This paper presents results from a 1997 joint industry project arranged by MIT and Marintek. The data was taken at the Marintek towing tank in Trondheim, Norway. The primary goal of the project was to evaluate how flow-induced vibration of a marine riser is affected by placement of staggered buoyancy along the riser. The riser was subjected to both sheared and uniform flow, and was tested with various configurations of buoyancy. Cross-flow measurements of acceleration were analyzed to determine dominant response frequency and vibration amplitude. Tension and flow velocity were also measured. It is shown that a riser with 50% staggered buoyancy or greater will have response dominated by vortex shedding on the large diameter buoyancy modules. It is also found that the addition of buoyancy may decrease fatigue damage rate, provided that measures are taken to minimize increases in the ratio of mass per unit length of the riser to mean tension. It is primarily the decrease in vortex shedding frequency due to the larger diameter that accounts for any reduction in fatigue damage rate. Unfortunately, the benefits gained from the addition of buoyancy may be undone, by the typical increase in mass and decrease in riser tension that occur when buoyancy is present.
The effect of buoyancy distribution on riser response and fatigue damage rate is not well established. It is known that an entirely bare riser will exhibit vortex-induced vibration (VIV) at a frequency governed by the Strouhal relationship,. In particular the frequency is inversely proportional to the diameter. A bare riser will vibrate at a higher frequency than a riser completely covered by buoyancy of a much larger diameter. But what frequency dominates in the case of a riser with both bare and buoyant regions? In this case design guidelines would be helpful in choosing optimum buoyancy coverage patterns. The experimental results presented in this paper provide some guidance that will prove useful in making design choices.
A model riser was constructed and tested. The model was fitted with various configurations of staggered buoyancy, and tested in the Marintek rotating rig. The rotating rig enabled testing in both uniform and linearly sheared current profiles. Measurements of riser tension, velocity, and cross-flow acceleration at multiple points on the riser, were included.
The ultimate objective is to understand the effect of buoyancy distribution on fatigue damage rate. In the analysis of the data a technique is shown which allows the measured acceleration response to be interpreted in terms of stress and ultimately fatigue damage rate.
The riser was made of commercial PVC pipe. The PVC portion of the riser was 11.340 meters long with wall thickness of 0.0023 m. The total riser length includes the attached load cell and came to 11.479 m. The mass of the bare riser was 5.3 kg, and the mass per unit length was 0.47 kg/m. Additional material properties of the PVC riser are given in .