Optimization of a Multi-Layered Target for a Pulsed Power Neutron Source

The ability to create neutron environments is critical fora wide variety of applications including national security, materials science, and medical research. It is also helpful in medical, material, and other research. The U.S. Naval Research Laboratory (NRL), Washington, DC, is interested in leveraging their pulsed-power facilities to produce an in-house, relatively inexpensive neutron source. NRL’s facilities produce 2 MeV (from Gamble II) and 5 MeV (from Mercury) protons and deuterons. This research explored combinations of target materials to produce the most neutrons in the forward direction. Models of the reactions involved were first validated against literature experiments, then target designs were optimized with Dakota, a robust software suite used for various aspects of design exploration. The nuclear reactions and targets were modeled with MCNP, a Monte Carlo neutral particle transport code. A code base for scaling and future design work was also developed. The materials used in the optimization were lithium, beryllium, carbon, and deuterated polyethylene (CD2). The validation revealed discrepancies between the cross section libraries and literature. The CP2020 and JENDL/DEU libraries performed relatively well. The TENDL-2019 library greatly underperformed for the deuteron reactions. The ENDF/B-VIII.0 library also deviated in some cases from the existing data. The optimizations resulted in four unique targets, but came with a large amount of uncertainty. Lithium was shown to be the better target for the proton reactions, while beryllium was better for deuteron reactions. There were some variations of beryllium-lithium combination targets that produced the most neutrons, but these need further investigation. The carbon and CD2 were found to produce neutrons at a much lower rate than beryllium or lithium and were not used in any target