PSI - Issue 22

Yaroslav Dubyk / Procedia Structural Integrity 22 (2019) 275–282

280

6

Yaroslav Dubyk / Structural Integrity Procedia 00 (2019) 000 – 000

Fig. 4. (left) Crack Opening Area for axial crack; (right) Variation of normalized pressure loss terms in modified Henry-Fauske model.

3. Results and discussion Monte-Carlo simulations were performed with number of samples 10 6 . First Ultimate, Yield Strength as well as Fracture toughness distributions were generated. For a Normal operation conditions FAD was established and critical crack lengths for axial and circumferential crack were iteratively calculated. Then, using Eq. (4), COA as well as COD were calculated, followed by leak rate estimations, according to modified Henry-Fauske model for different crack types (see Table 3.). Some results of these calculations are presented in Fig. 5-7. As we can see from Fig. 5, accounting for crack morphology strongly affects the results scatter, and it is more significant then mechanical properties scatter (see Fig. 6 and Fig. 7). Among the crack morphology characteristics n L (number of bends and protrusions per mm of flow length) is the most significant parameter, that’s why IGSCC crack type (see Table 3) is the most representative in failure analysis. It can be seen from Fig 6 that Fracture toughness has bigger influence rather than Ultimate (Yield) strength, which corresponds to Stefanini & Bloom (2018) results.

Fig. 5. Leak Rate for IGSCC cracks, Normal Operation conditions: (left) Axial Cracks (right) Circumferential Cracks