Parameter optimization of pulse current gas tungsten arc welding process using teaching learning based optimization and cuckoo search algorithm

The welding of aluminium alloy possess a great challenge to technologists. Due to ever increasing application in the industries the welding of similar and dissimilar aluminium required to be welded with rigour and success. The welding of 5xxx series aluminium alloys with 6 xxx series aluminium alloys requires most technical specific person as one is non heat treatable alloy and other one is heat treatable alloy. The difference in chemical, physical and mechanical properties enable the scientist to weld it concurrently as one property is needed in one area and another property needed in another area of weldment. Pulse current gas tungsten arc welding is applied to weld the two dissimilar aluminium alloys i.e. AA 5083-O and AA 6061-T651. The experimentation data in form of regression equation has been used as objective function for the fitness function. The two population based techniques teaching learning based optimization and cuckoo search algorithm (CS) has been employed to find the refined and homologous values. The optimized values obtained are put in predictive model and are compared with the experimental results obtained using design of experiment. The results indicate that the optimized values are within the permissible limits and are in line with the experimental values and higher than predicted values.