Nicolas Schunck | 18-ERD-008
The goal of this project was to produce modernized nuclear data libraries of light ion fusion reactions and actinide fission product yields. Such libraries are essential input in both stockpile science and fundamental science (e.g., nucleosynthesis, superheavy nuclei). When such libraries are used to model complex processes involving exotic, very short-lived nuclei that have not been measured experimentally, the underlying nuclear data come from theoretical models—the predictive power of which needs to be quantified—together with its uncertainties.
Working with a team of nuclear theorists and statisticians, we built comprehensive theoretical frameworks based on the most fundamental theories of fusion and fission, which we coupled for the first time with a Bayesian framework with machine learning techniques for uncertainty quantification (UQ). Our results include (1) statistical emulators for neutron-alpha scattering and the fission product mass yields of neutron-induced fission of plutonium-239, which will enable developers to generate random realizations of the corresponding data for UQ in applications; (2) the first complete, and most accurate, ab initio prediction of the S-factor for proton capture on beryllium-7, which, combined with UQ, may become the recommended value at solar energies in the next review of solar fusion cross sections; (3) statistical emulators for fission mass yields in the neutron-induced fission of plutonium-239; and (4) the first quantification of how uncertainties in nuclear forces impact fission properties that are relevant in nucleosynthesis simulations to explain how heavy elements are formed in the cosmos.
In addition to predictions from the most advanced theories of fusion and fission, we delivered quantified uncertainties of these predictions, as well as statistical emulators that could enable developers to efficiently generate random realizations of nuclear data for uncertainty propagation. This new capability could be applied to quantify the uncertainties of neutron elastic scattering and (n,n') reactions on light nuclei important for stockpile stewardship science and provide more reliable estimates of fission fragment distributions of actinide nuclei (with full uncertainties) where experimental data is not sufficient.
This project was an important milestone in integrating accurate, science-based nuclear data developed at Lawrence Livermore National Laboratory and data uncertainties in stockpile stewardship codes. This project demonstrated Livermore's leadership in the application of machine learning techniques to nuclear theory, and supports Livermore's mission research challenge in nuclear weapons science.
Publications, Presentations, and Patents
Gazit, D., et al. 2019. "Erratum: Three-Nucleon Low-Energy Constants from the Consistency of Interactions and Currents in Chiral Effective Field Theory." Physical Review Letters, 122, 029901. doi:10.1103/PhysRevLett.122.029901. LLNL-JRNL-758658
Gysbers, P., et al. 2019. "The Quenching Puzzle of Beta-Decays." Nature Physics, 15, 425–426. doi:10.1038/s41567-019-0483-y. LLNL-JRNL-745901