Cross-section generation questions

Questions and discussion about applications, input, output and general user topics
Post Reply
Posts: 13
Joined: Mon Apr 03, 2017 4:53 pm
Security question 1: No
Security question 2: 64
Location: Arlington, Virginia, US

Cross-section generation questions

Post by j0937065063 » Fri Mar 06, 2020 9:16 am

Hi Jaakko,

The scattering matrix generated by the detector is an "averaged" homogenized cross-section for the whole model, is that correct? e.g. INF_SCATT0, INF_SCATTP0 ... etc.

Is there a way to generate material dependent macroscopic cross-section, including total, absorption, nu*fission, and P0, P1, P2, P3 .... scattering matrices?

I am trying to generate some multi-group cross-section for transport code calculations.



Ville Valtavirta
Posts: 455
Joined: Fri Sep 07, 2012 1:43 pm
Security question 1: No
Security question 2: 92

Re: Cross-section generation questions

Post by Ville Valtavirta » Mon Mar 09, 2020 10:55 am

Hi Vince,

Serpent does the group constant generation on universe basis, which means that you can get material-wise homogenized group constants by specifying different materials as different universes. The most simple approach would probably be to create an inf surface and create separate infinite cells/universes for different materials. Then you can replace the materials in cell, pin etc. definitions by filling in the corresponding universe instead.

Finally, set group constant generation on for all relevant universes using the set gcu card.


Posts: 6
Joined: Wed Jul 06, 2016 3:31 pm
Security question 1: No
Security question 2: 40
Location: Warsaw, Poland

Re: Cross-section generation questions

Post by Margo » Wed Aug 05, 2020 2:33 pm

Sorry for somehow repetitive and basic question, however as I am not sure if I fully understood the answers, I'd like to ask again.

I would like to estimate the effective macroscopic cross section of a medium with few isotopes that have high microscopic thermal absorption cross section. I have an external neutron source, normalised to its strength (set src rate). All cells are 2D cylinders, I use black boundary condition (set bc 1)
Is it enough to define a detector like the following?

% flux detector in the material in 1 energy group, m_1 is the cell with my material, u03 the universe I used as well in set gcu u03
det Flux_1 de 1 dc m_1 du u03

%macroscopic reaction rate / flux
det Sigma_1 de 1 dc m_1 du u03 dr -1 void dt 3 Flux_1

When I compare the results with my total macroscopic cross section (Sigma) I calculate with transmission (for the same material, source and thickness - I1=I0exp(-Sigma*d) ), I get very different results and I can't figure out what am I doing incorrectly...

Thank you in advance for any help!

Ana Jambrina
Posts: 62
Joined: Tue May 26, 2020 5:32 pm
Security question 1: No
Security question 2: 7

Re: Cross-section generation questions

Post by Ana Jambrina » Wed Aug 05, 2020 7:10 pm

How do you calculate the total macroscopic XS? Off the top of my head, some stuff to consider:
- It could use gcu option or detector-based for group constant generation.
- If choose 'gcu' option, keep in mind that group constant generation is calculated on universe basis. To obtain material-wise group constant define cell/universes for each material; If multiple materials are define in the same universe (i.e. whole geometry, universe 0), the result will be homogenised group constant integrated over the universe/geometry.
- If choose to output macroscopic XS through detectors, there is no need to keep 'gcu' option active (set gcu -1)
- 'dm'/'dc'/'du' describes the spatial domain where the detector is scored (region over which the response is integrated). - If you have defined a single material/cell/universe, the definition seems redundant.
- 'void' option in the detector card means the response is not pre-assigned with a specific material (when the detector scored in a collision, the XS is taken from the material at the collision point - allowing calculating integral reaction rates over regions composed of multiple materials); if it is not the case, define material/nuclide and substitute 'void'.
- the results of detector are integrated values without division by volume; including 'dv' card with volume to consider it.
- Ana

Post Reply