PSI - Issue 14

Takashi Nakamura et al. / Procedia Structural Integrity 14 (2019) 978–985

983

Takashi Nakamura/ Structural Integrity Procedia 00 (2018) 000–000

6

unloading part. The reversibility of slip in vacuum is considered to be high due to the absence of oxides and adsorbates, which act as obstacles to reverse slip. Pelloux reported that the striation pattern was not observed in vacuum due to the high reversibility of slip ((Pelloux, 1969). Also, in our present work, the striation pattern was not observed in vacuum, while it was clearly observed in air as shown in Fig. 9. This fact indicates that the reversibility of slip was higher in vacuum than in air. The high reversibility of slip decreases the real increment of crack propagation in a whole cycle and results in a low crack propagation rate in vacuum. Regarding the similarity between the granular feature of the fracture surface of internal fracture and that in vacuum, Oguma proposed a model similar to cold-welding (Oguma and Nakamura, 2013). According to this model, the direct metal-metal contact of the upper and lower fracture surfaces plays a great role to form the fine convexo-concave pattern. In a very low vacuum pressure such as high vacuum or ultrahigh vacuum, re-welding of the fracture surfaces can occur due to the low fraction of surface coverage by gas molecules. The present study was conducted under R = 0.1; however, local and partial contacts could occur in some location in the fracture surface. Especially in the small crack regime, a large number of contacts likely occurred because of the low crack growth rate in vacuum. This phenomenon is considered to promote the formation of the granular region. According to the results in Section 3.2, internal cracks are considered to propagate through the vacuum-like environment. However, it is difficult to know the actual vacuum pressure in the space between upper and lower crack surfaces of internal crack. In the following discussion; therefore, we briefly introduce the results of our previous study (Yoshinaka, et al., 2016), which measured the internal crack growth rate by using synchrotron radiation µCT imaging provided at SPring-8 in Hyogo, Japan. And then, we compare the internal crack growth rate with that of surface crack in high vacuum (Fig. 6) to discuss the vacuum level inside small internal crack. SPring-8 is one of the world’s largest third-generation synchrotron radiation facilities and provides very high brightness synchrotron radiation one hundred million times brighter than the conventional industrial X-ray. This radiation enables a high spatial resolution of several µm or below. A small hourglass shaped specimen with a parallel part of φ 1.8× 3 mm were machined from the same material used in this study. A transportable servo fatigue testing machine was set up in the preparation room at Spring-8, and a sinusoidal stress of σ max = 650MPa was applied under R = 0.1 with a frequency of 400Hz. The specimen was periodically removed from the fatigue testing machine, and µCT imaging was conducted at the second hutch of the medium-length beamline BL20XU. The maximum X-ray energy available is 37.7 keV and the final pixel size was 1.45 µm. The specimen was set on a rotating table, and X-rays propagated perpendicularly to the loading axis of the specimen. µCT images were created from radiographs obtained by rotating the table about the longitudinal axis, and the 3D information inside of the specimen was acquired. For reconstructing a 3D image, a conventional back projection method was used. The obtained µCT images were analyzed with the ImageJ image processing program (Schneider, et al., 2012). The more detailed information of the experimental procedures are available from the literature (Yoshinaka, et. al, 2016). Fig. 10 shows examples of internal crack growth processes projected on a plane perpendicular to the loading axis. The dark 4.2. Comparison between crack growth rate in vacuum and that of internal crack measured by synchrotron radiation μCT

Air High vacuum Internal crack Surface crack

10 -6

10 -8

(a) N = 1.64 × 10 7

(b) N = 1.81 × 10 7

10 -10

10 -12

10 3 4 5 6 7 8 9

20 30

(c) N = 1.93 × 10 7

(d) N = 1.96 × 10 7

Fig. 11. Comparison between ⁄ − ∆ curves between internal crack, surface crack in air, and surface crack in high vacuum.

Fig. 10. Examples of internal crack propagation processes projected on a plane perpendicular to the loading axis.