Application of local approach to fracture of an RPV steel: effect of the crystal plasticity on the critical carbide size

Avoidance of brittle fracture is a key issue of the integrity assessment of the nuclear reactor pressure vessels (RPV). The microstructural heterogeneities of these heavy components are in direct relation with the scatter of brittle fracture toughness. A microstructure informed brittle fracture (MIBF) model is used here to capture the effect of some microstructural features. This local approach type model is based on the Griffith criterion for brittle fracture applied to micro cracks nucleated on carbides. The size distribution of the cleavage initiators introduces a source of scatter for the fracture stress. The originality of the MIBF model is to introduce a second source of scatter: the stress distribution inside a representative aggregate of the bainitic microstructure. These stresses are the average for each bainitic packet of the maximum principal stresses. Their non-uniform distribution is due to the incompatibilities of plastic deformation introduced by crystallographic misorientations between packets. Finite elements simulations on aggregates representative of the steel bainitic microstructure are performed to identify the stress distributions. The size distribution of cleavage initiators is the carbide size distribution as the carbides cracking induced by plastic deformation is considered to be the main damage leading to cleavage propagation in RPV steel. The only free parameter of the model is the effective surface energy gamma f. Applied to the Euro fracture toughness dataset, the MIBF model predicts the specimen size effect on fracture toughness as well as the temperature dependence of the fracture toughness up to 200 MPa root m. The analysis of the numerical results shows that taking into account the incompatibility stresses widens the size range of carbides potentially implied in the fracture process. However, approximatively half of these carbides have still a size larger than the largest observed sizes underlining the importance to perform the determination of size distributions on a very large number of precipitates. Copyright (C) 2016 The Authors. Published by Elsevier B.V.