Control of flow around a low Reynolds number airfoil using longitudinal strips

We suggest longitudinal strips attached to an airfoil surface as a new stall control device. Their effects on the stall characteristics and flow modifications are experimentally investigated. The airfoil considered is SD7003 and the Reynolds numbers are $\mathrm{Re}=60000$ and 180 000 based on the chord length and freestream velocity. The drag and lift forces on the airfoil are measured by varying the angle of attack from $\alpha =0^{\circ }$ to 16° with and without strips. The optimal strip configuration is determined using a response surface method. Without strip, abrupt stalls occur at $\alpha ={11}^{\circ }$ and 12.5° for $\mathrm{Re}=60000$ and 180 000, respectively, whereas broad stalls occur without much changing the stall angles by optimal strips. The lift coefficient and lift-to-drag ratio are significantly increased by the strips at post-stall angles of attack. A corner vortex is generated at each corner of strip near the leading edge. Clockwise and counterclockwise streamwise vortices are generated at the left- and right-facing corners, respectively, and they slowly move away from the corners while travelling downstream. These vortices provide additional momentum to the airfoil suction surface, resulting in fully attached flow above the strips and reattachment of flow above grooved surface. The longitudinal strips presented here are different from other devices such as the vortex generator, trip wire, burst control plate, and zigzag tapes used for the separation control of low Reynolds number airfoil, in that the strips are installed nearly on the whole airfoil surface but with their heights lower than the boundary layer thickness, and produce positive control effects at post-stall angles of attack but little affect the aerodynamic performance at prestall angles of attack.

Figure 1

Schematic diagram of the experimental setup: (a) overall view; (b) airfoil and longitudinal strips; (c) two-dimensional planes for PIV measurements; (d) surface pressure measurement locations.

Figure 2

Variations of the lift and drag coefficients with the angle of attack (α): , present ($\mathrm{Re}=60000$); , Selig et al. [5] ($\mathrm{Re}=60000$); , Galbraith and Visbal [32] ($\mathrm{Re}=60000$); , Selig et al. [37] ($\mathrm{Re}=61400$).

Figure 3

Contours of the lift coefficient with the strip parameters ($h$, $w$, $s$) at a post-stall angle of attack ($\alpha ={12}^{\circ }$) using a response surface method. Thin solid lines denote the contour levels from $C_L=0.9$ to 1.1 by increments of 0.01. The force measurement locations are denoted as (black and white) solid circles in this figure. The response surface method provides an optimal strip configuration where no measurement is conducted, and thus we take the strip configuration closest to the one suggested as the optimal strips (denoted as white dot).

Figure 4

Variations of the force coefficients and drag polar with the angle of attack by strips: (a) $C_L$ and $C_D$; (b) drag polar. , No strip ($\mathrm{Re}=60000$); , optimal strips ($\mathrm{Re}=60000$); , no strip ($\mathrm{Re}=180000$); , optimal strips ($\mathrm{Re}=180000$). Here, the optimal strip configuration is $h/c=0.003,w/c=0.03$, and $s/c=0.15$.

Figure 5

Surface oil visualization on the suction surface ($\mathrm{Re}=60000$): (a) $\alpha =9^{\circ }$ (prestall angle of attack); (b) $\alpha ={11}^{\circ }$ (near-stall angle of attack); (c) $\alpha ={12}^{\circ }$ (post-stall angle of attack). Here, the upper and lower panels denote the cases without and with strips, respectively. Red-dotted and white-solid lines denote the reattachment and PIV measurement locations, respectively.

Figure 6

Contours of mean streamwise velocity and velocity vectors ($\alpha ={12}^{\circ };\mathrm{Re}=60000$): (a) no strip; (b) with strips (at $z/c=0$); (c) with strips (at $z/c=-0.045$). Here, the thick black lines in (a) and (c) denote $\overline{\mathit{u}}=0$. Black- and gray-colored areas indicate the regions with no velocity measurement due to the shadows of the airfoil and strips, respectively.

Figure 7

Surface pressure distributions ($\alpha ={12}^{\circ };\mathrm{Re}=60000$): , no strip; , with strips ($z/c=0$); , with strips ($z/c=-0.04$).

Figure 8

Cross-flow velocity vectors at $x/c=0.033$ ($\alpha ={12}^{\circ };\mathrm{Re}=60000$): (a) mean field; (b) instantaneous field. Here, the laser sheet for PIV is nearly perpendicular to the airfoil surface.

Figure 9

Instantaneous cross-flow velocity vectors ($\alpha ={12}^{\circ };\mathrm{Re}=60000$): (a) $x/c=0.05$; (b) 0.067; (c) 0.1. Black- and gray-colored areas in this figure denote the regions of no velocity measurement due to the shadows of the airfoil and strips, respectively. Yellow arrows denote the locations of corner vortex center from the strip corner.

Figure 10

Locations of corner vortex center in the streamwise direction and surface-oil flow pattern on the suction surface ($\alpha ={12}^{\circ };\mathrm{Re}=60000$). Here, vortex-center locations are obtained from the mean cross-flow fields at $x/c=0.033,0.05,0.067$, and 0.1. Black dots and red arrows indicate the locations and rotational directions of corner vortices, respectively.

Figure 11

Contours of the Reynolds shear stress with strips ($\alpha ={12}^{\circ };\mathrm{Re}=60000$): (a) $z/c=0$; (b) $-0.02$; (c) $-0.045$; (d) $-0.07$.

Figure 12

Schematic diagram of the mechanism responsible for the lift increase by the strips at a post-stall angle of attack.

Figure 13

Variations of the force coefficients ($C_L$ and $C_D$) with the strip length $\left(l\right)$ on the suction surface ($\mathrm{Re}=60000$). Here, other strip parameters are fixed to be $h/c=0.003,w/c=0.03$, and $s/c=0.15$, and the strip length on the pressure surface is fixed at $l/c=0.97$.