PSI - Issue 7

Benoît Bracquart et al. / Procedia Structural Integrity 7 (2017) 242–247 B. Bracquart et al. / Structural Integrity Procedia 00 (2017) 000–000

245

4

φ s = 100 µm

φ l = 1000 µm

39

36

24 Stress amplitude σ a (MPa) 27 30 33

21

10 3

10 4

10 5

10 6

Number of cycles to initiation N i

Fig. 2: S-N curves for crack initiation, obtained for d l = 1000 µ m and two sizes of microstructure, φ s and φ l . Crack initiation criterion: 100 µ m-long surface crack.

varied. Stress amplitudes have been selected low enough to prevent macroscopic crack propagation and exclude interactions between defects. After fatigue testing, SEM observations of each defect have been carried out to determine whether a crack initiated or not, enabling a more accurate detection of cracks than with the in-situ monitoring used earlier, for which only surface cracks were detected. For each configuration d − φ , specimens have been tested at several stress amplitudes, and the proportion of cracked defects have been determined. With these results, the stress amplitude corresponding to a 50% probability of initiation has been estimated for each d − φ configuration (see Table 1). Table 1: Stress levels for 50% probability of crack initiation at the defect, in the 4 configurations d − φ . d s = 100 µ m d l = 1000 µ m φ s = 100 µ m 28.4 MPa 25.8 MPa φ l = 1000 µ m 26.6 MPa 25.7 MPa As shown in Figure 2, the initiation stage is found to be impacted by the average grain size: for a given stress amplitude, the mean number of cycles to initiation N i is reduced when the grain size is increased. This influence of grain size on fatigue crack initiation is also observed for specimens with multiple defects. Indeed, as shown in Table 1, the stress amplitude corresponding to a 50% crack initiation probability is lower for the large grain microstructure, when small defects are introduced. For large defects, however, the di ff erence between both grain sizes is negligible. This di ff erence with the in-situ crack monitoring approach presented above can be explained by the fact the initiation criterion was di ff erent, leading to higher stress levels. The small grain microstructure is thus more sensitive to defect size. This di ff erence regarding defect size sensitivity can be interpreted in terms of plastic activity. For each grain size φ and stress amplitude of Table 1, a plastic activity indicator as been determined. It is arbitrarily chosen as the macroscopic plastic strain range between the softening and the hardening stages observed 4. Discussion