On Nov 23, 2009, at 9:11 PM, JiangjunZheng wrote:
In the meep tourial, there is a 2d exmaple of PhC waveguide. I
modified it for W1 PhC slab waveguide. It works well. But problems
come when it is applied for waveguides of different width, for
example W0.8.
To simulate the W0.8 waveguide, two methods are used:
1, use the triangular lattice, and shift the lateral holes (three
holes on each side) along the two in-plane axes.
2, use the rectangular lattice (three lines on each side). It is
easier to set the position of the lateral holes this way.
By using mpb-data and h5topng, it is checked that the waveguides
constructed under both lattices are the same.
However, after run-even, only half the guided modes are founded by
using the rectangular lattice (kx=0.5, kx along the waveguide
direction), these modes share the same line at x=a/2 (a is the
period) with Hz=0 and some images of Hz for other modes show some
lines along the x direction which make them not as smooth as other
images. For the W0.8 waveguide, it seems that the first method works
by comparing the field images with those of W1 waveguide.
I'm not sure what you mean about getting only half the guided modes.
If you have the right structure, you should get all the guided modes.
If what you want is a line defect in a triangular lattice with varying
width, you're right that the best way is to use a rectangular
supercell. It should look something like the one I described in this
message:
http://thread.gmane.org/gmane.comp.science.photonic-bands/1262/focus=1376
Steven
_______________________________________________
mpb-discuss mailing list
[email protected]
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/mpb-discuss
