Calibration and prediction assessment of different ductile damage models on Ti6Al4V and 17-4PH additive manufactured alloys

Nowadays, metal additive manufacturing is becoming always more popular, being able to deliver complex shaped high quality products. Though many studies have been conducted on the high cycle fatigue behavior of these materials, yet ductile failure has still not been completely investigated, to identify the failure limits under static complex stress states. In the present study, the calibration of three ductile damage models on two popular additive manufactured alloys was carried out. The selected alloys were Ti6Al4V, processed via Electron Beam Melting, and 17-4PH fabricated with Selective Laser Melting technology; both broadly used in actual industrial applications. For each material a set of samples, was fabricated to perform a thorough static mechanical characterization, involving tensile tests on round smooth bars, notched bars, tests under plane strain conditions and torsion tests. The stress state in the critical points was retrieved relying on FEM simulations, and the data collected via the hybrid experimental-numerical procedure subsequently used to tune the damage models. Specifically, the selected models are the Rice and Tracey, the Modified Mohr-Coulomb by Wierzbicki and the one proposed by Coppola and Cortese. While the former does not take into account the effect of Lode parameter, the latter two consider its influence on fracture onset. A minimization algorithm was used for their calibration, and different optimization strategies were adopted to check the robustness of identified parameters. The resulting strains to fracture as a function of damage parameters were plotted for each formulation. The failure prediction accuracy of all models was assessed and compared to the others.