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

Scaling Problem in Diffusion Equation

  1. Feb 21, 2013 #1
    Hello everyone,

    I have a question that is bothering me a bit. I would be happy if you could give an idea or tell me a specific point to look at.

    Lets say that we have an arbitrary function that obeys diffusion equation:

    f = f(η), here η is the scaling parameter for pde which equals to x/√t
    (Diffusion eq. becomes [itex]df/d\eta=-2\eta d^{2}f/d\eta^{2}[/itex] after scaling)

    As I understood, diffusion eq. becomes an ode after scaling and it contains all the possible solutions for x and t pairs in η. Can we say that f(η) is invariant with respect to x and t? or in other words say ∂f/∂t=0 or not? What kind of physical information can we get just by lloking at f(η)?

    Thanks in advance!
  2. jcsd
  3. Feb 22, 2013 #2


    User Avatar
    Science Advisor

    No, [itex]\partial f/\partial x[/itex] and [itex]\partial f/\partial x[/itex] are NOT 0 because [itex]\eta[/itex] is itself a function of x and t.

    [tex]\frac{\partial f}{\partial x}= \frac{df}{d\eta}\frac{\partial \eta}{\partial x}= \frac{1}{\sqrt{t}}\frac{df}{d\eta}[/tex]
    [tex]\frac{\partial f}{\partial t}= \frac{df}{d\eta}\frac{\partial \eta}{\partial t}= -\frac{x}{\sqrt{t^3}}\frac{df}{d\eta}[/tex]
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Threads - Scaling Problem Diffusion Date
A Solving an ODE Eigenvalue Problem via the Ritz method Wednesday at 10:17 AM
Scaling when solving Schrodinger equation numerically Jun 23, 2015
Scale Analysis Mar 16, 2014
Scaling Helmholtz equation Apr 16, 2013
Scaling parameters in central difference solution Feb 5, 2011