As a supplier of system solutions and components for combustion, hybrid and electric drives, Vitesco Technologies from Nuremberg places the highest demands on the quality and reliability of the components it produces. "This is why the continuous monitoring of our resistance welding systems is so important," emphasizes Eduard Fackelmann, Head of FF2 Industrial Engineering at Vitesco Technologies. On two systems with several welding stations, a different number of pins are welded for each inverter - depending on the variant, this can be up to 30 connections. To do this, the copper busbars must be precisely aligned and lightly pressed together over their entire surface. The busbars are then contacted by the two electrodes and welded to the surfaces with high currents of up to 7,000 A and the resulting heat development. Due to the slight pressure during the welding process, the welding guns move minimally, between 400 and 800 µm depending on the pin type.

The current strength and the movement profile of the welding contacts during the welding process are key quality parameters. If the measured welding path exceeds a certain limit, for example, the wear on the electrode may be too great, indicating that it needs to be replaced. "We wanted to monitor these parameters online and evaluate them statistically in order to detect any drift in the relevant characteristic values in good time and be able to initiate countermeasures immediately - before quality-critical welds can even occur," says Eduard Fackelmann, outlining the project's objective.

Comprehensive data analysis and visualization

Since Vitesco Technologies has been automating with PC-based control from Beckhoff for over 20 years, these systems are also programmed with TwinCAT 2. "This naturally facilitated the integration of TwinCAT 3 Analytics as a statistical evaluation tool," says Christoph Rügheimer, application engineer at Beckhoff. With TwinCAT 3 IoT Data Agent, the data is retrieved cyclically from the controller and sent to a message broker running on a C6032 ultra-compact Industrial PC. This sends the data to a second C6032, on which the continuous 24/7 data analysis via TwinCAT Analytics and the web server for TwinCAT HMI run. This separation of data acquisition and data analysis was deliberately chosen in order to keep all options open for future expansions. "The experts from Beckhoff supported us in developing the algorithm and integrating it into the existing infrastructure," recalls Eduard Fackelmann.

The analysis evaluates the welding path for the last "N" inverter assemblies for each pin and assigns it to one of the three classes 1 (green), 2 (orange) or 3 (red). The correct data had to be identified and defined from the system's existing control program in order to enable the desired displays. In the detailed overview, the weld is displayed for each individual pin in the form of a smiley. On another dashboard, the results of each of the joints produced on each individual welding station can be displayed - also in the form of smilies, which were implemented in the dashboards at the request of Vitesco Technologies. A green peak indicates that an inverter assembly has been produced. The time interval between the peaks allows the machine operator to see at a glance what cycle time the machine is currently running at, whether the machine cycle is changing and, if applicable, how many minutes the machine has already been idle. "This simple visual display significantly increases the technical availability of the two welding systems and prevents errors," emphasizes Eduard Fackelmann.