An evolutionary approach to parameter optimisation of submerged arc welding in the hardfacing process

Abstract

The Submerged Arc Welding (SAW) process finds wide industrial application due to its easy applicability, high current density and ability to deposit a large amount of weld metal using more than one wire at the same time. It is highly emphasised in manufacturing especially because of its ability to restore worn parts. SAW is characterised by a large number of process parameters influencing the performance outputs such as deposition rate, dilution and hardness, which subsequently affect weld quality. An exhaustive literature survey indicates that five control factors, viz., arc current,arc voltage, welding speed, electrode stick-out and preheat temperature, predominantly influence weld quality. In relation to this, an attempt has been made in this study to analyse the effect of process parameters on outputs of welding using the Taguchi method. The relationship between control factors and performance outputs is established by means of nonlinear regression analysis, resulting in a valid mathematical model. Finally, Genetic Algorithm (GA), a popular evolutionary approach, is employed to optimise the welding process with multiple objectives.