Bug in PN scattering cross sections and diffusion parameters

Old bugs, solved problems and other topics that are no longer relevant are moved here
Post Reply
User avatar
Jaakko Leppänen
Site Admin
Posts: 2446
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Bug in PN scattering cross sections and diffusion parameters

Post by Jaakko Leppänen » Thu Sep 02, 2010 11:41 pm

There is a bug in the calculation of PN scattering cross sections (SCATT0, SCATT1, SCATT2 ... SCATT5) and P1 diffusion parameters (P1_TRANSPXS, P1_DIFFCOEF, P1_MUBAR). The problem has been identified before (see description) and corrected in update 1.1.3 for all the other group constants.

The magnitude of the problem can be estimated by comparing the values of SCATT0 and SCATTXS. They should be the same within statistics, but in many cases there is a clear difference when there are void regions in the geometry. The problem can be avoided by replacing void regions with low-density gas, but this usually results in large statistical errors.

I'll try to fix the problem in update 1.1.14.
- Jaakko

User avatar
Jaakko Leppänen
Site Admin
Posts: 2446
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Bug in PN scattering cross sections and diffusion parameters

Post by Jaakko Leppänen » Fri Sep 03, 2010 10:17 am

For a quick fix, add line:

AddStat(RES_FG_SCATTFLX, ng*ngu + i, flx);

after line 892 in collectresults.c: (version 1.1.13)
- Jaakko

User avatar
Jaakko Leppänen
Site Admin
Posts: 2446
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Bug in PN scattering cross sections and diffusion parameters

Post by Jaakko Leppänen » Tue Nov 16, 2010 12:05 pm

This problem was fixed in update 1.1.14.
- Jaakko

estewart
Posts: 20
Joined: Wed Apr 01, 2020 2:45 pm
Security question 1: No
Security question 2: 7

Re: Bug in PN scattering cross sections and diffusion parameters

Post by estewart » Tue May 04, 2021 11:25 am

Hello

I am modelling HTR systems with low density helium regions, and the errors on the cross-sections for those regions are quite large. I'm just wondering if there is a way to reduce these errors without increasing the population per cycle. I increased the population per cycle by an order of magnitude which reduced the error by roughly 1/3 but they are still quite large.

Best wishes,

Emma

Ana Jambrina
Posts: 666
Joined: Tue May 26, 2020 5:32 pm
Security question 1: No
Security question 2: 7

Re: Bug in PN scattering cross sections and diffusion parameters

Post by Ana Jambrina » Wed May 05, 2021 10:16 am

For regions within the core, maybe the straight-forward action would be increasing the population size, and tuning the ratio of neutrons per generation and number of cycles.
In low-density medium, you might have a high-rejection probability due to delta tracking mode. For a low-density medium you might want to increase the virtual collision frequency to get sufficient scores for the collision flux estimator (see ‘set cfe’ option) - changing the default setting on how delta-/surface- tracking is used in the calculation.
For further details on tracking modes and CFE, see the detailed descriptions on delta- and surface-tracking and result estimators.
- Ana

User avatar
Jaakko Leppänen
Site Admin
Posts: 2446
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Bug in PN scattering cross sections and diffusion parameters

Post by Jaakko Leppänen » Wed May 05, 2021 12:20 pm

This paper discussing the efficiency of the CFE may also be useful:

https://doi.org/10.1016/j.anucene.2017.03.006
- Jaakko

Post Reply