For the production of e-mobility components such as cable harnesses, battery cells, power electronics, etc., ultrasonic metal welding is well-established process of choice. These electrical applications require high quality for every single connection; single points of failure and no possibility of repair after installation or commissioning are state of the art. At present, the prevailing binding mechanisms and their sensitivity to the numerous process influencing variables like base material hardness, surface, and cleanliness are still the subject of research. In order to ensure sufficient quality despite the lack of process understanding, random destructive testing is carried out during ongoing production. The welding systems’ internal monitoring methods are currently not sufficient to make a prediction of the joint quality achieved. To determine process phases and extract features regarding the joint formation, the observation of process vibrations at the horn, anvil, and the components using laser-doppler-vibrometry, laser triangulation sensors or other suitable external measurement technology is common. These methods require external accessibility to the measurement position, not given in the industrial production environment. In this study, measurements of the high-frequency power signal of the welding system are conducted, and several machine learning models for quality prediction are set up. To ensure the robustness, several disturbances, e.g., changing material hardness and cleanliness, are taken into account. Thus, it will be evaluated to what extent an industrially suitable quality monitoring can be implemented by means of electrical measuring technology and how much more accurate such an external measuring system is compared to the possibilities already available in the welding system.

The production of electrical components for vehicles, battery systems, and power electronics components using ultrasonic metal welding is now well established. The industrial applications take advantage of the large joining surface and the associated excellent electrical and thermal conduction properties. However, this also prevents the weld quality from being monitored based on resistance measurements. For this reason, mechanical properties are used as a substituting quality criterion for process optimization and control. The type of mechanical testing depends on the component geometry, resulting in different quality features not necessarily comparable to each other. The ultrasonic metal welding process takes place in successive phases, which are characterized by the varying relative movements of the parts to be joined and the welding tools, prone to changes of material hardness, surface cleanliness, and geometric deviations. In this study, we investigate the influence of different mechanical quality features as target parameters for quality prediction on prediction accuracy and robustness. For this purpose, we consider the joint formation and the mechanical strength over the welding time for different quality parameters, fracture behaviors, and joint geometries. Supervised machine-learning quality prediction models are developed for the respective processes. The predictions are based on external and internal welding machine sensors. In addition to the sensor configuration, the importance of the individual process phases as a data source for the prediction accuracy is also investigated. We show that the overall process predictability strongly depends on the quality of the quality feature itself. Combining different sensor signals improves prediction quality. Furthermore, an analysis of the first process phases is sufficient for a robust quality prediction.

In ultrasonic metal welding, low specific resistances and large joining surface cross-sections require the use of mechanical testing to quantify the joint quality. In this study, different quality features of ultrasonically welded joints made of pure copper sheet are investigated during the successive phases of joint formation. Two test series with different workpiece geometries are examined. It is shown that mechanical quality features such as shear and peel forces behave differently over the formation of the joint and are not transferable. As an alternative to these, laser scanning microscopy is used to record images of the fracture surface that describe the growth of the joint area during formation. The study finds that shear tensile force growth and joint area growth are non-linear and comparable, with optimized welds achieving joint areas of 30 mm² out of 64 mm² and 6 mm² out of 16 mm². Although overall quality increases with increasing welding time, the material strength in the joint zone decreases. Depending on the original rolling condition, between 43% and 59% of the original material strength can be identified as the joint strength. The automatic analysis of fracture images is a suitable alternative to mechanical testing for similar joints.

Ultrasonic metal welding is a well-established solid state joining process for electrical applications. The process relies on the friction between workpieces and welding tools for joint formation. This friction is generated by the process force and the ultrasonic oscillation of the welding tools imposed on the workpieces. At such high frequencies, the occurrence of resonances in actual workpiece geometries is not surprising. It is known that critical dimensions in length and width lead to nearly no bond, depending on the welding frequency and the mechanical properties of the material. In real applications, this limits the possible designs of terminals and leads to extensive testing of clamping devices. It is also known that machine learning (ML) models for quality prediction based on power signals or tool oscillation can account for changes in welding position. In this study, we investigated the impact of part resonance and antiresonance on horn and anvil oscillation, power consumption and bond strength to identify typical behaviors induced by the workpieces. The influence of material thickness and roughness was considered, and numerical analysis of the natural frequencies of the workpieces was conducted. It can be shown that the results allow a distinction between the welding positions and workpiece geometries without directly measuring the oscillation patterns of the workpieces, allowing a simple validation of geometry weldability and clamping device in applications. Furthermore, the investigation allows the knowledge based specific deduction of signal parameters for future ML models, allowing a consideration of welding position and workpieces geometry with reduced test data.

Ultrasonic metal welding (USMW) is a widely used solid-state welding process for electrical components. Despite careful dimensioning, defective welds occur due to the large number of influencing variables, which sometimes remain undetected by existing monitoring strategies. One known major influencing factor is the surface condition of the components to be welded, which is controlled in industrial applications using mostly chemical or mechanical cleaning methods. Existing surface-based mechanisms in USMW are insufficiently described in the literature or contradict each other, so that corresponding cleaning methods are usually developed on an application-specific basis. To further complicate efforts to define an optimal process, the surfaces of the materials used are usually only standardized in general terms, but not for USMW. This work, therefore, investigates the suitability of the surfaces of commonly used copper and aluminium materials for USMW and then defines an optimal state. Based on this, a nanosecond-pulsed laser process is developed that best meets these requirements. In addition, the potential of optimizing the welding process of copper joints to the laser-treated surface condition to achieve a robust overall process chain and transfer the results to the specific requirements of aluminum-copper joints is demonstrated. The welded joints are characterized in detail mechanically, electrically and metallographically in order to quantify the surface influence on the welding result.

Ultrasonic metal welding (USMW) is a widely used solid-state welding process for electrical components. Since the joining zone remains largely hidden during the welding process and high-frequency oscillations are used to form the joint, there is only a limited range of technologies available for process monitoring. The strategies and sensor technologies used so far are not always able to detect process disturbances or defective welds. Consequently, intensive research is currently being carried out to obtain significant process information through additional sensor technology. One solution is to install sensors on the tools (horn and anvil) in the closest distance to the joining zone. Due to their flexibility and contactless measurement, laser vibrometers are frequently used in USMW research to determine the working and interfering frequencies and their amplitudes. However, the integration of such systems into production is limited by the high system costs, the lack of installation space, and the failure of sensors due to contamination. An alternative solution for determining amplitudes and frequencies is the use of piezoelectric shear force sensors in the anvil. This overcomes the challenges of limited installation space and contamination. At the same time, however, sensor integration poses new challenges such as reduced anvil stiffness and possible resonant excitation of the sensor, resulting in a sensor defect. In this study, we demonstrate a design process for a shear force sensor system for USMW, including calibration and validation, which provides high-frequency process information on the anvil side. It is shown in detail that piezo-based force sensors can be used as a cost-effective and integrable alternative to laser-based sensors in USMW processes. The correlation of the measurement data from 175 welding tests results in a mean coefficient of determination R ² ~ 0.986 with an error measure nRMSE of ~ 0.032 between signals of the shear force sensors and a vibrometer as a reference sensor.