Realtime Prediction of Self-Pierce Riveting Joints - Prognosis and Visualization Based on Simulation and Machine Learning

Machine learning is used in many fields nowadays to predict events, be it a pure classification or the prediction of certain values. Thus, these methods are also increasingly mechanical joining technology, for example joint strengths, defects and rivet head positions discrete result values such as interlock. This paper further shows how complete contour including output stresses, strains damage can predicted visualized real time self-piercing riveting with semi-tubular rivet. First, classical sampling carried out experiments steel aluminum sheets different types thicknesses. These basis qualification numerical simulations. For this validation simulations compared via force curves. validated way several tool variants variation calculations each material-thickness combination. The simulation meshes thus generated database standardized respect comparability (same number nodes) data reduction performed. After testing approximation approaches, best possible results developed software demonstrator.