Hostname: page-component-7b7567c8c-7pwfn Total loading time: 0 Render date: 2026-11-16T05:43:43.855Z Has data issue: false hasContentIssue false

Aerodynamics of a wing in turbulent bluff body wakes

Published online by Cambridge University Press:  03 March 2022

Z. Zhang
Affiliation:
Department of Mechanical Engineering, University of Bath, Bath BA2 7AY, UK
Z. Wang
Affiliation:
Department of Mechanical Engineering, University of Bath, Bath BA2 7AY, UK
I. Gursul*
Affiliation:
Department of Mechanical Engineering, University of Bath, Bath BA2 7AY, UK
*
† Email address for correspondence: ensiag@bath.ac.uk

Abstract

The aerodynamics of a stationary wing in a turbulent wake are investigated. Force and velocity measurements are used to describe the unsteady flow. Various wakes are studied with different dominant frequencies and length scales. In contrast to the pre-stall angles of attack, the time-averaged lift increases substantially at post-stall angles of attack as the wing interacts with the von Kármán vortex street and experiences temporal variations of the effective angle of attack. At an optimal offset distance from the wake centreline, the time-averaged lift becomes maximum despite of small amplitude oscillations in the effective angle of attack. The stall angle of attack can reach 20° and the maximum lift coefficient can reach 64 % higher than that in the freestream. Whereas large velocity fluctuations at the wake centreline cause excursions into the fully attached and separated flows during the cycle, small-amplitude oscillations at the optimal location result in periodic shedding of leading edge vortices. These vortices may produce large separation bubbles with reattachment near the trailing-edge. Vorticity roll-up, strength and size of the vortices increase with increasing wavelength and period of the von Kármán vortex street, which also coincides with an increase in the spanwise length scale of the incident wake, and all contribute to the remarkable increase in lift.

Information

Type
JFM Papers
Copyright
© The Author(s), 2022. Published by Cambridge University Press
Figure 0

Figure 1.(a) Schematic of the wing in the wakes of bluff bodies; (b) schematic of the experimental setup and the PIV measurements.

Figure 1

Figure 2.Time-averaged lift coefficient in freestream and comparison with the literature.

Figure 2

Figure 3.Time-averaged lift coefficient as a function of offset distance at ${x_{LE}}/c = 2.0$.

Figure 3

Figure 4.Time-averaged streamwise velocity (left), root mean square of cross-stream velocity (middle), and cross-correlation of cross-stream velocity (right) for (a) H/c = 0.1, (b) H/c = 0.2, (c) H/c = 0.3, (d) H/c = 0.4 and (e) H/c = 0.5. Black dots show the optimal locations.

Figure 4

Figure 5.(a) Normalised wavelength and Strouhal number based on chord length; (b) normalised circulation of the wakes as a function of the ratio of the thickness of the wake generator to chord length.

Figure 5

Figure 6.(a) Normalised $\langle u\rangle $ and $\langle v\rangle $ velocity components at wake centreline (red symbols and lines) and optimal location (blue symbols and lines), for $x/c = 2.0$ and H/c = 0.3; (b) root mean square of wake angle as a function of wake generator thickness.

Figure 6

Figure 7.Time-averaged lift coefficient as a function of angle of attack for ${x_{LE}}/c = 2.0$ and (a) H/c = 0.1, (b) H/c = 0.2, (c) H/c = 0.3, (d) H/c = 0.4 and (e) H/c = 0.5.

Figure 7

Figure 8.Instantaneous vorticity contours for α=20°, ${x_{LE}}/c = 2.0$, H/c = 0.5: (a) ${y_{LE}}/c = 0$ and (b) ${y_{LE}}/c ={-} 0.6$.

Figure 8

Figure 9.Phase-averaged vorticity superimposed with streamlines for α=20°, ${x_{LE}}/c = 2.0,{y_{LE}}/c = \; 0$, H/c = 0.5.

Figure 9

Figure 10.Phase-averaged vorticity superimposed with streamlines for α=20°, ${x_{LE}}/c = 2.0,{y_{LE}}/c = \; - 0.6$, H/c = 0.5.

Figure 10

Figure 11.Illustration of the two control volumes for circulation computation, α=20°, H/c = 0.2, at (a) ${y_{LE}}/c = 0$ and (b) at ${y_{LE}}/c = \textrm{optimal}\;\textrm{location}$.

Figure 11

Figure 12.Time history of normalised LEV circulation for α=20°, ${x_{LE}}/c = 2.0$, (a) H/c = 0.1, (b) H/c = 0.2, (c) H/c = 0.3, (d) H/c = 0.4 and (e) H/c = 0.5.

Figure 12

Figure 13.Local flow angle sampled at 0.05c above and downstream of leading edge for α = 20°, ${x_{LE}}/c = 2.0$, at the wake centreline and optimal location, (a) H/c = 0.1, (b) H/c = 0.2, (c) H/c = 0.3, (d) H/c = 0.4 and (e) H/c = 0.5.

Figure 13

Figure 14.(a) Cross-correlation contours for all wake generators; (b) cross-correlation of cross-stream velocity as a function of spanwise distance; (c) correlation integral scale (L/c) and spanwise length scale (Ls/c) of the wake at x/c = 2.0 as a function of wake generator thickness.

Figure 14

Figure 15.Time history of normalised LEV circulation (left column) and the phase-averaged vorticity contour at the instance with maximum LEV circulation (right column) for α=20°, H/c = 0.5, (a) ${x_{LE}}/c = 2.0$, (b) ${x_{LE}}/c = 3.0$ and (c) ${x_{LE}}/c = 4.0$.

Figure 15

Figure 16.Cross-correlation of cross-stream velocity when the wing was placed at wake centreline and optimal location, α=20°, ${x_{LE}}/c = 3.0$, at (a) leading edge and (b) trailing edge.