Using OpenMC for MGXS Creation, a question

[Jeremyb8707bigbrain]

Hey yall I was using OpenMC for my MGXS stuff so I can put that in Moltres and I had some questions regarding the methods to obtain certain constants.

Based on the moltres_xs page (https://smpark7.github.io/moltres/getting_started/moltres_xs.html), I see that FISSE, GTRANSFXS, NSF, and REMXS all have a non direct way of obtaining them, meaning you have to manipulate the numbers out of OpenMC to get them.

My first question, how would I exactly do that?
My initial thought was to just literally manipulate every single number with its corresponding number, but I want confirmation that is wrong/right. For example, GTRANSFXS, I have my scatter matrix and scatter probability matrix, would I literally just multiply based on my CSV file? Like row 2,1,1 from 1 matrix times row 2,1,1 from the other?

My 2nd question, for NSF creation, how do I divide by the flux? I am able to get the nu-fission constants but nowhere do I see flux, which makes sense because OpenMC calculates this stuff without flow or anything, so how would I get flux or do I just use nu-fission?

Thanks for any insight!

Also for anyone who does OpenMC, could I ask why my decay_constant file was not outputted?

energy_groups = openmc.mgxs.EnergyGroups("CASMO-4")
mgxs_library = mgxs.Library(reactor_model.geometry)
mgxs_library.energy_groups = energy_groups

mgxs_library.mgxs_types = [
    'chi', 
    'transport', 
    'scatter matrix', 
    'multiplicity matrix', 
    'absorption', 
    'fission', 
    'kappa-fission', 
    'beta', 
    'diffusion-coefficient', 
    'scatter probability matrix', 
    'nu-fission', 
    'inverse-velocity', 
    'chi-delayed', 
    'chi-prompt',]

mgxs_library.domain_type = 'material'
mgxs_library.domains = materials
mgxs_library.by_nuclide = False

delayed_groups = list(range(1,7))
decay_rate = mgxs.DecayRate(domain_type = 'material', domain = salt, energy_groups = energy_groups, delayed_groups = delayed_groups, by_nuclide = False)

mgxs_library.check_library_for_openmc_mgxs()
mgxs_library.build_library()

tallies = openmc.Tallies()
tallies += decay_rate.tallies.values()
mgxs_library.add_to_tallies_file(tallies, merge = True)

reactor_model.tallies = tallies
#statepoint_file = reactor_model.run(mpi_args = ['mpiexec', '-n', '6']) 

statepoint = '/workspace/moltensaltsim/Material_Data/OpenMC_Files/statepoint.100.h5'
summary = '/workspace/moltensaltsim/Material_Data/OpenMC_Files/summary.h5'

stpt = openmc.StatePoint(statepoint, autolink = False)
summ = openmc.Summary(summary)
stpt.link_with_summary(summ)
mgxs_library.load_from_statepoint(stpt)

outdir = "/workspace/moltensaltsim/Material_Data/MultiGroup_Libraries/2-Cm"
for domain in mgxs_library.domains:
    mat_name = domain.name

    for xs_type in mgxs_library.mgxs_types:
        mgxs_obj = mgxs_library.get_mgxs(domain, xs_type)

        mgxs_obj.export_xs_data(
            filename=f"{outdir}/{mat_name}_{xs_type}",
            format="csv",
            xs_type="macro"
        )

[smpark7]

moltres_xs.py automatically does all calculations for you. On that webpage you linked, Step 1 says

The user must include the openmc_xs.generate_openmc_tallies_xml() function in the OpenMC input file to generate a domain_dict.

Refer to readme and openmc files in https://github.com/arfc/moltres/tree/devel/tutorial/eigenvalue/openmc and https://github.com/arfc/moltres/tree/devel/tutorial/godiva for how to set up your openmc simulations and how to postprocess your openmc output.

[smpark7]

Also, https://arfc.github.io/moltres/ is the official up-to-date website. https://smpark7.github.io/moltres/ is where I test website changes before merging into the official website

[Jeremyb8707bigbrain]

Thank you so much I was missing that!
I already have my OpenMC setup to output all the stuff required and to turn them into txt files, but for quicker stuff in the future I will definitely use this!

[smpark7]

If you ever feel up to it, we have an open issue #319 to update moltres_xs.py to use mgxs.Library to create mgxs tallies rather than the current implementation of creating each tally individually.

[Jeremyb8707bigbrain]

Wait I can definitely help with that! Ill look at it when I get home, thats what I use for my tallies, its quite useful!