Hi Paul,
I’ve been experiencing this problem myself (to a lesser extent than the first time I encountered it last year). Here is a sample script that reproduces the problem.
hex_test.py (3.1 KB)
There’s 3 very simple cases that are executed by the attachment. Tests 1 and 2 execute successfully. Test 3 results in multiple lost particle warnings until the maximum allowable number is reached and it fails. Test 3 is identical to Test 2, other than that the lattice pitch has been multiplied by 10.
I executed this on OpenMC 0.12.0 (last pull Nov 2020). I confirmed that the fix from pull #1414 is present in our installation.
Thanks for taking a look.
Edit: My mistake. I do receive the negative distance error with more complicated geometry, but the simple case only results in multiple " WARNING: After particle * crossed surface * it could not be located in any cell and it did not leak." I will delete and create a new topic if it’s not the same problem.