Zhuojin Pipe Fitting Co., LtdVisual Sensing Identification of Dynamic Weld Pool Width in Aluminum Alloy Pulsed MIG Welding
Literature Overview
This paper by Shi Yu, Fan Ding, Huang An, and Chen Jianhong, published in Acta Metallurgica Sinica (2005, Vol. 41, No. 9, pp. 994-998), presents a visual sensing approach for dynamic monitoring of weld pool width in aluminum alloy pulsed MIG welding. The research was conducted at the Gansu Provincial Key Laboratory of Nonferrous Metal New Materials, Lanzhou University of Technology, and supported by Gansu Provincial Key Technology Projects (JK9742181b) and the Key Laboratory Open Fund (SKL04012). The work addresses the challenge of real-time weld geometry control in aluminum alloy welding processes.
Technical Methodology and Findings
The authors established a weld pool image detection visual system and employed morphological processing to remove noise and cathode雾化区 artifacts from the image signal, yielding satisfactory weld pool edge images suitable for dynamic shape information detection. Step response experiments were then designed, and curve fitting methods were used to identify models relating base current, wire feed speed, and pulse current duty ratio to weld pool width.
| Process Parameter | Effect on Weld Pool Width | Control Strategy |
|---|---|---|
| Base current | Positive correlation | Primary width adjustment |
| Wire feed speed | Moderate positive correlation | Deposition rate control |
| Pulse current duty ratio | Significant positive correlation | Penetration modulation |
The identified models provide a theoretical basis for closed-loop control of aluminum alloy pulsed MIG welding processes. The morphological image processing approach demonstrated that reliable weld pool edge extraction is achievable even in the presence of arc light interference and spatter noise, which are common challenges in aluminum alloy welding due to the high reflectivity of aluminum and the tendency for arc instability.
Engineering Practice Implications
For production welding of aluminum alloy structures, real-time weld pool monitoring enables adaptive control that compensates for joint fit-up variations, heat sink effects, and process drift. In pipeline manufacturing, where weld pool width directly affects penetration profile and mechanical properties, visual sensing systems can be integrated into automated welding stations to maintain consistent weld geometry across varying joint configurations. The step response identification methodology presented here is directly applicable to building control algorithms for automated welding systems, where the identified transfer functions can be incorporated into PID or model predictive controllers.
Study Insights and Concluding Remarks
This work represents a foundational contribution to visual-based weld monitoring technology, demonstrating that morphological image processing can effectively isolate weld pool boundaries from complex arc and spatter backgrounds. The systematic identification of process parameter effects on weld pool width provides engineers with quantitative models for process optimization. For aluminum alloy welding in particular, where thermal management is critical due to the high thermal conductivity of aluminum, real-time pool width monitoring enables timely adjustment of welding parameters to maintain adequate penetration without excessive heat input. This approach is directly transferable to automated welding cells in pipeline and pressure vessel manufacturing, where consistent weld quality across variable joint conditions is essential for structural integrity.