Monte Carlo study of photoneutron production in U-235 following perturbations in cross section data

Active photon interrogation systems may be employed to detect high-Z isotopes without significant spontaneous fission emissions. These systems induce photonuclear reactions with emissions (such as fission neutrons) that may be detected. However, there are inconsistencies in the literature reporting resonance photonuclear interaction data for many isotopes. Recent publications show variations as large as 20% between various measurements of photonuclear cross section data. A perturbation methodology utilizing the modular nature of the MCNPX/MCNP-PoliMi code system has been implemented and is applied here to highly-enriched uranium. Monoenergetic photon sources between 8 and 18 MeV were simulated; neutron detection was performed using the MCNP-PoliMi liquid scintillator model. At photon energies less than 12 MeV, the number of detected neutrons is approximately 70% sensitive to changes in the (γ, f) cross section and 30% sensitive to changes in the (γ, n) cross section. As gamma-ray energy increases the (γ, f) sensitivity increases and the (γ, n) sensitivity decreases. There is a small (γ, 2n) sensitivity at photon energies between 15 and 17 MeV. The ability of modern simulation tools to predict photonuclear responses is greatly limited in this energy region due to the high sensitivity of the simulated results to observed discrepancies in photonuclear cross section data.