METHODOLOGY FOR THE EXPERIMENTAL ASSESSMENT OF TRUE STRESS-STRAIN CURVES AFTER NECKING EMPLOYING CYLINDRICAL TENSILE SPECIMENS: EXPERIMENTS AND PARAMETERS CALIBRATION

摘要

Simulations and structural integrity evaluations including severe plasticity have undergone significant expansion during recent years (e.g. fracture mechanics FE models including ductile tearing and/or generalized yielding), which demand accurate true stress-strain data until fracture. This is a consequence of the use of high toughness ductile materials subjected to severe loadings and high levels of operational efficiency and optimization. However, tensile tests present one inconvenience when providing such data, since the occurrence of plastic instability (necking) complicates the direct assessment of true stress-strain curves until final fracture. Two main difficulties can be pointed out: i) the nonuniform geometry assumed by the cross sections along its length and; ii) the imposition of a complex triaxial stress state. The first occurrence can only be overcome by real-time physical measurements. The second occurrence demands a correction model to provide an equivalent stress including triaxial effects. Current authors recently demonstrated that even the well-known Bridgman's correction presents limitations, particularly for strains greater than ～ 0.50 - 0.60, which motivated proposals to better describe the geometrical evolution of necking minimizing the need for real-time physical measurements [1]. As a new step in this direction, this work presents three key contributions: i) first, experiments regarding the geometrical evolution of necking were largely extended incorporating 10 materials to corroborate the validity of the recently proposed model (including Carbon, stainless steels and copper); ii) second, and for the same materials, the necking region was investigated in more details to verify to which extent an osculating circle well describes the high deformation region. A new model could be proposed to better support future solid mechanics analyses regarding equilibrium and stress/strain fields; iii) finally, a modified Bridgman's model is proposed, followed by recommended practices for testing. The results provide further support to σ-ε assessment considering severe plasticity and demanding less physical measurements.