Fitting Prompt Fission Neutron Spectra Using Kalman Filter Integrated with Empire Code

Abstract

Prompt fission neutron spectra (PFNS) have proven to have a significant effect on criticality of selected benchmarks, in some cases as important as cross-sections. Therefore, a precise determination of uncertainties in PFNS is desired. Existing PFNS evaluations in nuclear data libraries relied so far almost exclusively on the Los Alamos model. However, deviations of evaluated data from available experiments have been noticed at both low and high neutron emission energies. New experimental measurements of PFNS have been recently published, thus demanding new evaluations. The present work describes the effort of integrating Kalman and EMPIRE codes in such a way to allow for parameter fitting of PFNS models. The first results are shown for the major actinides for two different PFNS models (Kornilov and Los Alamos). This represents the first step towards reevaluation of both cross-section and fission spectra data considering both microscopic and integral experimental data for major actinides.