Hi Matt,
I worked up some small diffusion problems in FiPy. These snippets of
code may answer your
question from your original e-mail. Yes? If not, do you mean
anisotropic diffusion? If so, then
the second snippet may answer your question.
Best,
Dan Lewis
**Snippet 1**
eqn1 = TransientTerm() == ImplicitDiffusionTerm(D11) +
D12*c2.getFaceGrad().getDivergence()
eqn2 = TransientTerm() == ImplicitDiffusionTerm(D22) +
D21*c1.getFaceGrad().getDivergence()
boundaryConditions=(FixedFlux(mesh.getFacesLeft(),0.0),
FixedFlux(mesh.getFacesRight(),0.0))
** Snippet 2**
# To build of off diagonal terms we need a vector pointing in the "j"
direction with magnitude
# of the gradient of "c" in the "i" direction. This python code does
this exactly.
dcdydx = c.getFaceGrad()._take((1, 0), axis = 1)
# NOTE: If you want a transient solution you build your equation
this way:
eqn1 = TransientTerm() == ImplicitDiffusionTerm(D) + (D12 + D12)
*dcdydx.getDivergence()
Dan Lewis
Assistant Professor
Materials Research Center, Room 110
Rensselaer Polytechnic Institute
110 8th Street
Troy, NY 12180
[EMAIL PROTECTED]
518-276-2297
http://www.rpi.edu/~lewisd2
