Monday 2:00 PM

February 28, 2022

Room: 202B

Location: Anaheim Convention Center

Actinide oxides with the fluorite structure represent the most widely deployed class of nuclear fuel materials. Consequently, understanding how their thermophysical properties evolve due to variables such as temperature, composition and radiation damage is important for achieving optimal usage. Atomistic simulation provides a powerful tool for predicting these properties and provides a level of detail which allows the atomic scale processes giving rise to changes in thermophysical properties to be made apparent. In this presentation simulations of specific heat capacity, melting behaviour, thermal conductivity and diffusion will be presented for actinide and mixed oxide fuel materials. All the simulations were performed using the same potential description namely the CRG many-body model, which celebrates it 8th birthday this year. This retrospective look will show the successes of this approach and where there is still room for improvement in the atomistic description of these materials using classical potential models.

We present a new approach for coupling experimental data, physical models, and machine learning algorithms to predict the effective thermal conductivity of UO2-BeO, UO2-Mo, and U-10Zr nuclear fuels. First, a physics-based mesoscale model that takes into account the effect of underlying microstructure, temperature, and interface thermal resistance is developed. Then the model is verified and validated against available experimental data. The validated model is then utilized to generate more data at conditions different from experiments. Existing experimental data and the additional model results are then used to train and test non-linear regression models in open-source machine learning software. The final outcome of this framework is validated, physics-based reduced order models of the effective conductivity of these fuels. This novel approach reduces the required number of experiments and at the same time minimizes the computational cost necessary for qualifying new materials.

As nuclear fuel burn-up increases, its thermal conductivity deteriorates due to the accumulation of defects that lead to increased phonon scattering rates. This results in a rise in the centerline temperature of the fuel pellet, whereby heat generation must be decreased to avoid melting of the fuel. Fuel performance codes are utilized to track this; however, they are based on physical principles that are empirically parametrized. Recently, an effort has been made to develop predictive fuel performance codes. In this presentation, we elaborate on the importance of careful analysis of point defect’s contribution to thermal conductivity. A rate theory model for point defect evolution coupled with the Klemens-Calloway model is used to calculate the change in thermal conductivity of UO2 under irradiation conditions. The model suggests that at higher doses of irradiation the centerline temperature increases, and point defects concentrations are highest near the rim of the fuel pellet.

Characterization of thermophysical properties of nuclear fuel is an essential step in understanding its behavior and predicting its performance under irradiation. For thermal conductivity, κ limited data is available in the literature for most advanced fuel concepts. This work provides an innovative approach by coupling machine learning and first-principles electronic structure theory to predict κ as a function of solid fission product concentrations. A machine learning model trained on (XO2) compounds was created using attributes parameters extracted from density functional theory (DFT). The model developed was cross-validated with experiments using the hot bridge transient method on SIMFUEL samples. The methodology developed here for UO2 can be extended to advanced fuels, where there is no available data of thermal conductivity degradation during irradiation. This effort is a step in an Accelerated Fuel Qualification methodology where separate tests and modeling can reduce the time needed to develop and qualify new fuel systems.

Understanding the thermal properties of α-U is essential for improving its performance and efficiency in nuclear reactors. Due to the metallic nature of α-U, it is necessary to examine the thermal transport by phonons, charge transport by electrons, as well as the interactions between them. In this work, we develop a computational code for solving the coupled electron-phonon Boltzmann transport equation using the Monte-Carlo (MC) method. The inputs to the MC simulations, including 3D phonon dispersions, electron band structure, and relaxation times of various phonon/electron scatterings, are all produced by density functional theory calculations. We conduct MC simulations of thermal and charge transport in pure and defected α-U with various defects. The effect of defects on the thermo-physical properties of α-U is investigated. The code developed herein is applicable to solving energy transport problems in a general metallic material with complex nanostructures.