Non linearity when using hi intensity fields

[mojv]

Hi, I'm using the gold material that is included in the materials.py, that as it says in the file, it allows one to use wave from .24797 μm to 6.1992 μm. On the other hand, meep allows also to create a source with a high spectral content, so if for example we create a pulse with spectral content of about 600 nm, that will translates to a temporal width of the order of 1 fs.

As far as I understand, the default materials included in materials.py are linear, but a pulse of 1 fs would generate a great intensity, therefore the macroscopic polarization of the system will be nonlinear. does meep take that into account, or the only solutions is to no simulate a source a big spectral content.

[stevengj]

If you want nonlinear materials you need to set the chi3 or chi2 properties yourself. Otherwise it defaults to linear materials (regardless of whether this is physical).

However, a typically reason to create an extremely narrow pulse in Meep is to Fourier transform the output to obtain the broadband response, i.e. when you really just want the linear frequency domain response over a broad bandwidth. In this case, using linear materials is entirely appropriate.

Note also that whether "a pulse of 1 fs would generate a great intensity" depends also on the amplitude of the pulse. On a computer (unlike in a real experiment), you can use as small an amplitude as you like (and with arbitrary units).

In general, for serious work, I would recommend defining your own materials rather than using the materials library. That way you can use a minimal model (as few polarization fields as possible) to fit your gold parameters in your bandwidth of interest, and you can also include nonlinearities etc as desired.