PSI - Issue 80

Alok Negi et al. / Procedia Structural Integrity 80 (2026) 203–211 AlokNegi / Structural Integrity Procedia 00 (2023) 000–000

208

6

Fig. 2: (a) Geometry, boundary conditions (symmetry exploited), and finite element mesh detail (biquadratic plain strain quadrilateral elements) of the tiny-pipe model used in numerical simulations. The pipe geometry corresponds to idealized geometry (infinitely long pipe with longitudinal crack-like imperfection on pipe OD) used for FAD calculations [API TR 5C3 (2019); API 579 (2007)]. The symbol R i denotes the inner radius of the pipe, h is the wall thickness, (b) Internal pressure versus time loading history applied during the pipe burst test [API PRAC II (2012)], (c) Contour plots illustrating crack growth through phase-field damage variable ϕ , (d) and hydrogen di ff usion through concentration C distribution in pipe at the end of a testing period t of 672 hours. 3.3.1. Influence of notch Experimental observations from the PRAC II program [API PRAC II (2012)] revealed that the actual machined notch geometries in full-scale pipe specimens deviate from the idealized sharp flaws typically assumed in FAD-based evaluations. These deviations, such as blunted tips, tend to reduce local stress intensification and alter hydrogen ingress patterns, potentially leading to incorrect failure pressure predictions through FAD. In the PRAC II tests, the observed notch geometry exhibited such double-bottom characteristics, prompting a detailed investigation into the role of realistic notch morphology on SSC susceptibility. To assess this influence, the full geometric profile of the pipe and notch—including the double-bottom features—was explicitly modeled, as shown in Fig. 3(a). The corresponding phase-field simulation results in Fig. 3(b) demonstrate that, under identical environmental exposure and loading condi tions, a blunt notch does not lead to crack initiation, even after prolonged exposure durations. This result supports the conclusion that the double-bottom notch morphology alone is not the primary driver of the reduced failure pressures observed in the PRAC II experiments. 3.3.2. Influence of residual stresses Explicitly simulating manufacturing steps to generate residual stresses is computationally intensive. Therefore, a simplified thermo-mechanical approach is adopted to generate representative hoop residual stress fields. As shown in Fig. 4(a), residual stresses are introduced by applying thermal gradients across a pipe segment, mimicking thermal expansion and contraction. The resulting hoop stress distribution, normalized by the yield strength σ y 0 , is shown in Fig. 4(b) for an example case of applied temperature di ff erential of ∆ T = 2 × T s = 77 ◦ C. The residual stress distribution in Fig. 4(b) is consistent with approximations used in methods like the split-ring test [Nawathe et al. (2016); Elkhodbia