Modeling mechanical properties of welds in plates of commercial titanium alloys

Abstract strength, ductility and hardness were modeled for multi-pass welds deposited by gas tungsten arc welding in plates of commercial titanium alloys. Models are developed for analysis and prediction those mechanical properties by applying artificial neural network (ANN). The input parameters of the neural network are alloy compositions and heat treatment conditions and the output of the neural network are mechanical properties of the weld metal of titanium alloys, namely ultimate tensile strength (UTS), yield strength, elongation, reduction of the area and hardness. The titanium alloys used in the paper include commercially pure titanium, alpha or near-alpha titanium, alpha-beta titanium and beta or near-beta titanium