Stresses in horizontal storage vessels - a comparison of theory and experiments

Abstract

This paper presents the salient features of the experimental work carried out on a series of horizontal storage vessels. These include a 3.66 m diameter gas receiver and two smaller vessels each of 0.91 m diameter. The saddles used to support the vessels represent the two extremes of saddle design where the cantilever section is either of 'flexible' or 'rigid' construction and in both cases welded to the vessels. All the vessels were extensively strain gauged in the region of the saddle supports and at the centre of the vessel. Hydraulic, pressure surcharge and pneumatic tests were carried out on the vessels. Typical experimental results have been compared with the theoretical analysis described in (1)t. It is found that a satisfactory measure of agreement is obtained for both the flexible and more rigid results for the water fill tests, but that the pressure results are sensitive to vessel imperfection which are not considered in the theoretical analysis (1).

UK Rules for Unfired Pressure Vessels

“…The procedure, based on the work of Zick [21] and implemented into PD5500 by Tooth [22] can be broken down into four main design elements. When the vessel is situated on twin saddles, the following high stress (f) areas must be investigated:…”

Numerical analysis of a sling support arrangement for GRP composite pressure vessels

“…Work by Tooth et al 7,8,9,10 employs a shell analysis and Fourier series approach that evaluates the form of the contact interface pressure profile. This is undertaken by subdividing the contact region into a number of discrete areas around the circumference and across the saddle width.…”

The design of twin-saddle supported multi-layered glass-reinforced plastic vessels—the development of a parametric equation for the maximum strain

“…The analysis given in PD 5500 is essentially based on the work of Zick [5]. More recent experimental work by Tooth and coworkers [6,7], on steel vessels, indicated that Zick's treatment for the vessel full of fluid underestimates the maximum stresses in the vessel when a 'rigid' saddle is employed. It was therefore considered that the Zick approach, given in PD 5500 [1], could well predict unconservative (i.e.…”

“…To improve the accuracy of these predictions the earlier theoretical analysis [6,7], appropriate to the isotropic vessel, was extended to analyse the multilayered GRP vessel. After extensive experimental analysis, using three full size vessels, a mathematical 'model' has been established, and a validated theoretical analysis developed, which can handle vessels where the lay-up is either symmetrical or asymmetrical with respect to the vessel wall centre-line.…”