First, thanks to Prof. Johnson for his advice regarding the Courant parameter (it worked!)
Now I have a general question about how MEEP sets up the dielectric function in space. I've been trying to simulate "infinitely thin" apertures ( i.e., one cell thick), and some strange things seem to be going on. I'm using a perfect metal for now, and I noticed that MEEP will not ramp epsilon from 1 in free space to infinity in the metal in the space of a single cell. Also, if I have a one-cell-thick slab of dielectric, epsilon won't even reach the value I input before it starts to drop again (if I want eps= 2.25, it might go 1-->1.15-->1.79-->1.15-->1 as you pass through the thin sheet).
I thought maybe enabling the sub-pixel averaging would help with this, but at least with my ideal metal, I got the error "Danger, singular tensor!!! inf" and the simulation stopped.
Is there any way to avoid these epsilon "gradients" or do I need to shy away from simulating features that are only a cell or two thick?
Thanks,
Bryce
Now I have a general question about how MEEP sets up the dielectric function in space. I've been trying to simulate "infinitely thin" apertures ( i.e., one cell thick), and some strange things seem to be going on. I'm using a perfect metal for now, and I noticed that MEEP will not ramp epsilon from 1 in free space to infinity in the metal in the space of a single cell. Also, if I have a one-cell-thick slab of dielectric, epsilon won't even reach the value I input before it starts to drop again (if I want eps= 2.25, it might go 1-->1.15-->1.79-->1.15-->1 as you pass through the thin sheet).
I thought maybe enabling the sub-pixel averaging would help with this, but at least with my ideal metal, I got the error "Danger, singular tensor!!! inf" and the simulation stopped.
Is there any way to avoid these epsilon "gradients" or do I need to shy away from simulating features that are only a cell or two thick?
Thanks,
_______________________________________________ meep-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
