Airfoil Design at Low-Reynolds Numbers

Christopher Ong
Tagliatela College of Engineering – Mechanical Engineering
Dr. Maria-Isabel Carnasciali – Assistant Professor, Mechanical Engineering
Abstract
Airfoils were typically expected to be used for large craft like what is seen in commercial or private airplanes for
transportation of peoples and goods across the globe. In the last ten years, there has been an increased look into how airfoils
can be designed for lower ranges of speed. Currently small, low-speed aircraft are being used to study large areas due to
relative cost compared to other methods. The purpose of this research was to see if incorporating aspects of bioinspired
design with airfoils would improve the overall flight performance. Computational Fluid Dynamics (CFD) simulations were
used to verify existing airfoil design with experimental data currently available. Various geometries (bumps, dimples,
triangles, etc.) were added to the airfoil designs at the leading or trailing edge and simulated to get an understanding of how
fluids interacted with those shapes.
Introduction variables can be used to determine the velocities that were
to be used. With the use of airflow close to sea level and
An airfoil is a streamlined geometry that interacts with a an airfoil length of one meter, the velocity range is within
moving fluid which produces aerodynamic forces around 1 to 3 m/s, or 2.2 to 7 mph.
the body. The fluid of interest in this case is air. There are
two particular forces of interest for determining the Simulation Verification
effectiveness of an airfoil: lift and drag forces. The lift
force acts perpendicular to the body of the airfoil body To verify the accuracy of the simulation setup there was a
creating an upward motion. The drag forces act in parallel comparison of published test data and the simulation
against the airfoil body hindering its performance. results of the airfoil. For test data, the University of
Illinois has published hundreds of wind tunnel test results
on their website that can be used by anyone who is
interested in airfoil selection or design[4][5]. Due to
hundreds of different airfoils for various purposes, only a
handful of airfoils were selected. The selection criteria for
the airfoils were geometries that were designed for
smaller aircraft such as R/C planes. The airfoils that were
selected had their results compared, and it was determined
that E193 airfoil had the highest lift coefficient. This
higher lift coefficient can be attributed to the higher
curved shape that the airfoil has.
Figure 1 – Aerodynamic Forces
The simulation software that was used was FLUENT,
For drag there are two contributing forces knowns as which is part of the ANSYS® package. ANSYS® has
pressure and skin friction drag forces. The pressure drag several components that allow for geometry and meshing
forces are the forces that result from the pressure force generation. A 2D E193 airfoil was imported into
distribution around the airfoil body, while the skin friction ANSYS® Design Modeler. The research involved the
is the viscous effects caused by the fluid moving against study of fluid interacting with the airfoil. The E193 is a
the rest of the airfoil body [1]. These forces are non- solid body, so a control volume had to be created
dimensionalized by a dynamic pressure force determined surrounding the airfoil in order to simulate the airflow
by the size of the airfoil and the speed of the air flowing around the body. Shown in Figure 2, the control volume
by. The resulting lift coefficient (CL) and drag coefficient can be seen with a rectangular shape that is 25 meters tall
(Cd) are given by equations 1 & 2. and 12.5 meters back. The rest of the control volume has
a circular area that has a radius of 12.5 meters. A surface
2𝐹𝐿 2𝐹𝑑 was then created for the control volume with an airfoil
𝐶𝐿 = (1) 𝐶𝑑 = (2)
𝜌𝜌𝑣 2 𝜌𝜌𝑣 2 shape cut out in the center from that surface.
The Reynolds number is an important dimensionless
number that helps describe the fluid flow that is being
studied. Reynolds number is a ratio between the inertial
and viscous forces of a fluid, which is density of fluid,
velocity of the fluid, reference length over the dynamic
viscosity of the fluid [1]. The airfoil study is at Reynolds
number of 6E04, 1E05 and 2E05, which using several
 1.200
1.000
0.800
0.600

Cl
0.400
0.200
0.000
Figure 2 – 2D Control Volume Figure 3 – 0.000 0.050 0.100 0.150
Meshed 2D Volume
cd
The control volume was then meshed as seen in Figure 3,
using mesh controls that allowed for a mesh biased
focusing more on the surroundings of the airfoil. The cell
density can be seen in Figure 3, as the darker part of the
mesh. The logic behind this is that the fluid interacting Figure 4 – Simulation Data
with the airfoil is the important focus to capture the most
accurate lift and drag forces acting on the body. It is
important to capture the pressure field in the surroundings
of the airfoil for best results when looking at either the
leading or trailing edge of the airfoil.
The simulation setup used a density based solver, with
boundary conditions being the component velocities at the
inlet, pressure outlets for the surroundings and using the
Spalart-Allmaras viscous turbulence model. In FLUENT,
there are three turbulence models, Spalart-Allmaras, K-
Epsilon and an inviscid model. All the models produced
similar results for the lift and drag forces, but the Spalart-
Allmaras was chosen as the model due to the shorter
simulation times. This allowed for a wider range of
simulations to be ran within a shorter amount of time.
FLUENT has several monitors that allow the
determination of lift coefficient, drag coefficient and
residual monitors. For a simulation to be determined as
useful the residual monitors has to converge under a
certain criteria and the criteria that was selected was 1E- Figure 5 – Wind Tunnel Data from University of
04. The residual monitors report the continuity equation, Illinois
the select model values and the velocity components.
The data published by the University of Illinois did not
The simulations were ran with the velocity boundary come with any uncertainty values, so it cannot be
conditions of 0.88, 1.46 and 2.92 m/s, which correspond determined what the exact values are. The published data
with the Reynolds numbers of 6E04, 1E05 and 2E05, at the higher Reynolds number can be compared to the
respectfully. Each velocity were broken up into simulation data with similar results. For Reynolds number
component velocities within the angle range 0 – 12 of 6E04 and 1E05 the lower lift to drag coefficient values
degrees in increments of 2. This changing angle is do not correspond with the simulation results. This can be
referred to as the angle of attack, which is the angle of the due to the inaccuracy of the tools used in the wind tunnel
airfoil acting against the moving fluid. This was done to trials or noise caused by other factors that could hinder
compare the results of the simulation and the data that the accuracy of the test. There seems to be enough
was published by the University of Illinois. In Figure 4, correlation between the wind tunnel data and the
the simulation values can be seen at the corresponding simulation data to give an idea of the accuracy of the
Reynolds numbers and angles of attack. In Figure 5, the simulation setup. At higher values there is an uncertainty
wind tunnel data produced by the University of Illinois of closer to +/- %5.
can be compared to the simulation data.
Results and Discussion
Various geometries were simulated to get a basic
understanding on how the fluid interacts with the basic
shapes. There were dimpling and recessed surfaces
simulated that dug into a flat plate alongside raised
surfaces that showed the contrast. It was determined that
surfaces digging into a body would create eddies, that
would apply a larger pressure load lowering the
aerodynamic properties of the body [2]. The raised
surfaces were chosen to be incorporated into the airfoil to
see how these shapes affected the aerodynamic properties
of the airfoil.
In Figure 6, it can be seen that the airfoil at a 0-angle of
attack has a higher pressure field compared to that of the
rest of the body. This can be considered the case for other
angles in which the airfoil’s leading edge is facing the
moving fluid. Due to the higher pressure field at all
simulated angles of attack, this area was chosen to
incorporate other geometries to understand how the fluid Figure 7 – Modified Airfoils at 0 Angle of Attack
flow would be affected, ultimately affecting the airfoil’s The five simulation trial results are compared in Figure 8,
aerodynamic properties. which shows how the lift coefficient compared to the drag
coefficient. These simulations were ran at a 1E05
Reynolds number with a corresponding velocity of 1.46
m/s. The residual criteria for the monitors were 1E-03; far
from an ideal case of 1E-06. This was done due to the
lower computational times for the results.
When comparing the data in Figure 8 to Figure 4, it
shows that the trials have overall smaller lift coefficients
and higher drag coefficients. The drag coefficient drastic
change can be attributed to the increased surface at the
leading edge causing a higher pressure buildup. This leads
to a higher coefficient drag and can lead to a lower
Figure 6 – Pressure Gradient of Airfoil at 0 Angle of coefficient of lift. It appears a streamedlined body
Attack performs better than a geometry that has was modified
(Red – Highest Pressure / Blue - Lowest Pressure) with bumps, dimples or raised surfaces.

The two basic geometries that were incorporated were

right triangles and half circles at the leading edge. The
half circles were approximately 0.1 meters in diameter or 1.00
about one-tenth of the total airfoil length. The triangles
0.80
were about 0.1 meters in length with varying heights of
0.02, 0.025 and 0.03 meters in height over the surface. 0.60
Two modified airfoils can be seen in Figure 7; these
Cl

airfoils are at 0-angle of attack. 0.40

0.20

0.00
0 5 10 15
Angle of Attack (Degrees)
 0.25 Acknowledgement
I like to thank my faculty mentor, Dr. Carnasciali, for
0.20
assisting me on this project. She guided me and worked
0.15 with me through the numerous questions I had during the
course of the project. I look forward to continuing the
Cd

0.10 research.

0.05 I would also like to thank Summer Undergraduate

Research Fellowship (SURF) of the University of New
0.00 Haven for providing me this great opportunity In
0 5 10 15 particular, I like to thank Carol Withers, Janice Sanderson
Angle of Attack (Degrees) and Mr. and Mrs. Carrubba for their continued support of
the SURF program.
Biography

Figure 8 – Modified Airfoil Results

Conclusion and Further Consideration
The focus of the study was the fluid flow over a 2D airfoil
and considering changing aspects of the geometry to
influence the aerodynamic forces acting on. By adding
elements to the geometry, the lift and drag forces were
influenced by the changes. The values from the modified
geometries are lower than that of the original airfoil. This
can be attributed to the changing of the thickness of the
body at points. With the airfoil being 2D, changing the
chord length or the thickness of the airfoil will add the
Christopher Ong, from Patterson, New York, is currently
surface of which the aerodynamics forces are acting on.
a senior studying Mechanical Engineering at the
This will change the amount of force acting on the airfoil.
University of New Haven. His current interests are in
The airfoil is a simplified 2D case neglecting any 3D thermodynamic and fluid mechanic studies. He plans to
considerations of the fluid flow. This would be the next further his education by pursuing a masters in aerospace
step for further studies. By bringing the airfoil to a 3D engineering with a focus in computational physics. A
geometry, there has to be consideration on the total wing career working on rocket propulsion systems would be a
geometry, angle of the wing and tips at the ends that can dream come true!
affect the fluid flow in other ways. From here locating
strategic points on the airfoil to influence the fluid flow
would be the next step. By adding geomtries to these
strategic locations will assist with keeping a stable flow
around the airfoil.
References
[1] Anderson, J. D. (2001). Fundamentals of
aerodynamics. Boston: McGraw-Hill.
[2] Vento, J. (n.d.). Analysis of Surface Augmentation of
Airfoil Sections via Flow Visualization Techniques.
Retrieved August 12, 2015.
[3] Mullen, B. (2014, February 9). FLUENT - Flow Over
an Airfoil. Retrieved June 4, 2015.
[4]http://m-selig.ae.illinois.edu/ads/coord_database.html
[5] http://m-selig.ae.illinois.edu/uiuc_lsat.html
[6]http://www.globalsecurity.org/military/systems/aircraft
/intro-performance.htm