Speaker
Description
ABSTRACT
In finite element analysis (FEA), improving the accuracy of the prediction capability of material cards to obtain the correct stress response of a specific material is still a major challenge in the field of forming technology. Furthermore, new aluminium alloys or steel grades are steadily released by the material manufacturers, differing in their attributes such as stiffness, strength or maximum elongation. And not only the material manufacturers, but also the FEA software tools developers are expanding their material libraries with increasingly complex constitutive models. This often requires extensive reverse engineering strategies and multiple experimental tests – primarily tensile tests – to achieve a sufficient quality of the material cards. As a result, more complex characterization strategies, time-consuming calibration processes and increased costs are inevitable.
The usage of optical measuring systems (DIC) during experimental tests is common practice in the area of material card calibration, since a lot of information of the material behaviour can be obtained, e.g. by means of grayscale correlation. But in most cases, material card calibration strategies are based only on the optimization of some of these information from the DIC, such as Lankford parameters or stress-strain curves. Previous publications have demonstrated how the data of the complete distortion field can be used via full-field calibration (FFC) for parameter identification of yield curve extrapolation approaches (Ilg et. al. 2019) or for the optimization of the yield locus (Hippke et. al. 2020). Grédiac et. al. 2006 utilizes the data from full-field measurements too and proposed the Virtual Fields Method (VFM), which identifies parameters based on the Principle of Virtual Work.
The theoretical investigation and software development presented by Liebold et. al. 2023 will be utilized within this work. Thereby, displacement driven simulations based on an optical measurement with the GOM/ARAMIS system are used for material card calibration. Therefore, the measured distortion field of a tensile test gets transformed into a LS-DYNA® input deck with the software Envyo®, allowing to locally provide the same or at least the interpolated displacements at discretized nodes in the FEA model and in the experiment. To investigate the improvements made by this approach and to show its applicability for different mesh sizes, a new tensile specimen providing a wide range of various stress states is presented. The iterative parameter identification process is performed using LS-OPT® and is applied to different material models that are implemented in the LS-DYNA® material library, including isotropic and anisotropic material models.
LITERATURE
Ilg, C./A. Haufe/K. Witowski/D. Koch/M. Liewald (2019): Parameter Identification using Full-Field calibration (FFC), 38th International Deep-Drawing Research Group Conference, Enschede, NL.
Hippke, H./B. Berisha/P. Hora (2020): A full-field optimization approach for iterative definition of yielding for non-quadratic and free shape yield models in plane strain, IOP Conference Series Materials Science and Engineering.
Grédica, M./F. Pierron/S. Avril/E. Toussaint (2006): The virtual fields method for extracting constitutive parameters from full-field measurements: a review, in: Strain: an International Journal for Experimental Mechanics, Vol. 42., pp. 233-253.
Keywords: Material Calibration; DIC; Optimization.