Speaker
Description
Authors:
Isabel Hernandez, Jennifer Jo Ressler, Bonnie Canion
Understanding gamma-ray production in a neutron environment is vital in many scientific concentrations. Emergency response, planetary exploration, safety and shielding designs – among many other applications – depend on analyzing neutron induced gamma-ray spectra from measurements. These research areas often require simulation capabilities to extend analysis and substitute as a low-cost and fast solution option to physical measurements. Due to the wide availability and applicability of neutron transport codes, many users depend on the status of current nuclear data libraries in order to produce accurate simulated results.
Discrepancies observed between simulations and measurements may be due to the availability of the nuclear data needed to model an interaction, the accuracy and precision of the nuclear data available, or the manner in which the simulation tool is incorporating the nuclear data to model an interaction. Therefore, it is of current interest to determine the condition of current nuclear data libraries and compare them to gamma-ray data. We focus on the ENDF/B VIII evaluation, due to the high volume of users. Proposed benchmark datasets, such as published IAEA compiled standards, will be used in its comparison. In this work, we developed a process for extracting neutron induced gamma-ray production data from the GNDS data libraries for discrete gamma-ray energies, as well as exploring best methods for including the continuum spectrum contribution.
We then compared these gamma-ray production cross sections in the GNDS library to experimental benchmarks. We saw that non-graphical library extraction methods provided a more accurate cross section compared to experimental results, though the continuum contribution that led to a comparable cross section was often non-trivial compared to its discrete counterpart, and the appropriate integration energy window size varied greatly. This is notable, since the importance of the non-discrete contribution to the total cross section is prevalent in all analyzed isotopes and may lead to interpretation issues in simulation codes. Lastly, we used popular neutron transport codes, MCNP, Mercury, and GEANT4 to extract the interpreted ENDF/B VIII data and provide comparisons to measurements. We found that isotopes with library storage issues found in the previous step had greatly misinterpreted cross sections across the board compared to the raw library data. These isotopes include 12C and 14N, two prominent isotopes relevant for applications such as non-destructive cargo screening techniques. Future work will include investigating why these Monte Carlo codes are accessing nuclear data differently and extending this process to other isotopes and interactions. Recommendations to the U.S. Nuclear Data Team have and will be made for further improvements where data are incorrect or missing.
Prepared by LLNL under Contract DE-AC52-07NA27344.
LLNL-ABS-827870