Reflection from PML boundary condition


#1

Hi,
im trying to simulate a 1D multilayer system such as a Bragg mirror.
in order to get efficient simulations i chose 2D simulation and reduced it to a single mesh cell in one of the simulation dimensions so that i get effectively a 1D simulation. The boundary condition for the dimension in which i have a single mesh cell are of course periodic.
Unfortunately i noticed that im getting a non-negligible reflection (around 5-10%) from the edge of the simulation.
I tried using all possible types of PML settings including stabilized and angle, and even greatly increased the number of PML layers to ~500, but nothing solves the reflection problem.
This reflection strongly affects the simulation results and makes them completely unreliable.

Is there anything i can do to fix this problem?

Thanks.


#2

Dear @nimrodg

Sorry for any inconvenience.

I am not sure if this question is related to the other topic that you created recently, but can you please upload your simulation file a review?

One things that I can tell now is to check your mesh and make sure that you do not have a high mesh aspect ratio which can cause poor PML absorption. This is explained in the links below:

https://kb.lumerical.com/en/index.html?sp_sers.html
https://kb.lumerical.com/en/index.html?rf_methodology.html

Also, you can check RT of 4 layer stack KB application example as your starting point which uses a 1D simulations.

Please keep me updated and I will be glad to be of a help.

Thanks


#3

It is related.
i was wrong in assuming i have reflections from the boundary conditions. What appeared to me as reflectiones was actually flactuations caused by the Kerr medium used in the simulation. so now i need to understant why the Kerr medium doesnt behave as it should.


#4

Thanks for updating me. I have responded you on this post and we can continue our discussion there.

Thanks