Contribution à la surveillance des structures épaisses en béton : proposition d'une démarche pour intégrer le suivi de la teneur en eau dans le pronostic

This research takes place within the framework of the structural monitoring of nuclear power plant containments that EDF operates. Our goal is to determine how the knowledge of the concrete water content these structures could improve the predictions of their mechanical behavior, which are periodically undertaken as part of the safety case for the facilities.As a first step, we have studied monitoring data uncertainties propagation through the models, in order to quantify the information brought by the water content measurement. We have also addressed the effects of these uncertainties on the extrapolations to the end of the structure lifetime. In a second phase, we have proposed a delayed deformation monitoring model including water content as an explanatory variable and based on a linear dependence with strains. This approach has been validated with data from the VeRCoRs mock up.Then, we sought to estimate more accurately the metrological requirement to qualify in-situ measurement systems for concrete water content. To do this, we used the Monte Carlo method to simulate the propagation of measurement uncertainties, by comparing the performances of the model that we proposed with those of a more conventional engineering approach.