Hi all,
I am working on modelling a reactor in OpenMC and I import the geometry inputs through the cubit/dagmc method generating a “.h5m” file. Regardless the selected number of batch and particles per generation, I am observing high leakage fraction with no empty cells present that could explain the leakage. Investigating further, I used OpenMC-Plotter and noticed regions with overlapping surfaces in areas where only a single volume should be present. I have also defined a graveyard area bounding the DAGMC universe. While I have not explicitly set boundary conditions for the imported geometry (by default reflective boundary), I have assigned boundary conditions only to the outer layer of the graveyard region.
In the figure below, the volumes colored purple, pink, and yellow represent regions defined as single-material areas without any additional intersecting surfaces or volumes.
For the “.h5m” file generation in Cubit I follow “imprint and merge” process prior to overlap check. Cubit reports 7 overlapping surfaces (over about 20000 surfaces in total) but only two of these appear in overlapping areas identified from OpenMC-Plotter. In addition it should be noted that all of the overlapping surfaces indicated in Cubit are significantly smaller than the areas indicated in OpenMC-plotter. This issue persists regardless the selected mesh density.
Given that overlapping surfaces are presented in about 45 deg, could this be related with DAGMC artifacts issue addressed in the past?
I am using the OpenMC version 0.15.0 and latest version of OpenMC-Plotter on a Windows 10 WSL environment.
Any thoughts/insights into why OpenMC identifies overlapping surfaces in single-material/volume regions would be much appreciated!
Thank you in advance.