Investigating Edge Cracking in Dual-Phase Steels: Testing, Mechanism, and Modeling

Sheet metal forming, particularly blanking, is crucial in manufacturing car parts, as it shapes flat sheets into precise geometries using tools and dies. Ensuring defect-free production requires accurate modeling to effectively predict and optimize the forming process. Localized necking is a critical failure mode in sheet metal forming and is typically assessed using the forming limit curve (FLC), which defines the strain conditions leading to failure. However, applying the FLC to Dual-Phase (DP) steels presents challenges due to edge cracking, particularly at shear-cut edges, occurring below the FLC. These edge cracks compromise part integrity and can lead to rejection. While alternative cutting methods, such as laser cutting, can mitigate edge cracking, shear cutting remains the most cost-effective and widely used approach despite its limitations. Various tests, such as the hole expansion capacity (HEC) test, can measure fracture strain at the edge. In this research, a novel in-plane bending test was used to investigate the edge ductility of DP800 after blanking. Unlike the hole expansion capacity test, this method eliminates the influence of punch friction, contact stress, and out-of-plane stress. Additionally, the strain gradient inherent in beam bending provides stable crack propagation, simplifying crack tracking compared to traditional tensile tests on sheared edges. Samples were prepared using shear cutting and subsequently deformed with the inplane bending test. Scanning electron microscopy, EBSD analysis, and microhardness measurements were employed to examine void distribution and non-uniform plastic deformation along the material edge. μCT scanning was used to reveal the extent of micro-cracks. The study investigated the effects of material orientation, cutting parameters, and the global strain gradient on edge fracture strain, showing correlations between edge ductility, material orientation, and cutting tool sharpness. The blanking process was found to induce an inhomogeneous void distribution in the thickness direction. During subsequent in-plane bending, micro-cracks initiated at the burr region and propagated towards the rollover region, followed by crack growth occurring away from the edge. The crack initiation location was correlated with high roughness, plastic deformation, and void volume fraction at the burr region. To model the two-stage edge forming process, which includes blanking and subsequent edge stretching, the Lou damage model was employed in Abaqus. To accountfor large deformations, remeshing was performed at predetermined intervals in the blanking simulation. The element deletion technique for crack initiation and growth was used. The model successfully captured the damaged area consisting of voids after shear cutting and replicated the fracture surface profile, including distinct regions such as the rollover and burnish zones. However, discrepancies in the exact size of these zones compared to experimental results suggest that the model needs additional input in the thickness direction. Subsequently, state variables were mapped from the 2D blanking simulation into a 3D model to reflect the initial state of the edge for the in-plane bending test. Furthermore, Gaussian scatter was introduced into the damage field, and the edge was perturbed to reflect the roughness induced by cutting, creating inhomogeneity along the edge. While the model successfully replicated edge fracture strain comparable to experimental results, accurately simulating edge crack initiation and growth remained challenging.