Modeling of pulsed TIG welding parameters and optimization of the mechanical properties using statistical approaches

In this paper, an attempt is made to compare the feasiblity and effectiveness of Taguchi method and Simulated Annealing (SA) Algorithm by means of regression analysis to find optimum welding parameters of the pulsed TIG welding process. The response considered is Notch tensile strength (NTS). To develop the mathematical models, various linear and curvilinear regression functions have been fitted to the experimental data. The best model is selected and Simulated Annealing (SA) Algorithm is then used to optimally determine input parameters levels in order to obtain any desired output. Finally, Taguchi method is applied to find the optimum parameters settings. The results show that Both the Taguchi method and regression + SA found the optimum solution in a reasonable time; therefore both of these methods are able to find the optimum solution in such problems.