For the last decade, EDF has carried out several important civil engineering R&D programs, mostly dedicated to the analysis and behaviour of prestressed pressure containment vessel (PCCV) reactor building of EDF’s nuclear power plant. These concern the elaboration of new constitutive laws for concrete, techniques of modelling and resolution algorithms which are implemented in the finite element program of EDF, Code_Aster®.
The validity of models and more generally the methodology in carrying out nonlinear calculation must be obtained by comparing their performances with experimental results (benchmarking). These are generally obtained based on simple tests on small size specimens, where elementary features of models are qualified. More complex tests are also essential to determine the capacity of the calculations to predict the structural behaviour of more realistic and industrially representative cases.
The PACE 1450 EXP campaign is an intermediate size experiment to validate the following features in cooperation between EDF R&D and the Institute of Reinforced Concrete Structures and Building Materials of the University of Karlsruhe:
- Concrete constitutive models
Several material constitutive laws are available in Code_Aster®. Mostly based on damage mechanics they are isotropic and anisotropic. In some cases they are also coupled with plasticity to encounter for inelastic strains. The second family of constitutive models concerns the delayed strains induced by shrinkage and creep.
- Nonlocalisation algorithms
Constitutive models exhibiting softening suffer mesh dependency. The localisation of damage within local portions of the mesh requires the use of nonlocal adaptation of concrete constitutive models.
- Bond slip model
Crack pattern prediction and their opening are quite sensitive to the interaction between concrete and reinforcements.
- Prestressing tendons
Prestressing is commonly represented using uni-dimensional elements (1D bar elements) which combined with 2D or 3D concrete meshes can raise stress singularities. For that reason, even for elastic linear analyses results vanish with mesh size.
- Leakage / permeability
Determining the leakage across the RPC specimen requires the coupling of mechanical with transportation models. This is realised based on several assumptions within very restricted validity domain.
The design of the PACE 1450 EXP specimen is inspired from a real PCCV of an EDF 1450 MW nuclear power plant. The minimum size of the specimen (in terms of height and arc-length) has been selected in order to comply with the following requirements:
- Both global (displacements, strains, crack patterns, efforts, etc) and local (area of crack initiation, concrete stress/strain state close to tendons, strain in rebars, etc) observations remain as similar as possible between the specimen and the real behaviour of the common area from a real PCCV structure.
- A sufficient observation area remains available for measuring purposes, and far enough from boundaries in order to avoid any disturbances from loading points (anchorages, sealing devices, etc).