PSI - Issue 13

Gyo Geun Youn et al. / Procedia Structural Integrity 13 (2018) 1305–1311 Author name / Structural Integrity Procedia 00 (2018) 000 – 000

1308

4

Fig. 3. Dimensions of the pipe used for virtual pipe test

Fig. 4. Applied input displacement for virtual pipe test (a) R = 0 (b) R = -1

3.2. Crack growth prediction for virtual pipe test The FE model used for virtual test is shown in Fig. 5. 3-D first order element (C3D8 in ABAQUS) is used with 0.6mm mesh size ( L e =0.6mm) at crack propagation area. Quarter (1/4) model is used with symmetric option. Number of elements and nodes are 112,135 and 121,627 respectively. The multi-axial fracture strain energy model is applied to perform FE analysis. The multi-axial fracture strain energy, W f is suggested in Eq. 6 and critical damage value, ω c is 0.35 for L e =0.6mm. FE prediction results of virtual test are shown in Fig. 6. Fig. 6a is the predicted load-LLD curve of unaged and aged pipe under monotonic loading condition. Fig. 6b and 6c are the predicted load-LLD curve of unaged and aged pipe under cyclic loading condition with load ratio R = 0 and R = -1, respectively. The comparison of crack growth behavior is given in Fig. 6d. From the results, it can be known that the maximum load values are similar between unaged and aged CF8A, regardless of loading type. Also, load-LLD of aged CF8A decreases faster after maximum load point regardless of loading type. However, the interesting point is that the crack growth rate becomes similar between unaged and aged CF8A when loading condition is changed from monotonic to cyclic.