improvement: normalisation of scattering matrices

[aidancrilly]

The current implementation of calculating scattering matrices is not normalized as it uses pointwise energy-in to energy-out. This is typically an issue only if wide energy ranges are used. Can use same approach as Minotaur (Heaviside weighting functions) as an option (more expensive).

[aidancrilly]

Using method from here:

http://www.iaea.org/inis/collection/NCLCollectionStore/_Public/20/032/20032072.pdf?r=1