PSI - Issue 79

H.G.E. da Silva et al. / Procedia Structural Integrity 79 (2026) 97–104

100

Table 3. Properties of the adhesive SikaForce ® -7710 L100.

Tensile strength [MPa] Tear strength [MPa]

13

Elongation at break [%] Shore hardness [D]

8

9

80

Table 4. Properties of the ROHACELL ® 71WF core.

Specific weight [kg/m 3 ] Young’s modulus ( E ) [MPa]

75

Compression strength [MPa] Tensile strength [MPa]

1.7 2.2 1.3

105 0.25

Poisson ’s ratio ( ν )

Shear strength [MPa]

2.3. Experimental procedure

The sandwich skins were fabricated using prepregs supplied by CIT in roll form, from which individual layers of 300×300 mm² were precisely cut. The layers were then manually stacked. The laminates underwent a curing cycle at 140°C for 28 minutes. The laminates were carefully removed from the press. For core fabrication, a PMI block measuring 300×300 mm² was disc cut to match the laminate dimensions. A spatula was used to apply an even layer of adhesive to the laminates, ensuring uniform coverage. The laminates were then stacked on the prepared core in the required orientations. The structure was cured at room temperature for one week. Once the sandwich structure was fully fabricated, the specimens were cut to the desired width. The core compression tests and flexural tests were performed using a Shimadzu Autograph AGX 100 kN testing machine. For the core compression tests, three specimens were prepared (Fig. 3 a). A flat plate press was used for the compression tests, with a testing speed of 0.4 mm/min. For the 4PB tests, shown in Fig. 3 (b), four specimens were tested (two for each configuration), with a test speed of 5 mm/min and a support span of 250 mm, following the ASTM C393 standard. The deflection at mid-span was measured using a dial comparator.

a)

b)

Fig. 3. (a) Core compression test setup; (b) 4PB test setup with specimen positioning and deflection measurement at mid-span.

2.4. FEM preprocessing

The two-dimensional specimens were modelled as deformable solids, whereas the supports and loading punches were represented as rigid discrete elements. Each skin was subdivided into eight plies, with orientations defined according to the layup. The material properties and solid sections were assigned based on the data outlined in Section 2.2. A static analysis incorporating geometric nonlinearity was performed. A surface-to-surface interaction was established between the punches and the specimen surfaces, preventing frictional effects and eliminating penetration at the contact regions. The supports were assumed to be fully constrained, while the loading punches were assigned a vertical displacement with restricted movement along the horizontal axis. Bias effects were applied to refine the mesh in critical regions, particularly around the adhesive layers and the punch-specimen contact zones (Fig. 4). The adhesive behavior is governed by a triangular traction-separation law, incorporating the quadratic stress (QUADS) damage initiation criterion, with failure progression evaluated based on a linear energy dissipation approach. The core was modelled as a crushable foam with isotropic hardening, using a compression ratio of 1 and a plastic Poisson’s ratio of 0. The Tsai -Wu failure criterion was employed to predict the onset of ply failure. The material properties, including tensile ( σ ut1 and σ ut3 ) and compressive strengths ( σ uc1 and σ uc3 ) in both fibre and transverse directions, as well as shear strength ( τ u13 ), were obtained from manufacturer specifications.