Development of Molecular Dynamics Potential for Uranium Silicide Fuels

摘要

Use of uranium-silicide (U-S_1) in place of uranium dioxide (UO_2) is one of the promising concepts being proposed to increase the accident tolerance of nuclear fuels This is due to a higher thermal conductivity than UO_2 that results in lower centerhne temperatures. U-S_1 also has a higher fissile density, which may enable some new cladding concepts that would otherwise require increased enrichment limits to compensate for their neutronic penalty. However, many critical material properties for U-Si have not been determined experimentally It is anticipated that modeling and simulation may deliver guidance on the importance of various properties and help prioritize experimental work In order to develop knowledge-based models for use at the engineering scale with a minimum of empirical parameters and increase the predictive capabilities of the developed model, inputs from atomistic simulations are essential. First-principles based density functional theory (DFT) electronic structure calculations may provide the most reliable information However, it is probably not possible to obtain kinetic information such as amorphization under irradiation directly from DFT simulations due to size and time limitations. Thus, a more feasible way may be to employ molecular dynamics (MD) simulation Unfortunately, so far no MD potential is available for U-S_1 to discover the underlying mechanisms. Here, we present our recent progress in developing a U-S_1 potential. The development is based on the Tersoff type potentials for single element U and Si. The Si potential is taken from the literature and a Tersoff type U potential is developed in this project With the primary focus on the U_3Si_2 phase, some other U-S_1 systems such as U_3S_1 are also included as a test of the transferability of the potentials for binary U-Si phases. Based on the potentials for unary U and S_1, two sets of parameters for the binary U-S_1 system are developed using the Tersoff mixing rules and the cross-term fitting, respectively. The cross-term potential is found to give better results on the enthalpy of formation, lattice constants and elastic constants than those produced by the Tersoff mixing potential, with the reference data taken from either experiments or DFT calculations. In particular, the results on the formation enthalpy and lattice constants for the U_3S_(12) phase and lattice constants for the high temperature U3S1 (h-U_3Si) phase generated by the cross-term potential agree well with experimental data. Reasonable agreements are also reached on the elastic constants of U_3Si_2, on the formation enthalpy for the low temperature U_3Si (m-U_3S_1) and h-U_3Si phases, and on the lattice constants of m-U_3Si phase. All these phases are predicted to be mechanically stable. The unary U potential is also tested for three metallic U phases (α, β, γ). The potential is found capable to predict the cohesive energies well against experimental data for all three phases. It matches reasonably with previous experiments on the lattice constants and elastic constants of αU.