1. Not finding help here? Sign up for a free 30min 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!

Integral arising from the inverse Fourier Transform

  1. Nov 14, 2015 #1
    1. The problem statement, all variables and given/known data

    I was using the Fourier transform to solve the following IVP:

    [itex] \frac{\partial^2 u}{\partial t \partial x} = \frac{\partial^3u}{\partial x^3} \\
    u(x,0)=e^{-|x|} [/itex]

    2. Relevant equations

    [itex]
    f(x) = \frac{1}{\sqrt{2\pi}}\int_{-\infty}^{\infty}\hat{f}(\omega)e^{i\omega x}d\omega
    [/itex]

    [itex]
    \hat{f}(e^{-|x|}) = \sqrt{\frac{2}{\pi}}\frac{1}{1+\omega^2}
    [/itex]

    3. The attempt at a solution

    After working out the PDE, I've gotten to the solution:

    [itex]
    u(x,t) = \frac{1}{\pi}\int_{-\infty}^{\infty}\frac{e^{-\omega^2 t}}{1+\omega^2}e^{i\omega x}d\omega
    [/itex]

    And, after doing the usual trick of completing the square, I've gotten the integral down to

    [itex]
    u(x,t) = \frac{1}{\pi}e^{-\frac{x^2}{4t}}\int_{-\infty}^{\infty}\frac{e^{-t(\omega-\frac{ix}{2t})^2}}{1+\omega^2}d\omega
    [/itex]

    But from here I have no idea how to proceed. Keep in mind that my professor often assigns problems by taking them from the textbook and tweaking random things without working through them himself. In a few of his assignments, he has made the problems unsolvable by doing so. If this integral has a closed form solution, I would really appreciate a hint on where to proceed from here. If it doesn't, then I would appreciate knowing just as much so I can bring it to my professor's attention.
     
    Last edited: Nov 14, 2015
  2. jcsd
  3. Nov 14, 2015 #2

    vela

    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper
    Education Advisor

    I didn't check that your solution so far is correct, though it looks like you messed up in completing the square. Your last expression doesn't look well defined for t=0.

    But assuming you end up with a similar integral, contour integration in the complex plane looks like a way to go.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted