PSI - Issue 47

Mohammad Reza Khosravani et al. / Procedia Structural Integrity 47 (2023) 454–459 Author name / Structural Integrity Procedia 00 (2023) 000–000

456

3

2. Experimental setup and procedure

2.1. Design and fabrication of specimens

In the present study, two groups of specimens were designed, printed, and examined: dumbbell-shaped samples, and CT specimens. We used PLA material to fabricate the above-mentioned groups of specimens based on the FDM process. In detail, dumbbell-shaped samples and CT specimens were designed and printed according to ASTM D638 (ASTM D638, 2014) and ASTM D5045-14 (ASTM D5045, 2014), respectively. All specimens were created using a computer aided design software and saved in ”.stl” format. The files of dumbbell-shaped samples, and CT specimens were then imported into Cura TM slicing engine which is an open source tool for slicing and configuring the printing parameters. Finally, the files were utilized to extrude and deposit material. In Fig. 1, the schematics of dumbbell shaped and CT specimens are illustrated. In both groups of the specimens, the layer thickness and nozzle diameter were 0.2 mm and 0.6 mm, respectively. In CT specimens, the e ff ective width of the specimen is w and the crack length ( a ) must be selected such that 0.45 < a / w < 0.55. Here, w = 75mmand a = 36 mm. As mentioned in ASTM D5045-14 (ASTM D5045, 2014), a sharp initial crack is required in the CT specimen prior to mounting in a tensile test machine. Therefore, a sawing machine was utilized to create a sharpened crack with a length of 2 mm. It should be noted that, four specimens were printed and examined for each raster orientation and printing speed.

18

90

6

93

0

R 76

90

0

z

y

93

x

57

- 45

+45

165

- 45

+45

19

z

y

x

36

75

18

Fig. 1. The schematics of dumbbell-shaped and CT specimens, and filament deposition pattern (dimensions in mm).

The specimens were printed with infill density of 100%. In all dumbbell-shaped samples and CT specimens, two di ff erent printing parameters were changed to determine the influence of the printing parameters on the fracture behavior of the 3D-printed components. Particularly, raster orientation and printing speed were changed in this study. Raster direction is defined as the printing direction relative to the loading direction. Printing speed defined as the distance traveled by the extruder along the bed per unit time while extruding. In the current study, all specimens were printed with 45 ◦ / -45 ◦ and0 ◦ / 90 ◦ filament orientations at printing speed of 20 mm / s and80mm / s. Although specimens were printed with di ff erent raster layups and printing speed, all other 3D printing parameters (e.g., extruder width, printing speed, and layer thickness) were kept constant to ensure consistent printing quality. In experimental tests on the dumbbell-shaped samples, a hydraulic machine is used at room temperature. The machine is fitted with 15 kN load cell, and it has cross-head speed range 0.01 mm / s to 30mm / s. In order to avoid the likelye ff ect of the displacement rate, all tests are performed under displacement control condition with a constant rate of 10mm / min according to ASTM D5045-14 (ASTM D5045, 2014). Electronic control unit of tensile test machine allows monitoring the applied load and movement of the top cross head. Here, we used an appropriate tensile grips in order to keep the specimen perfectly aligned in the vertical direction. A pre-load of 5 N is applied to specimen prior to test start to take up slack from gripping apparatus. It is noteworthy that ere was no fracture outside the gauge section. Thus, all failures are valid and have been used to determine fracture load and mechanical strength of the examined dumbbell-shaped specimens. In the current study, CT tests are carried out on a hydraulic machine tensile machine. It is worth mentioning that we have designed and fabricated a fixture for CT tests. The fixture has upper and lower grips 2.2. Details of experimental tests