PSI - Issue 33

Costanzo Bellini et al. / Procedia Structural Integrity 33 (2021) 824–831 Author name / Structural Integrity Procedia 00 (2019) 000–000

826

3

they were made of, and the adopted stacking sequence were defined too. In fact, the behaviour of such kind of material relies on these characteristics. Then, the numerical model for the simulation of the flexural behaviour was defined, by paying attention to the delineation of material properties and the boundary conditions. Once the different types of laminate and the tests had been defined, these laminates were manufactured through the vacuum bag process, and then the specimens were extracted from the produced laminates to be tested according to the chosen three-point bending procedures. Finally, there is the presentation and the discussion of the results. In particular, the comparison of the results obtained from the numerical activity and experimental tests is illustrated, in order to testify the effectiveness of the proposed model. Moreover, micrographs were taken to better understand the fracture mechanisms and the results were compared with numerical ones. 2. Materials and methods In this work, the three-point bending test was chosen as experimental procedure. This type of mechanical test is characterized by feasibility and simplicity; in fact, both long and short beams can be tested by varying the distance between the supports, the so-called span. In such a manner, it was possible to consider the flexural strength and the interlaminar shear strength. As prescribed by the ASTM D790 standard, that is the standard adopted to determine the flexural behaviour, the dimension of each specimen depended on the thickness of the laminate, that was 5 mm. So, each specimen had a length of 160 mm and a width of 20 mm. Instead, the specimen dimensions for the interlaminar shear characteristics were determined according to the ASTM D2344 and were 25 mm in length and 10 mm in width. The materials involved in the present work were an aluminium sheet and a CFRP (Carbon Fibre Reinforced Polymer) made of carbon fabric with a 2x2 twill weaving style. As concerns the stacking sequence, an FMLs with CFRP laminate as external layers was considered. In particular, it consisted of three layers: two of CFRP separated by the aluminium sheet, whose thickness was equal to 0.6 mm. As concerns composite laminates, each layer was composed of six prepreg plies; in such a manner, considering a single prepreg sheet thickness of 0.35 mm, the obtained thickness of the whole laminate was about 5 mm, as stated before. The interface between the composite material and the aluminium sheet was warranted by inserting a structural adhesive typically used in the aeronautical field, that was the AF 163 - 2k, in the lamination sequence. As concerns the numerical model, the mesh was constituted by brick elements whose dimensions were: length of 2 mm, width of 1.25 mm and thickness of 0.44 mm. For the long beam specimen, a mesh reproducing only a quarter of the specimen was prepared, thanks to the double geometrical and load symmetry. Therefore, as visible from Fig. 1a, the calculation domain was 80 mm long, 10 mm wide and 5 mm thick and it was constituted by 3840 elements. The short beam specimen was smaller than the long one, so only half of the specimen was meshed. In this case, the width and the thickness of the built mesh were the same as the previous one, while the length was 12.5 mm, as reported in Fig. 1b. A total of 672 elements were employed. As concerns the material properties, the aluminium was simulated as an isotropic material, with Young’s modulus of 70 GPa, Poisson’s ratio of 0.3, and yield strength of 100 MPa. The composite material was simulated as an orthotropic material, with in-plane Young’s modulus of 60 GPa, out-of-plane one of about 9 GPa, and maximum tensile strength of 700 MPa. As concerns the boundary conditions, for the long beam specimen two symmetry planes were considered: the XZ plane and the YZ plane, while only the symmetry in the YZ plane was applied for the simulation of the short beam, as visible in Fig. 2. The load was simulated by assigning a vertical displacement, along the Z axis, to the node in correspondence to the centre, while the support was modelled by blocking the vertical displacement of the node at a distance equal to half span from the centre. The FMLs to be tested for validating the numerical model are produced through the vacuum bag technology, schematized in Fig. 3. The first step consisted in preparing all the constituting materials, that are the aluminium sheet, the prepreg plies and the adhesive film, for the stacking operation. Then, 6 prepreg plies were laid down on the mould, that had been previously prepared, followed by an adhesive film, the metal sheet, another adhesive film, and the remaining 6 prepreg plies. Once the laminate had been constructed, it was wrapped with release film and breather cloth, and then it was put in the vacuum bag, that was sealed by using butylic tape. The laminate was cured in the oven at a temperature of 127 °C, reached with a ramp of 2 °C/min, and then cooled to room temperature before demoulding.