Issue with lost particles from DAGMC geometry

User Support
1

I’m new to OpenMC, following @Shimwell 's tutorial, specifically a variant of this notebook.

I have created a SphericalShell geometry in paramak and exported it as DAGMC: spherical_shell_example.h5m. DAGMC checks on the geometry show that it has no leaky surfaces or volumes.

Material assignments work fine (inner sphere as a D-T plasma) and I am placing a source in the middle:

my_settings.source = openmc_plasma_source.FusionPointSource(fuel="DT", temperature=20000.0, coordinate=(0,0,0))

Unfortunately, OpenMC only simulates a single batch because 100% of particles are lost. I’m baffled - suggestions welcome, please!

2

Welcome to the community

Happy to take a look at this, are you able to post the full edited example.

All the best

Jon

3

Thanks - here goes…

import openmc
import openmc_plasma_source as ops  
import openmc_tally_unit_converter as otuc
import neutronics_material_maker as nmm

filedir = '/path/to/file/'
filename = 'spherical_shell_example.h5m'

bound_dag_univ = openmc.DAGMCUniverse(filename=filedir+filename).bounded_universe()
my_geometry = openmc.Geometry(root=bound_dag_univ)

mat1 = nmm.Material.from_library(name='DT_plasma').openmc_material
mat1.name = 'sphere-plasma'
mat2 = nmm.Material.from_library(name='tungsten').openmc_material
mat2.name = 'sphere-fw'
mat3 = nmm.Material.from_library(name='lithium-lead',temperature=900.0).openmc_material
mat3.name = 'sphere-blanket'
mat4 = nmm.Material.from_library(name='SS_316L_N_IG').openmc_material
mat4.name = 'sphere-shield'

materials = openmc.Materials([mat1, mat2, mat3, mat4])

tally1 = openmc.Tally()
material_filter = openmc.MaterialFilter(mat3)  # the blanket material
tally1 = openmc.Tally(name='blanket_heating')
tally1.filters = [material_filter]
tally1.scores = ['heating']

my_tallies = openmc.Tallies([tally1])

my_settings = openmc.Settings(batches = 1, particles = 100, run_mode = 'fixed source')

my_settings.source = ops.FusionPointSource(fuel="DT", temperature=20000.0, coordinate=(0,0,0))

my_model = openmc.Model(
    materials=materials, geometry=my_geometry, settings=my_settings, tallies=my_tallies
)

statepoint_file = my_model.run()

sp = openmc.StatePoint(statepoint_file)

heating_tally = sp.get_tally(name='blanket_heating')
heating_tally.get_values().sum()

tally = otuc.process_tally(
    heating_tally,
)
print(tally)
4

I changed the material filter to mat1 and got a score

I can’t get a score for materials 2, 3, or 4

My only thought is that the surfaces are not being created correctly and particles don’t leave material 1 for material 2.

I think the dagmc geometry might need meshing with smaller triangles the defaults could be reduced a bit ( min_mesh_size: float = 5,
max_mesh_size: float = 20)

Also it might be worth increasing the particle and batch count

This sort of geometry might be easier and quicker as a native openmc CSG geometry. It looks like nested spheres to me which openmc.Sphere can do very well.

5

I changed the mesh size to min 0.01 and larger batch sizes, but the problem still occurs: new geometry. Not just with this example geometry, but also others from paramak. Will try the native openmc geometry…

6

Sorry to hear the paramak is misbehaving, if you want to throw the paramak scripts on an issue on the paramak GitHub or here I can double check.

Do you need any help with the native geometry