Abstract
To produce batteries for electric vehicles, various welding processes are used, of which ultrasonic welding is an important one. In recent years, the quality control of welded components has become increasingly important due to fires and frequent discharges caused by battery heat. In addition, ultrasonic welding is sensitive to the process environment, so quality monitoring is essential. In this paper, we introduce a study to predict the quality of ultrasonic welding by measuring the energy generated during the welding process, the vibration of the horn and anvil, and the temperature of the weld. In particular, we will introduce various methods to measure the temperature of the weld, which is difficult to measure in the ultrasonic welding process, and a study that performs machine learning by fusing the energy and vibration obtained in the process.
Publisher
The Korean Welding and Joining Society