Problem with Diffraction Grating Filter design

Hello everyone,

I am trying to replicate the diffraction grating filter response (at least the transmission) from the following paper (Fig. 2 of the paper) :
Particle swarm optimization and its application to the design of diffraction grating filters

Here is my lumerical file:

diffraction_grating_filter.fsp (275.2 KB)

Now I think I can get the correct wavelength at which my transmittance is lowest (reflectance is high) but my filter does not have a sharp cutoff as the one shown in the figure.

Can anyone help me in debugging as to what I might be doing wrong? Here is the transmission plot that I am getting:

transmission

Hi @sourangsu.banerji,

Thank you for the question. After looking at the paper and your simulation, I have a few suggestions to try:

  1. It seems to me like the paper uses a 2D simulation, you should do the same and use a 2D FDTD solver.
  2. To fully resolve the structure you should probably increase the mesh density. Use convergence testing to determine the correct mesh density to use.
  3. The width of your simulation region should match the periodicity of the structure, which I believe should be 0.3359 um instead of 0.40879 um.

Try those changes and let me know if it helps.

I have studied this paper and reproduced his results.
The following is my model, which contains all the script parameters, and defines the evaluation function. It can be optimized using the built-in PSO algorithm of FDTD.
GMR_test_FDTD_PSO.fsp (438.8 KB)
image

After doing some test (convergence testing as @kjohnson said ), I can get close results as the paper.


The evaluation function is based on the rms error function from the paper, but author does not tell us the exact wavelength points, so I determine them by myself.

2 Likes

Many thanks! Junyu

1 Like

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.