Effect of sized and specimen geometry on the initiation and propagation of the ductile fracture

Strength to the fracture of the pipe in PWR has to be justified with mechanical analyses. These tests are based on the strength to ductile fracture of steels which are tested in lab. The values of resistance to fracture are obtained through tensile tests on CT specimens (determination of J-R curves). The purpose of this study is to justify the sizes of the specimens which have to be used to characterize the strength to ductile fracture of steel in secondary pipes. Tests were conducted on 0,5T-CT, 1T-CT and 2T-CT specimens. Two materials with different suffer contents were studied. The test results show that the JO,2 values gotten from the different specimens are similar. But the strength to ductile fracture in 2T-CT specimens in lower than the one measured in 0,5t-CT and 1T-CT specimens. The surface of fracture of the different specimens displays splits perpendicular to the notch and parallel to the sheet surface. These splits are produced by the separation of the manganese sulfur inclusions. The effect notes on the J-R curves seems to be relevant to these splits. The reason why these splits might be responsible for a decrease of the tearing modulus are not clearly defined up to this point. The results which have been published show the importance of the geometry effects (presence or not of lateral notches...) and the loading mode on the strength to ductile fracture. We note that the curves determined from tests on CT specimens are conservative. A few preliminary studies showed that the geometry effects on resistance to fracture can be studied and explained by using local approach methods. The Rousselier modeling is useful to explain the behaviour of ferritic steels in ductile fracture. (authors). 20 refs., 7 figs., 5 tabs