Fatigue damage in marine structures is generally assessed by means of stress−number of cycles (S−N) curves. These curves are obtained from many experimental tests on structural components. Fracture mechanics (FM), using the crack propagation method is also a practical technique. In this approach, fatigue damage is estimated based on crack propagation from an initial crack size through the surface and depth of structural members. In this paper, S−N curves based on FM were compared with S−N curves obtained following two standards: DNVGL-RP-C203 and API-RP2A-WSD. The results showed that the curves did not agree well. Therefore, to obtain better fitting, calibration factors are needed. From S−N curves established for two stages, two calibration factors – α1 (for the first stage) and α2 (for the second stage) – are proposed for fitting S−N curves based on the FM approach with the DNV GL and the API S−N curves. The wall thickness, the initial crack depth and the critical crack depth were found to have a significant effect on determination of the calibration factors.

No abstract