Zero Defect Manufacturing: Detection of Cracks and Thinning of Material during Deep Drawing Processes

摘要

The progressive automation and consistent rationalization in production of automotive components requires the use of new sheet materials as well as innovations in press and tool technologies. Sheet metal forming is getting more and more complex. Thus several structural defects like thinning, cracks and creases as well as geometrical imprecisions due to spring back may occur. At the same time, consumer demands are rising concerning product quality and value for money. Therefore optimized processes and efficient methods of quality control are required. As the main objective a 100%-in-process quality control throughout the process chain is demanded in press shops. The aim of the experiments presented here was to detect structural faults in formed body components. Thinning is often considered a failure criterion in the metal forming industry. At high-strained parts even cracks may develop as a result of excessive structural thinning. Especially in advancement of high-strength steels, edge cracking may be a significant failure mode in many sheet metal stamping processes. Furthermore research activities were focused on the quality of formed threads: their presence and accurate shape. Inline measurements by the help of electromagnetic and micromagnetic NDE techniques offer the opportunity for online measurements of physical parameters while feeding acquired data into calibrated regression models. Therefore changes in mechanical material characteristics can be predicted for ferromagnetic magnetorestrictive materials and failures can be discovered. Eddy current testing in conjunction with multivariate analysis methods have been used for quality control on formed body components. The method proved to be suitable to discover cracks and, in some cases, thinning. The results show a significant correlation between eddy current signals and critical amounts of thinning. Future research may investigate how to develop these testing methods for inline control.