ETD

• nuclear fuel
• OFFBEAT
• PCMI
• uncertainty and sensitivity analysis
• validation

Within the OperaHPC project, aimed at improving the computational modelling of nuclear fuel behaviour, this work presents a first Uncertainty and Sensitivity Analysis (UaSA) of the multidimensional OpenFOAM-based Fuel Performance Code (FPC) OFFBEAT applied to Pellet-Cladding Mechanical Interaction (PCMI). Although multidimensional FPCs can capture local effects more accurately than traditional one-dimensional approaches, they also involve greater modelling complexity and a strong dependence on input parameters. The impact of the corresponding uncertainties on PCMI predictions is still poorly documented and is therefore addressed in this study.
The UaSA is performed using a simulation framework that couples OFFBEAT with the DAKOTA toolkit and custom Python scripts. The Risø AN8 bump test is adopted as the reference experimental case. In accordance with Wilks’ theorem, 93 simulations are carried out to guarantee a 95/95 tolerance interval. A total of twelve uncertain input parameters is considered, and their influence is evaluated on two Figures of Merit (FoMs): cladding outer diameter and hoop stress.
The results show that the modelling of the cladding yield stress plays a key role in the accuracy and physical consistency of the simulations. A constant yield stress value of 250 MPa reproduces experimental diameter measurements reasonably well, but is not consistent with the actual operating temperatures (600–650 K). In contrast, the FRAPTRAN-based yield stress model provides a more realistic description of the cladding mechanical response and avoids non-physical stress predictions. The sensitivity analysis further identifies the relocation recovery fraction (RecFrac) and the maximum number of pellet cracks (nCracksMax) as the most influential parameters, with nCracksMax having a dominant effect on hoop stress uncertainty. Overall, these results support the qualification of multidimensional FPCs for safety analyses and point to key aspects requiring further model development.