PSI - Issue 42

Diego F. Mora et al. / Procedia Structural Integrity 42 (2022) 224–235 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

234

11

critical energy release for the whole temperature range (case 1) is represented in Fig 10(a), while that for temperature dependent critical energy release (case 2) is shown in Fig 10(b). It is clear from this result that in case 1, the temperature does not affect the crack growth and even if the crack tip is in a hot region of the material, the crack does not arrest. Opposing to this, in case 2, the temperature influences the fracture toughness and therefore controls the crack advance.

(a)

(b)

Fig. 10. Crack growth and temperature contours (a) case 1; (b) case 2.

The results of the crack propagation of the circumferential defect are displayed in terms of the COD in Fig. 11. This figure compares the COD obtained in the experiment and the FE-simulation with the mesh 1 and 2 for case 2.

Fig. 11. COD evolution during the PTS transient.

Some differences can be seen in Fig. 11 between the experimental and the numerical results using XFEM. The experimental results show a rapid increase in the COD and the beginning of the experiment due to some ductility at the crack tip and crack blunting. In the simulation, the COD at the beginning of the transition is small, this is because the crack propagation is assumed to behave brittle at any time and neither blunting nor ductility are considered in the simulation. In the experiment, clear stepwise COD is shown during the crack growth indicating crack arrest and reinitiation cycles, while in the simulation there is no clear step in the COD response. The COD curve for both