DYNAMICS AND CONTROL OF HIGH-REYNOLDS-NUMBER FLOW OVER OPEN CAVITIES

TLDR

Cavity flow oscillations, known for four decades, remain difficult to suppress reliably; they are of interest for aerospace and as a canonical control problem, and recent views suggest they may be lightly damped resonances driven by external disturbances, with gaps in understanding around high‑frequency forcing, feedback limits, and model‑validity constraints. We review recent advances in understanding, modeling, and controlling oscillations in the flow past a cavity. The review covers recent advances in modeling cavity flows and controlling them with both open‑loop and closed‑loop techniques. Cavity oscillations may be lightly damped resonances sustained by external disturbances rather than self‑sustained, challenging the traditional view.

Abstract

We review recent advances in understanding, modeling, and controlling oscillations in the flow past a cavity. The fundamental mechanisms underlying cavity flow oscillations have been known for at least 40 years, but suppressing these oscillations in a reliable and robust way is still a challenge today. Interest in controlling the flow past a cavity is motivated by aerospace applications, but in addition, cavity flows provide an attractive canonical problem for exploring general flow control techniques. The focus is on recent advances in modeling these flows, and in controlling them, using both open-loop and closed-loop techniques. A relatively new perspective is that cavity oscillations may not always be self-sustained, but under some flow conditions may be lightly damped resonances, sustained by external disturbances such as boundary layer turbulence. Areas in which our understanding is incomplete, and which deserve further study, are discussed, in particular the effects of high-frequency open-loop forcing, fundamental limitations of feedback control for a given configuration of sensors and actuators, and the development of a feedback design methodology that respects the limited range of validity of the available dynamical models.