A study on the parameter identification and failure prediction of ductile metals using Gurson–Tvergaard–Needleman (GTN) model

This study provides an investigation on the parameter identification and failure prediction of two ductile metals, SA516 and S30408, using the Gurson-Tvergaard-Needleman (GTN) model. Focusing on the physical background of each GTN parameter, a joint use of microscopic observation, damage monitoring, mechanical response of the cell model, and the FE trial and error was performed to determine the GTN parameters. Obvious differences were found in the three nucleation parameters of the two metals, indicating parameters suggested in previous studies cannot be directly used but independently calibrated for more accurate failure predictions. Experiments on the notched tension proved that void volume fraction experiences a steady growth before it reaches its critical value fc. After which, the voids growth rapidly, leading to the final fracture. Therefore, the parameter fc was suggested to be calibrated through matching the sudden drop point on the load-displacement curve of a notched tensile specimen. Once all eight parameters been calibrated, the GTN model can well predict the failure of both cracked and uncracked bodies. This prediction is not limited to the maximum load, but also the reduction of loading capacity due to stable crack propagation. In addition, compared with the fracture morphology from SEM observations, the FE calculation can well predict the critical fracture strain of the matrix of porous metals under different stress states. This experimental investigation provides insights on how to calibrate all eight parameters in the GTN model, and proves the applicability of GTN model on failure predictions of both macro and micro mechanical behaviors.