In-situ detection of forging defects without sensors in the forging tool

The early detection of defects in forged parts offers economical advantages due to the possibility of sorting them out of the process chain betimes. Hence, no further production costs are occurred by those parts. However, the detection of defects during forging processes is not state of the art. Usually defects of forged parts - which are often not obvious and therefore cannot be identified by the press operator - are detected by measurements after cooling or later by the customer. In both cases costs occur. Due to the rough conditions in forging shop floors (e. g. high temperatures and forces) the integration of sensors monitoring the forging process is hardly viable. Within the presented research project an in-situ forging process monitoring system is developed which works without sensors in the forging tool and shows results on the process quality directly after forging. A system consisting of two independent concepts (defect-identification and filling-identification) is developed to detect typical forging process failures. On the one hand the defect-identification is able to detect e.g. wrong press energy or insert-position. On the other hand the filling-identification aims to detect, if the cavity is completely filled by the work-piece. Both concepts work in real-time with the help of a computer and a screen visualising the results. One the one hand this paper presents the results of forging trials using the defect-identification and evaluates its capability. On the other hand the concept for the filling identification is presented.