Development of a joining gap control system for laser welding of zinc-coated steel sheets driven by process observation

Laser deep penetration welding is already widely used in industry. However, a further increase in the number of possible applications is hindered by process instabilities leading to process defects for a variety of welding scenarios. To overcome this obstacle, sensing and control systems can be used to detect deviations in relevant process features and counteract by adjusting specific process parameters. Hence, there is a demand for process control systems to increase process stability and thus quality. To fulfil this demand, the development of a process sensor which acquires a broad variety of process emissions aiming to gain the full picture of the welding process is the ultimate goal. One way to find relevant relations between defects and process features is the correlation of a vast amount of data to a specific process defect. This approach is very time consuming and did often lead to unsatisfactory results if no correlations could be found between the sensor signals and the occurring process defects. For a more targeted development of a process control system, we propose the use of process observation methods with high temporal and spatial resolution. By the use of image and data analysis process features which characterize the evolution of defects can be extracted and the relation of specific factors to the characteristic process features can be found. Moreover, the mechanisms of the evolution of process defects can be observed and a tailored strategy can be designed to ensure a robust and reliable control of the process. By means of this strategy, sensors to measure the defect-related features, control interfaces, and actuators to change these features can be specifically chosen. In our paper, we illustrate these steps by the development of a joining gap control system for the laser welding of zinc-coated steel sheets in an overlap configuration. This process is a topic of current research and highly relevant for industrial application. In our study, we prove that the keyhole opening is a process feature which correlates with the joining gap between two sheets. Moreover, we show an exemplary design and build of a setup for adapting the joining gap with respect to the requirements given by the real-time applicability during the welding process. Afterwards, we merge sensor and actuator to a control system and validate it on different process situations. The validation shows that the gap can be measured and set during the laser welding process to reliably increase the process stability and quality. We conclude our paper with an outlook on how this system might be applied in an industrial environment.