Determination of normal and shear residual stresses from fracture surface mismatch. (15th October 2015)

• Record Type: Journal Article
• Authors: Araujo de Oliveira, J.; Kowal, J.; Gungor, S.; Fitzpatrick, M.E.
• Is Part Of: Materials & design. Volume 83(2016:Jul.)
• Page Start: 176
• Page End: 184
• Publication Date: 2015-10-15
• Subjects: Residual stress -- Fracture -- Contour method -- Digital image correlation -- Shear stress; Materials -- Periodicals; Engineering design -- Periodicals; Matériaux -- Périodiques; Conception technique -- Périodiques; Electronic journals; 620.11
• DOI: 10.1016/j.matdes.2015.06.014
• Languages: English
• ISSNs: 0264-1275

Abstract

Graphical abstract: Highlights: A novel method for normal and shear residual stress determination was developed. Residual stresses were retrieved after the failure of a sample with known stresses. The information was extracted from measurements of the fracture surface topography. Experimental results compared fairly well to analytical and finite element models. The new technique can be a valuable tool in failure analysis. Abstract: The contour method of residual stress measurement has recently been adapted to measure fractured, rather than cut specimens. The fracture contour method was capable of determining normal residual stresses acting prior to the plane-strain failure of a large aluminium alloy forging, but shear residual stresses could not be measured (Prime et al., 2014). We demonstrate that the application of digital image correlation to topographic measurements of a fracture surface pair allows the determination of shear residual stresses in addition to the normal stress component. Miniature compact tension samples were extracted at an angle from a bent beam to give a known variation in normal and shear residual stress on the fracture plane. The material used was a metal matrix composite, which could be deformed plastically to introduce a known distribution of stresses and also present limited plasticity upon fracture, allowing plane-strain condition in a small specimen. The samples were fractured at cryogenic temperatures to further restrict plasticity. Although the fracture surface was non-planar and evidence suggested the occurrence of plasticity near the edges, experimental results correlated fairly well with the calculated normal and shear residual stress profiles.