eigenmode solution convergence problem with sievenpiper stru

Hi,

I am doing simulation on sievenpiper-like structure, ie: periodic structure with open to air. I put a airbox above my DUT. bottom of DUT is PEC. then I select the top face of airbox as PML. while I am not clear what parameter I should set for "uniform layer thickness" and "minimum radiating distance" (does this equal to my air box length?).
all I use is the default value. The other 4 sides is set as periodic boundary condition.
After simulation, it does not converge. evethough at some run it seems converge: ~3-4%. but next run gu up to 100%. (I set 20 times run) Also I find the eigenmode with larger complex number. I am confused... any suggestion appreciated!

hello,
i'm doing exactly that and having exactly the same problem. Did you ever manage to simulate the dielectric slab over conductor?

best wishes
msokey