Peak broadening from experimental data

[nicriz]

Hi all,

I have a set of diffraction data for nanodiamonds (powder with sub-micrometer diamond crystals) and I would like to extend the plugin SANSND to overwrite the default coherent elastic scattering and use the experimental data instead. My data is the experimental equivalent of the default NCrystal plot of the cross-section as a function of angle and wl (where do I find the code that produces this?), example below for the current nanodiamond ncmat file:

My understanding is that the experimental data is different because it naturally takes into account the peak broadening from the finite size of the crystals in the powder, which is ultimately what I want to model.

As a first step, I would be happy to implement the experimental data as it is and check that I get the same data back if I run the original experiment with McStas+ NCrystal sample. The interaction with the existing SANS models above/below the original experiment wl range can come later.

My idea is to integrate the xs in angle, so that I get the xs as a function of lambda. Then when I need to sample mu I can sample it from the marginal theta distribution at a specific wl. I am not sure this is the smartest way though. Also in quantitative terms I am not sure this makes sense.

Any input is appreciated, thanks!