[SY-B7] Symposium B-7

[SY-B7] Why Multiscale Modeling of Nuclear Fuel is Absolutely Essential and Why it is so Challenging

Invited

^○Michael R Tonks¹, Yongfeng Zhang², David Andersson³, Simon Phillpot¹, Christopher Stanek³ (1.University of Florida, United States of America, 2.Idaho National Laboratory, United States of America, 3.Los Alamos National Laboratory, United States of America)

The performance of reactor fuel is historically determined using codes that rely on empirical fits to experimental data. However, as reactor operation becomes more variable and as we consider changes to the fuel and cladding system to increase accident tolerance, these empirical models become insufficient to meet our needs. The US Nuclear Engineering Advanced Modeling and Simulation (NEAMS) program is beginning to implement an alternative approach to fuel performance materials models that relies on representing the evolving state of the fuel microstructure using state variables that impact the material properties. However, developing the required models for this approach using only experimental data would be extremely difficult, time consuming, and expensive. Hierarchical multiscale modeling and simulation, ranging from density functional theory to mesoscale simulation approaches, provides a powerful means to obtain additional data to inform the development of these new materials models. However, applying this multiscale approach to reactor fuel has a number of challenges. In this presentation I will summarize this multiscale approach being taken in the US and discuss a number of the issues that make it difficult.