OpenMC stop running for shielding calculation in a specific batch

Problem Description

I’m encountering an issue with a simple shielding calculation for a fusion neutron source. I’m sampling the thickness of the shielding layer and filtering in a material_tissue defined as an operator. The shielding is being compared with two different materials: one is simply water, and the other is high-density borated polyethylene.

Sampling works fine, but when I attempt to run a simulation for each material with a larger number of particles to visualize the dose map, I face a problem. The simulation works correctly for high-density borated polyethylene, but for water, it gets stuck in a specific batch depending on the seed defined.

One brute-force solution could be trying different seeds until it works or running simulations with a low number of batches and then combining the results. However, I find these approaches somewhat archaic.

Talking with @Shimwell I’ve tried recompiling OpenMC with the suggestion provided in this link, but it still doesn’t work. Any suggestions on what might be happening would be greatly appreciated.

Environment

  • OpenMC version 0.14.1 (I did not encounter this problem with version 0.13.3)
  • Cross sections: ENDFB-8.0-NNDC and TENDL-2019 (ENDF with TENDL where ENDF cross sections are not available).
  • Cases run both locally (12 CPU, 32 GM RAM) and on a cluster (128 CPU).
  • Not likely a particle-per-batch issue, as I’ve run more complex models with many particles successfully.

Material Definitions

Water

water = openmc.Material(name="h2o")
water.add_nuclide("H1", 2.0)
water.add_nuclide("O16", 1.0)
water.set_density('g/cm3', 1.0)

borated_polyethilene = openmc.Material(name="borated_polyethilene")
borated_polyethilene.add_nuclide("C12", 61.20, percent_type='wo')
borated_polyethilene.add_nuclide("H1", 11.60, percent_type='wo')
borated_polyethilene.add_nuclide("B10", 1, percent_type='wo')
borated_polyethilene.add_nuclide("B11", 4, percent_type='wo')
borated_polyethilene.add_nuclide("O16", 22.20, percent_type='wo')
borated_polyethilene.set_density('g/cm3', 0.94)

Model geometry

Pink: Cylindrical material_tissue
Red: Cylindrical Neutron source
Grey: Cylindrical Shielding
White: Vacuum cell with “vacuum” boundary conditions

Regarding this problem, I have checked by changing the cross sections library and it persists. So keeping the same cross sections library and performing a simulation with the same parameters and the same seed, version 0.14.1 presents this error and version 0.13.2 does the simulation correctly. If someone can reproduce it, I can share the .xml files. Thank you very much

@ASureda welcome to the community, and sorry to hear you’re having trouble with this model. Please do share your input files if you can so that someone here can try to reproduce the issue you are seeing.

Hi Paul, thank you very much for the reply. Could you give me your email and I will send you the model?