STYLE: MODELLING OF THE EXTRACTION OF BEND SPECIMENS FROM REPAIR WELD AND THE INFLUENCE OF RESIDUAL STRESS ON FRACTURE TESTING

摘要

Within the EU 7th framework programme, STYLE, a number of large-scale tests have been performed. One of these tests, Mock-Up 2 (MU-2), was performed on a through wall crack located at a repair weld adjacent to a multi-pass narrow-gap weld. The aim of MU-2 was to investigate ductile crack growth under conditions with significant levels of residual stress. As part of the materials testing programme, low-constraint fracture specimens (three-point bend specimens with a/t=0.1) were extracted from the weld to test the weld materials fracture toughness. An overview of these tests is provided here. However, these low constraint tests demonstrated somewhat unusual fatigue crack growth on inserting the crack, leading to the crack depth being shorter in the centre of the specimens to the outside. Subsequently, although it has not been possible to use these specimens to determine the materials J-R curve, it does provide a features test for ductile modelling with the Gurson-Tvergaard-Needleman (GTN) local approach model for ductile crack growth. This paper provides an overview of the modelling associated to understand these observations, including an estimate of the retained residual stress, fatigue growth estimates and subsequent ductile modelling. An overview of the calibration of the GTN model is also provided using the weld material's tensile tests, high constraint compact-tension tests and MU-2.