Speaker
Description
The National Nuclear Data Center (NNDC) maintains a repository of physical and theoretical quantities in the Evaluated Nuclear Data Format/B library (ENDF/B), currently on the 8th release (ENDF/B-VIII.0). Many fields rely on the accuracy of these quantities and their uncertainties, from real world applications (such as nuclear power) to more theoretical ones. This work is part of a larger effort by the NNDC to supply nuclear data for Intentional Forensics (IF) applications. The aim of IF is to develop a system for tagging commercially produced reactor fuel to alleviate a great security concern to the United States and other nation states and organizations. Apparently unrelated questions concerning the nature of the \textit{r}-process, including abundances, reaction rates, and astrophysical sites, remain enigmatic and an extremely active field of study. Both of these pursuits rely heavily on neutron capture in stable (but understudied) and unstable nuclei. The NNDC is analyzing and expanding the available data for these efforts, based on the ENDF/B-VIII.0 neutron sublibrary, containing 557 species of long lived isotopes and isomers relevant to neutron applications.
For both IF and many astrophysical applications the most relevant nuclear quantities are unmeasured (or under measured!), and will remain so for some time. Theoretical estimates are employed with varying levels of success due to assumptions that, while necessary, might not be supported by the physics. Current methods attempt to connect the cross section to physical quantities such as mass, level density, and neutron separation energy in order to identify patterns. These often focus on a single nuclear quantity or bulk property (such as mass model), which heavily weights that quantity's importance by default. As such, new methods are always of interest. This work is to first produce a catalogue of Maxwellian Averaged Cross Sections (MACS) for neutron-capture for the isotopes in the neutron sublibrary and spanning from thermal energies to the neutron activation range. From this library we will attempt to improve estimates of neutron capture Maxwellian Averaged Cross Sections, especially off-stability where few or no measurements exist. As the neutron fluence in many applications can be treated as a weighted sum of Maxwellian spectra, inverse problem theory allows us develop a regression model for the temperature dependence of the MACS. By applying the predictive capability of machine learning (ML) we can couple these estimates to additional physical quantities where known, to highlight or emphasize previously unknown relationships. These results will be provided to the IF group to inform their efforts. The results will also be implemented in new \textit{r}-process simulations, implementing the most up-to-date astrophysical models and nuclear network methods.