PSI - Issue 14

Takashi Nakamura et al. / Procedia Structural Integrity 14 (2019) 978–985 Author name / Structural Integrity Procedia 00 (2018) 000–000

982

5

observed in stage IIb (Fig. 8). The differences in fracture surface between the environments were less in stage IIb. However, the detailed features were different in terms of the striation pattern. As shown in Fig. 9, the striation patterns were obviously formed in air, while there were no obvious striation patterns in vacuum. Based on the facture surface observation, it was clarified that crack growth processes in air and in vacuum were clearly different. And the difference was more evident in the small crack regime. Especially in vacuum, a few micrometer-size convexo concave pattern, which closely resembled the granular area in the fracture surface of internal crack, was observed near the crack initiation site. These findings suggest that the early stage of internal crack growth process is considered similar to that in high vacuum.

(a)

(b)

(c)

Fig. 8. Magnified views of the fracture surfaces: (a) in air, (b) in high vacuum, (c) internal fracture (the granular region in Fig. 1(b)).

(a)

(b)

Fig. 9. Fracture surfaces far from the artificial defect, which correspond to the region around ∆ = 18MPa √ m : (a) in air. (b) in high vacuum. 4. Discussion

4.1. Effects of vacuum on crack growth behavior

As mentioned in Section 3.1, the crack growth rate in vacuum was significantly lower than that in air (Fig. 6). In this section, the reason for the difference in crack propagation rate between air and vacuum is discussed to figure out the effects of vacuum on crack propagation. In the literature, the retarding effect of vacuum on various materials including an ( α + β ) titanium alloy has been reported. The literature has pointed out that oxidization or chemical adsorption on a newly-formed slip stage dominantly affects crack propagation in air (Duquette and Gell, 1971): a newly-formed slip stage at the crack tip in air is oxidized or chemisorbed in a very short time, while these phenomena hardly ever occur in vacuum. Oxides and adsorbates decrease the surface energy on a newly-formed surface; therefore, the crack can easily propagate in air compared to in vacuum. Namely, the decrease in surface energy in air can explain the relative retarding effect in vacuum. Oxides and adsorbates on a newly-formed slip stage can also act as obstacles for dislocation motion at the crack tip. Since these phenomena are almost negligible in vacuum, the plastic deformation at the crack tip is easier in vacuum. The high plasticity in vacuum can promote crack tip opening. According to linear fracture mechanics, the large crack tip opening results in a high crack propagation rate. However, plastic strain in vacuum tends to be highly homogenized, and this promotes more blunting of the crack tip than expected by linear fracture mechanics (McEvily and Gonzalez, 1992). Additionally, the high plasticity in vacuum causes severe crack closure (Sugano, et al., 1989). The high homogeneity of plastic strain, large blunting of the crack tip, and severe crack closure are considered to be the reasons for the retarding effect of vacuum. Another important factor affecting the crack propagation rate is reversibility of slip (Pelloux, 1969, Sriram, 1993, Shyam and Milligan, 2005). The dislocations moving forward during the loading part of the cycle can reverse during the following