannot pretend that L and sqrt(L) are the same for this purpose. Is there anyway to recover, or does this 'almost integral' correspond to anything?
     
    Last edited: Mar 25, 2009
  2. jcsd
  3. Mar 27, 2009 #2
    Would it not be much easier to treat it like a Fourier series? :-) (Mind you, I haven't tried so I don't know)
    Anywho I got a little confused when reading because you went from talking about sums of [tex]\sin(n\pi x/L)[/tex] terms to [tex]\sin(n\pi/L)[/tex], i.e. the value of the whole thing at 1. Anyway, assuming [tex]k_n=n\pi x/L[/tex] define [tex]c(k_n)=c_n\sqrt{n}[/tex], then each term in the sequence is

    [tex]\frac{c(k_n) k_n^{-1/2}\sin(k_n)\sqrt{\pi x}}{L}[/tex]

    which should give you an integral.. proving its existence might be tricky.. but you pays your money you takes your choice.
     
  4. Apr 5, 2009 #3
    Try the textbook by Pinsky: Partial Differential Equations and Boundary-Value Problems
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook