Temperature monitoring system of friction stir welding based on digital twin
Xiaohong Lu,Zhuo Sun,Yihan Luan,Le Teng,Steven Y. Liang
DOI: https://doi.org/10.1080/0951192x.2023.2278113
IF: 4.1
2023-11-05
International Journal of Computer Integrated Manufacturing
Abstract:Temperature field distribution of friction stir welding (FSW) influences weld quality directly, so real-time monitoring of the welding temperature is significant. However, it is difficult to monitor the temperature in the core zone due to mechanical obstructions, material plastic deformation and complex thermal-mechanical coupling. To tackle this issue, this study develops a digital twin-based temperature monitoring system of FSW. Initially, a five-dimensional integrated framework based on the digital twin concept is proposed, outlining the process of building a digital twin-based temperature monitoring system of FSW. Subsequently, a motion simulation model of FSW is established, and synchronous motion simulation of the welding process is achieved. Real-time temperature readings from the FSW workpiece surface are gathered via an infrared thermal imager and synchronously transmitted to the FSW temperature monitoring system using socket communication. Additionally, a predictive model utilizing Support Vector Regression (SVR) is incorporated, enabling real-time and precise prediction of extremum temperatures in the core zone and over-limit alarm functionality. Finally, the temperature monitoring system of FSW, grounded in the digital twin concept and integrating the outlined models and features, is developed and experimentally validated. The system enables operators to exercise timely control over the welding process to ensure weld quality.
engineering, manufacturing,computer science, interdisciplinary applications,operations research & management science