PSI - Issue 19

Akifumi Niwa et al. / Procedia Structural Integrity 19 (2019) 106–112 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

110

5

3.2. Effect of frequency on high temperature bending fatigue test with axial stress

23.4 MPa was set as the axial stress which value is higher than the stress level that can occur in the actual equipment, but was set to clarify the influence of axial stress. Fig. 6 (a) shows the results of high-temperature bending fatigue tests conducted at frequencies f = 5 Hz, 12.5 Hz, and 20 Hz by applying an axial stress of 23.4 MPa in this way. The vertical axis is the stress amplitude on the surface of the center of the notch calculated by FEM analysis using the Young's modulus corresponding to the temperature distribution in the furnace of testing machine. However, it is a virtual elastic stress that assumes that the specimen does not yield and cause the plastic deformation. The horizontal axis represents the number of cycles when the test piece is completely separated. From Fig. 6 (a), it was found that the fatigue life at lower frequency tended to be shorter than that of higher frequency. In general, it is known that high temperature fatigue can not ignore the effects of creep, and that frequency dependence appears [8]. In particular, because higher axial stress is applied in this test, it is inferred that the effect is more pronounced. Then, the result which the horizontal axis is changed to the time when the test piece is broken is shown in Fig. 6 (b). It is clear that the time dependency is strong, as the difference in fatigue life by frequency is almost eliminated when arranged by time to failure. From this result, it is thought that creep has a strong influence on the fracture mechanism. The test results also show that in the double logarithm graph of Fig. 6 (b), the virtual stress amplitude and the failure time are not in a linear relationship, and the graph bends downward as the stress amplitude decreases. This indicates that the stress amplitude dependence on the failure time is large on the higher stress amplitude case, and the dependence is small on the lower stress amplitude case. In order to investigate the fractured surface of the specimen, the fractured surface was observed by scanning electron microscope after high temperature bending fatigue test with axial stress of 23.4 MPa at f = 20Hz, and the result is shown in Fig. 7 . In this figure, it is shown that the fracture surface when the stress amplitude is (a) σ a = 17.8 MPa and (b) σ a = 82.9 MPa, respectively. From this result, at the low stress amplitude case such as (a) σ a = 17.8 MPa, intergranular fracture and void formation in the grain boundary caused by creep are observed. On the other hand, on the high stress amplitude case such as (b) σ a = 82.9 MPa, a creep fracture surface appears near the center of the thickness direction of the test piece. But near the upper and lower surfaces of the test piece, necking due to plastic deformation was seen. Such fracture surface morphology is clearly different from that observed in high temperature fatigue fracture in stainless steel, nickel-based ODS alloys and so on, suggesting that a different fracture mechanism is occurring [9-11].

Fig. 6. Result of bending fatigue test with axial stress at 1400°C. (a) Horizontal axis represents number of cycles to failure, (b) Horizontal axis represents time to failure.