PSI - Issue 19

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

523

4

3. Result and discussion

3.1 SSRT

Figure 3 (a) shows the results of the SSRT. The ultimate tensile stress in air was 2050 MPa. For the SSRT in hydrogen gas, the specimen broke before the peak stress in air. The reduction of the area was 26.8 % in air and 8.2 % in the hydrogen gas. It was confirmed that the material suffered from severe hydrogen embrittlement in the SSRT. Figures 3 (b) and (c) show the fracture surfaces in the air and hydrogen, respectively. The specimen tested in air showed a cup and corn fracture. This suggested that the fracture started from the interior of the specimen. Therefore, the center of the specimen was observed. The morphology of the fracture surface in air was a dimple. On the other hand, in hydrogen, the fracture started from the surface. Therefore, the fracture surface near the surface was observed. The morphology of the fracture surface was intergranular cracking. These observations also clearly showed that this material was sensitive to the effect of hydrogen.

Fig. 3. (a) Result of SSRT; (b) Fracture surface in air; (c) Fracture surface in hydrogen gas.

3.2 Results of smooth specimen

Regarding the smooth specimen, the S-N diagram is shown in Figure 4. The fatigue life in the hydrogen gas increased compared to that in air, particularly in the long-life region. In the short-life region, the difference in the fatigue life between the hydrogen gas and air decreased. The possible reasons for the extended fatigue life in the hydrogen gas were either the delay of crack initiation or deceleration of the crack growth or both. Regarding crack propagation rate, Somerday et al. (2013) showed the acceleration in their crack growth test in hydrogen gas. Oguma and Nakamura (2113) showed the deceleration of the crack growth in a vacuum due to the absence of oxygen and water vapor. Somerday et al. also revealed that the onset of hydrogen-accelerated fatigue crack growth occurred when the stress intensity factor range,  K , is higher than the critical value. Based on these experimental facts, the reasons of the extended fatigue life in hydrogen gas can be understood. The point is that the onset of hydrogen-accelerated fatigue crack growth depends on the stress level and oxygen and water vapor are absent in the hydrogen environment. In the short-life region, the critical  K for the onset of accelerated crack growth is easily achieved because the stress amplitude was high. In this case, the fatigue life in hydrogen should be shortened. However, in fact, the fatigue life in hydrogen was not shortened at least within the range of the stress amplitude in this study. This suggested that the crack initiation life in hydrogen gas was delayed. In the long-life region, it may be difficult to attain the critical  K for the onset of accelerated crack growth because the stress amplitude is low. Instead of the acceleration, deceleration of the fatigue crack growth was caused in hydrogen gas because of the absence of oxygen and water vapor. Similar to the above, the crack initiation life in hydrogen gas was delayed. Therefore, the extension of the fatigue life became significant with a decrease in the stress amplitude.