Mechanical analysis of the bow deformation of fuel assemblies in a pressurized water reactor core

The permanent bow deformation of fuel assemblies (FAs) in the core of pressurized water reactors (PWRs) during irradiation may cause both safety and handling problems. The evolution of the FA bow deformation is considered to be a complex process with a large number of influencing mechanisms and several unknowns due to the limited knowledge about the boundary conditions and processes inside an operating nuclear reactor core. Since the first occurrence of strongly bowed cores, computational tools to predict the FA deformation have been developed to optimize the FA design and the FA loading pattern in the core. However, significant prediction errors persist, both regarding the bow amplitude and direction. The objective of this work is therefore to approach the FA bow modeling from a novel point of view, namely setting the focus on sensitivity and uncertainty analysis to assess the predictability of the FA bow patterns. To perform these analyses, a finite-element FA structural model is built up and is finally extended to a coupled row model of 15 FAs in the reactor core. To estimate the distribution of lateral hydraulic forces within the core row, a two-dimensional CFD (Computational Fluid Dynamics) model is created with a porous-medium approach. In addition, creep, irradiation growth, and spring relaxation models are developed to predict the evolution of the FA deformation during irradiation. The obtained in-laboratory and in-reactor FA model response is in good qualitative agreement with what is observed for FAs deployed in nuclear reactors. The sensitivity and uncertainty analyses, performed for both single FAs and a row of FAs, demonstrate that the uncertainties about the creep rate and the hydraulic conditions have a considerable impact on the bow amplitudes and directions. They may therefore fundamentally modify the bow pattern predicted with best estimate methods. It is concluded that FA bow calculations should always be accompanied by an uncertainty analysis to estimate the variability of the model predictions. To improve the predictions in the future, a specific effort must be invested in decreasing the uncertainty range of the concerned parameters.