A unified framework for prediction of vortex-induced vibration based on the nonlinear data-driven identification of general wake oscillator modeling

TL;DR

This paper introduces a novel unified data-driven framework for predicting vortex-induced vibrations using semi-empirical wake oscillators, enhancing accuracy and physical realism through advanced identification strategies.

Contribution

It develops a semi-empirical wake oscillator model with new identification methods, improving VIV prediction accuracy and physical understanding compared to previous models.

Findings

High accuracy in VIV prediction demonstrated

Two identification strategies validated with CFD data

Overall identification better captures fluid damping effects

Abstract

In this paper, we present novel identification strategies to develop a unified framework for vortex-induced vibration (VIV) prediction based on the general semi-empirical wake oscillator. Greybox nonlinear system identification method accompanying high-fidelity computational fluid dynamics (CFD) and/or experimental data could be applied for the identification process. The proposed template of general wake oscillators contains low- to high-order damping terms to be identified for characterizing the possible flow dynamics. Two different strategies, including individual identification of single wake oscillator and overall identification of coupled VIV control equations, are proposed. VIV system consisting of an elastically-mounted circular cylinder submerged in laminar flow at Reynold number of 100 is considered. Both strategies have been tested and have exhibited high accuracy. The second strategy, i.e., overall identification of coupled VIV control equations, would be more suitable for the future framework owing that its training process considers the effect of fluid damping. A detailed mathematical introduction to future works on framework development covering the wide Reynold number range is addressed. The proposed unified framework is a landmark update of past wake oscillators both in terms of prediction accuracy and physical principles and has considerable research significance and practical engineering value.