Possible new entry point for NaN's in flux limiter version of the code

[davidjamesweir]

I think that there is a new entry point for NaNs to be generated in the code in the branches that proceed from the flux-limiters branch. The suspected entry point is in the calculation of r and phi in transport_Z_dir and tranport_E_dir. Currently investigating this in the branch NAN-INF-FL-check.

In the other branches this is not caught, and so will only cause a crash on creating a power spectrum.

[Issue created by DanielCutting: 2022-03-16]