MODELLING OF SLANT FAILURE USING SMALL SIZE SPECIMEN

摘要

The instability of a pipeline crack eventually leads to brittle or ductile crack propagation. The resistance to ductile crack propagation is assessed by the energy dissipated in the CVN test. However the Charpy specimen exhibits mainly mode I failure, with no small shear lips, while real failure is a combined mode often described as slant failure. In the present investigation, instrumented Charpy tests with nominal and reduced thickness down to 2.5 mm are carried out. Instrumented Battelle drop weight tear tests where also performed with nominal and reduced thickness, in order to vary the ligament versus thickness ratio. The results of the Charpy tests are simulated by the finite element method. The results are then discussed in terms of energy dissipated during crack initiation and crack propagation. It is shown that by reducing the size of the Charpy specimen, slant failure is promoted, which results in a decrease of the specific energy absorbed. However, most of the difference of absorbed energy is in the crack initiation mode, and only marginally in crack propagation. Consequently, the fraction of the total energy dissipated in crack propagation is increased by reducing the sample thickness, making it a possible tool to assess the resistance of a material to crack propagation, provided that brittle fracture is avoided and no separation is present.