Large scale fatigue tests on internally stiffened tubular welded joints have been carried out at the National Engineering Laboratory as part of the second phase of the United Kingdom Offshore Steels Research Project.
Tubular joints in the T configuration have been fatigue tested in air under constant amplitude loading conditions. The results, when presented in the conventional S-N format where S is the applied hot spot stress range and N is the number of cycles to failure, suggest that the introduction of heavy stiffening is detrimental to fatigue life while light stiffening enhances fatigue life, when compared with tubular joints of similar geometry with heavier chord section thickness.
The unconventional failures experienced by the joints in this investigation cast doubts on the applicability of the hot spot stress approach to ring stiffened tubular joints.
Structural performance requirements of fixed offshore structures being designed and built for operation in the North Sea, have increased over the past decade as exploration extends into deeper water. In order to achieve the rigidity and ultimate strength levels necessary, nodal joints have had to be designed with extremely large wall thicknesses. Such a necessity results in considerable penalties in terms of lower fatigue design curves, not to mention supply and fabrication constraints.
One alternative design option to overcome such limitations is that of internally ring stiffened nodal joints. Although, it is essential that the fatigue characteristics of these complex joints are fully understood, little work has been carried out in this area. However, such a paucity of information has not prevented the technique being utilised in current designs to capitalise on the improved static strength and considerable weight saving. Since the radial flexibility of the chord member has a direct influence on the stress/strain concentration factors (SCF/SNCF) and thus on the fatigue life of simple joints, internal stiffeners are often used to reduce this flexibility and hence reduce the SCFs.
The designer has two methods for assessing the fatigue performance of tubular joints: fracture mechanics and the empirically derived relationship between applied stress range and life, the S-N approach. The use of the former approach is still in its infancy at the critical design stage and thus the latter is more widely used. The S-N approach relies heavily on the establishment of the maximum stress range applied to a joint and this is found by multiplying the nominal stress range in the brace by an SCF to give the 'hot spot' stress range. Parametric formulae giving SCFs have been derived for some joint geometries under various loading conditions(1-3), but until very recently none have appertained to internally stiffened joints. In those cases where there are no appropriate formulae, the designer has to either adopt a conservative blanket value for the SCF or resort to numerical analysis. The fatigue life is then found from the application of the hot spot stress range and an S-N curve appropriate to the geometry in question.
