Learning Forum

[NLev]

In my current project I am trying to simulate the excitation of lattice resonance modes in a sub-wavelength metasurface made of elliptical amorphous silicon nanowires, by using a Kretschmann configuration with glass as the substrate below the nanowires.It doesn't seem to be possible to run a stable simulation using BFAST sources at 45 degree angle of incidence, with the source placed within the glass - the fact that the index of glass is higher than that of the air between the a-Si nanowires seems to make FDTD think that the simulation would be unstable. nI've been trying to read about other solution involving the Krestchmann configuration, but they don't seem to fit well enough to my setup - I have a 3D instead of a 2D structure, and I would like to be able to analyze the reflectance for a certain bandwidth of wavelengths rather than analyzing for a single wavelength.nnAny advice would be appreciated.nNirn

[Kyle]

nIf the incident angle is beyond the critical angle there is no stable time step for BFAST. In this case you will have to use Bloch boundary conditions and a sweep of single frequency simulations. There is more information on the limitations of BFAST here:nn

[NLev]

Should there be any concerns regarding validity of results when using symmetric / anti-symmetric boundary conditions with a 45 degree plane wave source? nThe transmission monitor (technically a reflection in the simulation) shows a lattice response with a power of above (-)1, I believe I just need to increase the simulation time. The result I show below is from a frequency sweep as @kjohnson recommended.nn

[Kyle]

nThe symmetry boundary conditions shouldn't be used with an angled plane wave source. Instead, Bloch (for narrowband simulations) or BFAST (for broadband simulations) should be used. These pages have more information on these boundary conditions:nPeriodic boundariesnBloch boundariesnBFAST boundariesnn