A comprehensive review on the integration of artificial intelligence in friction stir welding for monitoring, modelling, and process optimization

Abstract

Recent advancements in artificial intelligence (AI) technologies have expanded their applications across various industrial environments, particularly in the field of Friction Stir Welding (FSW), a relatively modern manufacturing technique. AI techniques are primarily employed for modeling, monitoring, optimization, and management of complex systems influenced by multiple parameters within industrial processes. This study systematically reviews and evaluates commonly utilized AI techniques in FSW, highlighting their effectiveness, accuracy, and comparative advantages. The discussion is organized into three distinct sections, each focusing on the critical roles of AI and machine learning (ML) in FSW. The first section addresses process prediction, showcasing how AI techniques predict welding outcomes using historical data and process parameters, which enhances decision-making prior to actual implementation. The second section examines process control, emphasizing how AI systems enable real-time monitoring and adaptive control of the welding process. This functionality allows for immediate parameter adjustments, thus significantly improving weld consistency and quality by minimizing defects. Lastly, the third section pertains to the optimization of FSW parameters, illustrating how AI-driven algorithms analyze complex interactions among multiple variables to determine the most effective process settings. By adopting this structured approach, the review articulates the comprehensive benefits of integrating AI into the friction stir welding process, ultimately contributing to enhanced joint quality and improved operational efficiency.