SIMUSSCopyright: © IGMR
Metal Ultrasonic Welding, short USMW, is a solid-state welding process, which is widely spread in industry. Due to the low heat input, the process is particularly suitable for demanding applications such as mixed material joints and electronic components. In the context of the increasing electrification of vehicles, the industrial application of the process in battery technology, for wire and conductor assembly as well as power electronics is gaining even more importance. Despite the industrial significance, the process is not yet fully understood. Because there is a lack of scientifically founded knowledge about the complex interactions of tools and joining parts during the welding process, process fluctuations cannot be scientifically explained and monitored.
Against this background, new quality monitoring methods are required for the USMW process, which can be used to draw conclusions about the weld quality based on process information. Overall, there is currently a lack of an industrially applicable monitoring method for recording the condition of the welding system and the oscillator system in the joining process. Therefore, the information required for weld quality prediction is at present not at hand. A vibration analysis of the welding system is a promising approach here.
Within the framework of this project, a vibration-based condition monitoring system for quality assurance is to be developed. Due to an increased variety of the measured data against classical monitoring methods, the vibration analysis of the welding process supports a comprehensive understanding of the process as well.
First, a test bench with USMW machine was set up and systematically analysed. The vibration behaviour of the entire system like a vibration system incl. machine frame as well as the necessary measuring technology and measuring positions were determined. In order to generate reference data for condition monitoring, welding tests with constant boundary conditions were carried out and then successively individual weld and machine dependent influences were examined and their measurement results evaluated. Using machine learning methods, the processed vibration-related measurement data are correlated with the results on the weld quality and a monitoring procedure is developed. At the end of the project, the monitoring method for the USMW process will be optimized for later industrial application and finally validated in a field test.