Proceedings of the ASME-JSME-KSME 2015 Joint Fluids Engineering Conference

AJKFluids2015
July 26-31, 2015, Seoul, Korea

AJKFluids2015-28124

OPTIMIZATION OF LOOPED AIRFOIL WIND TURBINE (LAWTTM) DESIGN

PARAMETERS FOR MAXIMUM POWER GENERATION

Subhodeep Banerjee Binhe Song

Washington University in St. Louis Washington University in St. Louis
St. Louis, MO, USA St. Louis, MO, USA

George Syrovy Ramesh K. Agarwal

EverLift Wind Technology, Inc. Washington University in St. Louis
Nassau, DE, USA St. Louis, MO, USA

ABSTRACT CF Coefficient of force in direction of track

The looped airfoil wind turbine (LAWT™) is a patented
new technology by EverLift Wind Tecnology, Inc. for
generating power from wind. It takes advantage of the superior
lift force of a linearly traveling wing compared to the rotating
blades in conventional wind turbine configurations. Compared
to horizontal and vertical axis wind turbines, the LAWT™ can
be manufactured with minimal cost because it does not require
complex gear systems and its blades have a constant profile
along their length [1]. These considerations make the LAWT™
economically attractive for small-scale and decentralized power
generation in rural areas. Each LAWT™ is estimated to
generate power in the range of 10 kW to 1 MW. Due to various
advantages, it is meaningful to determine the maximum power
generation of a LAWT™ by optimizing the structural layout.
In this study, CFD simulations were conducted using
ANSYS Fluent to determine the total lift and drag coefficient
for a cascade of airfoils. The k-kl-ω turbulence model was used
to account for flow in the laminar-turbulent transition region.
Given the lift and drag coefficients and the kinematics of the
system, an analytical formula for the power generation of the
LAWT™ was developed. General formulas were obtained for
the average lift and drag coefficients so that the total power
could be predicted for any number of airfoils in LAWT™.
The spacing between airfoils was identified as the key design
parameter that affected the power generation of the LAWT™.
The results show that a marked increase in total power can be
achieved if the optimum spacing between the airfoils is used.
NOMENCLATURE
c Chord Length (meters)
CD Coefficient of drag
CL Coefficient of lift
FT Force along track per airfoil (Newton)
n Number of airfoils along track
P Total power generation (Watts)
Re Reynolds number based on chord length
S Airfoil span (m)
U∞ Freestream velocity (m/s)
UT Track speed (m/s)
UW Wind speed (m/s)
α Angle of attack
β Forward track angle
θ Airfoil pitch angle from horizontal
ρ Density of air (kg/m³)
φ Freestream velocity angle from horizontal
ω Reverse track angle
INTRODUCTION
The looped airfoil wind turbine (LAWT™) is a novel
concept for a wind turbine with a looped triangular structure.
The LAWT™ system proposes to take advantage of the
superior aerodynamic performance of an airplane wing
compared to conventional wind turbines [1]. The LAWT™
employs simple airfoil blades wheeled inside a track and
connected by chains. The airfoils are installed evenly above one
another on the triangular structure with the leading edges
parallel, traveling along a triangular path. The lift and drag
forces on the airfoils in the ascending and descending sections
of the tracks drive the system as a whole. The chains can
transfer kinetic power to the generators. A simple sketch of
LAWT™ system is shown in Fig. 1.
The rotor blades in conventional wind turbines are very
complex and expensive to design and manufacture. In addition,
the conventional wind turbine is restricted to one location and

1 Copyright © 2015 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 cannot be moved. On the other hand, a LAWT™ utilizes
shorter wing segments instead of the traditional three 50 to
100m long rotating blades (e.g. in case of a HAWT). Each wing
segment has a constant profile along its length, which can
significantly reduce the manufacturing cost. Furthermore, a
LAWT™ can be designed to be flexible and can either be
adjusted to face the wind all the time or stay in a fixed position.
FIGURE 1. SIMPLE LOOPED-AXIS WIND TURBINE SYSTEM [2]
The wings can be either symmetrical or asymmetrical. The
ascending wing segments are positioned with an angle of attack
ranging from 0 degree to 18 degree. The LAWT™ structure has
three tracks: an ascending track, a descending track and a
horizontal return track. The optimal track angle is 60 degrees
and preferred range of values is 45–75 degrees. To find the
optimal configuration to maximize the power generation,
several design parameters must be considered, such as the
spacing between the airfoils, the track speed, and the angle of
the ascending and descending tracks. This paper develops an
analytical formulation for the power given the average lift and
drag coefficients for each airfoil based on the kinematics of the
LAWT™ system. In addition, various numerical studies are
conducted to investigate the effect of the ascending track angle
and the spacing between the airfoils on the lift and drag
coefficients in order to identify the most important parameters
for power optimization.
KINEMATICS OF AIRFOIL OPERATION ON
ASCENDING TRACK
The LAWT™ is set in motion by the components of the lift
and drag forces in the direction of the track. During start-up
operation, the only velocity component seen by the airfoil is the
wind (assumed in horizontal direction). As a result, the high
initial pitch angle also acts as the angle of attack and motion
along the track is initiated primarily by the drag. As the track
attains speed, the track motion induces a downwards velocity
on the airfoil, which reduces the angle of attack seen by the
2 Copyright © 2015 by ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
airfoil, bringing the airfoil to a more favorable operating
condition where the lift starts to contribute more to the resultant
force along the track than the drag. Steady-state operation is
reached when the track reaches its design speed and brakes
engage to prevent it from accelerating further. The design track
speed in this study is considered to be equal to the wind speed,
as shown in Fig. 2.
FIGURE 2. KINEMATICS OF AIRFOIL ON ASCENDING TRACK
WITH TRACK SPEED EQUAL TO THE WIND SPEED
The freestream velocity seen by the airfoils on the
ascending track can be determined from the wind speed, the
forward track angle, and the track speed using the cosine rule as
shown in Eq. (1).
2 2
𝑈∞ = 𝑈𝑊 + 𝑈𝑇2 − 2𝑈𝑊 𝑈𝑇 cos 𝛽 (1)
The freestream velocity acts at an angle φ from the
horizontal given by Eq. (2).
2 2
𝑈∞ + 𝑈𝑊 − 𝑈𝑇2
𝜑 = acos ( ) (2)
2𝑈∞ 𝑈𝑊
To ensure that the angle of attack seen by the airfoil during
steady-state operation is the design angle of attack α, the airfoil
must be pitched on the track at an angle θ with respect to the
horizontal where𝜃 = 𝛼 + 𝜑.
The lift and drag on the airfoil act perpendicular and
parallel to the direction of the freestream respectively. The
resultant force along the track as a result of the lift and drag is
what powers the track motion. The coefficient of this resultant
force along the track can be obtained from the airfoil lift and
drag coefficients as given in Eq. (3).
𝐶𝐹 = 𝐶𝐿 sin(𝛽 + 𝜑) + 𝐶𝐷 cos(𝛽 + 𝜑) (3)
As can be seen from Fig. 2, for the case that the track
speed is equal to the wind speed, the direction of the drag is
such that the contribution of the drag to the resultant force
along the track is negative. However, since the lift is much
greater than the drag during steady-state operation, the net force
along the track remains positive to power the track motion.
For a single airfoil, the coefficients of lift and drag can be
obtained from the literature. For multiple airfoils in a LAWT™
 configuration, some differences are expected from the literature
values due to finite spacing between the airfoils. Numerical
simulations are used to determine the average lift and drag
coefficients for a cascade of airfoils which are substituted in
Eq. (3) to calculate the resultant force coefficient along the
track. The force along the track per airfoil can then be obtained
using Eq. (4).

1 2
𝐹𝑇 = 𝜌𝑈∞ 𝑐𝑆𝐶𝐹 (4)
2

The total power generated along the ascending track with n

airfoils can be obtained by Eq. (5), excluding mechanical and
other inefficiencies in the system.

𝑃 = 𝑛𝐹𝑇 𝑈𝑇 (5)
FIGURE 3. MESH IN THE COMPUTATIONAL DOMAIN
AROUND THE AIRFOIL
FLOW FIELD SIMULATION OF A SINGLE AIRFOIL ON
ASCENDING TRACK
Numerical simulation is conducted with a single airfoil on
the ascending track to validate the CFD modeling approach by
comparing the simulation results against the empirical data for
the airfoil available in the literature.

NACA 64421 Airfoil

This study is based on the NACA66421 airfoil. The NACA
64-series airfoils are thick airfoils widely used for wind turbine
blade application. They have several advantages in meeting the
intrinsic requirements for wind turbines in terms of design
point, off-design capabilities and structural properties [3].
The maximum lift for the NACA64421 airfoil is generated
at an angle of attack of approximately 12 degrees. Accordingly,
this is chosen as the design angle of attack for the LAWT™.
The baseline forward track angle is chosen as 60 degrees and a
wind speed of 14 m/s is considered. For steady-state operation
with track speed equal to wind speed, the freestream velocity
seen by the airfoil from Eq. (1) is also 14 m/s. The chord length
used in the LAWT™ design is 0.3 m.
From the lift curve and drag polar data available in Ref.
[4], the CL and CD for the NACA 64421 airfoil at an angle of
attack of 12 degree are approximately 1.2 and 0.12 respectively.
It should be noted that the Reynolds number (Re) for the
LAWT™ is around one-tenth of the Re used in the experiment
to obtain these values of CL and CD. However, Ref. [4] shows
only a slight dependence of the lift curve slope on the Re at an
angle of attack of 12 degrees. As such, in the absence of data at
the actual Re, the above values of 1.2 and 0.12 are used to
validate the numerical approach.
FIGURE 4. ZOOMED - IN VIEW OF C-GRID AROUND THE
AIRFOIL
The single airfoil mesh shown in Fig. 3 is imported into
ANSYS-Fluent to solve for the flow field and obtain the lift and
drag coefficients. Accurate performance prediction in
simulation of low Reynolds number flows over airfoils is
difficult due to the tendency of the laminar flow to separate and
transition over the airfoil. After separation, the flow structures
become increasingly irregular and eventually, transition from
laminar to turbulent flow occurs [5]. To account for flow
behavior in the laminar-turbulent transition region, the k-kl-ω
turbulence model is used throughout this study [6]. The flow is
assumed to be steady and incompressible. In addition, adaptive
meshing is employed near the airfoil surfaces as required to
maintain y+ values below 1 to ensure the accuracy of the
numerical solution. The solutions are obtained on three meshes
from coarse to fine to determine a suitable mesh to ensure the
mesh independence of solutions.

CFD Modeling Results and Discussion

The commercially available software ANSYS-ICEM is
used to generate a structured mesh around a single airfoil. The
mesh has a C-grid topology wrapped around the airfoil. This
type of mesh allows accurate modeling of curved surfaces and
refinement of the mesh at the leading and trailing edge. The
mesh from ANSYS-ICEM is shown in Figs. 3 and 4.
The plot of the pressure coefficient for the single NACA
64421 airfoil is shown in Fig. 5. The computed values of CL
and CD from ANSYS-Fluent are 1.196 and 0.104 respectively.
These values match the empirical data from Ref. [4] and thus
validate that the numerical approach is accurate.

3 Copyright © 2015 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 The meshes around 4, 5, 10 and 19 airfoils are generated in
a similar manner in ANSYS-ICEM. Each simulation is run with
a freestream velocity of 14 m/s with an angle of attack of 12
degrees. The k-kl-ω turbulence model is used in the simulations
and the flow is assumed to be steady and incompressible in each
case.

Results and Discussion

Samples of the pressure coefficient distribution over each
airfoil for the 3 and 4 airfoil cascades are shown in Figs. 7 and
8 respectively.

FIGURE 5. PRESSURE COEFFICIENT DISTRIBUTION FOR

SINGLE NACA 64421 AIRFOIL ON ASCENDING TRACK

FLOW FIELD SIMULATION OF CASCADE OF AIRFOILS

ON ASCENDING TRACK
For a given LAWT™, there is a certain number of airfoils
on the ascending track spaced some distance apart. This section
investigates the effect of stacking the airfoils along the track for
the LAWT™ configuration. In this section, simulations of
cascades of 3 to 19 NACA 64421 airfoils are conducted in
ANSYS-Fluent. The goal of this study is to establish trends in FIGURE 7. PRESSURE COEFFICIENT DISTRIBUTION FOR
average lift and drag coefficients versus the number of airfoils CASCADE OF 3 AIRFOILS ON ASCENDING TRACK
so that the average lift and drag can be predicted for any given
number of airfoils. The cases consist of 3, 4, 5, 10, and 19
airfoils respectively. The spacing between the airfoils along the
track direction is set equal to the chord length of 0.3 m.

CFD Modeling of Cascade of Airfoils

The mesh for the cascade of airfoils is generated in
ANSYS-ICEM using a similar blocking approach as for the
single airfoil repeated for each airfoil in the cascade. The mesh
has a C-grid topology wrapped around each airfoil. Fig. 6 shows
the mesh around three airfoils. In this mesh, there are 410,777
quadrilateral cells and 419,372 nodes. The far field boundary is
set at 15 chord lengths. Once the mesh is imported into ANSYS- FIGURE 8. PRESSURE COEFFICIENT DISTRIBUTION FOR
Fluent, adaptive meshing is employed to ensure that the y+ CASCADE OF 4 AIRFOILS ON ASCENDING TRACK
value is always less than 1.
The lift and drag coefficients for each airfoil in the 3, 4, 5,
and 10 airfoil cascades are given in Tables 1 and 2 respectively,
while Table 3 shows the average lift and drag coefficients for
all 5 cases. Airfoil 1 refers to the bottom-most airfoil in each
case.

TABLE 1. LIFT COEFFICIENT FOR EACH AIRFOIL IN CASCADE

Number of airfoils in cascade

3 4 5 10
1.246 1.255 1.190 1.013
1.138 1.078 1.121 1.013
1.138 0.992 0.989 0.984
- 1.046 0.956 0.967
FIGURE 6. DETAILS OF C-GRID AROUND THREE AIRFOILS - - 1.037 0.931
- - - 0.883

4 Copyright © 2015 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 - - - 0.818 formula for average lift and drag for the case of 30 airfoils
- - - 0.795 along the track and assuming a span of 1 m and standard air, the
- - - 0.786 power generated by the LAWT™ is approximately 8.45 kW.
- - - 0.976

TABLE 2. DRAG COEFFICIENT FOR EACH AIRFOIL IN CASCADE

Number of airfoils in cascade

3 4 5 10
-0.040 -0.054 -0.068 -0.075
0.041 0.027 0.011 -0.027
0.084 0.059 0.045 0.005
- 0.088 0.066 0.028
- - 0.094 0.044
- - - 0.056
- - - 0.061
- - - 0.074
- - - 0.079
- - - 0.101 FIGURE 9. AVERAGE LIFT COEFFICIENT VERSUS NUMBER
OF AIRFOILS IN CASCADE FOR BASELINE CASE
For each cascade, the lift coefficient is the highest for the
bottom-most airfoil and then decreases along the cascade,
except for the top-most airfoil, which displays a slight increase
again. On the other hand, the drag coefficient starts out at a
negative value and increases along the cascade such that the
top-most airfoil has the highest drag coefficient. The negative
drag coefficient on the bottom-most airfoils in the cascade is
caused by the large static pressure on the bottom surface that
cannot be balanced by the pressure on the top surface due to the
presence of the additional airfoils. This unbalanced pressure
overwhelms the viscous drag to produce a net force back
towards the freestream velocity on the bottom-most airfoils.

TABLE 3. AVERAGE LIFT AND DRAG COEFFICIENTS FOR

DIFFERENT NUMBERS OF AIRFOILS IN CASCADE

# airfoils Average CL Average CD FIGURE 10. AVERAGE DRAG COEFFICIENT VERSUS NUMBER
OF AIRFOILS IN CASCADE FOR BASELINE CASE
3 1.174 0.0284
4 1.093 0.0299
FLOW FIELD SIMULATION OF A SINGLE AIRFOIL ON
5 1.063 0.0312
DESCENDING TRACK
10 0.914 0.0344
Based on the design and kinematics of the LAWT™, the
19 0.766 0.0540
airfoils on the descending track may contribute an additional
force along the track. If the airfoils are fixed with respect to the
Based on the average lift and drag coefficients in Table 3,
track, the descending track only contributes to the power when
the trends in average lift and drag coefficient versus the number
the track speed is below a certain threshold. If the track speed
of airfoils can be identified. These are plotted in Figs. 9 and 10
exceeds this threshold, the net angle of attack becomes negative
respectively. From Fig. 9, it can be seen that the average lift
and the net resulting force detracts from the power generated by
coefficient has a negative power law dependence on the number
the ascending track. Alternatively, mechanical pitch
of airfoils. On the other hand, Fig. 10 shows that the average
adjustments can be added to the LAWT™ design to ensure that
drag coefficient increases linearly with the number of airfoils. It
the contribution from the descending track remains positive [1].
should be noted that although the average lift per airfoil
The present study considers an airfoil with a fixed pitch
decreases as the number of airfoils is increased, the total lift
that has been flipped to a trailing-edge-first configuration on
(and hence, power) still increases.
the descending track. The angle of attack for the flipped airfoil is
The results in this section can be used to predict the
20.2 degrees, which corresponds to an airfoil pitch angle such
average lift and drag coefficients for any number of airfoils and
that the ascending angle of attack is the design value of 12
calculate the total power using Eqs. (3) – ( 5). Using the general

5 Copyright © 2015 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 degrees when the track speed is a quarter of the wind speed and
both forward and reverse track angles are 60 degrees, as shown
in Fig. 11. It should be noted that in actual practice, the
descending track will lie in the wake of the ascending track and
will likely see a reduced wind speed. No attempt has been made
to include the effects of the wake velocity reduction in this
study.
OPTIMIZATION METHODOLOGY FOR LAWT™
DESIGN
Given some value of CL and CD, the effect of the various
angles in the LAWT™ design have already been accounted for
in the analytical formulation of power via Eqs. (1) – (3).
However, it is expected that the CL and CD are also affected by
the track angle and airfoil spacing. In this section, various
numerical studies are conducted by varying the forward track
angle and the spacing in turn around the baseline values of 60
degrees and one chord respectively (case 0). The objective of
this section is to identify the design parameter that has the
biggest effect on the power generation of the LAWT™ for
further optimization. The four cases investigated are
summarized in Table 4. It should be noted that for each of these
cases, a cascade of 3 airfoils is considered.

TABLE 4. SUMMARY OF CASES RUN

FIGURE 11. KINEMATICS OF AIRFOIL ON DESCENDING TRACK
WITH TRACK SPEED EQUAL TO 1/4TH OF THE WIND SPEED Case 0 1 2 3 4
Spacing 1c 1c 1c 1.25c 0.75c
The mesh generation and modeling setup are similar to the β (deg) 60 75 45 60 60
single airfoil on the ascending track mutatis mutandis. The
pressure coefficient distribution obtaining using ANSYS-Fluent Results and Discussion
is plotted in Fig. 12. From Fluent, the lift coefficient is 0.535 For each case, the lift and drag coefficients for each airfoil
and the drag coefficient is 0.1661. The coefficient of power are obtained from the numerical simulation. The average lift
along the track can be determined using Eq. (3) with the and drag coefficients, and the coefficient of the resultant force
forward track angle β replaced by the reverse track angle ω. along the track direction are computed and are shown in
Table 5.

TABLE 5. RESULTS OF OPTIMIZATION STUDY

Case 0 1 2 3 4
Spacing 1c 1c 1c 1.25c 0.75c
β (deg) 60 75 45 60 60
φ (deg) 60 52.5 67.5 60 60
Average CL 1.174 1.240 1.083 1.237 1.043
Average CD 0.028 0.027 0.040 0.030 0.030
Average CF 1.003 0.968 0.986 1.056 0.889
Average CF per
unit chord of 1.003 0.968 0.986 0.845 1.182
spacing
FIGURE 12. PRESSURE COEFFICIENT DISTRIBUTION FOR
SINGLE AIRFOIL ON DESCENDING TRACK
For cases 0, 1, and 2, the spacing and hence the number of
airfoils remains constant. From Eq. (4) and (5), the total power
The lift coefficient for the flipped airfoil is considerably
is therefore directly proportional to the average coefficient of
less than the single airfoil on the ascending track; on the other
force along the track. Thus, it can be seen that the forward track
hand, the drag coefficient is much higher. This is expected
angle of 60 degrees is the optimal case. The average CF for both
because the flow encounters the sharp trailing edge of the
cases 1 and 2 are lower than the average CF for case 0. It should
airfoil and separates leading to a stall condition, which reduces
be noted that although the average CL for case 1 is greater
lift and increases drag. This is verified from the velocity
compared to case 0, the lift acts along a direction such that it
contours on the airfoil showing a prominent zone of separation
has a smaller component in the direction of the track. This
after the sharp edge. These results indicate that the drag is also
result verifies the original patent that the optimal track angle for
a significant contributor to the resultant force along the track for
the ascending track is 60 degrees.
the descending track.
The results for the cases with varying spacing are of
greater consequence. From Table 5, case 3 with spacing of
1.25c has the highest average CF whereas case 4 with spacing

6 Copyright © 2015 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 of 0.75c has the lowest average CF. However, the spacing also
affects the number of airfoils that can be placed along the same
length of track. Assuming a track length that fits 30 airfoils
with a baseline spacing of 1c, increasing the spacing to 1.25c
would reduce the number of airfoils to 24 while reducing the
spacing to 0.75c would increase the number of airfoils to 40.
Taking this into account, the average CF per unit chord of
spacing shows that case 4 would have the greater power
generation according to Eq. (5) by 18% compared to the
baseline case.
These results show that significant increases in power
generation are possible by optimizing the design of the LAWT™,
particularly the spacing between the airfoils. This highlights the
scope for future research using a more advanced optimization
method such as Genetic Algorithm to find the optimum spacing
for maximum power generation for a LAWT system.
It should be mentioned here that the LAWT concept can
also be used for several hydro-kinetic applications, for example
for tidal and ocean currents energy conversion [7].

REFERENCES

[1] Syrovy, G.J., 2013, “Looped airfoil wind turbine,” U.S.

Patent 8618682.
[2] Teschler, L., 2013, “Looks like a conveyor, works like a
wind turbine,” from
http://machinedesign.com/batteriespower-supplies/looks-
conveyor-works-wind-turbine
[3] Chen, X., 2014, “Optimization of Wind Turbine Airfoils/
Blades and Wind Farm Layouts,” Ph.D. thesis, Washington
University in Saint Louis.
[4] Abbott, I.H., and von Doenhoff, A.E., 1959, Theory of Wing
Sections: Including a Summary of Airfoil Data, Dover, New
York, NY.
[5] Castonguay, P., Chunlei, L., and Jameson, A., 2010,
“Simulation of Transition Flow over Airfoils using the
Spectral Difference Method,” AIAA 40th Fluid Dynamics
Conference and Exhibit, The American Institute of
Aeronautics and Astronautics, Chicago, IL.
[6] ANSYS, 2012, Fluent Theory Guide, Canonburg, PA.
[7] Syrovy, G.J., 2014, “Looped airfoil wind turbine,” U.S.
Patent 8,866,325.

7 Copyright © 2015 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use