METHODS FOR THE EXPERIMENTAL EVALUATION OF TRUE STRESS-STRAIN CURVES AFTER NECKING OF CONVENTIONAL TENSILE SPECIMENS: EXPLORATORY INVESTIGATION AND PROPOSALS

Numerical elastic-plastic simulations have undergone significant expansion during the last decades (e.g. refined fracture mechanics finite element models including ductile tearing). However, one limitation to increase the accuracy of such models is the reliable experimental characterization of true stress-strain curves from conventional uniaxial tensile tests after necking (plastic instability), which complicates the direct assessment of the true stress-strain curves until failure. As a step in this direction, this work presents four key activities: i) first, existing correction methods are presented, including Bridgman, power law, weighted average and others; ii) second, selected metals are tested to experimentally characterize loads and the geometric evolution of necking. High-definition images are used to obtain real-time measurements following a proposed methodology; iii) third, refined non-linear FEM models are developed to reproduce necking and assess stresses as a function of normalized neck geometry; iv) finally, existing correction methods are critically compared to experimental results and FEM predictions in terms of potential and accuracy. The experimental results evaluated using high-definition images presented an excellent geometrical characterization of instability. FEM models were able to describe stress-strain-displacement fields after necking, supporting the exploratory validations and proposals of this work. Classical methodologies could be adapted based on experiments to provide accurate stress-strain curves up to failure with less need for real-time measurements, thus giving further support to the determination of true material properties considering severe plasticity.