Representative Structural Volume for French containment buildings: from numerical computations to experiment
Résumé
In high power French nuclear power plants (1300 and 1450 MWe), the internal confinement vessel represents the third passive barrier in case of accidents. That is why it is carefully monitored, specially by performing integrity tests to ensure its durability.
As mechanical experiments on full scale models are too complicated to carry out, numerical tools have been developed to carry out simulations capable to study the mechanical behaviour of the prestressed concrete containment vessels (PCCV). For the preparation of these simulations, a Representative Structural Volume (RSV) has been designed, to reproduce, as accurately as possible with acceptable computational costs, the behaviour of the common area of the internal vessel of a 1300 MWe reactor plant. The model contains concrete, steel reinforcement bars in all three directions and prestressed cables in both hoop and meridional directions.
The RSV was considered as a numerical example to evaluate the ability of available constitutive laws of concrete cracking (damage theory, with or without plasticity, etc) to be applied to a preindustrial model. First the global mechanical behaviour of the structure (pressure – displacement) is studied and analysed, where several specific observations are underlined. Especially, the side effect of prestressing tendons which seem to initiate micro damages and to propagate them. Mesh dependency is also studied to highlight the need for a regularisation technique. The need for several specific experimental validations is demonstrated.
An experimental setup has thus been designed to validate both qualitatively and quantitatively, the observations made thanks to simulations. With a special care for the representation of the prestressing. Dimensions and experimental processes to be realised are presented. Preliminary computations are proposed to evaluate the expected mechanical behaviour of the experimental model. The mode of fracture is especially highlighted and a parametric study is also proposed to emphasize the mesh dependency of calculation results. On this basis, numerical and experimental results will be complementary, since numerical studies will help the calibration of the experiment (maximum bearable pressure), while experiments will help to validate, or not, the role of the prestressed tendons and the necessity for this type of structures to use regularised techniques.