Large-Scale Tests for Vortex Shedding Effects on Marine Risers

Abstract

INTRODUCTION The interest increasingly being shown in floating production systems has drawn renewed attention to the problems of riser design. The production riser is clearly a crucial item which requires a design procedure in which a high degree of confidence can be placed. In this respect the requirements of a production riser are more exacting than those of the drilling risers used in the exploratory phase. A reliable prediction of service stress history is essential to the prediction of fatigue life. This is usually obtained from one or more of the computer simulation packages available for riser simulation. Any computer simulation is only as accurate as the mathematical modelling it contains, and in writing such a program assumptions are made as to which effects are significant enough to include, the manner in which to simulate them, and in the choice of various empirical constants involved. Many riser simulations are purely two-dimensional but experimental data, such as that presented here shows that a uni-directional wave input does not result in a uni-directional response. Vortex shedding effects are a significant result of wave loading on a marine riser. THE TEST PROGRAM Research teams from the Departments of Mechanical Engineering at Heriot-Watt University and University College London began a programme of model basin testing in 1980 to supplement the scarce data available for verification of marine riser simulation computer programs. Some early results were published in reference (1). Combinations of floating platform and riser models were subjected to regular and random wave excitation and the resultant displacements and bending moments of the riser were recorded as digital time histories on computer discs. Three riser configurations were chosen for study: Two, designated A and B, were of circular cross-section but of different diameter to length ratios, while the third model C, was of a multi flow line pattern. THE TEST FACILITIES AND TECHNIQUES The tests reported here were carried out in the No.3 towing tank of National Maritime Institute (NMI) Ltd., Feltham, England by a team from Heriot-Watt University and from University College London. This tank is 25 ft (7.6 m) deep, 50 ft (15.2 m) wide and 1300 ft (396 m) in length and has a vertical wedge wavemaker which can provide unidirectional waves, either regular or to a given energy spectrum. A tower, consisting of two vertical columns to which 12 television cameras were attached was erected in the tank approximately 300 ft (91 m) from the wavemaker. The signals from the cameras viewing the riser were analysed on-line providing real time digital displacement data. The surface vessels used were not designed to model specific floating platforms. Their function was merely to support the top of the riser and produce realistic surge and sway inputs. The vessel used in conjunction with riser A was a semisubmersible design moored fore and aft by catenary moorings. For the experiments described here it was fitted with a tensioning device which produced a top tension of 13.6 lbf + 4% (60.6N).