Crash behaviour of vehicles can be improved by the use of partial presshardened components . Although partial presshardening is a common technique, little understanding about the forming behaviour of the transition areas between the weak and the hardened zone has been gained yet. In crash simulations, these transition zones often are not considered, resulting in an abrupt change in mechanical properties with corresponding stress concentrations. Therefore, the results of these numerical simulations must be treated with caution. For crash simulations, in particular the forming behaviour at high strain-rates is of interest because presshardened components behave different at dynamic loadings compared to the quasi-static loadings . Material characteristics such as yield strength and elongation at break are affected by the loading speed. Even the start of instability and necking depends on the strain hardening coefficient and strain rate sensitivity. Therefore, the strain rate dependency of materials for strain-rates up to 1000 1/s  and the failure behaviour  is taken into account in crash simulations. For the description of failure behaviour, models like CrachFEM  or GISSMO  can be used. They consider different failure modes and multi-axial stress states. Especially the failure model CrachFEM uses the forming limit diagram (FLD) to describe the beginning of instability in crash simulation . In general, FLD can be used for the description of the material behaviour at multi-axial loadings. In the specific case of CrachFEM, the dynamic FLD used is theoretically calculated by means of a high-speed tensile test without any experimental verification. However, also in the FAT guideline  the importance of experimental characterizations at multi-axial loadings and high speeds is pointed out.
The results discussed in the present work reveal the forming behaviour at quasi-static and dynamic loadings of specimens with different microstructures, existing in the transition areas of partial presshardened steel. In a first step, characteristic cooling-rates (at partial quenching) and characteristic austenitisation-rates (at partial austenitisation) occurring in the transition areas were determined. Afterwards, the tempering of specimens with five different characteristic cooling-rates and austenitisation-rates, respectively, has been performed. By this tempering, the same microstructures as existing in the transition areas of components can be replicated homogeneously within the entire specimens. These specimens were tested in tensile tests at different strain rates up to 160 1/s. The results were used to calculate the FLD of the different microstructures existing in the transition areas with CrachFEM. Furthermore, an experimentally determined FLD of presshardened Usibor at a loading speed of 10 m/s is compared to a FLD at quasi-static loading. The changes in forming behaviour due to the increased loading speed are discussed. In addition, the dynamic FLD is used to verify results from CrachFEM calculations, which have been deduced from the experimental data of the high speed tensile tests. As a result, concrete recommendations for the consideration of transition areas in crash simulations could be derived.