API RP2A Edition 11 thru 21 welded joint X and X' fatigue design curves have been replaced by a basic SN curve with slope m=3 changing to 5 at ten million cycles. In this paper, the new curves are compared with test data from both API-sponsored and European research -- including inter-related corrections for seawater, thickness, and use of weld profile control. Corrections for grinding and peening are also introduced, as well as SN curves and fatigue design requirements for cast nodes and grouted joints.
The simple Alpha Kellogg SCF equations have been replaced by the more rigorous Efthymiou equations, with new adjustments for chord stress effects and weld toe position.
The API fatigue revisions have achieved better alignment with the ISO 19902 draft design procedures, without sacrificing the economy of existing practices. However, they are more complex. Designers will have a lot of new refinement to deal with, and having this background will help them understand it.
Fatigue provisions in API RP2A were last updated in 1992, taking into account API-sponsored research from 1976–89 and a larger body of European work done primarily in 1978–87. That research re-affirmed the hot spot stress approach which had been in wide use since 1972, but refined it to include a more rigorous definition of the relevant stress, and to include the effects of random loading, of sea water environments, size, and profile.
During the 1990s, API established the Offshore Tubular Joint Research Center jointly with the Edison Welding Institute and the University of Ohio. OTJRC continued to review data on fatigue and static strength, and conducted new research in these areas. Unfortunately, their proposals for revision lay dormant because RP2A itself lay dormant for seven years, with most standardization efforts going into ISO.
