Stochastic modelling of cross-flow vortex-induced vibrations

Abstract

Semi-empirical models are commonly used to predict vortex-induced vibrations (VIV). Empirical parameters are often based on forced harmonic motion tests of short rigid cylinders in one degree of freedom, which is a significant simplification of the three dimensional VIV experienced by flexible cylinders. Still, the models may provide satisfactory estimates of the response of slender marine structures exposed to current. At least in terms of root mean square (r.m.s.) of displacements and strains, dominating frequency and dominating mode. However, VIV of long slender beams, especially in sheared current, show large response variability in time and space, which appears to be of a random nature. Semi-empirical models that predict steady-state VIV are hence unable to reflect the non-stationary response observed in experiments, even though the time and space averaged result might be well represented.In this work, a previously published semi-empirical time domain model for cross-flow vortex-induced vibrations is modified to describe the stochastic nature of the response of long slender beams subjected to stationary current. The mean non-dimensional frequency of the synchronization model (previously constant) is taken to be a simplified Gaussian process, where standard deviation and spectral frequencies are input. The stochastic synchronization model allows the response to jump between different eigenfrequencies and corresponding modes, and still predict mean and r.m.s. values close to the deterministic model. Its performance is verified through simulation of an experiment with a long riser in sheared flow. The response sensitivity of standard deviation and spectral frequencies is investigated. It indicates that for a proper choice of empirical coefficients, the chaotic response of the riser can be quite realistically simulated in terms of frequency variation and, to some extent, amplitude modulation.

Empirical sensitivity of two-dimensional nonlinear wake–cylinder oscillators in cross-flow/in-line vortex-induced vibrations

“…They showed that the low-order riser modes are very sensitive to the measurement noise. Recently, a stochastic modelling of cross-flow VIV in time domain has been presented by Ulveseter et al (2017) who introduced a Gaussian random process through a synchronization between vortex shedding and cylinder frequencies. They predicted a chaotic VIV response depending on the specified empirical coefficients and the assumed random process spectrum.…”

Derivation of a slow phase model of vortex-induced vibrations for smooth and turbulent oncoming flows

“…Monte Carlo simulations are the simplest way to tackle non deterministic problems since the solution just requires the generation of synthetic realizations of the turbulent flow and the statistical analysis of the response (Tagliaferri and Srinil, 2017). This approach has also been used to determine the response of structures with many more degrees-of-freedom (Li, 2019) or in more complicated geometries than the 2-D case of simple wake-oscillator models (Ulveseter et al, 2017). Although being simple to implement, the Monte Carlo approach does not offer the same depth of understanding as explicit closed form solutions which are only valid in some simple configurations, e.g.…”

A New Energy Recovery Device by Utilizing the Merchant Ship Rolling

“…J.V. Ulveseter et al [33] mentioned that the equation of motion was solved by the time integration scheme Newmark-β method. According to Ref.…”