Reflection / transmission larger than 1 for nanohole array


#1

Hello, sorry for restoring this thread.

Can anybody help me with this simulation? It is similar to the nanohole array.fsp provided on this link,
https://kb.lumerical.com/en/sp_nanohole_array.html

The only few changes i made are the period of the array which is 900nm, increasing the source of the wavelength to within 500-1500 nm, changing the position of the reflection/transmission monitors and increased the auto shutoff from 1e-5 to 1e-6; but i am getting unreasonable results. I have no idea where did i go wrong.

This is the simulation file. sp_array.fsp (273.9 KB)


transmission이 1 이상의 값을 가지는 경우
Transmission for finite number of cells
Steps of Creating a Compact model for a Taper Waveguide
Ring Resonator simulations with phase change material
#2

Anyone knows what did i not consider correctly?


#3

Hi @wxlim1,

The problem seems to be the PML absorption. The performance of the PML layers can be affected when you have a thin layer that requires a fine mesh in a small region of the simulation.

One solution is to increase the number of PML layers. I did this in the attached simulation file: sp_array_modFG.fsp (261.1 KB). I also used the steep angle profile, which is appropriate for periodic structures where you can get scattering in different directions (see this post for more information). Note that I used a slightly coarser mesh to speed up the simulation and reduced the number of frequency points to decrease the memory requirements.

The transmission and reflection after these changes look more reasonable now:


#4

Hi @fgomez,

May i assume that whenever my reflection/transmission seems way from the norm, the second step i should do is attempt to perform the simulation with a steep angle profile?

This is because it is not easy to tell if a simulation will experience the light passing through the PML boundaries at steep angle.


#5

Following my previous query.
How do i know what are the indications to use a steep angle profile?


#6

Hi @wxlim,

Sorry for not replying earlier. The steep angle profile is recommended when you have periodic boundaries and scattering at an angle. As you mentioned, it is not always easy to know if there will be scattering at a steep angle, so in general it is safer to use this profile whenever you have a periodic structure than can scatter light in different directions.