PSI - Issue 54

T. Fekete et al. / Procedia Structural Integrity 54 (2024) 314–321 Investigation on geometric imperfections… Structural Integrity Procedia 00 (2019) 000–000

318

5

Dog-bone shaped flat tensile test specimens with an active zone of nominal dimensions 8 × 2 × 30 [mm×mm×mm] were used –see ASTM E8/E8M-22. The transition zone edge radius between the active zone and the gripper heads was R= 57 [mm]. The manufacturing tolerance was 0.02 mm in the active zone and 0.05 mm outside the active zone. A sketch of the specimen is shown in Figure 3. The specimen material was S460 general purpose structural steel, and they were manufactured on an EML-1200SV 5-axis CNC machine. The dimensions of each specimen were measured with the CNC machine's built-in high-precision measuring system, with an accuracy of 0.0005 mm. Thus, the coordinate maps of the surfaces for each sample could be determined. A square grid with dimensions of 0.5 × 0.5 [mm×mm] spacing was printed on the gauge length’s surface. Tensile tests were carried out on a Gleeble 3800 thermomechanical simulator, which provided force and crosshead displacement data in addition to the strain data measured by a mechanical strain gauge. For full-field observation of the test specimen geometry, an in-house optical data acquisition system, consisting of fine resolution USB cameras, their supports, the illumination system, and the control software, was developed –see Fekete et al. (2022). This way, the time evolution of the deformation field during the tests could be observed (Figure 4). The crosshead speed was 1 mm/min during the tests. The time synchronized data acquisition had a 0.004 s sampling period. Photographs were taken at every 0.02 s from the specimens. Figure 4 shows that necking is not always positioned in the centre plane of the gauge length, as would be expected for a geometrically perfect specimen.

Fig. 4. Tensile test specimen at the start (left) and at the end (right) of the test

3.2. Coordinate maps of the gauge length

The coordinate maps introduced in Section 3.1 were used to determine the thickness of the specimens along the gauge section. Figure 5 shows the thickness profile for samples F1 and F2 along three cross-sections in length. The centre of the gauge section is noted by 0 ; the coordinate maps are established in the [- 15 mm; 15 mm] interval. As it can be seen neither specimens’ thickness is constant along the axial direction. The average thickness of F1 is 1.96 mm and it is 0.02 mm thinner at the specimen’s neck region compared to the centre. Similarly, F2 ’s average thickness is 1.98 mm and its centre is 0.02 mm thicker than the neck region. It was assumed that these geometric imperfections also significantly contribute to the asymmetric necking of the specimens, therefore, the production geometry represented by the coordinate maps was implemented in the DT s of the measurements.

Fig. 5. Thickness along the gauge lengths of specimen F1 and F2