I’m considering implementing range rejection in my BEAMnrc simulation to make it viable to use as a direct particle source in lieu of phase spaces and had a question about a frequent number I see cited.

Based on a study by Sheikh-Bagheri, it seems like EGS was altered to be able to tag any Brem event that didn’t originate in the target when they concluded that an ESAVE value of ~1-2 MeV in a 6 MV simulation only caused the photon fluence to be underestimated by about 0.2%.

Is this study possible to reproduce using the existing LATCH system or would it be prohibitively difficult? Thanks!

When using a transformed phase space source (or using the EGS-feature-dynamic-source) in egs_chamber on a parallel computing cluster, each individual job appears to use the entirety of the phase space rather than just a chunk of it.

Is that intended by design? If the seeds are the only things different between the individual jobs, then I would expect the results to be highly correlated and the uncertainty to be way too optimistic.

If I directly call the phase space as a particle source (and even apply a transformation directly in the first scoring geometry), then the log files seems to indicate that the phase space file is being broken into chunks and sampled correctly in that every job starts with a different unique particle from a different section of the phase space.