home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cs.utexas.edu!sdd.hp.com!caen!destroyer!cs.ubc.ca!unixg.ubc.ca!jeffv
- From: jeffv@physics.ubc.ca (Jeff Vavasour)
- Newsgroups: sci.physics
- Subject: Need suggestions for numerical diffusion problem
- Date: 13 Jan 1993 00:15:53 GMT
- Organization: The University of British Columbia
- Lines: 30
- Distribution: world
- Message-ID: <1ivmvpINNiit@iskut.ucs.ubc.ca>
- NNTP-Posting-Host: physics.ubc.ca
-
- Hi, I've been working on a problem that involves diffusion in a potential
- V(x,y,z) that has the symmetry of a diamond lattice. My probability
- P(x,y,z,t) is given by
-
- __2
- K \/ P - V P = dP/dt
-
- which I would like to solve numerically for some smooth V with its values
- known only on some grid.
-
- This problem easily reduces to a one-dimensional problem in cylindrical or
- spherical symmetries, but I've been having some difficulty in the discrete
- symmetry of the diamond lattice. A general 3-D solution with appropriate
- boundary conditions would work, but a 100 interval 1-D numerical solution in
- the cylindrical or spherical case becomes a 100x100x100 3-D solution in the
- diamond symmetry requiring much too much computation time. There are also
- some problems with stability (not of the PDE solution, but of the
- self-consistent field that it is a part of).
-
- Does anyone know of any references that might provide clue for simplification?
- I would expect that there's probably some work out there where Schroedinger's
- equation was solved in this symmetry which have an adoptable method, but I
- haven't been successful in finding one. (Perhaps a Hartree-Fock treatment of
- the electron distribution in a diamond lattice?)
-
- Please e-mail me. Any help would be appreciated.
-
- - Jeff
-
-
-