Research on the instability of stiffened and unstiffened cylindrical shell components of the type used in offshore oil rigs is described. The research concentrating on the non-linear geometric and material behaviour as influenced by initial distortions and residual stresses, is being carried out by the London Centre for Marine Structures and Materials. The paper reviews the scope of the experimental and theoretical research todate, and presents the results of computer based parametric studies. These results are compared with existing design recommendations. It is concluded that the API rules for unstiffened cylinders are conservative, whereas the corresponding DnV rules are less so and may be unconservative for relatively stocky cylinders. The DnV rules for cylindrically curved panels are examined and appear to be acceptable.
The discovery of hydrocarbon resources in the North Sea, and the subsequent construction of several steel jacket structures offshore has let to an intensive research effort into aspects of the behaviour of marine structures in Great Britain. A series of research centres has been set up by the U.K. Science Research Council, each centre being allocated a particular theme1. The London Centre, comprising both Imperial College and University College, is concerned with marine structures and materials. One of the major projects underway is a study of the inelastic buckling of steel shells of the types used in conventional marine structures and likely to be used in future generation structures.
Conventional braced steel jacket structures are constructed mainly from cylindrical elements. The legs may be orthogonally stiffened or ring stiffened, whereas the bracing members are normally unstiffened and relatively stocky cylinders, Fig. 1. Although a vast amount of literature pertaining to shells already exists 2 much of it is related to aerospace rather than civil engineering structures. Major differences in material properties, construction techniques, slendernesses, levels of geometrical imperfections and residual stresses exist between the shells used in offshore construction and those used for other purposes. The potential pitfalls of extrapolating research information or design rules to areas for which they were never intended cannot be over emphasised, as most experienced civil engineers know that such an expedient may often be a recipe for disaster.
For the above reasons the research at the London Centre, although generally conducted across a broad front, has concentrated in the first instance on the strength and stiffness of cylindrical shell elements. The switch to limit state design codes has focussed the attention of researchers on ultimate behaviour and remarkable progress in our understanding of the inelastic behaviour of flat plated structures subjected to destabi1ising forces has been made in recent years 3. These same techniques are now being used to examine curved plated structures.
This paper aims to summarise the work in progress and planned for the future in this area at the London Centre. It also presents some preliminary results pertaining to cylindrical components which are compared with currently available design rules.