Developpement and validation of a new methodology to assess ratcheting of PWR components

During the design, construction and operation of a nuclear component, it is necessary to ensure its integrity whatever the operating conditions: nominal or accidental. The demonstration of the components' resistance of the primary and secondary circuits to failure modes is necessary in order to validate the design of these structures. Among the possible failure modes is the phenomenon of ratcheting. The ratcheting check of nuclear power plant structures is mainly investigated by means of simplified methods or a complete inelastic analysis. Nevertheless, these methods are either conservatives or complex to use and implement. In this context, the aim of this thesis is to develop a new industrial design rule and/or new calculation methodology that is applicable to complex structures. The first chapter is addressed to the state of the art, in particular to the ratcheting phenomenon. The second chapter presents the modeling of metallic materials and the numerical methods to simulate cyclic calculations. We propose a new method for accelerating cyclic calculations in order to make the step-by-step integration method faster. The third chapter is devoted to the modeling of COTHAA tests. Constitutive models are evaluated in order to propose a robust model capable of simulating ratcheting. Results predicted by a simplified version of Chaboche model are found in good agreement as compared to experimental measurements. We also show the ability of the new acceleration method to simulate these tests. The fourth chapter is dedicated to the experimental study. We propose a new structural ratcheting test: the DEFPROG test. Secondly, we validate the model proposed in the third chapter on these experimental results. The fifth and last chapter is devoted to the proposal of the new design rule to forecast the risk of ratcheting. We propose and validate a new simplified method, while relying on experimental results and modeling. (author)