This paper discusses the recent experimental and analytical studies related to buckling of fabricated steel cylinders subjected to combinations of axial compression load and external pressure. The effects of initial imperfections and residual stresses are being investigated. A description of the test facilities and model geometries are given. The D/t ratios of the models range from 48 to 1000. The test programs include ring and stringer stiffened as well as ring stiffened cylinders.
Stiffened cylinders are used extensively as primary members of offshore drilling and production platforms. In addition to providing support, tubular members may be used for ballasting during installation, provide buoyancy during operation for floating production platform and possibly for storage of fluids during operation.
As offshore structures become larger and more expensive to build it is necessary to have a solid base of experimental data that is representative of actual structures. A good experimental base is necessary for development of design criteria that provides an economical design as well as a reasonable degree of confidence in the margins of safety provided for various conditions that may arise.
Various codes have provided or are in the process of providing rules for offshore structures. tm0ng these are American Bureau of Shipping (ABS), DNV and API RP 2A2 Other codes that have rules for Stiffened and unstiffened tubular members are ASME, British Standard4, ECCS5 andDAST6. ABS is presently developing design criteria for tension leg platforms.
At the present time there are significant differences in the stability criteria given by the different rules. The approaches used are very similar, however. For all except the British Standard, the buckling stresses for cylinders under axial compression or external pressure loads are determined by applying knockdown factors to the theoretical elastic buckling stresses to account for such effects as initial imperfections, residual stresses, and nonlinear material properties in the inelastic region. For combinations of these loads interaction equations are given. The differences in the rules are due principally to a lack of sufficient test data on fabricated cylinders from which mathematical models for knockdown factors and interaction curves can be developed.
This paper discusses several of the test programs that are now in progress to provide the necessary experimental data base. Primary emphasis is now directed toward stiffened cylinders subjected to simultaneous longitudinal and hoop compression loads.
The interaction equations given by Refs. 1-7 for fabricated steel cylinders subjected to combinations of axial compression and hoop compression are summarized below. A comparison is made of the shapes of the curves in Fig. 1. This figure shows the obvious need for a resolution of the differences.
The American Petroleum Institute, API RP 2A2 provides recommended rules for the design of fixed offshore platforms. Design rules are given for unstiffened and ring stiffened cylinders under external pressure and unstiffened cylinders under axial compression and bending. Interaction equations are given for combined axial tension or compression and external pressure.