PSI - Issue 66

B. (Bo) van Schuppen et al. / Procedia Structural Integrity 66 (2024) 412–418 Author name / Structural Integrity Procedia 00 (2025) 000–000

414

3

2. Experiments The steel specimens used for this research are of steel grade S275JR, with an average yield strength of y = 307 N/mm 2 and an average tensile strength of = 431 N/mm 2 . These specimens are extracted from the same material batch as in Baarssen et al. (2022), where the monotonic tensile properties are reported. All specimens have a nominal plate thickness = 8 mm, a nominal width = 110 mm and a nominal hole diameter of 18 mm. The bolt-hole configurations with their corresponding dimensions are shown in Fig. . The tests are carried out in two phases: (1) the pre-cracking of the specimens under fatigue loading and (2) the tensile testing. For every specimen type, 3 specimens were only subjected to the second phase, to be able to estimate the effect of the cracks. These specimens are named A1-A3, B1-B3, and D1-D3. The rest of the specimens are subjected to both phases. These specimens are named A4-A5, B4-B5, and D4-D7. In these cases, the cracks will be induced by fatigue loading until cracks are detected using the broken wire electronic binary sensor method. For both test phases, a hydraulic universal testing machine from Schenck with a capacity of 400k N is used. The cyclic loading is applied as a sinusoidal waveform with a frequency of 4 Hz, a load ratio R=0.05, and a load range estimated in such a way that the fatigue crack nucleates within a reasonable amount of loading cycles in the high cycle fatigue range, i.e. N>10 4 cycles. This results in a load range of 190 kN, 148 kN, and 152 kN for types A, B, and D, respectively. The broken wire electronic binary sensor method Raeisi et al. (2017) is used to detect cracks in their early stages, i.e. less than 1 mm. In this method, an electrically insulated copper wire with a diameter of 0.1 mm is glued along the edge of the holes on the steel plate by cyanoacrylate. When a crack is initiated, this will cause the fracture of the copper wire. The failure of the copper wire is detected and used to terminate the cyclic loading phase. The application of this circuit to a type D specimen can be seen in Fig. . The tensile tests are performed controlling the displacement of the actuator at a rate initially set to 1.2 mm/min and successively increased to 3.6 mm/min when the load-displacement curve becomes non-linear, i.e. in correspondence with gross plasticity developing in the net cross-section. A Linear Variable Differential Transducer (LVDT), with an initial gauge of 110 mm, is installed over the holes on each specimen, spanning over the center of the plate, to more accurately capture the elongation over the holes. The measured force, the signal of the LVDT, and the displacement of the actuator are acquired with a sampling rate of 10 sample/s.

Fig. 1. Bolt hole configurations and specimen dimensions