Development of a Damage Prediction System for Bending and Cutting of High Strength Steels

Abstract. The application of modern high strength low alloyed steels (HSLA) and advanced high strength steels (AHSS) for structural and safety relevant components in the automotive industry offers the advantage of combining low specific weight with high material strength. Typical manufacturing processes for these steel grades are bending and cutting operations. The forming and cutting potential of these innovative steel grades is different to conventional steels as the process and the damage behaviour is changing. In bending operations cracks occur at the outer bending edge, whereas in cutting operations delamination can appear at the sheared edge. These damages, even though they are small, can initiate the component to fail. For a reliable use of such materials in industrial application a method for the process design is essentially needed. In particular, damages have to be predicted at an early stage. In industrial application damage is detected by a trial-and-error approach causing significant work and a high failure rate. A system for an offline assessment of the risk of failure is unknown so far. In the scope of this work, a method is presented to describe the damaging behaviour of both, bending and cutting operations, by theoretical metamodels. In order to generate a database experiments were carried out using different high strength steels. The main influence factors have been varied, such as the rolling direction, the punch-to-die clearance and the cutting contour in the cutting operation. The bending was investigated using an air-bending process varying the bending angle, the bending radius and the rolling direction. To calculate further sampling points a finite element model has been developed and validated against the experimental data. The damage criterion of Lemaitre has been applied. The necessary parameters were determined by reverse identification by means of the major strain for the bending operation and by the punch force-punch stroke curve for the cutting operation. To build up a system for the prediction of the damage the gained data basis was approximated by mathematical functions. An error analysis was carried out showing good accordance. In doing so, a metamodel for the occurrence of damages could be established. The functions are implemented in a software tool which allows the user to determine the failure probability for a given parameter set.