Discussion Forum

Hi everybody. I am currently trying to set up a simulation on light transmission through a single hole in a gold film. Previously, I used a 3D model of 5 layers as periodic nanoholes arrays (attached picture 1), and set the incident light to be plane wave, y-polarized H-field, x-polarized E-field travelling down the negative z-direction. So by setting the side boundaries which are perpendicular to the H -field as PMC, and those perpendicular to the E-field as PEC, I can simulate for a periodic nanohole array. This model worked well for the total field formulation (harmonic propagation).

And now I would like to simulate for a SINGLE hole. So I think additional PMLs are needed to cover up my whole model, including the lateral dimensions. But I realized that I can't define the inner boundaries of PML as continuity, so my result was deemed to be wrong. Another WRONG way to do it is by using scattered field formulation (scattered harmonic propagation) without a port (attached picture 2) -- it is wrong because scattered field formulation only works well for the case where the scatterer (in this case the gold film) is completely enclosed by one type of medium.

Eventually, COMSOL support suggest me to do it in two separate models: First, run the simulation for the periodic array with the gold film but WITHOUT hole (by using the total field model with PEC and PMC), then import the solution as an initial wave setup for the same model (with hole) using the scattered field formulation with PMLs at all sides and scattering boundary condition.

So my question is: how to "import a solution to be the incident wave of a scattered field model"? Did anyone come across this method before? Or does anyone has a better suggestion? Your help is highly appreciated haha...

Thanks in advance!