PSI - Issue 13

Kai Suzuki et al. / Procedia Structural Integrity 13 (2018) 1065–1070 Kai Suzuki et al. / Structural Integrity Procedia 00 (2018) 000 – 000

1069

5

3.3. Fatigue crack growth behaviour: A cause of the scatter in crack growth rates Figure 6 shows replica images of the fatigue cracks of the HEA and LEA. In the LEA, the crack length at which the scatter in crack growth rates decreased was approximately 500 μm. In contrast, in the HEA, the scatter in crack growth rates did not decrease with increasing crack length in the present study, as mentioned above. To understand the difference in the scatter in crack growth rates between the HEA and the LEA, using the replica method, we observed the crack propagation paths at approximately 500 μm in crack length of the HEA and LEA. A difference between the fatigue cracks of the HEA and the LEA was observed in the degree of crack deflection. In Fig. 6, the yellow dashed lines indicate the peak positions of the deflection of cracks on the specimen surface at the respective number of cycles, namely, the widths of the region between the dashed lines indicate the degree of crack deflection. The cracks shown in Fig. 6 are the representative cracks of the respective alloys, and the lateral distance between the dashed lines in the HEA and LEA at 500 μm was 133 and 70 μm, respectively. The same measurements were conducted for eight other cracks in each alloy. The average lateral distance between the peak positions for nine fatigue cracks of the HEA and LEA was 127 and 64 μm, respectively. The difference in the lateral distance between the peak positions of the deflected fatigue cracks was visually evident in terms of the propagation path, as shown in Fig. 6. That is, the fatigue crack of the HEA propagated diagonally to the load direction across two grains. In contrast, the propagation path of the fatigue crack of the LEA showed a relatively fine zigzag morphology. Since the crack propagated parallel to the slip lines in both alloys, the crack deflection was attributed to the crack propagation along the slip plane. Accordingly, the difference in crack deflection might be related to the dislocation substructure evolution along the slip plane (Awatani et al., (1978)). Investigation on the dislocation substructure evolution is a key to understanding the underlying characteristics of crack growth in the HEA and, therefore, will be performed in a future work. Moreover, since a larger crack roughness enhances the effects of crack closure (Suresh, (1998)) and stress shielding (Ritchie et al., (1988)), the larger degree of crack deflection in the HEA can cause temporal deceleration or even non-propagation. The effect of crack deflection on fatigue crack growth also needs to be discussed in further investigations.

Fig. 6. A set of replica images showing a fatigue crack at (a) 4×10 5 , (b) 7.5×10 5 , (c)9.5×10 5 , (d)1.0×10 6 (e) 1.05×10 6 cycles, and (f) the fatigue crack length in the HEA tested at 270 MPa. Another set of replica images showing a fatigue crack at (g) 2 × 10 5 , (h) 2.8 × 10 5 , (i) 3.4 × 10 5 , (j) 4.0 × 10 5 , (k) 4.3 × 10 5 cycles, and (l) the fatigue crack length in the LEA tested at 220 MPa. The arrows indicate crack tips. The loading direction corresponds to the horizontal direction. The yellow dashed lines indicate the peak positions of the deflection of cracks on the specimen surface at respective number of cycles.