I am looking into how spectrum-averaged neutron capture branching ratios are calculated within the OpenMC data processing pipeline, specifically for the ENDF/B-VIII.1 library.
My specific concern is that ENDF/B-VIII.1 does not appear to provide the explicit isomeric cross sections (separating ground vs. metastable states) required to derive these ratios directly. While this data is available in libraries like JEFF-3.3 (via EAF format/TENDL), it seems to be missing from the primary ENDF/B-VIII.1 files.
Since we need branching ratios for custom spectra—beyond the standard fast or thermal averages provided on the OpenMC data page—we need to generate our own. Could you clarify the methodology used to handle this in the openmc-dev/data scripts when the source ENDF file lacks these specific cross sections?
Specifically:
Does your processing rely on a fallback physics model or an external supplemental library to determine the branching fractions for ENDF/B-VIII.1?
Do you utilize FISPACT-II (or a similar external tool) as part of your data-generation workflow to perform these spectrum-averaged calculations?
Any insight into the logic used for these “missing” isomeric production channels would be greatly appreciated.
Hi @gert1 and thanks for the question. Unfortunately, the situation with capture (and other reaction) branching ratios is not in a very good state. As you noticed, the depletion chains we distribute for OpenMC include either thermal- or fast-spectrum capture branching ratios. You can find a brief note on where the values come from here.
As it says, the thermal values are effectively borrowed from values that were used for Serpent. The fast values were computed using a simplified calculation on an SFR assembly. However, both of those sets of values correspond to older data libraries. When we’ve produced updated libraries such as for ENDF/B-VIII.1, we have just taken the branching ratios as is, without recomputing them, which means there may indeed be differences in cases where the libraries changed the ground and/or metastable production cross sections. Part of the reason we’ve done this is because even recent data libraries are often missing the metastable production cross sections. I’d love to find a better solution here but unfortunately haven’t had the time or resources to address this.
To answer the other question, no, we do not currently use FISPACT-II in any part of our generation workflow.
Hi Paul, I am not sure if it will work or not, but in ONIX, Julien uses some branching ratio tables that cover some neutron energy as you can see in his eaf-2010-multiplicities isomeric data. Is it possible to use this data to make the openmc reevaluate its branching ratio based on the calculated spectrum in their model? Could that simplify the depletion chain and become flexible to the model spectrum (not locked to PWR or SFR)?