Is it possible to simulate a thick film in FDTD?
To eliminate the thin-film interference (in the visible spectral range), the thickness of a film has to be >hundreds of microns, but this is impossible due to constraints on calculation costs.
How do I simulate a thick film with nano-structure on both sides? The use of a TFSF or plane wave source is preferable.
Placing monitors just beyond the nano-structure will allow you to define a source in the far field for a second simulation, possibly using the same simulation, except with the source being injected from the opposite side. Depending on whether you are interested in the frequency or time domain, you may want to define a custom time signal using outputs from one simulation.
What is the slab material? Is it dispersive and/or lossy? This will change the time signal pulse shape and may require other techniques to determine; however, in frequency domain it is straight forward using transfer matrices or far-field projection operators.
I hope this answers your question.