PSI - Issue 34

Veronika M. Miron et al. / Procedia Structural Integrity 34 (2021) 65–70 Miron et al. / Structural Integrity Procedia 00 (2021) 000 – 000

67

3

(“Silastic”, Dow, Midland, Michigan) a proven setting was chosen, while for Dragon Skin 10 Slow (“Dragon Skin”, Smooth-On, Macungie, Pennsylvania) three different print settings were compared regarding dimensional, visual, and mechanical properties. All specimens were printed with a 0.4 mm nozzle, a bed temperature of 70°C, and a layer height of 0.25 mm. The print speed and speed of the heat source for Silastic and 1 st and 2 nd Dragon Skin setting was 10 mm/s and was doubled to 20 mm/s for the 3 rd print setting of Dragon Skin. The waiting time of the heat source was 2 s for Silastic and Dragon Skin’s 1 st print setting and was increased to 4 s for the Dragon Skin’s 2 nd and 3 rd print setting. As commercially available material for 3D printing silicones, Silastic was printed and characterized using uniaxial tension as well as compression, biaxial tension, and pure shear tests. Additionally, Dragon Skin was tried for the printing process and compression test cubes could be fabricated as well as tested. All tests were performed on a Bose test bench with an axial actuator and a 450 N load cell at room temperature. For each test type and speed, at least three specimens were fabricated and tested. Tensile test specimens with the ISO 572-2 5A geometry were punched out of a printed rectangular plate with 120 mm x 80 mm x 1 mm to avoid the influence of printed perimeters. The printed plate was filled fully with ±45° top/bottom layers. The specimens were tested displacement controlled with 1 mm/s and 10 mm/s until the force limit of the load cell. For the biaxial tensile tests, 2 mm thick octagonal plates with an outer diameter of 140 mm were printed with full ±45° top/bottom layers from which cross specimens with 35 mm broad arms and 20 mm radii were punched out to avoid the influence of perimeters. The tests were performed at 0.1 mm/s, 0.5 mm/s, and 1 mm/s. The pure shear specimens had the dimensions of 10 mm x 100 mm x 1 mm with half-cylinders of 10 mm diameter added on the long sides for clamping. The pure shear specimens were directly printed fully filled with ±45° top/bottom layers and tested at 1 mm/s and 10 mm/s. The compression tests were performed on cubes of 20 mm x 20 mm x 20 mm printed with one top- and one bottom layer and one perimeter. The cubes were filled with a 100% rectilinear infill pattern. Additionally, twelve Silastic cubes were printed with 10% grid infill. Three loading and unloading cycles were performed up to 12 mm displacement for all but the 100% filled Silastic cubes, which were tested up to 7.5 mm displacement due to the force limit of the load cell. The testing speeds were 0.1 mm/s and 1 mm/s for the compression and 1 mm/s for the unloading. To check the dependency on the print orientation, some specimens were also tested in x-direction. For fitting a hyperelastic material model, the experimental data were averaged over all loading rates. From the compression tests, only the results for the 100% filled specimens were considered. The stress-strain data were input into the software MCalibration for the different loading types and all available hyperelastic material models implemented in the CAE software Abaqus were fitted and evaluated for their reliability to describe the material behavior. Incompressibility was assumed for all fits. The strain energy density function of the respective models are given in the Abaqus help manual (Abaqus, n.d.). The best fits were then evaluated in Abaqus, and their stability limits were extracted. 3. Results and Discussion All 100% filled Silastic specimens could be printed with satisfactory quality. However, for the 10% grid infilled cubes the one perimeter and low infill amount led to small holes in the walls and the top layer was not closing due to insufficient support from the lower layers. This lead to spread in the compression test results of these specimens. The Dragon Skin cubes were printed with three different print settings. An increase in heating time between layers and further on doubling the print speed increased the dimensional and optical qualities. The sides for the 1 st and 2 nd print settings with less heat influx and print speed than for the 3 rd print setting show wavy side walls and holes inside the cube. Overall, the 3 rd print setting would be chosen to keep printing Dragon Skin parts. Overall, the results of the material characterization show non-linear behavior that can be approximated with hyperelastic material models. The stress-strain curves of the uniaxial and biaxial tensile tests as well as the pure shear tests are summarized in Figure 1. The 100% module of the printed Silastic ranges between 2 to 2.5 MPa which is in accordance with the manufacturer ’s data sheet. In compression, both tested materials show hysteretic behavior. Testing the Silastic specimens at different strain rates showed no significant influence on the stiffness. This is in accordance to results found for PDMS and Ecoflex as presented in (Cakmak, 2021). The independency of Silastic and Dragon Skin silicone rubbers on the loading rate