Empirical determination of full true stress-strain relations for ductile metal bolts using photogrammetry with computer vision

Abstract

Uniaxial tensile testing of ductile metal bolts or cylindrical bars is commonly used to obtain engineering stress-strain relations, true stress-strain relations until the onset of necking, and values of true stress and strain at fracture. However, true stress-strain relations between the onset of necking and fracture are not typically obtained, and thus not reported, in experimental studies. This presents a challenge for detailed numerical modelling of such components, which require equivalent stress-strain relations – determined from complete true stress-strain relations – as inputs for constitutive models. This study presents an efficient method for empirically determining full true stress-strain relations for ductile metal bolts or cylindrical bars from standard uniaxial tensile testing, using photogrammetry and computer vision, without the need for additional specialised equipment and expensive software. The method is used to obtain empirical full true stress-strain relations for small-diameter M4 and M5 bolts. The empirical results are compared to theoretical post-necking true stress-strain relations obtained using a widely used extrapolation method through a detailed numerical example. The empirical method presented is shown to obtain more accurate true stress-strain relations from standard uniaxial tensile testing than the extrapolation method. These findings suggest that the proposed method can improve the quality of material testing results reported by future experimental studies in support of more realistic numerical modelling and more efficient steel bolted connection design.