I am modelling nano particles embedded in multilayer thin films. So far i have calculated the field inside the thin film stack analytically (using TMM theory), injected the results into COMSOL (scattered field formulation) and applied appropriate PMC/PEC boundary conditions. This approach works satisfactory for normal incidence.
Moving on to oblique incidence, it seems from the litterature that periodic FLOQUET boundary conditions are the ideal choice for my type of problems. I tried implementing them in COMSOL, but so far i can only get reasonable results in two cases:
1) No layers, e.g. just a column of air. In this case oblique incidence "works".
2) Normal incidence. In this case any number of layers "work".
I tried to create a MWE illustrating my problems. The MWE contains just an air-glass interface and for simplicity i have used full-field formulation with periodic ports*. I am well aware that the problem presented in the MWE can be solved analytically (and in lower dimensions in COMSOL), but i need 3D for futher calculations. A few comments on the mesh setup:
1) To resolve the field oscillations, a mesh resolution of 5 points per wavelength (material dependent) is used.
2) The mesh of the upper/lower zx (and zy) planes is forced identical (using the "Copy Face" function) to ensure that the periodic boundary conditions work as intended.
Any comments and/or good references on FLOQUET theory are much appreciated :)
* This choice allows COMSOL to setup the Floquet periodicity k-vectors automatically.
Moving on to oblique incidence, it seems from the litterature that periodic FLOQUET boundary conditions are the ideal choice for my type of problems. I tried implementing them in COMSOL, but so far i can only get reasonable results in two cases:
1) No layers, e.g. just a column of air. In this case oblique incidence "works".
2) Normal incidence. In this case any number of layers "work".
I tried to create a MWE illustrating my problems. The MWE contains just an air-glass interface and for simplicity i have used full-field formulation with periodic ports*. I am well aware that the problem presented in the MWE can be solved analytically (and in lower dimensions in COMSOL), but i need 3D for futher calculations. A few comments on the mesh setup:
1) To resolve the field oscillations, a mesh resolution of 5 points per wavelength (material dependent) is used.
2) The mesh of the upper/lower zx (and zy) planes is forced identical (using the "Copy Face" function) to ensure that the periodic boundary conditions work as intended.
Any comments and/or good references on FLOQUET theory are much appreciated :)
* This choice allows COMSOL to setup the Floquet periodicity k-vectors automatically.