Abstract:
Background With the development of neutron calculation methods and improved modeling capabilities, the errors introduced by model approximations and discretization methods in nuclear reactor physics calculations have gradually decreased. However, nuclear data, due to the challenges in measurement, have become the key input parameter affecting computational accuracy.
Purpose In this study, a nuclear data adjustment module based on sensitivity analysis and the generalized linear least squares algorithm was developed within the self-developed sensitivity and uncertainty analysis platform, SUPES.
Methods First, sensitivity analysis was used to determine the relationship between system responses and input parameter variations. Next, similarity analysis was applied to select experimental setups with high similarity at the neutron physics level. Finally, the generalized linear least squares algorithm was employed to minimize the error between computed and measured values, resulting in nuclear data adjustments.
Results The adjustment of the ACE format continuous energy database was performed on 22 cases from the critical benchmark HEU-MET-FAST-078. The numerical results show that the root mean square error of the effective multiplication factor (keff) was reduced from 3.10×10−3 to 1.53×10−3.
Conclusions The comparison and analysis verified the correctness of the developed nuclear data adjustment module.
下载: