Small scatterers with large periodicity


I’m trying to simulate the periodic structure in the simulation file and measuring the reflection from it.
When the periodicity goes beyond a certain amount (340 nm), strange fluctuations and negative transmission appear in the reflection spectrum and simulation does not reach the autoshutoff level.

I decrease the mesh size and increase the simulation time but it sometimes results in diverging and the issue persists. I am interested in measuring the reflection spectrum for periodicity = 350-400 nm.

How can I eliminate this fluctuation? Reflection spectrum (for periodicity = 350, 400 nm) and simulation file are attached in the following.

Thank you in advance,

Al-AlO-Al-Si.fsp (238.6 KB)

Dear @soroosh

Your simulation file was set fine, but I found a few things that needed to be modified.

Most importantly, the FDTD span on the z direction is too small. We recommend to leave at least half the max wavelength (400nm here) between PML layer and edge of critical object (Al circle here).
I also used a mesh override over the circle disk to resolve it properly. At the end, I had to perform convergence tests to make sure that results are reliable.

After the above mentioned modifications, here is the plot if I choose periodicity to be 400nm:

As you can see, plots are smoother and negative transmission problem are resolved. For convergence testing, I ran simulations with finer mesh, and thicker PML layers. You can see that all the plots overlap.

Here are the results when periodicity is 350nm:

Same as previous plot, it does not have ripples and there is no negative reflection.
Please note that convergence testing is an essential part of every simulation and I strongly recommend you to continue working on the simulation file (see attachment below).
Al-AlO-Al-Si_Modified.fsp (241.9 KB)


Dear @bkhanaliloo,

Thank you for your time and consideration. I really appreciate your comments.