Multi-objective optimization of the hydroforming process considering different plastic yield criteria

The present work aims at determining the optimal working conditions for the manufacturing of the AA6061-T6 Al alloy by the hydroforming process. As case study a stepped geometry was used. A numerical model was created using the commercial explicit Finite Element code LS-DYNA. The plastic behaviour of the investigated alloy was modelled implementing experimental data (flow stress curves, Lankford coefficients and Forming Limit Curves) and using two different yield criteria: an anisotropic one (Barlat '89) and the conventional isotropic one (Von Mises). Finite Element models were tuned using experimental data from warm hydroforming tests: comparing both the sheet thinning and the die cavity filling, quite different friction conditions had to be supposed for obtaining a good fitting with both the yield criteria. Finite Element models were finally used for evaluating the working range of the hydroforming process: results from a Central Composite Design simulation plan were imported within an integration platform (modeFRONTIER) to evaluate the optimal hydroforming conditions based on a multi-objective genetic algorithm optimization. Even if both numerical models (implementing different yield criteria) were tuned using experimental data, quite different results were obtained from the optimization procedure: the adoption of the anisotropic criterion was thus proved to be the suitable choice for better catching the material behaviour, as also confirmed by experimental hydroforming tests aimed at verifying the robustness of numerical data.