Extracting energy from a flow: An asymptotic approach using vortex-induced vibrations and feedback control

Philippe Meliga, Jean Marc Chomaz, François Gallaire

Research output: Contribution to journalArticlepeer-review

Abstract

This paper considers vortex-induced vibrations of a cylinder in water streams for renewable energy production. We use an analytical model recently obtained by the authors from the asymptotic analysis of a coupled flow-cylinder system, and assess the ability of a control velocity applied at the cylinder wall to optimize the magnitude of dissipated energy at disposal to be harvested. The retained approach is that of proportional feedback control. When the system evolves on its limit cycle, we show that the control yields an increase in the mean dissipated energy by 3.5%, as well as a significant improvement of the robustness with respect to small inaccuracies of the structural parameters. However, we also show that the system is susceptible to converge to cycles of lower energy when subjected to external disturbances, as a result of the simultaneous existence of multiple stable cycles. Consequently, we propose a transient control algorithm meant to force the return of the system to its optimal cycle. Its efficiency is assessed for two feedback approaches relying on distinct types of measurements: we find significant differences in the time needed to reach convergence to the optimal cycle, which ultimately results in energy being spent when feedback is designed from cylinder measurements, and in energy being harnessed when feedback is designed from flow measurements.

Original languageEnglish
Pages (from-to)861-874
Number of pages14
JournalJournal of Fluids and Structures
Volume27
Issue number5-6
DOIs
Publication statusPublished - 1 Jul 2011

Keywords

  • Energy production
  • Flow control
  • Limit cycles
  • Transient dynamics
  • Vortex-induced vibrations

Fingerprint

Dive into the research topics of 'Extracting energy from a flow: An asymptotic approach using vortex-induced vibrations and feedback control'. Together they form a unique fingerprint.

Cite this