13th International Conference on Fracture June 16–21, 2013, Beijing, China -4- A close examination to this figure shows that, in all cases, the introduction of a hold time into the fatigue cycle produces an acceleration of the fatigue crack growth rate which can reach up to three orders of magnitude for long hold times and for elevated temperature (650 °C). Two regimes are distinguished in this acceleration of (da/dN)tm. For short hold times (typically 1200 s at 550 °C and 10 s at 650 °C), the slope of the normalized FCGR versus hold times, tm is of the order of 0.20 - 0.40. In this regime (named A), the slope of (da/dN)tm versus tm is assumed to value 0.25 for the reasons detailed in the following. This regime A extends over hold times lower than a transition time named ti hereafter (~ 1200 s at 550 °C and ~ 10 s at 650 °C). This transition time is thus dependent on temperature. It is also dependent on ΔK but this load dependence is not introduced here because it is out of the scope of the present work. Figure 3 shows also that when tm > ti, the slope of the (da/dN)tm – tm curves is close to 1, which means that in this regime, called hereafter regime B, the crack growth rate is purely time dependent. The a-N records reported by Gustafsson et al. [10] showed that in regime B, the crack propagated during the hold time (see their figures 7 and 8). A comparison of our results at 550 °C with those reported previously by Gustafsson et al. [10] shows that both materials exhibit similar properties although they experienced different heat treatments. The same comparison at 650 °C with the results reported by Pédron and Pineau [5] indicates that, for the same testing conditions: ΔK = 25 MPa.m1/2, same temperature, same hold time, our material tends to exhibit better fatigue crack propagation properties. These variations in FCGR in the presence of a hold time are related to differences in the microstructure of the materials. In particular our material has a much smaller grain size as compared to those investigated by Pédron and Pineau (d ~ 50 µm) [5]. SEM observations showed significant differences in the fracture modes depending on test conditions (Figure 4). At low temperature and for small hold times the fracture mode is purely transgranular with the presence of fatigue striations (Fig. 4a). At increasing temperature and for longer tm, i.e. in the regime B, the fracture mode is purely intergranular, as shown in figures 4c. In the intermediate regime A, i.e. close to the transition with regime B, the fracture surface is also predominantly intergranular, although less brittle than that observed in regime B. In this particular condition as shown in figure 4b, the fracture mode is mixed, with areas showing features of intergranular and transgranular fracture. Figure 4. SEM observations of fracture surfaces. a) 450 °C, 2 Hz; b) 550 °C, 10/300/10; c) 600 °C, 10/1200/10.
RkJQdWJsaXNoZXIy MjM0NDE=