Determination of remnant residual stresses in fracture toughness specimens extracted from large components

Material fracture toughness data are required to undertake fitness-for-service assessments of engineering components containing cracks. Calculations of crack driving force in the component are compared with material fracture toughness values to assess the likelihood of subsequent failure. Experimental measurements of fracture toughness are usually made on small specimens extracted from a larger 'parent' component following strict experimental guidelines, formulated to ensure measured toughness values in the fracture specimens are appropriate for use in the full-size component. Implicit in this procedure is the assumption that the extracted fracture specimens contain no residual stresses, with any residual stresses in the full-size component being accounted for in the crack driving force calculation. This paper considers a recent conjecture within the structural integrity community that the extracted fracture specimens may themselves contain a residual stress field which may influence measurements of fracture toughness. This could potentially lead to a degree of 'double accounting', i.e. the effect of residual stresses may be included in both the material toughness and the crack driving force. This, in turn, could lead to unnecessary conservatism in safety assessments. To explore this conjecture, the results of numerical modelling and neutron diffraction measurements of residual stresses in fracture specimens extracted from two different welded parent components are presented. One of the components is significantly larger than the extracted specimens, with the other being marginally larger than the extracted specimens. Results confirm the intuitive expectation that the residual stresses in specimens extracted from much larger components are negligible, whereas if the dimensions of the extracted specimens are comparable with the larger component then significant residual stresses may remain.