Multiscale Modeling of the Mechanical Performance of the Fuel-Clad Assembly in Light Water Reactors

The mechanical performance of the fuel clad assembly largely conditions the safety envelope of light water reactors. Due to the complex environments to be experienced by the fuel assembly, it is practically impossible to comprehensively test and characterize the response of the system -this is particularly true for off-normal operations-. Modeling efforts and tools can therefore largely be used to predict the expected response of the system. To this end, we present an original multiscale modeling framework enabling the simulate the response of the fuel-clad assembly under off normal conditions (e.g. Loss of Coolant Accident Scenario) as a function of the materials pedigree. Specifically, the approach consists of (i) developing, calibrating and validating a model for the mechanical response of Zircaloy 4 under tensile and creep loading scenarios, (ii) deriving surrogate constitutive laws from the use of the high-fidelity model and (iii) integrating these surrogate laws in finite element solvers. Polycrystal simulations are achieved via the use of LApx, a polycrystal plasticity solver. LApx relies on Fast Fourier Transforms (FFT) to compute the micromechanical fields within a polycrystal during loading. FFT based methods are particularly efficient from a numerical performance standpoint; thereby enabling the generation of simulation large databases. This work will present a new constitutive model, embedded in LApx, to capture the relative and absolute contributions of slip, climb and diffusion mediated deformation as a function of the materials texture, dislocation content, precipitate content and grain size. Critically, the model can simultaneously predict the tensile and creep response of Zircaloy 4 with great accuracy vis a vis experimental data.
LApx is then used to generate a large (i.e., 20,000 creep simulations) synthetic database of the materials response under creep loading scenarios. This database is then mined to extract simple constitutive laws for the polycrystal. This is achieved via the use of simple interpolation laws and mathematical mapping transformations of variables. The surrogate law is then integrated in the MOOSE finite element package. Following this, finite element simulations of the mechanical response and performance of a fuel clad assembly are performed.