TWI Bulletin, July - August 2008
Case study: Comparative fatigue lives in question - so ask TWI
When a leading offshore operator weighed up the relative merits of available cathodic protection options for its pipelines it took the added precaution of checking out the effect of competing systems on the fatigue performance ofthe pipe. Would the sacrificial anodes proposed compromise the life of the pipeline?
The client decided to undertake a fatigue test programme to assess the relative performances of two different types of anode attachment, one using an exothermic weld and another using a pin braze system where electrical resistiveheating is used to make a brazed joint.
Four of the exothermically welded attachments were installed on a 20 inch diameter, 18.3mm thickness specimen of the proposed pipeline material. It contained two girth welds 430mm apart in the centre and the attachments were made at90 degree intervals around the centre section. The test ended satisfactorily with a girth weld failure shortly after one and a half million cycles.
However post test examination of the pipe section carrying the sacrificial anodes revealed that fatigue cracking in the pipe wall had initiated at all four attachment welds to a depth of up to half the wall thickness. Clearly theirremaining life was short.
To gain more comparative fatigue data flat plate strip specimens were prepared, each with a pair of anode connectors attached on opposite sides of the plate centre. Four attachments were made with exothermic welds and four with pinbraze lugs following, to the letter, the manufacture's rigorous attachment procedure.
Fatigue testing in axial tension in a 600kN capacity servo hydraulic fatigue test machine was performed, and all tests were conducted to failure. In other words the crack was allowed to develop to around 85 to 95 per cent of theplate thickness and the surface crack was easily visible.
For its client TWI was able to conclude that the fatigue strength of the pipe with the pin braze attachment was found to be higher than that with the exothermically welded attachment. Also it deduced that the strip tests gave aconservative result compared with the full scale tests.
Finally TWI was able to put its client's mind at rest by concluding that, based on the small number of tests performed, the fatigue strength of the pipe with either attachment system was greater than that demanded by the designapproach used by the client.
