The objective of the Fuels Product Line (FPL) is to deliver an integrated set of increasingly predictive computational tools for nuclear fuel performance analysis and design. A multiscale approach to modeling and simulation has been adopted in developing the FPL toolset in which simulations of fuel performance at the engineering scale are informed by material property and behavior models developed from mesoscale simulations of microstructure evolution under irradiation, which are themselves guided and enabled by inputs of fundamental materials parameters obtained from atomistic simulations. The multiscale approach has been adopted with the conviction that by deriving fuel behavior from an instantaneous knowledge of fuel microstructure, rather than by correlating it to an empirical metric such as “burnup”, an unprecedented degree of predictability can be achieved in the nuclear fuel performance arena. BISON and MARMOT are the FPL components under development for the NEAMS ToolKit. BISON is the tool for fuel performance simulations at the engineering-scale. MARMOT is the tool for performing mesoscale simulations of microstructure evolution under irradiation, from which material property and irradiation behavior models can be derived. The atomistic simulations used to obtain direction and input data needed for the mesoscale simulations are performed by a broad spectrum of researchers using a variety of codes, most of which are publicly available.
The development and validation of BISON will enable predictions of nuclear fuel performance at the engineering scale under normal, off-normal, and accident conditions. The approach in developing BISON makes use of computational methods that allow for high-fidelity geometrical representation of fuel pins/pellets. It also makes use of highly efficient solvers to enable three-dimensional, fully-coupled multiphysics simulations (as appropriate). It is executable both on high-end desktop workstations (for engineers and designers) as well as high-performance supercomputers (for researchers).
BISON can couple with both the mesoscale microstructure evolution tool (MARMOT) as well as the reactor core and system tools under development by the RPL. In the early phases of development for any particular fuel system, BISON makes use of existing models for material properties and irradiation behavior that may be largely empirical in nature and have a somewhat limited range of applicability, but it is designed to easily and continuously incorporate new mechanistic, physics-based models produced by the mesoscale model development activities in order to extend its predictive capability to new fuels and operational regimes. At present BISON is most mature for simulations of oxide fuels in light water reactors, but it has rapidly developing capabilities for metallic and MOX fuels in fast reactors, TRISO particle fuels in gas reactors, and plate-type fuels for research/test reactors.
The development and validation of MARMOT will enable simulations of the microstructure evolution of nuclear fuels and materials under irradiation. MARMOT simulations are performed at the mesoscale, which is the scale at which microstructural features of materials are explicitly resolved. Development and use of MARMOT is supported with atomic-scale simulations to guide and enable the mesoscale simulations by identifying and prioritizing important mechanisms and calculating needed input parameters.
A primary objective of MARMOT development is to provide a tool that can be used to generate fundamental fuel and cladding material property and irradiation behavior models that can be up-scaled to inform the engineering-scale fuel performance tool (BISON). This capability should reduce the dependence of BISON on empirical correlations to describe the irradiation performance of nuclear fuels with a view to enabling true predictability, even into compositional or operational regimes where little or no experimental data exists. Thus, MARMOT development activities are primarily focused on supporting the development of BISON, over time making it increasingly predictive.