Optimization of pulse current gas tungsten arc welding parameters using artificial bee colony optimization and simulated annealing

The welding of dissimilar aluminium alloys in 5xxx and 6xxx series are difficult to weld as defects are produced. The one of the cause is higher heat input. Therefore to reduce the heat input, pulse gas tungsten arc welding (PCGTAW) is employed. The pulsing character of welding process thus reduce the defects and increase the properties of weldment. The fine tuning of parameters such as peak current (P), base current (B), pulse frequency (F), pulse on time (T), and welding speed (S) is carried out using simulated annealing (SA) and artificial bee colony optimization (ABC). The parameters selected has higher percentage contribution in effecting the mechanical properties. Optimizing them will improve the properties of the material. As they possess the approach of reaching the global optima as compared to algorithms which has early converge or converge to local optima. The paper utilizes the data collected using experimentation and is used for homologous of process variables of parameters of PCGTAW using ABC and SA. The empirical relation equation obtained acts as problem objective for optimization through ABC and SA. The confirmation test indicates the values obtained using SA is accurate by maximum of 3.35 J and maximum improvement percentage is 1.09% of impact toughness and in artificial bee colony the maximum accuracy is of 5.6 J and maximum improvement percentage is 1.83%. The higher mechanical properties are obtained as compared to the optimization in design of experiments.