Hostname: page-component-7b7567c8c-7pwfn Total loading time: 0 Render date: 2026-11-16T00:44:37.846Z Has data issue: false hasContentIssue false

Passive oscillations of inverted flags in a uniform flow

Published online by Cambridge University Press:  17 December 2019

Yao-Wei Hu
Affiliation:
Fluid Mechanics Key Laboratory of Education Ministry, Beijing University of Aeronautics and Astronautics, Beijing100191, China
Li-Hao Feng*
Affiliation:
Fluid Mechanics Key Laboratory of Education Ministry, Beijing University of Aeronautics and Astronautics, Beijing100191, China
Jin-Jun Wang
Affiliation:
Fluid Mechanics Key Laboratory of Education Ministry, Beijing University of Aeronautics and Astronautics, Beijing100191, China
*
†Email address for correspondence: lhfeng@buaa.edu.cn

Abstract

The passive oscillations of inverted flags are investigated both experimentally and theoretically in this paper. First, the force and energy distributions of inverted flags, which contain elastic and inertia components, are analysed based on the experimental data. Two main differences between inverted and conventional flags are found: (1) the elastic energy of a conventional flag is concentrated near the free end, while the fixed end of an inverted flag presents the largest elastic energy; and (2) the elastic component is several orders of magnitude greater than the inertia component for an inverted flag, while they are of the same magnitude for a conventional flag. Second, a linear analysis shows that the critical flow velocities obtained from the experiments at small mass ratios are scattered around the theoretical curve of wavenumber $k=1.875$, which is in contrast with $k=4.694$ of a conventional flag. For large mass ratios, the mass ratio has a certain influence on the critical velocity rather than being irrelevant. For two parallel inverted flags, both the experimental and theoretical results indicate that the range of the in-phase flapping mode becomes smaller with an increase in the separation distance, and a multiple flapping state may occur. For $n\geqslant 2$ parallel inverted flags, the theoretical results show that two of all coupled flapping modes are dominant with most parameters. These findings could contribute to a better understanding of the passive oscillations of inverted flags.

JFM classification

Information

Type
JFM Papers
Copyright
© 2019 Cambridge University Press
Figure 0

Figure 1.Schematic diagram of the experimental set-up.

Figure 1

Figure 2.Theoretical model of $n$ parallel inverted flags, where $i=1,2,\ldots n$.

Figure 2

Figure 3.Variation of the tip spanwise displacement at $U_{0}^{\ast }=2.88$, where the circles and the curve represent the experimental and fitting results, respectively.

Figure 3

Figure 4.A comparison of the fitting and experimental results of the flag profiles at $U_{0}^{\ast }=2.88$: (a) spanwise displacement at each point along the inverted flag; (b) dynamic profiles of the inverted flag. The left-hand and right-hand panels of each pair represent the fitting and experimental results, respectively.

Figure 4

Figure 5.Variation of the force distribution in a period: (a) the elastic force $f_{1}$ ($\text{N}~\text{m}^{-1}$), (b) the inertia force $f_{2}$ ($\text{N}~\text{m}^{-1}$) and (c) the distributed force $f_{y}=f_{1}+f_{2}$ ($\text{N}~\text{m}^{-1}$). (d) Variation of the tip displacement in a period.

Figure 5

Figure 6.Variation in the energy distribution in a period: (a) the strain energy $e_{p}$ ($\text{J}~\text{m}^{-1}$), (b) the kinetic energy $e_{k}$ ($\text{J}~\text{m}^{-1}$) and (c) the distributed energy $e=e_{p}+e_{k}$ ($\text{J}~\text{m}^{-1}$). (d) Variation of the displacement of the tip in a period.

Figure 6

Figure 7.Variation of (a) the total force $F_{y}$ (N) in the y direction and (b) the total energy E (J) in a period.

Figure 7

Figure 8.Variation in (a) the average energy $\overline{E}$ and amplitude $A/L$ and (b) conversion ratio R at different flow velocities.

Figure 8

Figure 9.The curves represent the critical flow velocities of an isolated inverted flag at different wavenumbers k, and the symbols represent the results of previous experiments.

Figure 9

Figure 10.Contour plot denoting the disturbance growth rate ratio $\unicode[STIX]{x1D6FE}$ in the ($M_{S},U_{0}^{\ast }$)-plane under different separation distances: (a) $d^{\ast }=0.0001$, (b) $d^{\ast }=1$, (c) $d^{\ast }=2$ and (d) $d^{\ast }=4$. Region I corresponds to the straight mode ($\unicode[STIX]{x1D6FE}$ is not a number); II corresponds to the anti-phase flapping mode ($\unicode[STIX]{x1D6FE}<1$); III corresponds to the in-phase flapping mode ($\unicode[STIX]{x1D6FE}>1$); and IV corresponds to the indefinite mode ($\unicode[STIX]{x1D6FE}=1$).

Figure 10

Figure 11.Contour plot denoting the disturbance growth rate ratio $\unicode[STIX]{x1D6FE}$ in the ($d^{\ast },U_{0}^{\ast }$)-plane at $M_{S}=0.42$. The definitions of regions I, II, III and IV are presented in figure 10. Symbols ▴ (in-phase) and ○ (anti-phase) represent the experimental results in a wind tunnel $(M_{S}=0.42)$ conducted by Huertas-Cerdeira et al. (2018).

Figure 11

Figure 12.(a) Observation of the coupled flapping modes of $n=3$ and $d^{\ast }=2$: $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{03}=1$, $\unicode[STIX]{x1D702}_{02}/\unicode[STIX]{x1D702}_{03}=-p_{31}$ (mode 1); $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{03}=1$, $\unicode[STIX]{x1D702}_{02}/\unicode[STIX]{x1D702}_{03}=p_{32}$ (mode 2); and $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{03}=-1$, $\unicode[STIX]{x1D702}_{02}=0$ (mode 3). (b) Observation of the coupled flapping modes of $n=4$ and $d^{\ast }=2$: $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{04}=-1$, $\unicode[STIX]{x1D702}_{02}/\unicode[STIX]{x1D702}_{03}=-1$, $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{02}=-p_{41}$ (mode 1); $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{04}=1$, $\unicode[STIX]{x1D702}_{02}/\unicode[STIX]{x1D702}_{03}=1$, $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{02}=p_{42}$ (mode 2); $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{04}=-1$, $\unicode[STIX]{x1D702}_{02}/\unicode[STIX]{x1D702}_{03}=-1$, $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{02}=p_{43}$ (mode 3); and $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{04}=1$, $\unicode[STIX]{x1D702}_{02}/\unicode[STIX]{x1D702}_{03}=1$, $\unicode[STIX]{x1D702}_{01}/\unicode[STIX]{x1D702}_{02}=-p_{44}$ (mode 4).