The literature contains little useful design information on discontinuity stress decay in cylindrical shells near intersecting axial plate or tubular members. A widely used rule -of-thumb is that the stress will become regular at about one and one -half to two cylinder diameters from the discontinuity, but neither the validity nor the range of application of this rule has been established. This study uses the computerized finite element technique to define the decay rate and pattern in a cylinder near an insert plate, and in a cylinder near the intersection of a small branch. Additionally, the feasibility of using ring stiffeners for stress redistribution near the cylinder intersection is explored.
Today-'s fixed and semi submersible offshore drilling platforms are being designed for use in deeper water and stormier conditions than their forebears. As tubular member diameters and thicknesses are increased to withstand more severe loads, structural weight optimization becomes a necessity. One region that is often overweight, and should therefore receive care- ful investigation, is the tubular joint. Also, as the platforms grow in cost along with size, both owners and insurers are requiring more accurate and comprehensive stress and fatigue analyses in assessing the reliability of the rig.
In recent years, the computerized finite element method has crone to the fore as a powerful and cost-effective numerical analysis technique for platform design in general and for tubular joint design in particular. (1) The soundness of the method has become widely accepted, and it has been used extensively in stress analyses of tubular joints in platforms currently under construction. Inmost of these analyses, the joint design and construction was already complete, including the reinforcements at anticipated critical stress regions; the finite element method was used primarily to validate the design and to locate areas where additional reinforcement was required, i. e., t'hot spots".
The above use of the finite element method is often the only alternative, sinceconstruction deadlines can force fabrication and stress analyses to proceed concurrently. There is a built-in inefficiency in this approach, however. One invariably finds regions of low stress in those designs wherein metal could be deleted without significant effect on strength. Unfortunately, such weight optimization is often uneconomical at an advanced stage of design and/or fabrication. Clearly, the finite element method must be implemented at an early design stage, e. g., before selection of reinforcements such as rings, gussets, and longitudinal stiffeners is complete, if weight is to be optimized. one current program at Southwest Research Institute (SwRI) involving components of a pipe lay barge pontoon.