A Proposal for Benchmarking ²³⁵U Nuclear Data

Two experiments performed at the IPEN/MB-01 reactor are suggested to serve as a benchmark problem to verify mainly the adequacy of the ²³⁵U nuclear data for criticality analyses and for the isothermal reactivity coefficient determination of thermal reactors. The experiments are claimed to be well-defined, and they are suitable for a benchmark problem partially due to their small uncertainties and partially due to the lack of any sort of calculated correction factors or any quantity that comes either from the calculational methodologies or from another experiment. The isothermal experiment fulfills a specific need to introduce a reactor response that is sensitive to the ²³⁵U cross-section shape below 5 meV. This feature could be accomplished due mainly to the very precise control bank system characteristics of the IPEN/MB-01 reactor. The MCNP-4B calculational analyses reveal that the most recent ²³⁵U evaluation (Leal-Derrien-Larson's evaluation) incorporated in ENDF/B-VI release 5 performs well in the theory-experiment result comparisons of the aforementioned experiments. Particularly in the isothermal experiment, ENDF/B-VI release 5 produces results that even considering the deviations inherent to the Monte Carlo method meet the desired accuracy (±1.0 pcm/°C) for the isothermal reactivity coefficient determination in contrast to the JEF-2.2 and JENDL-3.2 libraries, which produce unacceptably high k_eff results. The main reasons are the ²³⁵U nuclear data in the case of JEF-2.2 and the nuclear data of both ²³⁵U and iron in the case of JENDL-3.2.