Results don't look periodic

[JamesR]

Hello,

I’ve got a system of rods here:

…that I want to model as an infinite array of rods. I’ve place Periodic BCs in the x and y directions, and PML in the z directions. The source is a standard plane wave. Mesh refinement in the middle

When I run the simulation the results don’t look periodic on the sides. It looks as if the E field doesn’t recognize that there should be another pillar to interact with directly to the left/right of it. Am I misunderstanding something in how these simulations work?

Thanks for any help

[aalam]

It looks to me that the extent of your simulation region is responsible for this non-periodic result. In the plot, we can see that the gap between the pillars is about 0.1 micron. However, at the edges you can see that the result extends almost 0.3 micron in both direction. In order to model the periodic nature of the pillar array, your simulation edge should be placed halfway between the pillar at the edge and the next one (i.e. 0.05 um away from the end pillar). Also, you can see that the substrate ends before the simulation edge as well.

In order to model the infinite periodic array, you should place the simulation volume in such a way so that it occupies a unit cell that can be repeated to create the entire array (which is not the case here). The simplest solution is to center your simulation region around the center pillar and then make the extent in x and y direction equal to the period (pillar diameter + gap). Let me know if this helped.

[JamesR]

Thank you! Yes this solved it, thank you very much.