Modeling and Optimization of A-TIG Welding Process Using Taguchi Method and Statistical Analysis

Document Type : Research Article


Ferdowsi University of Mashhad


In this paper, orthogonal array Taguchi (OA-Taguchi) method, regression modeling and analysis of variance (ANOVA) have been used to model and optimize the response parameters in A-TIG welding process. Welding current (I), welding speed (S) and welding gap (G) have been considered as process input variables in fabricating of AISI316L austenitic stainless steel parts. Depth of penetration (DOP) and weld bead width (WBW) have been taken in to account as process response parameters. In this paper SiO2, nano-particles have been considered as an activating flux. To gather required data for modeling and statistical purposes, OA-Taguchi based on design of experiments (DOE) has been employed. Then, process response parameters have been measured and their corresponding signal to noise (S/N) ratio values have been calculated. Different regression equations have been applied to model the DOP and WBW. Based on ANOVA results, the most fitted models have been selected as an authentic representative of the process. Furthermore, welding current has been determined as the most important parameter affects DOP and WBW with 68% and 88% percent contribution respectively using ANOVA. Next, S/N analysis, in such a way that WBW minimized and DOP is maximized has been used. Finally, experimental performance evaluation tests have been carried out, based on which it can be concluded that the proposed procedure is quite efficient (with less than 7% error) in modeling and optimization of A-TIG welding process.


Main Subjects