egs++ phsp scoring bug when scoring plane intersects regions

[rtownson]

It seems that particles are scored incorrectly for egs++ phsp scoring when the scoring plane intersects more than one region. I reproduced the issue both for egsphsp and IAEA (constant Z) formats.

See issue and example input file here: https://www.reddit.com/r/EGSnrc/comments/uxvu5b/issues_with_scoreplane/

[blakewalters]

@rtownson: I finally had a chance to look at this issue posted to reddit > 1 yr ago (!). The documentation for egs_phsp_scoring states (maybe not clearly enough) that when using the "phase space geometry" method, scoring surfaces must coincide with region boundaries. Considering this then, if I move the scoring plane, "flatPanel", from Z=5.002 cm to Z=5.0019 cm, I get a photon X-Y distribution that looks more reasonable. There can, indeed, be roundoff issues when attempting to match a "phase space geometry" to a region boundary that can then lead to erroneous particle scoring, but I don't think that's what's happening here.