PSI - Issue 73

Dominik Gřešica et al. / Procedia Structural Integrity 73 (2025) 27 – 32 Dominik Gřešica, Petr Lehner, David Juračka / Structural Integrity Procedia 00 (2025) 000 – 000

30 4

3. Boundary conditions and material As mentioned above, the whole process of preparing and setting up the numerical model was adopted from the previous study (Lehner et al., 2024). For the purposes of the numerical model, the input material characteristics were initially drawn from the technical data sheets of the 3D printing material manufacturer or from available publications. The investigated 3D printed part, modelled and intended for production from PC Blend (polycarbonate) (Priore, 2016; Prusa i3, 2022). Given that this was a study of the behaviour of the joint and the joined materials, it was necessary to include the properties of wooden prisms with a cross section of 50 x 50 mm, with wood class C24. Since in the modelled arrangement the wooden part is inserted into the 3D printed joint, pins made of S235 steel were used to ensure stability. The numerical model was symmetric and therefore it was possible to create a half of the physical specimen (see Figure 3). For high fidelity of the results, the contact areas for friction were set according to the respective material pairs. A further idealization was to set a fictitious wood beam up to the support, as minimum stress values were expected in that part. At the point above the support, a zero displacement and a turning radius corresponding to half the height of the wooden prism was set. A gradually increasing displacement was introduced into the symmetry surface, divided into several parts for better analysis behaviour. The simulation of three-point bending was prepared with a displacement range to 20 mm. The steel pins have an overlap of approx. 5 mm over the edges of the 3D printed element, which prevents them from digging in.

Fig. 3. Example of (V04) FEM model - symmetrical half of the structure - and boundary conditions of the calculation.

At the interfaces between the materials, the simulations were configured to allow only compressive forces and friction to be transmitted. The finite element mesh employed an average element size of 8 mm, with finer meshing applied around critical areas like holes and points of interest. The mesh setup led to 38 786 nodes and 19 321 finite elements. The output of the calculation were graphs of stress, strain and force in the probes. These could be included from a single force-displacement diagram. The resulting force-displacement diagrams revealed limit points, representing the failure criteria for each material: a shear stress of 3 MPa for wood, an equivalent plastic strain of 5 % for the 3D print, and a yield stress of 235 MPa for steel.