Abstract
The nuclear reaction code EMPIRE has been extended to provide evaluation capabilities for neutron cross section covariances in the thermal, resolved resonance, unresolved resonance and fast neutron regions. The Atlas of Neutron Resonances by Mughabghab is used as a primary source of information on uncertainties at low energies. Care is taken to ensure consistency among the resonance parameter uncertainties and those for thermal cross sections. The resulting resonance parameter covariances are formatted in the ENDF-6 File 32. In the fast neutron range our methodology is based on model calculations with the code EMPIRE combined with experimental data through several available approaches. The model-based covariances can be obtained using deterministic (Kalman) or stochastic (Monte Carlo) propagation of model parameter uncertainties. We show that these two procedures yield comparable results. The Kalman filter and/or the generalized least square fitting procedures are employed to incorporate experimental information. We compare the two approaches analyzing results for the major reaction channels on 89Y. We also discuss a long-standing issue of unreasonably low uncertainties and link it to the rigidity of the model.
Original language | English |
---|---|
Pages (from-to) | 2752-2761 |
Number of pages | 10 |
Journal | Nuclear Data Sheets |
Volume | 109 |
Issue number | 12 |
DOIs | |
State | Published - Dec 2008 |
Externally published | Yes |
Funding
The work at BNL was sponsored by DOE-NNSA, DOE-NE, and also by DOE-Office of Science under Contract No. DE-AC02-98CH10886 with Brookhaven Science Associates, LLC.
Funders | Funder number |
---|---|
DOE-NE | |
DOE-NNSA | |
DOE Office of Science |