PSI - Issue 53

Martin Matušů et al. / Procedia Structural Integrity 53 (2024) 29 – 36 Author name / Structural Integrity Procedia 00 (2019) 000–000

32

4

a)

b)

Fig. 2. (a) Amsler resonant pulsator with the tested specimen (left) and with the reference specimen (right, with a blue tape) measured via infrared thermal camera Fluke RSE600; (b) Tensile test set-up with the specimen fixed in the Hegewald & Peschke testing device.

Fatigue experiments that were conducted in this campaign were stopped when any of these conditions was true: (a) frequency drops by 10 Hz, (b) load amplitude changes by ± 0.5 kN, (c) static load changes by ± 0.5 kN, or (d) 10,000,000 cycles reached. The last condition was classified as a run-out. Such specimens were then subjected to 1.36 times higher load amplitude at least. Fig. 3 illustrates the fatigue performance of the different series. As expected, the NoHT series exhibits the worst fatigue performance. All series subjected to a subsequent heat treatment resulted in superior performance across all the cases examined in this study. Specifically, T200 and T240 series demonstrate almost identical fatigue performance within the range from 100,000 cycles to the fatigue limit domain. In the low cycle fatigue (LCF) region below 50,000 cycles, the T200 series displays higher fatigue strength. The T300 series exhibits the highest fatigue limit. However, it is worth noting that this series shows the worst fatigue performance when transiting to the LCF domain. This behavior can be attributed to the decomposition of the silicon network. The T300 series exhibits the lowest scatter of experimental data. To analyze the fatigue regression, the Kohout-V ě chet model [14] is employed, which effectively represents the transition from LCF to high cycle fatigue (HCF) and the transition to the fatigue limit domain, as depicted by the curves in Fig. 3.

Fig. 3. S-N curves of the four groups of specimens based on the heat treatment.