Residual stresses can be a main cause of fractures, but forensic failure analysis is difficult because the residual stresses are relaxed after fracture because of the new free surface. In this paper, a method is presented for a posteriori determination of the residual stresses by measuring the geometric mismatch between the mating fracture surfaces. Provided the fracture is not overly ductile, so that plasticity may be neglected, a simple, elastic calculation based on Bueckner's principle gives the original residual stresses normal to the fracture plane. The method was demonstrated on a large 7000 series aluminum alloy forging that fractured during an attempt to cut a section into two pieces. Neutron diffraction measurements on another section of the same forging convincingly validated the residual stresses determined from the fracture surface mismatch. After accounting for closure, an analysis of the residual stress intensity factor based on the measured residual stress agreed with the material's fracture toughness and fractographic evidence of the failure initiation site. The practicality of the fracture surface method to investigate various failures is discussed in light of the required assumptions. © 2013 Elsevier Ltd.

Residual stress measurements using neutron diffraction and the contour method were performed on a valve housing made from 316 L stainless steel powder with intricate three-dimensional internal features using laser powder-bed fusion additive manufacturing. The measurements captured the evolution of the residual stress fields from a state where the valve housing was attached to the base plate to a state where the housing was cut free from the base plate. Making use of this cut, thus making it a non-destructive measurement in this application, the contour method mapped the residual stress component normal to the cut plane (this stress field is completely relieved by cutting) over the whole cut plane, as well as the change in all stresses in the entire housing due to the cut. The non-destructive nature of the neutron diffraction measurements enabled measurements of residual stress at various points in the build prior to cutting and again after cutting. Good agreement was observed between the two measurement techniques, which showed large, tensile build-direction residual stresses in the outer regions of the housing. The contour results showed large changes in multiple stress components upon removal of the build from the base plate in two distinct regions: near the plane where the build was cut free from the base plate and near the internal features that act as stress concentrators. These observations should be useful in understanding the driving mechanisms for builds cracking near the base plate and to identify regions of concern for structural integrity. Neutron diffraction measurements were also used to show the shear stresses near the base plate were significantly lower than normal stresses, an important assumption for the contour method because of the asymmetric cut.