This paper describes procedures which can be used when applying fitness-for-purpose concepts to flaws in welded constructions for the avoidance of failure by fatigue and brittle fracture. These procedures are based upon well established fracture mechanics methods which are already employed in offshore engineering to assess reliability of structures. Examples are given of the application of engineering critical assessments (ECA) to a number of typical post-fabrication problems. IC is shown through these examples how laboratory test data are combined with detailed numerical analyses to arrive at possible critical conditions. Such criteria are then compared with the results of extensive on-site inspection surveys to enable the performance of the structure to be assessed. The decision to instigate the appropriate improvement or repair action can then be made, in the light of the original design requirements.
The paper arrives at the conclusion that the judicious use of fracture mechanics based methods can assist the owners, operators and fabricators of offshore structures in obtaining reliability in an economic manner.
The Engineering Critical Assessment '(ECA) procedures described in this paper can be used to quantify the potential effects of various types of planar weld defect or, alternatively, of variations in weld dimensions. Defects, or dimensional geometry variations may be introduced into a structure during fabrication at which time they may be innocuous as far as brittle fracture or some other failure mode is concerned. However, in an offshore structure defects may grow by fatigue under cyclic loading and ultimately prove to be unacceptable at some later stage during the service life of the structure. It is therefore necessary to ensure that the fatigue life of the fabricated welded joint .is greater than the minimum service life and that any defect present will not grow to such a size as to cause either large scale brittle fracture or local ligament collapse during this life.
A detailed background description of ECA is contained in the British Standards Institution Published Document BSI PD 6493:980 (I). This document is not a standard or code, but was published to provide guidance and to simulate further research and discussion. It is at present under review.
ECA procedures are similar in nature to the defect assessment method (2,3) which has been employed during the design of offshore welded structures. The major difference is that the defect assessment method is used during design to produce 'target' values for toughnesses, acceptable initial defect sizes and weld dimension tolerances. Within these 'target' values will be a certain degree of conservatism and if they are achieved during fabrication, an efficient and safe structure will result.
For many reasons, the actual structure in the fabrication yard may differ from the design drawings and requirements. Most of the differences will be unimportant but some, such as achieved toughnesses, real weld dimensions and planar weld defects may compromise the reliability of the structure.
