Background Accurate evaluated nuclear data libraries are fundamental inputs for reactor physics simulations, directly determining calculation precision. In October 2024, the ENDF/B-VIII.1 library was officially released, featuring extensive updates in nuclear reaction models, experimental measurements, and evaluation methodologies compared to its predecessor, ENDF/B-VIII.0. To ensure its reliability in advanced nuclear energy systems, its applicability to fast spectrum reactor calculations must be thoroughly validated.

Purpose This study aims to systematically evaluate the performance and applicability of the newly released ENDF/B-VIII.1 evaluated nuclear data library in Monte Carlo reactor physics simulations for various fast spectrum experimental devices.

Methods The nuclear data processing program NECP-Atlas was utilized to process the raw ENDF/B-VIII.1 library and generate continuous-energy nuclear data in ACE format. A comprehensive suite of domestic and international fast spectrum benchmarks—including HMF057, PMF035, BN-600, SNEAK-7A/7B, CEFR, and ZPPR-9—was selected to cover diverse fuel types (HEU, Pu, MOX), core geometries, and coolants (sodium and lead). High-fidelity simulations were executed using the Monte Carlo code RMC to calculate key parameters such as the effective multiplication factor k_eff and control rod worth. The results were compared against calculations using the ENDF/B-VIII.0 and ENDF/B-VII.0 libraries.

Results Numerical results demonstrate that the processed ENDF/B-VIII.1 library can be successfully applied to Monte Carlo calculations for fast spectrum devices. For simple uranium assemblies and the SNEAK benchmarks, ENDF/B-VIII.1 exhibits improved accuracy, yielding k_eff values closer to the experimental benchmarks than previous versions. Conversely, for complex plutonium-bearing cores (e.g., CEFR and ZPPR-9), the calculated k_eff values are notably lower, leading to larger negative deviations from experimental data. This trend stems from the significant re-evaluation of the Pu-239 radiative capture cross-section in the new library. For control rod worth simulations in ZPPR-9, ENDF/B-VIII.1 generally matches the accuracy of older libraries across most cases, though a distinct deviation occurs in Case 5 due to the localized high-plutonium concentration in that experimental zone.

Conclusions The ENDF/B-VIII.1 evaluated library is fully capable of conducting reactor physics calculations for fast spectrum devices, providing reliable predictive performance for core criticality and control rod worth. However, the re-evaluated Pu-239 data introduces non-negligible discrepancies in plutonium-fueled systems, necessitating localized database adjustments, mixed-library approaches, or highly refined geometric modeling in future applications.