Learning Forum

[w.k.sky.x]

Hello,

I am working on a structure, which has nanostructures (0.1um) on the surface, but the substrate has multiple stacked materials (~1um).  The simulation is not converging well under the autoshutoff level, as such, I cannot confirm its accuracy. 

Is there any suggestions on how I should work on multiscale simulation and has less autoshutoff level. Thank you!

[Guilin Sun]

This usually means the structure has strong resonance. you may try those two measures:

1: increase the simulation time significantly, and let the autoshutoff min to terminate the simulation. In case the result is not smooth, such as with ripples in the transmission/reflection, you will also need to reduce autoshutoff min

2: sometimes the strong resonance is not inside the bandwidth you desire. in this case you can set the source in time domain: use 100fs pulse length and 200fs offset. You can check the resonance using a time monitor and view the spectrum.

occasionaly, some resonance is caused by improper PML. If this is the case, you can increase the distance between the PML and the structure, and increase the number of PML layers, or use steep angle PML. Maybe you do not know if it is due to PML. You can try to increase PML layers and the distance can be half-wavelength or longer and see if it improves.

Please try those modifications.