PSI - Issue 13

Yanzeng Wu et al. / Procedia Structural Integrity 13 (2018) 890–895

893

Yanzeng Wu et al. / Structural Integrity Procedia 00 (2018) 000 – 000

4

c

b

Fig. 4. (a) Schematic representation of image capturing points in a complete loading cycle. (b) Vertical displacement field of crack tip. (c) Magnification of the region around the crack tip for CTOD measurement.

3. Results and discussion 3.1. Crack closure characterization

As for two cases previously mentioned, the crack opening displacement V as a function of fatigue load P is shown in Fig. 5 for the three pairs of tracking points, located 10, 50, 100 pixel from the crack tip. It can be seen that the variation tendency of crack tip opening displacement is similar for different macrostructures (HAB and non HAB) and distances from crack tip. For lower loads, the displacements have very little variation and they are almost equal to zero, which indicates the crack is closed. As load increases, the crack starts to open gradually and a nearly linear variation is exhibited for higher loads. There is a critical value of load, P op , before the linear variation, which means that the crack starts to fully open. This load is known as the crack opening load and it is essential to estimate the effective stress intensity factor range Δ K eff . Moreover, as the tracking points are moving away the crack tip, the CTOD at the maximum load is found to increase and the crack opening load seems reduced slightly. In addition, a sensitivity analysis is performed to explore the effect of HAB on the crack closure behaviour. In Fig. 5, it can be observed that the CTOD at the maximum load for the HAB case is smaller than the non-HAB case. However, the crack opening load P op for the HAB case is higher compared with the non-HAB case. Accordingly, the effective load range for fatigue crack growth is lower, which leads to a lower fatigue crack growth rate for the same applied load level. Lu et al. (2017) also found that the crack growth rate decelerated at the HABs, which is in agreement with the crack closure characterization. 3.2. CTOD variation In this section, CTOD variation for a full loading cycle is quantitatively measured for the same crack length ( a =2.40mm), where the crack tips are located in HAB and non-HAB zones respectively. Results for different distances from the crack tip are plotted in Fig. 6. The arrows indicate the loading or the unloading paths. It can be found that the CTOD values for unloading branch is larger than that of loading branch when the crack is opening, which is related with crack tip plastic deformation. Thus, the portions of the loading and unloading branches make up a CTOD hysteresis loop which corresponds a full loading cycle. In Fig. 6, it is observed that the width of the CTOD loops decreases with the distance from the crack tip, which means that the influence of crack tip plastic deformation becomes smaller. In addition, by comparing Fig. 6a and 6b, the width of the CTOD loop with respect to HAB zone is slightly smaller than that of non-HAB zone. To quantify the CTOD loops, the areas of the CTOD loops are all calculated by integration and the results are reported as shown in Tab. 1. For non-HAB zone, the area of the CTOD loops decreases as the distance from crack tip increases. For HAB zone, the areas are comparatively close. Moreover, the areas of loops with respect to HAB zone are also smaller than that of non-HAB zone for the same distance value, and the trend is consistent with the value of CTOD for maximum load as shown in Tab. 2. For HAB zone, the value of CTOD for maximum is lower, which is attributed to the lower plastic deformation. In-situ fatigue crack growth results appear that the crack growth rate for HAB (1.05×10 -4 mm/cycle) is lower that non-HAB (3.10×10 -4 mm/cycle) at the same crack length ( a =2.40mm), where the stress intensity factor range Δ K , a linear