Validation of axial and transverse force–displacement responses and principal strain rate ratios in the critical zone as a precursor to anisotropic damage prediction in metal sheets

Most ductile damage models account for the triaxiality η-dependence of fracture strains, provided by the triaxiality failure diagram (TFD). η-values are commonly obtained from FE analysis. By contrast with von Mises materials, for anisotropic materials, η depends on yield criterion and material. The common procedure of matching numerical to experimental axial force–displacement (FD_L) responses may be insufficient because the transverse response (FD_W) is not validated. Further, the use of extensometers blurs local events. In this study, assuming different anisotropic yield criteria, we investigate for highly anisotropic Zirlo sheets (i) on a global level the FD_L and FD_W curves and (ii) on a local level the evolutions of the principal strain rate ratio β’, which is related to η. Tests with uniaxial tension, notched and shear specimens were carried out. Strains were measured by DIC. The pre-critical region reveals that the yield model needs to account for anisotropy regarding r-values and flow. The extended Barlat89 yield criterion is therefore chosen for damage modeling. For selected specimens, representative η-values are determined as input to the anisotropic damage model MAGD. The other specimens serve validation. MAGD properties are obtained by optimization. Despite some deviation for shear, comparison of FD_L and FD_W responses with experiment shows favorable agreement for the combination of anisotropic yield and damage models used here and that more accurate prediction with strain-dependent r-values. β’ agree well with experimental data up to fracture, indicating accurate prediction of η up to fracture. By comparison with “undamaged” FE results, the single contributions of damage and the strain-dependence of r-values to β’ become apparent.