Hi Daniel,Thanks for the fast response! Actually I should mention, the full problem I'm looking to simulate is a two-field coupled PDE -- voltage and temperature, with charge/energy flux, which would involve the anisotropic conductivity with anisotropic thermoelectric/thermal coefficients as well. Something like was done in this paper:
http://arxiv.org/abs/cond-mat/9806042 But I don't want to run slow Jacobi iterations like they did.
After I sent my email I found this bug report (yours?) on the FiPy repository:
https://github.com/usnistgov/fipy/issues/361 (and, your reply in this email long ago: http://permalink.gmane.org/gmane.comp.python.fipy/2840 )So it sounds like FiPy is not quite built to support the full problem. I have since been toying with writing my own sparse matrix population code just for the square grid, which would be very easy for isotropic but it does not seem quite trivial in the anisotropic case. :-) If you have some updates to FiPy along these lines, I'd be happy to be a bug tester.
(Indeed the problem I have in mind involves highly discontinuous internal boundaries, e.g., a sharp p-n junction in a magnetic field, but hopefully it will still be numerically well conditioned. At least, the matrix of transport coefficients should be positive definite and so it will still be a proper diffusive problem.)
Cheers, -Mark
smime.p7s
Description: S/MIME Cryptographic Signature
_______________________________________________ fipy mailing list [email protected] http://www.ctcms.nist.gov/fipy [ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]
