Boundary conditions for multiple sources (normal and angled)


I am having problems in setting the boundary conditions. My simulation needs to have both normal-incident source and angled-source at the same time. Since I need to use Bloch boundary conditions, i am having problem with the k-values since i have two sources. Do you have any idea how can i apply boundary conditions in this case?

I will very much appreciate helping me with this.

multiple TFSF sources,with different illumination angle


In this case it could be possible to run two separate simulations with one of the sources in each simulation. Then only 1 k vector is needed in each simulation so there won’t be any conflict in the boundary conditions settings. You could then sum the fields from the two simulations to get the result as though both sources were injected at once.

This is demonstrated in the following example which shows how to simulate a circularly polarized plane wave source:

In the example, two methods are used to obtain the same final result. In one method two plane wave sources with orthogonal polarization and a 90 degree phase shift between them are included in the same simulation file. In the second method each source is simulated separately and the sum of the results is the same as the first method.


Thank you very much for your suggestion. I am actually doing a pump-probe experiment.
But in my case, the pump is incident normally and the probe is oblique, so the probe must sense the changes in the media due to the pumping.
What I am currently doing is inputting the final population levels into the second simulation ( the one with only oblique source).
But I think this is not enough, i think i should also be including the induced polarizations Pa and Pb, right? and how do you think this can be achieved?



In this case since the materials are nonlinear, it requires simulating both sources at the same time. Since the Bloch periodic boundaries can’t deal with two different source angles in one simulation, perhaps it would be possible to use PML boundaries to simulate a finite span of the structure and use two TFSF sources which can inject plane waves with a finite span:

Let me know if you think this would work.