JOURNAL OF AIRCRAFT

Vol. 48, No. 4, July–August 2011

Low-Reynolds-Number Effect on Aerodynamic

Characteristics of a NACA 0012 Airfoil

Dong-Ha Kim∗
Korea Aerospace Research Institute, Daejeon 305-333, Republic of Korea
Jo-Won Chang†
Korea Aerospace University, Goyang 412-791, Republic of Korea
and
Joon Chung‡
Ryerson University, Toronto, Ontario M5B 2K3, Canada
DOI: 10.2514/1.C031223
The boundary-layer properties and aerodynamic characteristics for the NACA 0012 airfoil were investigated at
low Reynolds numbers (Rec
2:3  103 , 3:3  103 , and 4:8  103 ) and low angles of attack (

0 to 6 ). Boundary-
layer visualization and static pressure measurements were performed to show the abrupt increase of lift coefficients
between

2 and 3 in low-Reynolds-number cases. This nonlinear lift variation is due to the abrupt formation of
the attached boundary layer. Moreover, the angle-of-attack range where the pressure drag coefficient is decreased
was observed due to the pressure increase starting from the trailing edge. This demonstrated the different
aerodynamic characteristics for below Rec
5:0  103 .

Nomenclature
C = airfoil chord length
Cd = drag coefficient
Cl = lift coefficient
Cp = pressure coefficient
Rec = Reynolds number based on airfoil chord length
x = freestream direction from airfoil leading edge

= angle of attack

I. Introduction
HE flow properties in the low-Reynolds-number range, which
T are applicable to small-sized aerovehicles, wind turbines,
propellers, and human-powered aerovehicles, have generally
shown significant aerodynamic performance issues due to various
boundary-layer events such as laminar separation [1–3]. In the
subcritical region where laminar separation is dominant, the flow
properties are ultimately influenced by Reynolds number, and
unusual aerodynamic characteristics are observed [4]. Jacobs and
Sherman [4] conducted wind-tunnel tests with various airfoils and
suggested the Reynolds number effect Rec from 4:0
104 to
3:3
106 . Their results for Reynolds numbers over 1:0
105
showed the linear variation of lift coefficient, whereas lift coefficient
showed the nonlinear pattern below Rec  5:0
104 in the low
angle-of-attack range. Such a nonlinear lift coefficient variation is
clearly different from classical results, although its nonlinearity is
small. An interest of flow properties in low Reynolds numbers has
been raised due to the emergence of small-sized biomimetic flights

Presented as Paper 2009-1472 at the 47th AIAA Aerospace Sciences

Meeting Including The New Horizons Forum and Aerospace Exposition,
Orlando, FL, 5–8 January 2009; received 3 September 2010; revision
received 1 March 2011; accepted for publication 1 March 2011. Copyright ©
2011 by the authors. Published by the American Institute of Aeronautics and
Astronautics, Inc., with permission. Copies of this paper may be made for
personal or internal use, on condition that the copier pay the $10.00 per-copy
fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,
MA 01923; include the code 0021-8669/11 and $10.00 in correspondence
with the CCC.
∗
Senior Researcher, Advanced Aerodynamic and Structure Department.
†
Professor, Department of Aeronautical Science and Flight Operation;
[email protected]. Senior Member AIAA.
‡
Associate Professor, Department of Aerospace Engineering. Member
AIAA. Fig. 1 Lift coefficients and drag polar in a NACA 0012 airfoil.
1212
 KIM, CHANG, AND CHUNG 1213

[5]. However, there are insufficient reliable data below Rec  Jacobs and Sherman [4]). Lift coefficients between
 1 and 2 are
5:0
104 . Accordingly, more refined results and physical under- dramatically increased in all Reynolds number cases, and the
standing about the nonlinear variation of the lift coefficient are increase rate between
 1 and 2 significantly follows the increase
necessary to apply various flowfields. of Reynolds number. However, from
 2 , the increase rate of the
The objectives of this study are to gain physical knowledge about lift coefficients starts to decline with angle of attack, and the increase
the aerodynamic characteristics in the range of Rec  2:0
104 rate between
 2 and 3 is decreased with the Reynolds number
5:0
104 . In the subcritical region, because the boundary layer is a increase. Thus, from
 1 to 6 , the lift coefficients are nonlinearly
transitional regime, its properties are ambiguous and dramatically varied with the angle of attack, and their variations fitted by cubic
varied, depending on Reynolds number [6]. Therefore, more refined spline are shown in Fig. 1a. Such a nonlinear lift coefficient variation
and systematic data are necessary for the understanding of was also observed in the study of Jacobs and Sherman [4] (Fig. 1a,
phenomena, which is different from the classical pattern. Tests were NACA 0015, Rec  4:29
104 ), even if its nonlinearity was small.
performed for Rec  2:3
104, 3:3
104 , and 4:8
104 from
 0 Thus, the nonlinear variation of the lift coefficient is attributed to the
to 6 with an interval of
 1 . The test conditions with narrow low-Reynolds-number effect.
Reynolds number ranges were selected, because the boundary-layer Figure 1b indicates the drag polar (pressure drag) variation with
properties are very sensitive with respect to the Reynolds number. respect to the Reynolds number. In this figure, angle-of-attack ranges
Boundary-layer visualization and static pressure measurements were where the drag coefficients are decreased are observed for a given lift
performed using a NACA 0012 airfoil having a 0.18 m chord length. coefficient. These ranges are observed from
 2 to 6 (Rec  4:8

The experimental apparatus is well described by Kim et al. [7]. The 104 ), from
 3 to 6 (Rec  3:3
104 ), and from
 3 to 6
pressure on the airfoil surface was measured by 24 pressure holes and (Rec  2:3
104 ). Thus, the drag coefficients start to be decreased at
a pressure transducer (model 239, Setra). The pressure range of the lower angles of attack in higher Reynolds numbers. At a constant lift
pressure transducer was 0:25 inH2 O (about 60 Pa) and could coefficient, the drag coefficients are significantly decreased between
measure very low pressure. Its sampling rate was 200 Hz, and the Rec  3:3
104 and 4:8
104 . The decrease in drag coefficients is
uncertainty pressure measurement was less than 3.5%. caused by the flow reattachment on the airfoil upper surface.
Figure 2 shows the boundary-layer visualization using the smoke-
wire technique. For this process, a wire with a 0.1 mm diameter was
II. Results and Discussion installed at 0:06C behind the leading edge and at a distance of 1.0 mm
Lift coefficients, calculated from the airfoil surface pressure, above the airfoil surface. Under this experimental condition, besides
exhibit a linear increase up to
 1 , as shown in Fig. 1a and, in this buoyancy due to wire heating, the boundary-layer flow may expe-
region, the slopes of lift coefficients are between  and 2. It is rience a flow disturbance caused by the wire because the wire was
constantly increased up to the case of Rec  1:7
105 (study of installed inside the boundary layer. However, the Reynolds number

Fig. 2 Boundary-layer visualization in a NACA 0012 airfoil.

1214 KIM, CHANG, AND CHUNG

based on the wire diameter was favorably less than 30 in all cases [8]. lift coefficient is attributed to the Reynolds number increase in the
This indicates that the flow disturbance caused by the wire could be low angle-of-attack range.
neglected in this study, and its detail items were discussed by Kim At Rec  4:8
104 , as the angle of attack increases from
 1 to
and Chang [9]. 6 , the laminar separation point rapidly and nonlinearly moved from
In this study, the boundary-layer visualization elucidates the the trailing edge to the middle region of the airfoil. Simultaneously,
reason for the appropriate increase in lift coefficients. At Rec  the transitioned area seemed to be increased rapidly on the airfoil
2:3
104 , the boundary layers were similar between
 1 and 2 surface. For the low angle of attack in this study, the separation point
with respect to the formation of the attached boundary layer, whereas moved rapidly from the trailing edge toward the middle region of the
they showed a significant difference between
 2 and 3 . That is, airfoil. However, for the high angle of attack, the separation point
the boundary layer was well attached when the angle of attack hardly moved around the leading edge. This was approximately
became
 3 . This illustrates the rapid increase of lift coefficient in confirmed by the boundary-layer visualization on the upper surface
Fig. 1a, and it is applicable to the case of Rec  4:8
104 . However, of the airfoil. As a result, the transition degree and the transitioned
the boundary layer started to be attached at lower angles of attack as area in the boundary layers are nonlinearly increased with the angle
the Reynolds number increased from Rec  2:3
104 to 4:8
104 . of attack. Therefore, the transition from the laminar to the turbulent
Accordingly, for Rec  4:8
104, the lift coefficient was rapidly flow is an important flow phenomenon for the nonlinear variation of
increased between
 1 and 2 . As a result, the formation of the lift coefficients in the low-Reynolds-number range.
attached boundary layer is the main reason for the abrupt lift increase Figure 3 shows the pressure coefficients on the airfoil surface at
in the low-Reynolds-number range; moreover, the rapid increase of angles of attack of
 1  6 . In Figs. 3b–3f, the solid lines indicate

Fig. 3 Pressure coefficients in a NACA 0012 airfoil.

 KIM, CHANG, AND CHUNG 1215

the pressure coefficients calculated under inviscid flow conditions boundary-layer visualization. Such an abrupt lift increment induces
using XFOIL. From about
 1 , the pressure coefficients on the nonlinear variation with respect to the angle of attack, and the trend is
upper surface show a constant variation from the middle position to significantly influenced by the Reynolds number, despite its small
the trailing edge, indicating laminar separation on the airfoil surface. variation. The angle-of-attack range where the drag coefficient is
The laminar separation is a principal factor in the increasing drag decreased was observed due to the pressure increase starting from the
coefficient. However, starting from Rec  4:8
104 at
 2 trailing edge. As a result, the aerodynamic characteristics in the
(Fig. 3b), the increase in pressure coefficients is launched from the Reynolds number range below Rec  5:0
104 are evidently
trailing edge. At this test condition, the drag coefficient starts to be different from the classical pattern and are significantly varied with
decreased (Fig. 1b). This physical trend is enhanced with an the Reynolds number.
increasing angle of attack, resulting in a decreased drag coefficient in
all cases. On the other hand, for Rec  2:3
104 and 3:3
104 , the
pressure coefficients observed over the trailing edge begin to be Acknowledgment
increased at
 3 . This indicates that, as the Reynolds number This work was supported by the Korea Research Foundation Grant
increases, the drag coefficients start to be decreased at a lower angle funded by the Korea Government (Basic Research Promotion Fund,
of attack. Actually, such a drag coefficient reduction is also observed KRF-2007-313-D00124).
in other types of airfoils in the low-Reynolds-number range [10,11].
Consequently, the pressure increase starting from the trailing edge
plays an important role in drag decrement for the cases with low
freestream velocities. Such a phenomena observed in the low-
References
Reynolds-number range had been used to enhance the aerodynamic [1] Lissaman, P. B. S., “Low-Reynolds-Number Airfoils,” Annual Review
of Fluid Mechanics, Vol. 15, No. 1, 1983, pp. 223–239.
performance in low-speed airfoils [12], achieving lift enhancement
doi:10.1146/annurev.fl.15.010183.001255
and drag reduction. [2] Carmichael, B. H., “Low Reynolds Number Airfoil Survey,” NASA
The position at which the pressure coefficient after laminar CR 165803, 1981.
separation increases on the upper surface of the airfoil is well known [3] Lin, J. C. M., and Pauley, L. L., “Low-Reynolds-Numbers Separation
as a sign of reattachment. The reattachment point can be identified at on an Airfoil,” AIAA Journal, Vol. 34, No. 8, 1996, pp. 1570–1577.
a cross point between the pressure coefficient calculated under an doi:10.2514/3.13273
inviscid flow condition and the pressure coefficient measured on the [4] Jacobs, E. N., and Sherman, A., “Airfoil Section Characteristics as
airfoil surface for a sufficiently large Reynolds number [13]. How- Affected by Variations of the Reynolds Number,” NACATR 586, 1937.
ever, this does not correspond with the result in this paper due to the [5] Muller, T. J., and DeLaurier, J. D., “Aerodynamics of Small Vehicles,”
low-Reynolds-number effect. The “A” in Figs. 3c–3f shows the Annual Review of Fluid Mechanics, Vol. 35, No. 1, 2003, pp. 89–111.
doi:10.1146/annurev.fluid.35.101101.161102
positions with maximum pressure fluctuations Prms along the airfoil [6] Huang, R. F., and Lin, C. L., “Vortex Shedding and Shear-Layer
surface coordinate. A pressure coefficient is always increased, ac- Instability of Wing at Low Reynolds Numbers,” AIAA Journal, Vol. 33,
companied with a maximum pressure fluctuation. For Rec  4:8
No. 8, 1995, pp. 1398–1403.
104 (Fig. 3f), the cross point between the calculated and measured doi:10.2514/3.12561
pressure coefficients corresponds well with the A point around [7] Kim, D. H., Yang, J. H., Chang, J. W., and Chung, J., “Boundary Layer
x=C  0:5. The boundary-layer measurement suggested in the study and Near-Wake Measurements of NACA 0012 Airfoil at Low Reynolds
of Kim et al. [7], who investigated the boundary layer under the same Numbers,” AIAA Paper 2009-1472, 2009.
condition, does not indicate reattachment at x=C  0:5 but rather [8] Batill, S. M., and Muller, T. J., “Visualization of Transition in the Flow
over an Airfoil Using the Smoke-Wire Technique,” AIAA Journal,
the transition point in the boundary layer. Therefore, in this case, the
Vol. 19, No. 3, 1981, pp. 340–345.
boundary-layer development is not enough to characterize the doi:10.2514/3.50953
pressure coefficient development due to low freestream velocity. For [9] Kim, D. H., and Chang, J. W., “Unsteady Boundary Layer for a Pitching
Rec  2:3
104 and 3:3
104 , and
 6 (Fig. 3f), position A Airfoil at Low Reynolds Numbers,” Journal of Mechanical Science and
precedes the cross point, and this position appears to correspond with Technology, Vol. 24, No. 1, 2010, pp. 429–440.
the transition point over the boundary layers shown by Kim et al. [7]. doi:10.1007/s12206-009-1105-x
As a result, the cross point follows the boundary-layer transition in [10] Selig, M. S., Lyon, C. A., Giguere, P., Ninham, C. P., and Guglielmo,
the low-Reynolds-number range. J. J., Summary of Low-Speed Airfoil Data, Vol. 2, SoarTech Publ.,
Virginia Beach, VA, 1996, pp. 95–99.
[11] Selig, M. S., Lyon, C. A., Giguere, P., Ninham, C. P., and Guglielmo,
III. Conclusions J. J., Summary of Low-Speed Airfoil Data, Vol. 3, SoarTech Publ.,
The boundary-layer properties and aerodynamic characteristics Virginia Beach, VA, 1998, p. 105.
for the NACA 0012 airfoil were investigated at low Reynolds [12] Somers, D. M., “Effects of Airfoil Thickness and Maximum Lift
Coefficient on Roughness Sensitivity,” Rept. NREL/SR-500-36336,
numbers (Rec  2:3
104 , 3:3
104 , and 4:8
104 ) and low National Renewable Energy Lab., Golden, CO, 2005.
angles of attack (
 0  6 ). The lift coefficients are abruptly [13] Yarusevych, S., Sullivan, P. E., and Kawall, J. G., “Coherent Structures
increased between
 2 (1 ) and
 3 (2 ) for Rec  2:3
104 in an Airfoil Boundary Layer and Wake at Low Reynolds Numbers,”
(4:8
104 ). It is due to the abrupt formation of the attached boundary Physics of Fluids, Vol. 18, No. 4, 2006, Paper 044101.
layer with respect to the angle of attack, and it was demonstrated by doi:10.1063/1.2187069