PSI - Issue 19

Akira Ueno et al. / Procedia Structural Integrity 19 (2019) 494–503 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

500

7

Fig. 9 Relationship between RRA and Ni eq (Yamada, T. et. al .(2012)).

4.2. Effect of mean stress on fatigue limit

Fig. 10 Relationship between  w and  mean .

Figure 10 shows the relationship between fatigue limit  w and mean stress  mean . Three straight lines inclined right downward show the modified Goodman diagram of each conditions. For cases of aluminum alloy A7075 (Ueno, A. et. al. , (2012)), data plots obtained under R = -0.5 and 0.0 in H 2 gas located rather lower than dashed-and-dotted line estimated from  w at R = -1.0 and tensile strength. For case of low Ni eq SUS316L, data plots obtained under R = -0.5 in H 2 gas located on the dashed line. And also, difference between solid line(in air) and broken line(in H 2 ) is little. Therefore, as compared with H 2 sensitivity under high mean stress of A7075, that of low Ni eq SUS316L is slight problem.

4.3. Effect of Ni eq content on stability of the austenite phase

For cases of low Ni eq SUS316L, stress-induced martensitic transformation are promoted. And then, martensitic phase easily influenced by hydrogen embrittlement. In this section, distribution of the stress-induced  ’ martensitic phase are analyzed by means of the EBSD. Figure 11 shows a schematics of analyzed position. One is a near fracture origin region((a)), the other is a near specimen surface region((b)) on the vertical section of fatigue fractured specimen. The  ’ martensitic phase analyzed