Matthias Eder

For the numerical description of the material behavior of sheet metals, there is a large diversity of models. These models vary strongly in terms of complexity and calibration effort. Before using a material model for numerical analyses, it is essential to validate it. The paper presents an innovative tool that allows generating a well-founded database for a comprehensive validation of the yield locus description with few and simple tests: the MUC-Test (Material Under Control). The special geometry of the punch and die generates complex strain distributions in sheet metal samples during forming. The experiment covers the strain range between uniaxial and equibiaxial tension. The contribution shows the potential of this innovative tool by using the test results of three different material classes. Using numerical analyses, sensitivities of different material model parameters were investigated. Thereby, the focus lies on the curvature of the yield locus in the area between uniaxial and equibiaxial stress state.

To evaluate the prediction accuracy of the anisotropic yield function, we propose an original cruciform hole expansion test. Displacements on two axes were applied to the cruciform specimens with a hole in the center. The thickness strain in the region near the hole was compared to the simulation results. Because this forming test is free of friction and bending, it is an appropriate method to assess the anisotropic yield function without the influences of friction or the Bauschinger effect, or the need to consider the stress-strain curve in high-strain region. Hill1948, YLD2000-2D, and spline yield function which is an improved version of the Vegter model were selected, and 6000 series aluminum alloy sheets (A6116-T4) were used in this study. The parameter identification method of the spline yield function also proposed in this paper using the pseudo plane strain tensile test and optimization software. As a result, the spline yield function has better predictive accuracy than the conventional anisotropic yield functions Hill1948 and YLD2000-2D.