PSI - Issue 68

A. Ceci et al. / Procedia Structural Integrity 68 (2025) 372–378

375

4

A. Ceci et al. / Structural Integrity Procedia 00 (2025) 000–000

Once the muffle furnace reaches a temperature of 900°C, the AA 6082 aluminum alloy is melted in an electric furnace. Such a high temperature is chosen to ensure that the metal solidification in the plaster mold does not occur too quickly, as the pouring is done at ambient temperature. After pouring and subjecting the mold to mechanical vibrations during the solidification stage, it is quenched in water, and the metal structure is cleaned of plaster and excess metal, resulting in the same structure of the starting one made of PLA. 3. Mechanical Characterization After manufacturing, the mechanical characterization of the lattice structures has been performed through compression tests using specific plates (Fig. 4 a), followed by a comparison with numerical and image analysis. To perform the tests and then analyze the results through image analysis, in addition to the testing machine (MTS Insight - Electromechanical - 100kN Standard Length), a camera has been adopted (Fig. 4 a). At first, red markers were drawn on the specimens to be analyzed (Fig. 4 b). This choice was made because the image is divided into the RGB colour channel in DIC (Digital Image Correlation) for marker control.

a)

b)

Fig. 4 : a) Compression test, b) Specimen with markers.

4. Numerical Analysis and Experimental Results The experimental results were compared with Finite Element Analysis (FEA) to identify the discrepancies between numerical modelling and experimental data. This comparison is important for a future optimization of the unit cell using numerical algorithms, aiming to improve stress distribution and, consequently, the overall strength of the lattice structure. The finite element analyses accounted for nonlinearities due to large displacements (geometric nonlinearity) and the material's elastic-plastic behaviour. Tetrahedral elements with 10 nodes and quadratic shape functions were employed, as they are well-suited for modelling elastoplastic behaviour in irregular meshes. The total mesh size consisted of approximately 2 million elements; the mesh used for the individual unit cell is shown in Fig. 5. The elastic-plastic behaviour was modelled using isotropic hardening, as no load cycles were considered in the analysis, with the yield surface computation based on the von Mises equivalent stress. The uniaxial stress-strain curve for the material, implemented in the FEM analysis, was obtained from a tensile test performed on a dog-bone specimen manufactured using the same procedure shown in Fig. 3. The FE analysis was performed by keeping fixed the nodes at the base and applying vertical displacements to the top lattice surface (Fig. 6).