Search Results

Forming Limit Diagrams (FLD) have been widely and successfully used in sheet metal stamping as a failure criterion to detect localized necking, which is the most common failure mechanism for conventional steels during forming. However, recent experience from stamping Dual-Phase steels found that, under certain circumstances such as stretching-bend over a small die radius, the sheet metal fails earlier than that predicted by the FLD based on the initiation of a localized neck. It appears that a different failure mechanism and mode are in effect, commonly referred to as “shear fracture” in the sheet metal stamping community. In this paper, experimental and numerical analysis is used to investigate the shear fracture mechanism. Numerical models are established for a stretch-bend test on DP780 steel with a wide range of bend radii for various failure modes. The occurrences of shear fracture are identified by correlating numerical simulation results with test data.

Die wear is a significant issue in sheet metal forming particularly for stamping Advanced High-Strength Steels (AHSS) because of their higher strength and microstructure composition. Reliable predictions of the magnitude and distribution of die wear are essential if cost-effective wear-protection strategies are desired in the early stages of tooling development. A die Wear Severity Index (WSI) is introduced in this paper to quantify the magnitude of die wear, which in essence characterizes the frictional energy dissipation per unit area on the die surface throughout the entire forming cycle. It can be readily obtained as part of any finite element simulation of stamping process utilizing incremental solution techniques.

A detailed investigation of influence of friction on drawbead restraining force modeling is presented in this paper. It is motivated by the need to accurately correlate line bead strengths, which are usually the output of an optimized draw development for controlling materials flow and achieving desired formability, and the physical drawbead geometries required for die face engineering. A plane-strain drawbead model with linear Coulomb friction is first established and the restraining forces corresponding to a range of bead penetration depths are obtained. The comparison of the simulation results with experimental data indicates that, while a larger Coefficient of Friction (COF) has better correlation for smaller bead penetrations and smaller COF does better for deeper bead penetrations, no single COF matches satisfactorily for overall range of bead penetration depths.

Flanging is a secondary operation in sheet metal forming processes. Traditionally, the design of flange shape and trim line is based on an engineer's experience. It takes several iterations to achieve the desired flange geometry because of potential splits. In this paper, an efficient CAE-based tool is developed to quickly predict the formability of a given flange design and enable the optimization of trim lines. A numerical algorithm is formulated in this CAE tool to convert the 3D flanging process into an equivalent in-plane deformation problem. The developed CAE tool is also integrated with the optimization software LS-OPT for trim line design.

The “adequate” number of integration points (NIP) required to achieve accurate springback simulation results is studied in this paper in an effort to clarify confusions reported in the literature and shed light on the origin of the confusion. A bending-under-tension model is adopted where springback solution can be obtained with analytical integration through metal thickness. Numerical integrations are then performed and compared with analytical solution to assess associated errors. A crucial distinction is made in the paper that, the model can be posed either as a displacement-value problem where both tension strain and bending radius are prescribed or as a mixed-value problem where the tension force and bending radius are prescribed. Although they are physically equivalent due to the uniqueness of solution, the numerical solutions are different. The associated errors in springback respond differently to the number of integration points employed.