Lower-Length-Scale Model Development

To develop mechanistic models for fuel thermal conductivity, the fuel team used supercells up to 55 nm long to determine the thermal conductivity of UO2 with Xe incorporated. Atomistic simulations were used to determine thermal resistance values for four different types of grain boundaries, and these values have been used in mesoscale simulations of heat transport through representative fuel microstructures. [LANL]

Density functional theory techniques, previously applied to diffusion of Xe in UO2, have now been extended to Kr. Thus, both major gaseous fission products are now included in the simulations, which have identified the transport mechanism as being vacancy mediated. Activation energies have been calculated for each gas element, and these values are now used in MARMOT. [LANL]

A phase-field model was used to perform parametric studies of fuel restructuring due to void and grain boundary migration under temperature and stress. These studies are elucidating the roles of bulk diffusion, surface diffusion, grain boundary mobility, and grain size on microstructure evolution. Atomistic studies are underway to understand the interaction of dislocations with He bubbles of varied sizes and pressures. Results indicate that, except for very large bubbles under high pressure, bubbles are not obstacles for screw dislocations. [INL]

The following modifications to MARMOT were initiated: the fully-coupled solution algorithm in the phase-field code is being replaced with a staggered solution algorithm, the new algorithm was tested for Allen-Cahn formalism (in which the order parameters are not conserved) and is being extended to Cahn-Hilliard formalism, and the overall structure of MARMOT was modified to make it easier to develop and implement new models. Now, generic kernels for the split solution to the Cahn-Hilliard formalism have been introduced so that a model can be created by only coding in the problem-specific chemical potential. [INL]