Weld quality characterization by vibration analysis for ultrasonic metal welding processes

• DOI: 10.1016/j.jajp.2023.100149
• Publication Date: 2023-07-05T01:28:25Z
• Authors: Florian W. Müller; Christian Mirz; Sascha Weil; Alexander Schiebahn; Burkhard Corves; Uwe Reisgen
• Keywords: Fixture; Robustness; Ultrasonic welding

Ultrasonic metal welding is an efficient, solid-state process that well established in the production of electrical equipment. However, isolated and untraceable failures occasionally occur industrial processes. According to state art, these are due a multitude influencing variables, such as material surface condition joining partners, tool fixture wear. This study presents quality prediction models based on measurements mechanical vibrations machine using laser triangulation sensors, sonotrode penetration depth measurements, internal power signals. The acquired data processed transformed into random forest model estimate shear strength welds. robustness has been successfully validated by experiments with significant external disturbances, roughness, contamination hardness. Within framework this study, development stable industrially applicable concept for monitoring demonstrated regression achieves mean relative estimation error 4.30% R2 value 0.964. Furthermore, classification determines disturbances individual welds was validated, achieving micro F1 94% macro 95%.