Influence of film-like contaminants on ultrasonic metal welding and their in-situ detection

Ultrasonic metal welding (USMW) is a solid-state welding process that is primarily used for joining electrical and electronic components. Common areas of application include series production of battery systems, power electronics and wire harnesses. The large number of disturbances in real-world applications, as well as a lack of basic knowledge about their effect on the welding result, cause difficulties to develop methods for quality prediction. In critical combinations, disturbances lead to faulty welds that can remain undetected by established monitoring strategies. Organic residues on the materials caused by production and handling have proven to be potentially critical, depending on the welding configuration. This study investigates the influence of various film-like, organic contaminants on USMW of copper sheets. For this purpose, a selection of contaminants is applied to the sheets to be joined and the welding process is monitored using a variety of sensors. In addition to evaluating their influence on tensile shear strength, suitable process signals are presented that can be used to reliably identify contamination and predict quality. For a more comprehensive analysis of the effects of contamination, differences in the thermal decomposition processes of the organic layers and their residues in the connection zone are considered. Particularly with a structure-borne sound signal at the anvil, reliable detection of film contamination is possible regardless of the welding parameters, even without complex algorithms. Maximum R² of 0.76 and a mean absolute error (MAE) of 150 N or ~ 6% in quality prediction can already be achieved using exemplary Extra Trees Regression with only 140 welds.