High strength steels are widely used for structural applications in which their fatigue behavior is a crucial design factor. During product development, various material characterization and qualification experiments are mandatory. For a faster and more cost-efficient development, data driven methods promise to replace or to complement the material testing by prediction of the fatigue strength. Therefore, a comprehensive database must be created, considering a multitude of processes and parameters, such as the metallurgical route and heat treatment, the chemical composition of the material, its surface condition, load and geometry parameters and many others. With an ontology-based, semantically linked knowledge graph, representing the manufacturing history of the material, such relationships can be accounted for.

In this presentation, our approach towards digitalization of the fatigue behavior is described, which includes assembling a database mostly from literature values, extension and implementation of a suitable domain ontology into EMMO, development of a process graph (A-Box template) for specific material variants, an appropriate data mapping workflow as well as ML-based predictions of the material properties. The talk is from a rather practical perspective, also highlighting the challenges we faced and are still facing in this comprehensive digitalization project.