SMITHERS: An object-oriented modular mapping methodology for MCNP-based neutronic–thermal hydraulic multiphysics

Abstract

A novel object-oriented modular mapping methodology for externally coupled neutronics–thermal hydraulics multiphysics simulations was developed. The Simulator using MCNP with Integrated Thermal-Hydraulics for Exploratory Reactor Studies (SMITHERS) code performs on-the-fly mapping of material-wise power distribution tallies implemented by MCNP-based neutron transport/depletion solvers for use in estimating coolant temperature and density distributions with a separate thermal-hydraulic solver. The key development of SMITHERS is that it reconstructs the hierarchical geometry structure of the material-wise power generation tallies from the depletion solver automatically, with only a modicum of additional information required from the user. Additionally, it performs the basis mapping from the combinatorial geometry of the depletion solver to the required geometry of the thermal-hydraulic solver in a generalizable manner, such that it can transparently accommodate varying levels of thermal-hydraulic solver geometric fidelity, from the nodal geometry of multi-channel analysis solvers to the pin-cell level of discretization for sub-channel analysis solvers.The mapping methodology was specifically developed to be flexible enough such that it could successfully integrate preexisting depletion solver case files with different thermal-hydraulic solvers. This approach allows the user to tailor the selection of a thermal-hydraulic solver to the requirements and limitations of the specific problem under consideration, without needing to modify their existing depletion code input files. To enable support of a wide range of existing depletion solver input decks, SMITHERS can accommodate arbitrarily detailed geometry segmentation for the depletion calculation with a minimum of additional information required from the user. This new implementation was motivated by the desire to enable easier multiphysics modeling of a wide range of reactor types, from power reactors to research reactors.The initial implementation of this modular mapping methodology incorporates the neutronic depletion solver MCNP/MONTEBURNS and a multi-channel analysis two-phase flow thermal hydraulic solver. Two verification test problems were evaluated to verify that the code’s routines were operating as intended. Preexisting generic BWR and nontraditional PWR cases were selected to ensure maximum code coverage and evaluate the operation of all implemented mapping routines. The obtained power, coolant temperature, and coolant density results verified that SMITHERS was correctly performing on-the-fly mapping combinatorial-basis MCNP/MONTEBURNS-calculated material powers to the thermal-hydraulic solver’s nodal geometry and that nodal coolant temperature and densities were correctly returned to the combinatorial geometry of MCNP.