Deterministic neutron transport in FUSE

[kevinm387]

I've added some capabilities to do 2D deterministic neutron transport calculations in FUSE. I've created a new file, neutron_transport_actor.jl, that at this point is just a collection of functions that couple the FUSE dd to NeutronTransport.jl. I've made the capability to do neutron transport arbitrary concentric circles using concentric_circles(), a 1D midplane build using neutron_transport_1d(), and a 2D FUSE geometry using construct_2d() and run_2d(). These last two functions can produce meshes of FUSE geometries and maps of neutron flux like these:

I added two example notebooks as well for 1D and 2D simulations. @bclyons12 got my previous ones running, would you be willing to try running these too Brendan?

Unfortunately this still needs a good deal of improvement. Here's a list of outstanding things, roughly in order of importance:

• In simulations with many layers, I'm not getting converged results.
• After getting better convergence, this needs lots of V&V/benchmarking, probably against Monte Carlo results. This can be made easier using paramak_build to construct equivalent Monte Carlo simulations.
• The plasma source is evenly distributed throughout the inner layer, rather than in a realistic plasma shape.
• It is still too slow
• Including the first wall layer causes the meshing to hang because of several points that are very near each other. Currently I'm just ignoring the first wall layer.
• It currently assumes DT neutrons and doesn't distinguish DT vs DD
• This PR still needs to get merged. Additionally, I have a branch of NeutronTransport.jl that implements some speed enhancements that me and @bclyons12 have worked on.

Apologies for the long PR and for not having this in a better state prior to submitting. But since my time is running short at GA I just want to share as much as possible.