Hostname: page-component-7b7567c8c-tnkh4 Total loading time: 0 Render date: 2026-11-17T00:45:26.925Z Has data issue: false hasContentIssue false

Vortex formation on a pitching aerofoil at high surging amplitudes

Published online by Cambridge University Press:  27 October 2020

Luke R. Smith*
Affiliation:
Department of Aerospace Engineering, University of Maryland, College Park, MD 20742, USA
Anya R. Jones
Affiliation:
Department of Aerospace Engineering, University of Maryland, College Park, MD 20742, USA
*
†Email address for correspondence: lsmith1@umd.edu

Abstract

In many applications, conventional aerofoils are subject to a number of simultaneous motions that complicate the prediction of flow separation. The purpose of this work is to evaluate the impact of a large-amplitude free-stream oscillation on the timing of vortex formation for a simultaneously surging and pitching wing. Experimental flow field measurements were obtained on a NACA 0012 aerofoil over a wide range of surge amplitudes ($1.50 \leq \lambda \leq 2.25$) and reduced frequencies ($0.1 \leq k \leq 0.3$). Particular attention was paid to how various mechanisms of flow separation, specifically the velocity induced by the trailing wake and unsteady effects in the boundary layer, were impacted by a change in the properties of the surge motion. In the regime where $k \leq 0.3$, a change in the surge kinematics primarily manifested as a change in the relative strength of the trailing wake. Boundary layer unsteadiness was found to have a negligible influence on the timing of vortex formation in the same flow regime. Thus, the timing of leading-edge vortex formation was well predicted by a combination of an unsteady inviscid flow solver and a quasi-steady treatment of the boundary layer, a promising result for low-order predictions of vortex behaviour in unsteady aerofoil flows.

Information

Type
JFM Papers
Copyright
© The Author(s), 2020. Published by Cambridge University Press
Figure 0

Figure 1.Problem statement: an aerofoil with a rounded leading edge is subject to an unsteady free-stream velocity ($U(t$)) and incidence ($\alpha (t$)). The onset of flow separation is seen as a function of (1) the instantaneous wing kinematics, (2) the velocity induced by the trailing wake and (3) unsteady effects in the boundary layer.

Figure 1

Figure 2.Photograph (a) and sketch (b) of the experimental set-up and water-filled tow tank facility.

Figure 2

Figure 3.Sample particle image (a) and phase-averaged vorticity field (b) for the baseline surging and pitching condition of the aerofoil.

Figure 3

Figure 4.Surging kinematics for each of the four surge amplitude cases.

Figure 4

Figure 5.Surging kinematics for each of the four reduced frequency cases.

Figure 5

Figure 6.Pitching kinematics for the current work.

Figure 6

Figure 7.Basic stages of LEV formation for the baseline kinematics ($\lambda = 1.75$, $k = 0.165$, $\alpha _0 = 15^{\circ }$, $\alpha _1 = 8^{\circ }$). The aerofoil exhibits stages of (1) attached flow, (2) separation near the leading edge, (3) vortex rollup and (4) vortex shedding.

Figure 7

Figure 8.The flux of clockwise (blue) vorticity measured at the leading edge of the wing throughout the baseline case. Note that this measurement is shown with an increasing number of runs included in the phase-averaging process.

Figure 8

Figure 9.Illustration of the calculation of final vortex strength $\varGamma _f$. A flux plane is drawn beyond the rounded edge of the wing, and the LEV strength is determined by integrating the flux of clockwise (blue) vorticity through this plane.

Figure 9

Figure 10.The effect of reduced frequency ($k$) on the timing of vortex formation ($\psi _{vor}$) for a constant surge amplitude ($\lambda = 1.75$) and set of pitching kinematics ($\alpha _0 = 15^{\circ }$, $\alpha _1 = 8^{\circ }$, $\phi = {\rm \pi}$).

Figure 10

Figure 11.Illustration of the control volume used to estimate the strength of the near wake.

Figure 11

Figure 12.The effect of reduced frequency ($k$) on the strength of the near wake ($c/4$ beyond the trailing edge) at constant surge amplitude ($\lambda = 1.75$).

Figure 12

Figure 13.The expected magnitude of time-acceleration effects in the boundary layer relative to the contribution of viscosity and the pressure gradient (both of which are assumed to be $O(1)$ in this figure) at constant surge amplitude ($\lambda = 1.75$).

Figure 13

Figure 14.The effect of surge amplitude ($\lambda$) on the timing of vortex formation ($\psi _{vor}$) at constant reduced frequency ($k = 0.165$) and pitching kinematics ($\alpha _0 = 15^{\circ }$, $\alpha _1 = 8^{\circ }$, $\phi = {\rm \pi}$).

Figure 14

Figure 15.The distribution of vorticity along a line normal to the $c/10$ chordwise position on the aerofoil suction surface. The white dot represents the ‘edge’ of the boundary layer. Note that this profile is taken from the baseline case ($\lambda = 1.75$, $k = 0.165$) at $\psi = 182^{\circ }$.

Figure 15

Figure 16.The effect of surge amplitude ($\lambda$) on the instantaneous height of the boundary layer at a position roughly $c/10$ from the leading edge of the wing.

Figure 16

Figure 17.The expected magnitude of time-acceleration effects in the boundary layer relative to the contribution of viscosity and the pressure gradient (both of which are assumed to be $O(1)$ in this figure) at constant reduced frequency ($k = 0.165$).

Figure 17

Figure 18.The effect of surge amplitude ($\lambda$) on the strength of the near wake ($\varGamma _w$) for four surge amplitudes at constant reduced frequency and pitching kinematics.

Figure 18

Figure 19.Illustration of how the wake-induced velocity, which is reasonably constant across the four surge amplitudes, increases the effective incidence near the leading edge.

Figure 19

Figure 20.The value of the maximum non-dimensional pressure gradient ($\beta _{max}$) computed in a region near the leading edge ($0 \leq x/c \leq 0.25$) for a NACA 0012 over various surge amplitudes.

Figure 20

Figure 21.The predicted timing of vortex formation ($\psi _{vor}$) compared to experimental measurement for a sweep of surge amplitudes ($\lambda$, a) and reduced frequency ($k$, b).