Title: Using fuzzy logic to predict the influence of the tool shoulder geometry of friction stir welded Al 6082 T6 alloy

Authors: Mohamed Rafik Noor Mohamed Qureshi; Dhairya Vyas; Saumil K. Joshi; Karishma M. Qureshi

Addresses: Industrial Engineering Department, College of Engineering, King Khalid University, Abha, 62529, Saudi Arabia ' Computer Science and Engineering Department, Faculty of Technology & Engineering, The Maharaja Sayajirao University of Baroda, India ' Mechanical Engineering Department, Sigma Institute of Engineering, Gujarat, India ' Department of Mechanical Engineering, Parul University, India

Abstract: Friction stir welding (FSW) and its variants are important welding processes in many industries, including aerospace, railway, robotics and computers. Since welding plays a vital role in enhancing production and productivity, the effect of tool shoulder geometry on weld quality must be investigated. The weld quality is affected by tool geometry, welding speed, tool traverse speed, tool inclination angle, and so on. Consequently, the interaction of such parameters influences the weld quality, which becomes difficult to predict. In this research, welding was performed on Al 6082 T6 alloy using two separate shoulder geometries (raised and recessed shoulder) at three different welding rates and tool transverse speeds. Further, the ultimate tensile strength (UTS) and the microhardness of the material were used in weld quality evaluation. Two adaptive network-based fuzzy inference systems (ANFIS) were used to train and evaluate the UTS and microhardness, respectively. The Takagi-Sugeno fuzzy inference system was used to find the effect of tool shoulder geometry on the weld quality.

Keywords: adaptive network-based fuzzy inference systems; ANFIS; artificial intelligence; artificial neural network; ANN; friction stir welding; FSW; genetic algorithm; Al 6082 T6 alloy.

DOI: 10.1504/IJMATEI.2024.137536

International Journal of Materials Engineering Innovation, 2024 Vol.15 No.1, pp.1 - 16

Received: 05 Jul 2022
Accepted: 08 Aug 2022
Published online: 25 Mar 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article