Analyses of the radiation fields from deuterium-tritium (D-T) fusion reactions are crucial for the success of fusion pilot plant designs. For this effort, Monte Carlo methods have been the standard choice for neutronics analysis of fusion reactors due to their ability to handle complex geometries and highly anisotropic fluxes. In order to support blanket development in Europe, two mock-up experiments were developed based on the helium-cooled pebble bed (HCPB) and the helium-cooled lithium lead blanket designs.

The focus of this study is the Frascati Neutron Generator (FNG) HCPB mock-up benchmark, available in the Shielding Integral Benchmark Archive and Database (SINBAD). Originally, the experimental data from the FNG HCPB were compared to the computational results from MCNP-4C to validate MCNP and quantify nuclear data uncertainties.

The open-source Monte Carlo code OpenMC is a powerful and flexible tool for simulating neutron and photon transport and analyzing nuclear systems. This study aims to benchmark it against experimental measurements and computational results. An OpenMC model was generated by converting the MCNP input file into geometry and material files using the openmc_mcnp_adapter. Additionally, a method has been developed by our team to build a FNG neutron source with neutron source characteristics that are importable as an OpenMC source object. Here we use OpenMC to compute the tritium production rate (TPR) in lithium carbonate (Li2CO3) pellets within the HCPB and the activation foil reaction rates and compare the results to experimental and computational (TPR only) data available in SINBAD.

Additionally, nuclear data uncertainty quantification is performed via the Total Monte Carlo method for various cross sections in 6Li, 7Li, and 9Be using the SANDY (Sampler of Nuclear Data and uncertaintY) package. The TPR results agree well with the computational results from MCNP-4C, and underestimate the experimental data by 9 on average due in large part to 9Be cross-section uncertainties. The reaction rate results agree well with experimental data with the exception of the 197Au(n,γ) reaction, which showed a consistent discrepancy among all nuclear libraries applied to this study.