PSI - Issue 7

Taizo Makino et al. / Procedia Structural Integrity 7 (2017) 468–475

471

4

Taizo Makino et Al./ Structural Integrity Procedia 00 (2017) 000–000

Fig.2. Process of RCF crack initiation and propagation.

3. Discussion on the RCF damage originating from artificial defect and inclusion RCF cracks initiated at sulphide inclusions lead to flaking failure through the same damage process as those initiated at artificial defects. However, a significant difference in fatigue lives was indicated. We discuss the reason for the difference from the mechanical perspective. 3.1. On the difference in the flaking lives Vertical crack initiation was indicated in the early stage of the damage process as shown in Fig.2. Vertical cracks were observed at N =6.00×10 6 in the case of originating from inclusion as indicated in Section 2.3. In contrast, the Mori-type RCF test conducted in our previous study (Makino et al., 2014, 2016) demonstrated that vertical cracks were observed at N =1.00×10 4 in the case of originating from artificial defects. There is a significant difference in the lives for vertical crack initiation originating from between artificial defects and inclusions. However, as there is also a difference in the RCF test method and the specimen geometry, direct comparison of the lives is not always accurate. Therefore, we introduced an artificial defect, of the same size as that of the inclusion detected by laminography, on the surface of a plate specimen. And the specimen was evaluated by a reciprocating-type RCF test. The material used in this test is high carbon-chromium bearing steels with normally low sulphur concentration (JIS SUJ2). Table 1 summarizes the size of origin, life for vertical crack initiation N i , flaking life N f and life ratio of vertical crack initiation to flaking N i / N f in cases originating from an artificial defect and inclusions. Two cases originating from inclusion (Inclusion A/B) are the result of specimens of S49. The introduced artificial defect was approximately the same size as that of Inclusion B. Flaking life originating from an artificial defect was 30 – 40 % of that originating from an inclusion (both cases of A and B). However, the life of vertical crack initiation and the life ratio of vertical crack initiation to flaking showed significant differences between an artificial defect and an inclusion. One can infer from these experimental data that the difference of flaking life between an artificial defect and an inclusion is caused by the phenomenon that the life for vertical crack initiation and the life ratio of vertical crack initiation to flaking originating from an artificial defect are much shorter than those from an inclusion respectively.