PSI - Issue 75

Per-Olof Danielsson et al. / Procedia Structural Integrity 75 (2025) 572–580 Per-Olof Danielsson et al. / Structural Integrity Procedia (2025)

579

8

As obvious from Table 2, fatigue lives predicted using the ENS method fail to meet required life targets in the four locations in this use case. This discrepancy arises primarily because the ENS method includes stresses that do not contribute to crack propagation, leading to overly conservative fatigue life estimates (id 1 and 2 in Table 2). Such issues are difficult to resolve within the ENS framework. In contrast, at locations where non-damaging stress components — such as stresses parallel to the weld — are less significant (IDs 3 and 4 in Table 2), the difference in predicted fatigue life between the ENS method and the SFM approach is correspondingly smaller. The inclusion of weld residual stresses (WRS) in SFM calculations significantly affects predicted fatigue life. Fig. 8.a-d. illustrates the stress intensity factor, K I , transformed to represent crack-opening stresses along the anticipated crack path, both as a function of time at crack initiation and as an equivalent stress intensity factor after rainflow counting,  K I,eq . The practical advantages of the SFM approach become evident when reviewing Table 2. Without the SFM method, design revisions, extensive physical testing, or referencing analogous existing solutions would be required. The SFM approach, however, quickly provides clear decision support, enabling confident engineering choices with reduced effort and uncertainty.

Fig. 8a-d. Stress intensity factor K I as a function of time (top row) and equivalent stress intensity factor range,  K I,eq as a function of the crack length (bottom row) for positions Id 1-4 from Table 2. Both K I and  K I,eq are defined in directions that promote crack opening along the prescribed crack path. These graphs illustrate how local loading conditions and weld geometry affect crack-driving forces along the predicted crack path.