A fast methodology for the accurate characterization and simulation of laser heat treated blanks

• Published in: International journal of mechanical sciences 2021-02, Vol.192, p.106134, Article 106134
• Main Authors: Lattanzi, Attilio, Piccininni, Antonio, Guglielmi, Pasquale, Rossi, Marco, Palumbo, Gianfranco
• Format: Article
• Language: English
• Subjects: Digital image correlation; Gleeble test; Inverse method; Local properties; Numerical simulation
• ISSN: 0020-7403; 1879-2162
• DOI: 10.1016/j.ijmecsci.2020.106134

•Gleeble tests for capturing strength alteration introducing annealing function (Ann).•Laser treatment designed by inversely tuned FE model and (Ann).•Hardness related to local peak temperature correctly predicted by (Ann).•Hardening properties evaluated through a single DIC assisted tensile test.•Implemented hardening behaviour precisely reproduced the real deformative behaviour. [Display omitted] In this work, a new methodology for the material characterization of locally heat treated blanks is proposed. Laser treatments on the central part of dog-bone specimens, extracted from a strain hardenable Aluminium sheet (AA5754) in wrought condition (H32), were simulated through an inversely tuned Finite Element model (FE) and performed using a CO2 laser. In order to establish a correlation between the temperature history in each point and the corresponding material properties after the heat treatment, preliminary local annealing tests were conducted by means of the Gleeble 3180 system; such heat treated samples were then subjected to Vickers micro-hardness tests along the longitudinal direction in order to obtain data for defining a simple but effective annealing function based on the maximum temperature (Tpeak) experienced by the material. Tensile tests on the laser heat-treated specimens were assisted by a Digital Image Correlation (DIC) system, which allowed to obtain the full-field deformation history and, as a consequence, to extract different flow stress curves, since each region of the sample was subjected to different temperature levels. The material behaviour in terms of stress vs. strain curves extracted by DIC could be related to the same variable (Tpeak) and finally implemented into a FE model. The proposed methodology revealed to be effective in extracting the flow stress curves (yielding plus hardening) and, as demonstrated by the simulation of the tensile test, it allowed to predict the deformation behaviour of the material with graded properties more accurately than the qualitative approach based on the assignment of different mechanical properties by partitions.