In-process monitoring and control of weld nugget geometry for resistance spot welding process

Abstract

The spot welding process has been used extensively for joining of sheet metal, which is subject to many process parameter variation and is a good plant for an active process control. Among many variables associated with the welding process, electrode movement and dynamic resistance were examined for quality monitoring and control. Since the electrode movement is a mechanical response of the process, it was found to be more indicative of the weld quality and more adequate for weld quality control. And the correlation between the electrode movement and the weld quality was investigated by a series of experiment. The shunt effect and the electrode wear which have been a long-standing process disturbance were studied to examine their effects on the weld quality in connection with the electrode movement variable. Especially, numerical model was developed to predict the shunt effect and the electrode wear phenomenon was investigated by a series of extensive experiments. This study presents a conventional PI feedback control system that utilizes the electrodemovement as a feedback signal. The control algorithm was implemented to the actual spot welder with a micro-computer for real-time process control. The performance of this control system was evaluated through real-time compensation of the shunt effect and the electrode wear. Self-learning control of the welding process was also developed for more intelligent welding control system and its advantage over the conventional feedback control system was experimentally evaluated and discussed in detail.