Measurement and Control of Dynamic Weld Pool Surface in Gas Metal Arc Welding

The goal of this project is to establish foundations for building next generation intelligent gas metal arc welding (GMA W) machines so that the highly productive GMA W can compete with the much slower gas tungsten arc welding (GT A W) in critical applications and precision joining which currently require GT A W. The method is to emulate skilled human welders who acquire their process feedback through their observation on weld pool surface but enhance with machine powers their capabilities in conducting rapid adjustments and manipulation needed to sufficiently control the relatively complex GMA W. To this end, an innovative measurement method is proposed to measure the highly dynamic weld pool surface and metal transfer process simultaneously; the measurements are used to understand, analyze and model how the weld pool surface is determined by arc heat, arc force, droplet impact speed, droplet mass, and droplet detachment frequency; and a two-layer control structure is proposed to synergistically feedback control the weld pool surface and metal transfer. Although certain risks and possible modifications may not be avoidable, the PI has high confidence on the merits for the principles of the basic ideas proposed. If successful, this project will re-establish the foundations for GMA W process control. Intellectual Merit: The proposed research includes very challenging issues: (1) simultaneous imaging of dynamic weld pool surface and metal transfer process; (2) real-time image processing; (3) characterization of weld pool surface geometrically; (4) modeling of the optical system; (5) computation of weld pool surface; (6) modeling of dynamic response of weld pool surface under heat, force, and mass inputs; (7) modeling of dynamic response of metal transfer to current waveform; (8) model based adaptive predictive control of weld pool surface; (9) model based adaptive predictive control of metal transfer; (10) multiple layer control system. The completion of the proposed multi-disciplinary exploratory research requires the use of knowledge in a broad range of areas. It will not only establish foundations for building next generation intelligent GMA W machines so that GMA W can compete with GT A W in critical applications and precision joining, but will also provide data and ways to better study the complex GMA W process and help develop more trustable numerical models. Broader Impact: Welding is the final stage of manufacturing and US can not afford to give up its leadership. The completion of this proposed research will significantly help better understand the process and provide foundations for the development of next generation, intelligent welding machines. The multi-disciplinary research and collaborative activities will provide good opportunities for training the participating PhD students toward academic researchers and educators. The participation of undergraduate students will help them gain opportunities to lead the change of manufacturing industry from being "low-tech" to a high-tech industry. Also, the participating high school senior would learn the challenging part of engineering research so that he/she can be attracted to engineering field. Of course, the proposed research would also provide excellent materials for senior design projects and the development of the special website will provide a step-by-step procedure to guide researchers around the world to build the sensor for their research. In addition to GMA W, it will also provide a way to study the very complex laser-GMA W hybrid process.