13th International Conference on Fracture June 16–21, 2013, Beijing, China -7- (a). Reference model without disturbance (b). Case 1 (c). Case 2 (d). Case 3 (e). Case 4 0 t s 10 t s 20 t s 100 t s Figure 6. The crack growth process under different boundary condition Fig. 6 shows the simulation results. From case 1 to case 4, the crack growth process becomes gradually closer to the one of reference model. The most significant difference happens in the case 1, whose average value of boundary displacement on the top surface is the largest. While the smallest difference is observed in case 4, since the disturbance is added node-wisely, the average value tends to be zero in a relative smaller area compared with case 2 and 3, and it is numerically proved here that the disturbance of this kind has smaller effect on the far field crack growth as expected. 4.2. Effect of distance The Saint-Venant principle tells that if the distance from the area with disturbance increases, the effect will decrease. The accuracy of observation data is constrained to a certain range. This error could be ignored, since the additional strain becomes smaller as the size of analysis model becomes bigger. However, since the crack growth is a non-linear process, numerical estimation is needed to decide the size of analysis model to ignore the disturbances in boundary condition. This target is similar with the reference model in section 3. The only differences are made in the length in Z direction and boundary condition. The model height becomes 80mm, 160mm and 320mm. As the model size becomes bigger, the distance between the notches and the top surface, where disturbance exists, becomes longer. In order to keep the same strain rate of these models, the
RkJQdWJsaXNoZXIy MjM0NDE=