SUMMARY

In this work, a thermal neutron scattering data evaluation framework is presented that combines measured scattering data and computer simulations to evaluate the dynamic structure factor (DSF) and its uncertainties. The original parameter set of a given interaction model is randomly sampled according to interaction parameters’ prior probability distribution function. For each set of perturbed parameters, a corresponding DSF and double differential cross section (DDCS) are computed, and a weight associated with this set of perturbed parameter is obtained using a Unified Monte Carlo (UMC) method from differences between simulated and measured data. Using these weights, the best estimate of the DSF is computed as a weighted average of DSF values of all perturbed parameters sets. This evaluation framework is demonstrated on the TIP4P/2005f light water interaction model combined with DDCS data measured at the Spallation Neutron Source (SNS) Fine Resolution Fermi-Chopper Spectrometer (SEQUOIA) at Oak Ridge National Laboratory (ORNL). Molecular dynamics trajectories computed for TIP4P/2005f by the GROMACS code were processed to yield thermal neutron scattering kernel DSF and DDCS. The UMC model was used to compute weighted average of DSF, its uncertainty, corresponding DDCS, and a total scattering cross section. Cross sections computed from this DSF were then validated against independent experimental data, as well as relevant benchmarks in the International Handbook of Evaluated Criticality Safety Benchmarks (ICSBEP), including the PU-SOL-THERM-033, LEU-COMP-THERM-079, and HEU-COMP-THERM-006 benchmarks.