On Aug 29, 2012, at 10:58 PM, Roman Zubov wrote: > in the release notes you write that now Meep supports anisotropic dispersive > materials. The Meep reference says that the sigma-conductivity tensor must be > symmetric. But as I know its not the case for magnetised plasmas. For example > in the cold collisionless plasma at a fixed frequency the susceptibility > matrix is Hermitian. Which implies that the Meep sigma-conductivity matrix > must be anti-symmetric (proportional to the imaginary part of the > susceptibility matrix). So it seems to be impossible to model such a medium > in a narrow bandwidth as you suggest.
Meep currently only supports reciprocal media, corresponding to symmetric tensors. You are talking about non-reciprocal/gyromagnetic media, which are not supported. (Also, the conductivity tensor may only be diagonal. However, the Lorentzian sigma tensor may have symmetric off-diagonal components.) > Am I get it right, it is still not possible to model an arbitrary medium in > Meep in this release? There are various materials that Meep still does not support. Non-reciprocal (gyromagnetic) media. Bi-anisotropic (chiral) media. Nonlocal media (materials in which an electric field at one point creates a polarization somewhere else, e.g. metals at small scales where you take ballistic charge transport into account). Various kinds of nonlinearities (anisotropic Kerr media, off-diagonal anisotropic second-order nonlinearities, saturable gain, ....). And one can come up with more examples... Truly "arbitrary" media is a tall order for any code, I'm afraid. --SGJ _______________________________________________ meep-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
