Optimization of weld bead geometry in GTAW of CP titanium using imperialist competitive algorithm

This paper aims to use a novel optimization algorithm called imperialist competitive algorithm (ICA) in order to optimize the weld bead geometry in the gas tungsten arc welding process. This algorithm offers some advantages such as simplicity, accuracy, and time saving. Experiments were conducted in order to collect welding data and obtain a relationship for the bead geometry as a function of welding current, arc voltage, welding speed, and arc length. Furthermore, a regression equation for depth of penetration and bead width was obtained using the least squares method, and the equations were optimized using ICA. Ultimately, the value of the input variables to obtain minimum bead width and maximum depth of penetration was calculated using ICA. Computational results indicate that the proposed algorithm is quite effective and powerful in optimizing the cost function.