Optimization of a new interatomic potential to investigate the thermodynamic properties of hypo-stoichiometric mixed oxide fuel U_1−y Pu _y O_2−x

Abstract

The behaviour of stoichiometric U1-yPuyO2 compounds used as nuclear fuel is relatively well understood. Conversely, the effects of stoichiometry deviation on fuel performance and fuel stability are intricate and poorly studied. In order to investigate what affect these have on the thermophysical properties of hypo-stoichiometric U1-yPuyO2-x mixed oxide fuel, new interaction parameters based on the many-body CRG (Cooper-Rushton-Grimes) potential formalism were optimized. The new potential has been fitted to match experimental lattice parameters of U0.70Pu0.30O1.99 (O/M = 1.99) and U0.70Pu0.30O1.97 (O/M = 1.97), where M represents the total amount of metallic cations, through a rigorous procedure combining classical molecular dynamic and classical molecular Monte Carlo simulation methods. This new potential provides an excellent description of the U1-yPuyO2-x system. Concerning lattice parameter, although fitted on only one Pu content (30%) and two stoichiometries (1.99 and 1.97), our potential allows good predictions compared to available experimental results as well as to available recommendations in wide ranges of O/M ratio, Pu content and temperature. For the U0.70Pu0.30O2-x hypo-stoichiometric system (30% Pu content and O/M ratio ranging from 1.94 to 2.00), some direct properties (lattice parameter and enthalpy) and some derivative properties (linear thermal expansion coefficient and specific heat) were systematically investigated from room temperature up to the expected melting temperatures and a good agreement with experiments is found. Moreover, our potential shows good transferability to the plutonium sesquioxide Pu2O3 system.

Phase-field modeling with the TAF-ID of incipient melting and oxygen transport in nuclear fuel during power transients

“…(44c), the thermal conductivity k and the heat capacity C p depend on temperature but also on the phase composition and the oxygen content. In fuel performance codes, the heat capacity is often calculated according to empirical correlations or given by laws resulting from atomic-scale simulations [53,54,55]. It is interesting to keep in mind that C p could be calculated at run-time with a thermochemical solver coupled with a CALPHAD database such as, for instance, OpenCalphad with the TAF-ID.…”

Thermophysical properties of liquid (U, Zr)O2 by molecular dynamics

“…Mixed (U, Zr)O 2 oxides compositions were generated by randomly substituting U with Zr on cation sites in a F m 3m cubic uorite structure; no specic random structuregeneration procedure such as SQS (special quasirandom structures) was used, as recent work proved it had little to no impact for determining density or heat capacity in large cells for mixed oxides systems [14]. A single purely random conguration was used for each composition, as previous work from Cooper et al showed only small variations between multiple random congurations above 2500 atoms [10].…”

“…Covalent bond distances from the Morse and Buckingham part (ρ, r 0 ) are tted rst because of their important sensitivity: ensuring their minimization is key to avoid artefacts later in the tting run simulations; A αβ and D αβ are tted next. Finally, EAM parameters (G α , n β ) are tted last, since they bear no physical meaning relative to the system, and only serve to adjust the collective behaviour in the cell [14].…”