On Feb 14, 2008, at 12:57 PM, Mani Chandra wrote: > When one wants to define an arbitrary wave guide with metallic > boundaries, why does one have to use a very large negative eps > instead of a very large positive eps?
Because that's what metal is: eps < 0, and a perfect metal is the limit eps -> -infinity. When epsilon < 0 (for positive mu), the index of refraction is purely imaginary and hence the fields are evanescent in the metal. For a very large positive epsilon, on the other hand, you still have propagating modes within the material. (However, if you let epsilon - > +infinity then the effect is very similar to a metal, because the Fresnel reflections at the interface with a bounded-epsilon material go to unity.) > Also when the boundaries of the computational cell have been set by > boundary_region pml, does using set_boundary to set the boundary > condition as metallic for a particular side of the computational > cell override the pml settings ? No, because PML is not a boundary condition, it is an absorbing material that happens to be adjacent to the boundaries. Steven _______________________________________________ meep-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
