On Wed, 13 Jun 2007, adrian wrote: > I have one question about PML. I ran one case with 2-D waveguide > structure consisting of three layers. I know the correct description for > PML is just to set the thickness of PML and meep will automatically put > it inside the computation cell. However, the result was not expected. > The light just disappeared after short distance even I set long > waveguide length. However, if I add the thickness of PML in cell size, > the result showed light would propagate for long distance.
I think you have a misunderstanding here. When you specify PML, Meep puts it *inside* the computational cell. It does *not* add the thickness of the PML to the size of the computational cell. In your case, you have a waveguide propagating along the x direction, with a computational cell size of 2 in the y direction. If you then set the PML thickess to 1, it means that the PML completely covers the y direction - the whole computational cell is filled with absorbing material, and hence your waveguide mode decays rapidly. (In your actual file, you set the PML thickness to 3, which is thicker than the computational cell...I don't even remember what Meep does in this case, but again the computational cell will be filled with absorbing material.) You need to make the computational cell big enough so that the PML layers do not overlap your waveguide mode everywhere and cause it to decay where you don't want it to. One way to do this is to just add twice the PML thickness to your cell size, which is what you did in your second try. Steven _______________________________________________ meep-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
