Tim Gurney read Mechanical Sciences at Cambridge University where subsequently he also studied for his doctorate. Following four years with a firm of consulting engineers, he joined The Welding Institute, as it was then known. He has worked primarily on the fatigue of welded structures and is now Chief Research Engineer at TWI.
A decade ago it was realised at TWI that the state-of-play with regard to British fatigue design rules for steel welded joints was becoming very confused. As Tim Gurney clarifies in this review, there were previously at least four major sets of rules ...............
- BS 5400, Part 10, dealing with the design of bridges;
- BS 2573, for the design of cranes;
- BS 5500, for pressure vessels;
- BS 6235 and The Department of Energy Guidance Notes for Offshore Structures.
Of these many would consider the bridge design rules to be the pre-eminent version, and indeed they formed the basis for writing BS 5500, Enquiry Case 79, and the Department of Energy Guidance Notes. Meanwhile, BS 2573 was based largely upon the earlier set of bridge design rules in BS 153, but they were never amended when BS 153 was superseded by BS 5400. Those particular rules were, therefore, clearly outdated.
However, even though these various sets of rules had a clear parentage in the bridge rules, they contained significant differences and additions. These arose partly from the incorporation of more recent research information and partly as a result of a need to cover additional subjects.
For example, the Offshore Guidance Notes included rules for the design of tubular joints and for allowing for the effects of corrosion, neither of which even existed in BS 5400. In addition, they included rules for the use of toe grinding and for 'thickness effect', on both of which subjects much of the research had been carried out after BS 5400 was produced.
There were, however, other significant differences between the various rules and it was beginning to become possible to choose different design stresses for the same joint merely by selecting the 'right' design standard. This was clearly a ridiculous situation. After all, a joint does not 'know' what sort of structure it is in, nor how it was designed; it merely 'knows' that the loading and environment may be different from that to which a joint in another structure is subjected.
The situation was further confused by two other factors. In the first place there are many types of welded 'structure', notably in the mechanical engineering field, for which there are no formal design rules at all, but which are still likely to suffer fatigue failure in service. The designs of many of these were based upon the brief rules produced by TWI in 1976 and which, to a very large extent, were subsequently incorporated in BS 5400.
Others specifically use BS 5400, but in some instances find that a difficult exercise simply because the 'fatigue' and 'bridge loading' parts of BS 5400, are inextricably intermingled - fine for bridges, for which that Standard was written, but unhelpful for everyone else. Still others do not seem to use any rules at all.
Secondly, looming over the horizon, is yet another set of rules in EC3. At the moment it is not by any means clear what their final form will be or even precisely to what types of structure they will actually refer. Clearly, however, it would be helpful if the British could help to make those rules as sensible and intelligible as possible.
In the light of all the above, it was therefore proposed to the British Standards Institution (BSI) that a unified set of fatigue design rules should be produced which, it was hoped, would supersede all the individual sets and, in addition, provide a basis for the fatigue design of structures for which no rules currently existed. To cut a long story short, BSI accepted this suggestion and set up Committee WEE/44 to produce them; the initial results of its labours have now appeared as BS 7608: 1993.
In this first stage the objective was to incorporate the best features of all the existing fatigue design rules but to avoid introducing major revisions that were not in any of them, except in those instances where it was considered that the existing rules were totally wrong or unintelligible. Nevertheless, people who are used, for example, to BS 5400 will certainly find new features, simply because they already existed in another Standard, such as the Offshore Guidance Notes. They will also find that, in some places, the actual rules have not changed but the relevant explanations have, hopefully, been clarified. However, in this respect the author would very much like to hear from users about any clauses where further attempts at clarification are needed.
Taking BS 5400 as a base, it may be useful to summarise briefly the main changes that have been introduced. The following are listed roughly in their order of occurrence in the document, and not necessarily in order of their significance:
- There are some additions to the tables of joint classifications, which have incidentally reverted to their old format, covering, in particular, various details related to welded joints in tubular structures and vessels.
- Upper limits on flaw size, either of slag inclusions or planar flaws, are specified for transverse butt welds in terms of the applied stress range. It should be noted in particular that simply because a transverse butt weld can for example, be in Class D (depending on its form and how it is fabricated) it does not necessarily mean that any defects which it contains have to be consistent with that class. Thus, if the butt weld is situated close to a fillet weld, the applied stresses may be restricted to those for Class F, in which case the butt weld quality requirement could also be relaxed to Class F.
- The hot spot stress concept is introduced for the first time, together with a Class T design curve for designing tubular nodal joints. It has to be noted, however, that the hot spot approach can extend outside the design of tubular nodal joints.
- There have been significant changes to the design rules for bolts under axial loading. Although the rules are now believed to be more intelligible, it has to be said that they are probably still in need of modification. As they stand, the design stress parameter is stress range divided by the ultimate tensile strength of the bolt, but recent work has suggested that an approach based simply on stress range would probably be nearer the truth.
- As indicated previously, the thickness effect criterion has been introduced from the Department of Energy Guidance Notes. This is different from the rule promulgated by the Department of Transport for bridge design (but never introduced into BS 5400).
- There is now a rule covering the influence of seawater corrosion, again taken from the Department of Energy Guidance Notes.
- An increase in design stress is now allowed for transverse welds which have been toe ground in a specified manner. It is, however, recommended that this allowance should not be used ab initio but reserved for use when the need to increase fatigue life is discovered, for example, at a late stage of fabrication or when the structure is already in service.
- Use of fracture mechanics is endorsed for the first time in a British fatigue design standard (as opposed to a defect assessment standard), in situations where the normal fatigue strength assessment methods may be unreliable or inappropriate. This includes such problems as fitness-for-purpose assessment of flaws, exploration of the influence of relatively minor geometrical or stress variations, definition of the required frequency of in-service inspections and assessment of remaining fatigue life of a structure known to contain cracks.
As noted previously, the current Standard does not really include anything that does not currently exist somewhere else; it merely tries to bring everything together. There is, however, now a need to start considering potential revisions based on more recent research. These should certainly include, for example:
- Updating of the 'thickness effect' rule;
- Modification of the design rules for bolts;
- Some relatively minor revisions to the joint classification system which would simplify the eventual transition to EC3;
- nclusion of design rules for orthotropic decks;
- Introduction of rules covering the use of other weld improvement methods, such as peening and TIG dressing.
While it is hoped that these new rules will prove to be helpful to users, it would assist the author to hear from them of any other topics that they would like to see covered and also of any problems that need to be addressed.
