Hello,
I ran into an issue with hitting a seg fault in OpenMC version 15.4 while running transport-dependent depletion. The seg fault would generally occur after 10+ inactive cycles, but the exact point was not consistently the same. I noticed that it seemed to happen after a particle crossed a surface. I have checked the geometry - I did not have overlapping cells, and I wasn’t getting any lost particle warnings. I ended up rolling back to OpenMC 15.3, and that version can run the model without any issues. Because my model is working in 15.3, this issue isn’t critical for my personal work - but I wanted to make a post here to see if an issue should be made, or if it’s an error on my end. Thanks!
Notes from investigating the issue:
While hunting for the solution, I found this post/issue ( Depletion with temperature interpolation causes segfault · Issue #2336 · openmc-dev/openmc · GitHub ) about using temperature: interpolation during depletion causes a seg fault. I tried an isothermal version of my model, explicitly setting only one material to depletable = True and the rest explicitly False, using temperature:nearest, a larger tolerance, a default temperature, and defining a range that included all temperatures in my model - but all of these still encountered a seg fault.
The model is of a pebble-bed reactor, so I wonder if it is an issue similar to the one noticed here: Segmentation fault on simple model after many particles , but this may have already been addressed.