PSI - Issue 19

Masanobu Kubota et al. / Procedia Structural Integrity 19 (2019) 520–527 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

524

5

For the validation of this assumption, a high-cycle fatigue test of the smooth specimen was carried out in nitrogen gas. The fatigue life in the nitrogen gas was also plotted in Figure 4. The fatigue life in the nitrogen gas was the same as that in hydrogen gas. In nitrogen gas, no acceleration of the fatigue crack growth was certain and delayed crack initiation was most likely to occur. Therefore, it is presumed that the same things occurred in the fatigue test in hydrogen gas at low stress amplitude. Either way, it could be considered that the crack initiation in hydrogen gas was delayed. To verify this deduction, the results of the high-cycle fatigue test in the 10 MPa hydrogen gas led by Kubota and Kawakami (2014) are available. They observed crack initiation and propagation behavior of a smooth specimen made of medium-carbon steel in 10 MPa hydrogen gas and air and showed the delay of crack initiation in the hydrogen gas. In the strain-controlled fatigue test, hydrogen accelerates the crack initiation as shown by Kubota et al. (2018) and other researchers. In the high cycle fatigue regime, the effect of hydrogen on the crack initiation was not significant. Instead, the effect of the absence of oxygen and water vapor on the crack initiation appeared. The results of the staircase method for the smooth specimen are shown in Figure 5. The fatigue limit of the smooth specimen determined by the staircase method was 666 MPa in air and 714 MPa in the hydrogen gas. Similar to the fatigue life in the long-life region, the fatigue limit in the hydrogen gas increased. The material exhibited severe hydrogen embrittlement during the SSRT, however, there was no reduction in the fatigue limit in the hydrogen gas. The possible reason why the fatigue limit in hydrogen was higher than that in air was the same as that for the extended fatigue life in the hydrogen gas, that is, the absence of oxygen and water vapor. Nakamura et al. (2000) showed an increase in the fatigue limit in a vacuum environment. To verify this consideration, the fatigue limit in nitrogen gas was obtained and shown in Figure 4. The fatigue limit in the nitrogen gas was the same as that in the hydrogen gas.

Fig. 4. S-N curves of smooth specimen.

Fig. 5. Results of staircase testing of smooth specimen; (a) In air; (b) In hydrogen gas.

Regarding the crack origin, as shown by Murakami and Yamashita (2014), non-metallic inclusions can be crack initiation sites in the high-strength steel. However, it was confirmed that the crack originated from the specimen surface regardless of the environment in this study. There were no non-propagating cracks in the unbroken specimen