Huazhong Wang wrote: > I have a question about resolution. In a meep code, the resolution is > 10, i.e. the space grid is 0.1. I make a cylinder with radius > 2/3=0.666 (see code below). What is the real radius of cylinder when > running the program, 0.6 or 0.7?. > > (define resolution 10) > ... > (make cylinder (center 0 0)(radius (/ 2 3))(height infinity)...) >
You have to remember that FDTD is using the Yee grid lattice where different field components lie on different positions within the grid. If eps-averaging? is set to false (it is true by default), the epsilon of each field component is determined by its location on the Yee grid to rounding error given by machine epsilon. In other words, either the electric fields are within the cylinder or not so this will depend on how your various geometric objects are arranged within the Yee grid lattice thus it is hard to say a-priori what the "radius" of your cylinder is. Also, you shouldn't rely on the hdf5 output of the dielectric since that is also using an interpolation scheme. If, on the other hand, eps-averaging? is turned on, sub-pixel averaging will tend to smooth the boundaries of the cylinder so as to exactly cancel the first order perturbation of the dielectric. Ardavan _______________________________________________ meep-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
