Discussion Forum

Hello, I am calculating the resonances of electromagnetic waves caused by periodic structures in metasurfaces. Specifically, I set periodic boundary conditions in the xy direction and compute the electromagnetic waves of light incident from the z direction. In the z direction, PML (Perfectly Matched Layer) is set up to ensure that light passing through the metasurface is absorbed. However, the problem is that reflections from the PML cannot be ignored. Due to the periodic structure and reflections from the PML, resonances occur that shouldn't inherently exist. In actual metasurfaces, there's no such thing as PML on the top and bottom. I've observed that changing the thickness or the number of layers of the PML significantly affects the absorption efficiency, but the influence of reflection has not been mitigated. How can I set up a PML that completely absorbs in the z direction while maintaining the periodic boundary conditions in the xy direction? Currently, I am using polynomial stretching. Rational stretching doesn't seem to work well in terms of coordinate transformation. I would appreciate any suggestions. Thank you.