Response Variability in Flexible Cylinder VIV Model Test Data

Abstract

This paper presents some results from the 2011 SHELL tests at the MARINTEK basin. The tests involved towing densely instrumented 38m long flexible cylinders at Reynolds numbers up to 220,000. The main objective it to present the experimental results in a manner that describes the response variability that exists in the measured response data. Despite the fact that VIV is known to be a stochastic process, this is rarely addressed in the literature and currently there is no framework or 'best-practice' in the VIV community that can address statistically non-stationary data. In this paper, the experimental measurements are treated like non-stationary time-series and all statistical quantities which are typically of interest are computed with the use of short duration moving windows (or time-gates). A novel way of plotting and presenting VIV response data for flexible cylinders is introduced that is capable of revealing the inherent variability that exists in the cylinder's response.