PSI - Issue 68

Ivan Senegaglia et al. / Procedia Structural Integrity 68 (2025) 610–618

613

4

Ivan Senegaglia at al. / Structural Integrity Procedia 00 (2025) 000–000

Figure 2: Test setup: a) Test equipment: 1) LVDT displacement sensor, 2) Specimen testing assembly, 3) Testing machine; b) Load history: the elastic phase of every macrocycle is colored for clarity purposes.

Table 3. Loading history applied to the tested specimens. Loading parameters [kN] S1 S2 S3 S4 S5 Maximum (phase 1) 20 30 40 45 60 Mean (phase 2) 15 25 35 40 55 Amplitude (phase 3) 5 5 5 5 5

2.3. Computational modeling The gyroid lattice structure was modeled using a FE framework. The gyroid geometry was specified by a 3D cartesian space field equation, commonly employed to depict periodic minimal surfaces (Yoo, (2011)), ensuring the coherent distribution of the material throughout the volume. The gyroid unit cell was expanded under PBCs to replicate a larger RVE of the structure, enabling an approximation of the specimen's mechanical behavior with minimal computational resources. As shown in Figure 3, two distinct sets of boundary conditions were implemented to define the extremes of the specimen’s mechanical response, which, in contrast, has a finite 3D distribution of cells. In more detail, the cells closer to the machine interface are more constrained compared to the ones far from these regions. This discrepancy in constraint depends on the friction and the physical attachment to the stiffer interface along the loading direction. This constrained zone was modeled with an RVE featuring a full set of PBCs (AH XYZ ). Vice versa, the central part of the specimen has a more uniform stress distribution, and with constraint-free lateral surface. In this case, a model with PBCs only in the load direction (AH Y ) mimics with more accuracy the condition of maximum compliance. In both models, the loading direction corresponds to the Y-axis, as indicated in Figure 1 and Figure 3. The base material of the gyroid structure in the FE model was modeled using a non-linear material with multilinear isotropic hardening formulation. The constitutive law, used as input for the model, was derived from the elaboration of the results of the tensile tests. A 3D geometry correction procedure was conducted for both numerical models, considering a more realistic TPMS lattice thickness obtained by measurement. The testing conditions were simulated to reproduce the load history discussed in paragraph Error! Reference source not found. , limiting the elastic phase to a single cycle instead of 10.