Problem with OpenMC Simulation Using Cubit Geometry – “More than 95% of external source sites sampled were rejected”

[fakhar]

I exported my geometry from Coreform Cubit to use in OpenMC, but when I run the simulation, I get repeated errors:
ERROR: More than 95% of external source sites sampled were rejected.
Please check your external source’s spatial definition.
Abort(-1) on node 0 (rank 0 in comm 0): application called MPI_Abort(MPI_COMM_WORLD, -1)
Caught signal 11 (Segmentation fault)
export dagmc “model.h5m” overwrite
Could you please advise what might cause this issue?
Is there a specific export tolerance or geometry cleanup step I should apply before exporting to ensure proper source sampling in OpenMC?

I’ve attached a screenshot of the terminal output for reference.

[makeclean]

Could you share your inputs, ideally the python, xml and h5ms that would be needed to reproduce the problem?