PSI - Issue 42

C. Boursier Niutta et al. / Procedia Structural Integrity 42 (2022) 1449–1457 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

1452

4

densification. The resulting force-displacement curves were integrated to calculate the absorbed energy and the SEA, known the mass of the specimen. Fig. 3 shows the results of the first (Fig. 3a) and second (Fig. 3b) factorial plan.

(a) (b) Fig. 3. Experimental results of a) first factorial plan and b) second factorial plan [5].

As shown in Fig. 3a, both the diameter and the cell size affect the SEA. In particular, the denser the unit-cell, that is the smaller the size and the bigger the diameter, the higher the SEA. On the contrary, the results of Fig. 3b, which refers to the second factorial plan, show that for the retained material and unit-cell geometry, the number of cells n does not influence the SEA of the structure. This means that the absorbing capabilities of the investigated cell can be properly assessed through tests on specimens with a limited number of cells (thus permitting to reduce the manufacturing time) and can be used for the design of components with lattice structures. 3. FE model of the lattice structure and defects modeling In this section, the FE model of the crushing test of the lattice structure is firstly presented. The modelling of the AM defects is then detailed, as well as their random insertion in the unit-cell and the construction of the defected specimen. Increasing percentages of defects are here considered. 3.1 Crushing test of the lattice structure: FE model The crushing test of the lattice structure is simulated in LS-Dyna environment through transient nonlinear FE analyses. The lattice structure is modelled with 1D beam elements with the Hughes-Liu formulation. The Hughes Liu formulation with cross section integration is based on the transformation of the isoparametric 8-nodes element [6]. A mesh size of 1 mm is considered. It is worth noticing that, even though the use of 1D elements prevents a detailed description of the geometrical variations of the structure, as resulting from the 3D printing process, e.g., diameter variations, presence of pores etc., the computational effort is consistently decreased with respect to the more accurate 3D elements. As such, the 1D elements are more suitable for the design phase of components made of lattice structures. In order to guarantee the self-contact of the beams during the crushing phenomenon, the contact type *CONTACT_AUTOMATIC_GENERAL is retained. The mechanical behavior of the lattice structures is simulated with an elastoplastic material law, namely *MAT_PIECEWISE_LINEAR_PLASTICITY. The stress-strain curve of the material in the plastic field is obtained from [7], where tensile tests on specimens made with the same Carbon nylon were performed. According to [7], the Young ’s modulus is set to 3510 MPa and the yield limit is equal to 20 MPa.