On Feb 14, 2008, at 11:08 PM, Mani Chandra wrote: > I'm confused in one aspect of pml. I understand that they are > artificial boundaries bascially used to truncate the computation and > that they facillitate the exponential decay of EM waves. But what if > I define a waveguide with metallic boundaries inside the > computational cell? Then the E fields get killed at the boundaries > of the waveguide anyways right?
Right, you don't need absorbing boundary layers in directions where the field can't go anyway. So, in your example, you don't need them along the sides of a metallic waveguide, since it is not possible to scatter fields in those directions anyway. On the other hand, you will probably need them at the *ends* of the waveguide, if you want to simulate a structure where the waveguide conceptually extends out to infinity. Steven _______________________________________________ meep-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
