On Apr 2, 2008, at 8:26 AM, shiv chawla wrote: > Dear Steven and Meep Users > > If I define a periodic dielectric with period (0.24um) then, is it > necessary that the computation at the period boundary should take > place? ANd what epsilon shall it take at the boundary?
I'm not sure what your question is. When you tell Meep that the boundaries are periodic, it takes care of the boundary conditions for you. Just specify your structure in a periodic fashion (e.g. if you have a hole overlapping one boundary, you should also specify a periodic hole overlapping the other boundary). > I am asking this because initially my spatial step size was 0.05um > and all the simulation with varying dielectric constants remained > the same. I'm not sure what this sentence means. However, you should always output the epsilon.h5 file (run output-epsilon, as shown in the Meep tutorial) and look at the resulting dielectric file (with h5topng or whatever visualization program you like), to make sure you have defined the structure correctly. > Now I am doing it with 0.02um and the process has gone very slow. Is > this step right or i need to something else. Please do reply as I've > been trying this for long and i need results desperately. Usually you should start by running your computation at a low resolution (just enough pixels so that your structure looks crudely correct) and then keep doubling the resolution until the answers stop changing to whatever accuracy you want to achieve. Regards, Steven G. Johnson _______________________________________________ meep-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
