PSI - Issue 28

Michael Jones et al. / Procedia Structural Integrity 28 (2020) 2078–2085 Author name / Structural Integrity Procedia 00 (2019) 000–000

2079

2

1. Introduction When analysing ductile failure, it is necessary to define how the ductility varies in relation to the level of stress triaxiality exhibited. This relationship is often defined by means of a failure locus, constructed through experimental tensile testing on multiaxial notch bar specimens. A variety of notch geometries are required to examine a range of levels of triaxiality. Historically, triaxiality values have sometimes been determined using expressions derived by Bridgman (1964), which only require knowledge of the sample’s cross-sectional radius and the radius of curvature of the outer profile. However, the validity of these expressions should be tested by comparison with values determined from Finite Element Analysis (FEA). In this work, notch bar tensile tests were performed on specimens of 316H stainless steel at 550°C, and the values for stress triaxiality and strain distribution determined across the notch throat from the tests using Bridgman’s estimation are compared with the results from FEA. 2. Bridgman Theory and Experimental Testing 2.1. Characterisation of the strains and stress state in a notch Three different notch profiles were used in the experimental testing as shown in Figure 1, Webster et al. (2004). According to Bridgman, the axial true strain in the notch region, ���� , is assumed to be uniform and given by ���� � 2 ln � 0 � (1) where 0 is the initial cross-sectional radius and is the instantaneous cross-sectional radius. Bridgman’s expression for stress triaxiality, , was modified by Wierzbicki & Bao (2004) to give � 1 3 � � √ 2 ln � 1 � 2 � (2) where is the notch radius of curvature. (a) Small notch geometry (b) Medium notch geometry (c) Blunt notch geometry

Figure 1 (a) to (c) – Notch geometries (dimensions in mm).