Fix Seltzer-Berger max xs for positrons #922
Conversation
amandalund
added
bug
| static double const expected_avg_engine_samples[] = {4.0791015625, | ||
| 4.060546875, 5.13623046875, 4.6572265625, 4.43115234375, 4.35791015625, | ||
| 9.3740234375, 4.65478515625}; | ||
| 137.044921875, 4.060546875, 15.74169921875, 5.103515625, 5.26953125, |
There was a problem hiding this comment.
That's a lot of rejection, especially considering the comment above:
Note: the first point has a very low cross section compared to ionization so won't be encountered in practice. The differential cross section distribution is much flatter there, so there should be lower rejection.
Do you have any plots of the maximum cross section versus the actual distribution?
There was a problem hiding this comment.
Good point, I'll update the comment since it only applies to the electron DCS -- for low incident energies the positron DCS deviates a lot. These are cross sections for a 1 keV incident particle (same as that first point in the test) and a few different cutoff energies (0.9 keV is used in the test):
So, the sampling efficiency for the positrons here will be quite poor. For higher incident energies though the positron DCS will be much closer to the electron DCS.
There was a problem hiding this comment.
Yuck, that rejection rate is high indeed, but it should be a very unusual case indeed. Thanks for the plot!
* Add failing SB test * Revert to using Seltzer-Berger electron max DCS for positrons * Update comment
This reverts to using the max electron Seltzer-Berger DCS for positrons instead of the first DCS value (which I had mistakenly based on a zero production cut in #401). The xs scaling factor depends on the production cut, and will approach 1 as the exiting gamma energy approaches the cutoff value -- so in the extreme case the max positron xs will be the same as for electrons.
This should resolve the failure in #892, which occurs for low energy positrons when the sampled exiting gamma energy is very close to the production cut.