Coupled differential and integral data analysis for improved uncertainty quantification of the ⁶³,⁶⁵Cu cross section evaluations

Abstract

A new methodology has been developed that couples differential cross section data evaluation with integral benchmark analysis for improved uncertainty quantification. The new methodology was applied to the two new copper evaluations and resulted in an improved evaluation with smaller covariance data. Copper is a structural material in many nuclear applications, particularly those dealing with criticality safety. The current standard for the resonance evaluation of the two copper isotopes, ⁶³,⁶⁵Cu, has been determined to result in poor modeling performance. Therefore a new resonance evaluation of the two copper isotopes is vital to nuclear criticality safety applications. Performing a new resolved resonance region evaluation for copper has served as a backdrop to this work on developing new techniques for resolved resonance region evaluation. For the new evaluations, experimental cross section measurements have been carried out in the thermal energy region where no experimental data had previously been measured. Along the way, an automated routine was developed to aid with the determination of the quantum angular momentum of newly identified resonances. The impact of differential scattering cross sections with respect to angle was determined in the benchmarking process. The implications of the study of the impact of differential cross sections on criticality suggest a necessity for detailed treatment of the angular distributions during the evaluation process, as well as temperature broadening of the angular distributions for simulation applications. The formalism for temperature broadening of angular distributions has been derived and tested. The new evaluations were compared against the current ENDF/B-VII.1 standard on a set of 23 criticality safety benchmark models and displayed improved performance. In the new methodology developed for coupling of the differential and integral data evaluation, resonance parameters are directly and systematically adjusted based on feedback from integral benchmark experiments. Coupling this feedback directly to the resonance parameters gives the new method the advantage of implicitly adjusting all of the cross sections simultaneously, including the double differential cross sections. Based on integral feedback, the new methodology provides a way of updating the reported covariance of the resolved resonance region to reflect true state of knowledge.