A series of fatigue tests is described in which part-wall fatigue cracks in tubular joints were repaired by 'remedial grinding', that is the cracks were removed by grinding and the resulting excavation left unrepaired. The usefulness of this repair technique is examined by comparing the number of cycles endured after repair with the fatigue performance of unrepaired joints.
Fatigue cracking in offshore structures is an inevitable consequence of the arduous environmental loading imposed, especially in the North Sea. While every care is taken to design structures with good fatigue resistance, errors in design or fabrication, or inadequate definition of environmental loading can lead to premature cracking. Such problems are arising with increasing frequency as many of the structures currently in operation are approaching their design lives.
Offshore repairs are invariably difficult and expensive, especially for submerged members, and may result in significant downtime and consequential loss. Repair methods which have been adopted include repair welding and mechanical or grouted clamps. The choice of the most appropriate repair technique depends on many factors such as water depth, ease of access, member size and thickness, suitable weather 'windows', etc. Repair weldingunderwater requires a hyperbaric chamber in which divers operate in dry conditions. The added complexity of constructing a hyperbaric chamber and using qualified divers for welding and inspection, etc adds significantly to the cost. The ideal repair in most cases is that which can be carried out with a minimum of diver time.
A further important consideration is clearly the fatigue performance of the repaired member during its remaining service life, and this has been addressed in an industry - sponsored programme conducted at The Welding Institute. A previous publication (1) described fatigue tests on tubular T-joints in which weld toe fatigue cracks had been repaired by welding. The study described here examined an alternative repair technique in which part-wall fatigue cracks were removed by burr grinding, and the resulting excavations left unrepaired. The particular advantage of this 'remedial grinding' technique with respect to offshore structures is that it can be applied 'wet', ie without construction of a dry habitat underwater. In addition, the level of skill required of the diver is considerably less than that for welding.
Removing material from the vicinity of weld toes by grinding is a well established technique for improving fatigue strength. When performed correctly, the process removes the sharp slag intrusions that are intrinsic at all weld toes, and which are responsible for the inherently low fatigue strength of welded joints. In addition, grinding changes the local geometry at the weld toe and provides a lower stress concentration. When grinding is carried out to remove fatigue cracks however, the resulting excavation depth is likely to be a significant propagation of the wall thickness which, together with the notch effect of the excavation, may reduce fatigue strength rather than improve it. A previous study (2) on joints in flat plates had established that useful fatigue life was restored for excavation depths up to one third of plate thickness.
