C05 - Met­ro­logy for join­ing pro­cesses and joints

Mechanical joining processes produce assemblies that are mechanically stressed due to the process and can therefore already be damaged during the manufacturing process. Furthermore, geometric deviations can occur in the function-relevant features of a joint. In this context, both a reliable in-process measurement of the joining process parameters and a non-destructive geometric post-process inspection for subsequent quality control are required. A method should be developed that contributes to an increase in the robustness of established joining processes by improving the in-process measurement. Another method contributes to a reliable assessment of the effects of process changes with computed tomographic post-process measurements, for which a crack-measurability limit is to be specified and shall add to the variability. For in-process measurement technology, dynamic real-time measurement uncertainty estimators based on a Bayesian approach are developed in cooperation with subprojects applying joining processes, which allow an improvement in accuracy and a measurement uncertainty statement for joining process parameters. For post-process inspection, this sub-project focuses on increasing measurability through artefact reduction and specifying a crack resolution limit for X-ray computed tomography (CT) measurements. Here, the focus of the investigations is non-destructive testing for qualitative statements on the formation of joining points and a dimensional measurement for checking the geometry of the joined component. For non-destructive testing, the question is adressed which surface defects can be clearly identified taking into account the metrological structure resolution as well as the interface structure resolution. For dimensional measurement with CT, multi-material components from the joining process represent an unsolved challenge in terms of measurability or achievable accuracy due to the error-prone surface determination both at the inner interfaces and in mono-material areas. By Using intelligently chosen multiple measurements as well as simulative prior knowledge, a detection of the interfaces and a more accurate measurement of the surfaces can be achieved. Finally, the geometry and damage of joints, also for modified joining processes, can be detected more accurately by CT measurements and the detectability can be predicted a priori. Bayesian filters can be used to measure process variables more accurately and provide real-time measurement uncertainty information.