Fuel Simulations
The benchmarks on this site are intended to answer a fundamental question: is HeavyMetal solving the governing equations correctly? The simulations presented here ask a different question: what can those verified capabilities teach us about fuel behavior? They use HeavyMetal either to develop improved physical submodels for fuel performance analysis or to explore mechanisms that may govern the evolution of fuel microstructure.
Each study begins with observations of irradiated fuel. The cavitation study examines how localized tearing and pore growth can degrade heat conduction and raise temperatures in annular metallic fuel. The homogenization work examines how regions of open and closed porosity, with different internal gas pressures, can be represented in an engineering-scale constitutive model. The pore migration and surface instability studies use the verified surface diffusion framework to investigate whether thermomigration can explain observed redistribution of porosity and changes in fuel surface morphology. Together, these studies connect observed microstructures with physical mechanisms and their consequences for fuel behavior.
The present studies focus on metallic fuel, but much of the underlying computational framework is considerably more general. The same capabilities used here to model evolving pores, surfaces, material transport, thermal gradients, stress, and contact can be adapted to related problems in other fuel systems, including the amoeba effect in TRISO fuel and pore migration in UO₂ fast-reactor fuel. An important long-term objective is to use such detailed calculations to develop simpler, mechanistically based models suitable for engineering-scale fuel performance analysis; the pore-migration study, for example, explicitly develops the connection between pore-scale transport and a homogenized material flux.
These should therefore be viewed as demonstration and exploratory calculations rather than finished predictive models. Additional physics, material property development, sensitivity studies, and comparison with irradiation observations will be required before the models are ready for engineering application. The surface instability study, for example, identifies several such next steps and treats its present result as support for a hypothesis rather than proof of one. With further development, we expect these studies to provide both improved fuel-performance submodels and results suitable for peer-reviewed publication.
