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

Fourier transform of Langevin equation (integral cancellation problem)

  1. Apr 5, 2013 #1

    (To cut a long story short, can I cancel the integrals in Eq. 6 to leave me with Eq. 7?)

    I am trying to follow the method for modelling the motion of a tethered bead from a couple of papers ("Te Velthuis, A. J. W. et al. (2010) Biophys. J. 99 1292–1302" and "Lansdorp, B. M., & Saleh, O. A. (2012) Rev. Sci. Instrum. 83 025115"), but am getting stuck going from the Langevin equation (where γ is the friction coefficient and k is the tether stiffness):

    [1] [itex]\gamma \dot{x} (t)+kx(t)=F_{therm}[/itex]

    to the power spectrum:

    [2] [itex]P(\omega )\equiv \left |x^{2}(\omega )\right| =\frac{2\gamma k_{B}T}{\gamma ^{2}\omega ^{2}+k^{2}}[/itex]

    I've followed a few lecture handouts I've found online and think I understand the majority of the problem. From what I understand, I need to take the three components of the first equation and individually apply the Fourier transforms (all in the range ∞→-∞):

    [3] [itex]\dot{x}(t)=\frac{d}{dt}\left|\frac{1}{2\pi }\int x(\omega )e^{i\omega t}d\omega \right|=\frac{1}{2\pi }\int x(\omega ) i\omega e^{i\omega t}d\omega[/itex]

    [4] [itex]x(t)=\frac{1}{2\pi }\int x(\omega )e^{i\omega t}d\omega [/itex]

    [5] [itex]F(t)=\frac{1}{2\pi }\int F(\omega )e^{i\omega t}d\omega [/itex]

    If I substitute all these back into the original equation I get:

    [6] [itex]\gamma \frac{1}{2\pi }\int x(\omega ) i\omega e^{i\omega t}d\omega +k\frac{1}{2\pi }\int x(\omega )e^{i\omega t}d\omega = \frac{1}{2\pi }\int F(\omega )e^{i\omega t}d\omega[/itex]

    From the lecture handouts I've found, I should be able to rearrange this to get the following (or something similar):

    [7] [itex]\gamma i\omega x(\omega) +kx(\omega) = F(\omega)[/itex]

    Now, here's my question. Is it possible for me to cancel the parts of the integral that appear in all three components of in Eq. 6 (this is [itex]\frac{1}{2\pi}\int e^{i\omega t} d\omega[/itex]) to leave me with Eq. 7?

    Any help on this would be greatly appreciated, since I've been stuck on this for a couple of days now!


    p.s. I also seem to end up with a rogue i, which will become -1 when I square P(x). Have I done my Fourier transform equations wrong?
  2. jcsd
  3. Apr 5, 2013 #2
    Yes, but it is not the way I would do it.

    Instead, I would just take the Fourier transform of the original equation.

    instead of

    use this
    [itex]x(\omega )=\int x(t )e^{-i\omega t}dt [/itex]
    [itex]i\omega x(\omega )=\int x'(t )e^{-i\omega t}dt [/itex]
    [itex]F_{therm}(\omega )=\int F_{therm}(t )e^{-i\omega t}dt [/itex]

    the FT is linear:
    [itex]y(\omega ) + z(\omega )=\int (y(t) + z(t))e^{-i\omega t}dt [/itex]

    Just apply the FT to both sides original equation.
  4. Apr 5, 2013 #3
    As for the question about power spectrum, I think the equation is wrong. It should be
    [itex]P(\omega) = \left|x(\omega) \right|^2 = (x(\omega))^* x(\omega)[/itex]
  5. Apr 5, 2013 #4
    That's brilliant, thanks for your help.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook