Damage Evolution of 6005A Aluminum Alloy Sheet Based on Gurson-Tvagaard-Needleman Model: Experiment and Finite Element Simulation

A Gleeble-3800 thermal simulation machine was used to conduct high-temperature tensile test on 6005A aluminum alloy at 350-450 degrees C and a strain rate of 0.01-1 s(-1). True stress-strain curve and microstructure morphology of the material are obtained. The microstructure morphology is analyzed under different temperatures and strain rates. The effect of damage parameters on the mechanical behavior of materials is investigated using backpropagation (BP) neural network modeling and genetic algorithm optimization to determine Gurson-Tvagaard-Needleman (GTN) model damage mechanical parameters of f(0), f(N), f(c), and f(f) of 6005A aluminum alloy. Scanning electron and energy-dispersive x-ray spectroscopy methods combined with imaging Pro Plus software are used to measure void volume fraction. The findings are compared with those of neural network modeling and genetic algorithm optimization results. Results showed that model damage mechanical parameters of 6005A aluminum alloy determined via BP neural network modeling combined with genetic algorithm optimization are reliable.