Kerr Nonlinearities - choosing the correct input values

[Nikolaos-Matthaiakakis]

edit: problem fixed, I edit the information here so that it can be used as an example.

Here I replicate https://doi.org/10.1364/OPEX.13.006587 which shows the power dependent transmission and reflection through a thin Kerr medium slab (red line is Transmission or T and blue line reflection or R)

For the generated image in the publication they have
E_0,average=1.927x10⁸ V/m
λ=1.53μm
d=λ (the thickness of the slab)
n=2.46
n₂=1.25x10^-18 m²/V²

In my meep simulation i use alpha=1μm, and follow https://meep.readthedocs.io/en/latest/Units_and_Nonlinearity/ to convert the input values.

For the source field amplitude:
Since alpha=1μm, I convert E_{0, average} to the field amplitude E₀=E_{0, average}/sqrt(2)=1.3626e+08 V/m or E₀=1.3626e+02 V/μm

Thus I adjust the current of a CW source to give me a field amplitude of E₀ at steady state

For chi3:

I convert n₂=1.25x10^-18 m²/V² to n₂=1.25x10^-6 μm²/V²

and then I do

χ₃= 4nn₂/3 giving χ₃= 4.1x10^-6um²/V²

thus

χ_3,meep=4.1x10^-6

Here I compare the results with an analytical code

The difference arises because the analytical code handles the problem linearly and completely ignores the energy that is transferred to the higher harmonics. A_meep is not really absorption but corresponds to the energy transferred to the higher harmonics. Results might not be perfectly converged.

example with pulsed source:
https://github.com/Nikolaos-MAtthaiakakis/share/blob/main/ATR_nl_par.ipynb
example with CW source:
https://github.com/Nikolaos-MAtthaiakakis/share/blob/main/ATR_nl_CW_par.ipynb

[Nikolaos-Matthaiakakis]

This is now answered (solution in original question)