Large scale laboratory experiments on vortex-induced vibration

Predictions of the occurrence of vortex-induced vibration (VIV) of risers and other circular cross-section offshore components, either from experiment or CFD, may be placed in doubt because of a frequent mismatch between model and full scale Reynolds number. Assumptions that are sometimes used to support a lack of modelling similarity are that for a responding circular cylinder the flow is independent of Reynolds number and that under VIV cylinder flow is forced to be two-dimensional. The latter one is convenient if a 2D code is being used and the former might be equally applicable to experiments and CFD. However, how valid are these assumptions and, for the prediction of VIV of risers, how important is the scale effect? To try to answer the scale question a number of experimental investigations have been carried out in the DHL Delta Flume in the Netherlands, using circular cylinders with diameters up to 0.75m. Waves up to 2m high can be generated and there is a motorised carriage from which cylinders can be attached to simulate a current flow. With a 0.5m diameter cylinder the maximum Re obtainable is around 5 × 10⁵. VIV has been studied for a flexible cylinder in waves and a flexible cylinder in a current, and in this paper it is mainly the results in a current that are discussed.. The majority of the measurements of VIV response in a current at large scale are restricted to in-line oscillations. These experiments were conducted across a range of reduced velocity, which encompassed a range of Reynolds number from 4 × 10⁴ to 1.2 × 10⁵. Experiments repeated in a small flume at Imperial College; identical apart from a reduction in the Reynolds number values to a range from 3.5 × 10 ³ to 1.15 × 10⁴, produced very similar results for VIV amplitude as a function of reduced velocity. These results at an order lower Reynolds number are close enough to be used in full-scale design studies for risers.