13th International Conference on Fracture June 16–21, 2013, Beijing, China -8- increasing of cyclic numbers, one crack was formed again along the interface of the lathy martensites in the nearby grains, as can be seen in Fig. 8c. This would provide an easy path for the crack passing through the grain boundary since the cyclic plastic deformation has been extended to this area. Figure 9. Coalescence of micro-cracks on the grain boundaries at the loading angle of 60° Moreover, the damage area was dependent on the da/dN, i.e., the higher the FCG rate, the larger the damage area. This is supported by the results in Fig. 9a, where a large number of micro-cracks has been formed at a distance about 30 μm in front of the main crack tip. Both branch and shear micro-cracks were occurred on the grain boundary. With the fatigue life increased to 1739 cycles, some micro-cracks were coalesced with each other, resulting in relatively longer cracks. Finally, all the micro-cracks were interconnected to form the main crack at N of 1753 cycles. The interconnection of grain boundary cracks increased the FCG rate considerably, which was plotted in Fig. 3. By comparing with the crack path at the loading angles of 30° and 60°, it can be observed that the length of branch cracks at α of 30° are longer and more shear micro-cracks at α of 60°. This was related to the branch versus shear crack competition in the mixed mode loading conditions. 4.2. Prediction of crack growth deflection angles The initial deflection angles of FCG were predicted based on the SED criterion [2] in Eq. (4), the MTS criterion [1] in Eq. (5) and the criterion of Richard [4] in Eq. (6). In Eqs. (4) - (6), KI and KII are the mode I and II components, ν is Poisson’s ratio, θ0 is the critical deflection angles, and β is the angle between external loading and the crack surface. In this work, ν is set as 0.3, and the corresponding results are shown in Table 2. It is obvious that compared with the experimental data, all the predicted deflection angles for α of 30° and 60° are higher than the experimental data. Fortunately, the calculation based on MTS model was close to the experiments. Thus, the MTS criterion is suitable to predict crack growth direction in the HAZ under combined loadings. Nevertheless, all of the criteria cannot estimate the deflection of crack growth direction under pure mode I loading. This can be ascribed to the limitation of the models to homogeneous materials, and thus results in the disparities of crack deflection angle in mode I loading.
RkJQdWJsaXNoZXIy MjM0NDE=