PSI - Issue 19

Akifumi Niwa et al. / Procedia Structural Integrity 19 (2019) 106–112

111

Author name / Structural Integrity Procedia 00 (2019) 000 – 000

6

Fig. 7. Fracture surface after bending fatigue test with axial stress. (a) σ a =17.8 MPa, (b) σ a =82.9 MPa.

Next, Fig. 8 shows the result of cross-sectional observation of the fractured part of the test piece after the test. In this figure, (a) shows tested at σ a = 17.8 MPa and (b) shows tested at σ a = 82.9 MPa. From this result, when the stress amplitude is small such as σ a = 17.8 MPa, a clear intergranular fracture is observed, and voids are observed at the grain boundary in the entire region in the test piece. However, the voids found here are somewhat different in morphology from the fine void formation and connections in grain boundaries found in high temperature fatigue of other alloys [12, 13], but the wedge and round type voids seen in creep damage can be seen. On the other hand, when the stress amplitude is large such as σ a = 82.9 MPa, grain boundary fracture is limited near the center of the thickness direction of the fracture surface, and necking due to plastic deformation occurs near the upper and lower surfaces. In addition, it was found that although there were many creep voids similar to the case of low stress amplitude near the center of the thickness direction, creep voids were hardly generated in the part where plastic deformation occurred in the upper and lower surfaces. Since the failure mechanism differs between the low stress amplitude case and the high stress amplitude case in this way, it is thought that the virtual stress amplitude and the failure time are not in a linear relationship in the double logarithm graph shown in Fig. 6, and the graph bends downward as the stress amplitude decreases.

Fig. 8. Cross sectional observation after bending fatigue test with axial stress. (a) σ a =17.8 MPa, (b) σ a =82.9 MPa.