Inverse Identification of a 3D Anisotropic Yield Function Through an Information-Rich Tensile Test and Multi-sDIC

In the case of thick high strength steels, 3D anisotropic yield criteria are required to accurately capture the plastic material response. Material testing 2.0 aims at extracting material behaviour from an information-rich mechanical experiment with aid of inverse methods and full-field measuring techniques. In this context, inverse identification of 3D anisotropy through a single experiment is particularly challenging because of the large amount of anisotropy parameters. The difficulty lies in the specimen design that ideally ensures a simultaneous activation of all sought anisotropy parameters with sufficient sensitivity towards the observed strain field. In this contribution, we scrutinize the effectiveness of a specimen designed by shape optimization to inversely identify a 3D anisotropic yield function. This is done via a so-called Digital Virtual Twin (DVT) mimicking the actual measurement including the Digital Image Correlation (DIC). Given that the synthetic data is generated with a known material model used by the FE model, the identification accuracy of the Finite Element Model Updating (FEMU) approach can be objectively assessed. The novelty here is that the synthetic data consists of the data captured by two stereo-DIC (sDIC) systems. Moreover, the FEMU approach accounts for the multi-sDIC data sets. The DVT-results show that the adopted specimen design enables to simultaneously identify the sought anisotropy parameters. The proposed methodology is experimentally validated on S700MC steel with a thickness of 12 mm.