Mesoscale modeling to inform Bison models of accident tolerant fuel concepts

Abstract

Listen

Translate article

U3Si2 and doped UO2 have been under investigation in recent years as potential accident-tolerant fuel concepts. In this report, lower-length scale studies of these fuel concepts carried out during Fiscal Year 2020 are detailed. U3Si2 is a potential accident-tolerant fuel that shows promise due to its high thermal conductivity and higher uranium density relative to UO2. However, its swelling and fission gas release behavior in light water reactor (LWR) conditions is relatively unknown. To provide mechanistic insight and determine parameters for engineering-scale fuel performance modeling of pellet-form U3Si2, phase-field simulations of the growth, interconnection, and venting of intergranular fission gas bubbles were performed. The fractional coverage of the grain boundary and the fraction of bubble area that is vented were calculated as a function of time. From the simulation data, the fractional grain boundary coverage at saturation, an important parameter needed in engineering-scale modeling of swelling and fission gas release, was determined. Multiple simulations were run to determine the uncertainty in the calculated value. The effect of model assumptions and input parameters that are not well known was evaluated. Simulation results are compared to related theoretical and computational work. Based on the simulation results, a value of 0.60 for the fractional grain boundary coverage at saturation is recommended for U3Si2 fuel. Chromium-doped UO2 is a widely-studied near-term deployable accident-tolerant fuel concept because it results in a dense, large-grain structure that increases the fuel resistance to densification, swelling, and fission gas release. Densification occurs during reactor operation as pores remaining in the fuel from manufacturing shrink and are gradually eliminated. However, grain growth also occurs concurrently with densification and the resulting grain size change affects the densification process. Thus, it is necessary to have a mechanistic, quantitatively accurate model for grain growth as a basis for a densification model. In this report, the development a mechanistic model of grain growth in undoped UO2 fuel during reactor operation. The model development builds on published experimental data on UO2 grain growth, as well as atomistic and mesoscale simulation results. We begin by developing new fits with temperature T (in K) for the average grain boundary (GB) energy and mobility in UO2 using literature data, where the GB energy ¯? = (1.56-5.87×10-4T)±0.3 J/m2 and the GB mobility M¯ = (2.14±0.15×10-7) exp(-(290±22kJ/mol)/RT) m4/(Js). We then discuss the pinning of grain boundaries by porosity, including porosity left over after sintering and fission gas bubbles that form during operation. We present our mechanistic model and validate it using existing grain growth data. Finally, we implement the model in the BISON fuel performance code and quantify its impact for constant and transient power cases. Our model produces similar results to an empirical model for the constant power case, but it predicts more grain growth than the empirical model for the transient power case. We attribute this discrepancy to the new mechanistic model’s ability to account for the impact of the temperature and power history. Building on this grain growth model, a coupled model for grain growth and densification in doped UO2 will be developed.