Prediction and Optimization of TIG Weldment Macrostructure Zones Using Response Surface Methodology

The present work discusses the modeling and optimization of tungsten inert gas arc welding (TIG) and graphical representation of weldment macrostructure zones' shape profiles based on full factorial design of experiments and response surface modeling. To accomplish the task, an in-house developed low cost arc image magnifying system was used. The welds were produced based on a full factorial design matrix of TIG process parameters such as welding current, arc length and traverse speed. Interaction effect plots of control factors indicated weld width is affected more by the welding arc length and current. Similar effects were also observed for depth of penetration and heat affected zone (HAZ). The full factorial experimental data was further analyzed for response surface regression modeling and the response surface regression equations were tested for a number of test case data for their adequacy. The response surface regression relations were further used to optimize the TIG welding process parameters using multi response optimization technique. A mapping scheme was also developed to map the TIG weldment cross-sections' macrostructure zones including that of HAZ. The responses predicted by the surface response regression equations for the test cases were further utilized for graphical representation of the welment shape profiles including that of HAZ. The modeling methodology, regression relations and weldment mapping scheme used in the present investigation were found to be adequate for the estimation of fusion and heat affected zone boundaries.