Testing framework for variable k_in

[delongchamp]

The CyRSoXS library that is part of the NRSS has an untested feature to vary the k-vector into the scattering simulation, which should have an effect similar to tilting the sample in experiments as is done for critical-dimension RSoXS.

This feature should be tested. A preliminary suggestion is to create a 512x512x512 or 1024x1024x1024 morphology of a single sphere with lamellar stripes. The morphology could simulated with either rotating the morphology using the normally incident k (the only supported k in CyRSoXS currently):

or by rotating the k-vector:

The results should quantitatively match, within the limits of interpolation algorithms used for the tilting process in the former.

The first step would be to test this with composition scattering. Next would be to test this with added orientational scattering to verify that variable k handles the tilt of the electric field correctly.

Additional comparisons on more conventional CDSAXS style morphologies could be made to analytical descriptions of scattering composed by @dsunday in MatLab and some work done by Christopher Liman using rigorously coupled wave.