Combined <i>ab initio</i> and empirical model of the thermal conductivity of uranium, uranium-zirconium, and uranium-molybdenum

In this work we developed a practical and general modeling approach for thermal conductivity of metals and metal alloys that integrates ab initio and semiempirical physics-based models to maximize the strengths of both techniques. The approach supports creation of highly accurate, mechanistic, and extensible thermal conductivity modeling of alloys. The model was demonstrated on \ensuremath{\alpha}-U and U-rich U-Zr and U-Mo alloys, which are potential fuels for advanced nuclear reactors. The safe use of U-based fuels requires quantitative understanding of thermal transport characteristics of the fuel. The model incorporated both phonon and electron contributions, displayed good agreement with experimental data over a wide temperature range, and provided insight into the different physical factors that govern the thermal conductivity under different temperatures. This model is general enough to incorporate more complex effects like additional alloying species, defects, transmutation products, and noble gas bubbles to predict the behavior of complex metallic alloys like U-alloy fuel systems under burnup.