NET-CHRG vs. charge weighted yield

Versions

FLUKA: 4-5.1
Flair: 3.4-5.2

Description

Dear FLUKA experts,

due to reasons, I have compared the net charge deposited in a region obtained via usrbin → NET-CHRG and via charge-weighted usryield as shown below:

For reference, the primaries are 4.03 GeV protons, so I think the usryield limits are safe to include eveything. Also SPUHR is completely encapsulated within SCHOBPI.

The results are:
NET-CHRG: 9.1980E-04 ± 3.4245%
SPU_RI (in): -5.2575001E-07 +/- 9.0420411E-15 %
SPU_RO (out): -9.8692499E-07 +/- 3.7282799E-14 %

The usryields are integrated over charge and still must be multiplied by the 2000 GeV energy range. If I do this and subtract (out) from (in) I get a net charge passing the surface of

9.2235e-04 ± O(1e-8)%

The results are within the uncertainty of NET-CHRG, but since they are from the same run, I wonder if the discrepancy in value is physical, numerical or due to a misunderstanding on my side?

If it is numerical, can I rely on the usryield results to be significantly more accurate?

Input files

protons_vloss_76_sum.lis (1.5 KB)
protons_vloss_77_sum.lis (1.5 KB)
protons_vloss_78.bnn.lis (517 Bytes)
halo_collimation_sector_one.flair (61.4 KB)
halo_collimation_sector_one.inp (40.7 KB)
source_halo_coll_sector_one.f (11.5 KB)

Best regards
Roman

Dear Roman,
nice work!
Note that the USRYIELD statistical errors make no sense here, since yield weighting by charge brings in negative values that unfortunately ruin their calculation.
As for the 0.3% difference on the charge deposition you are left with, this may be due to sub-threshold recoils that are not intercepted by USRYIELD while they are by USRBIN. Hence the latter is more accurate for net charge estimation purposes.