Attaining optimized A-TIG welding parameters for carbon steels by advanced parameter-less optimization techniques: with experimental validation

• Published in: Journal of the Brazilian Society of Mechanical Sciences and Engineering 2019-06, Vol.41 (6), p.1-19, Article 261
• Main Authors: Vora, Jay J., Abhishek, Kumar, Srinivasan, Seshasai
• Format: Article
• Language: English
• Subjects: Algorithms; Arc welding; Carbon steels; Chemical composition; Engineering; Gas tungsten arc welding; Heat affected zone; Machine learning; Mechanical Engineering; Multiple objective analysis; Optimization; Optimization techniques; Organic chemistry; Parameter identification; Penetration; Process parameters; Response surface methodology; Technical Paper; Welding current; Welding parameters
• ISSN: 1678-5878; 1806-3691
• DOI: 10.1007/s40430-019-1765-0

Enhanced penetration potential of activated TIG (A-TIG) welding process over conventional TIG welding has made the former process preferred for welding in recent times. The quality and shape of the weld in A-TIG are not only influenced by the chemical composition of the flux, but also by the selection of welding parameters. As a variety of process parameters influence the results, a proper understanding of process performance and identification of favorable welding conditions (optimum setting of process parameters) are indeed necessary to improve quality. The present work highlights the application potential of multi-response optimization route by integrating response surface methodology with the JAYA optimization algorithm, particularly for optimizing the A-TIG welding process parameters for carbon steels. Systematic experiments were carried out considering welding current, arc gap and travel speed as input parameters, whereas the depth of penetration, depth-to-width ratio, heat input and total width of heat-affected zone were considered as output performance characteristics. An attempt has been made in the current study to recognize the precise setting of selected input process parameters for simultaneous optimization of the aforementioned performance characteristics. The result of JAYA algorithm has also been compared with the teaching–learning-based optimization technique. While fairly similar results were achieved, the implementation of JAYA algorithm was computationally efficient. Experimental validation of the single-objective as well as multi-objective optimization results indicates that the empirical models for the response parameters as well as the proposed optimization framework are accurate tools in A-TIG welding research.