Modelling optimisation of hot processing parameters of Ti-6Al-4V alloy using artificial neural network and genetic algorithm

Optimum condition of hot processing parameter with the premise of achieving the desired mechanical properties is often carried out from the corresponding references and checked subsequently using the trial-and-error approach, which consequently costs large amount of time and is highly associated with the experience and skills of the materials scientists. As the key branch of soft computing, the technique of combined artificial neural network (ANN) and genetic algorithm (GA) has been widely utilised to predict and optimise the processing parameter. In this work, the forging and heat treatment experiments of Ti-6Al-4V alloy were conducted with various experimental parameters, including forging temperature, deformation degree and heat treatment. The mechanical properties including the ultimate tensile strength, yield strength, elongation and reduction in area of this alloy were obtained by the tensile tests at room temperature. On the basic of obtained experimental data, the optimal model of hot processing parameters of Ti-6Al-4V alloy was established with the help of the combination of ANN and GA. It was found that the final research result is consistent with the actual experimental value, suggesting the fact that the developed optimisation model is an available and powerful tool to deal with multiple variables and nonlinear relationship problems, and can be widely utilised in the area of materials science and engineering.