You are on page 1of 12

Aerodynamics of Wind Turbines:

An Unsteady Blade Element Momentum Theory Model


AME 90934 Course Project

Chris Kelley

December 13, 2010

Introduction
In the study of wind turbines, blade element momentum (BEM) theory is useful in that it can predict the
torque of a wind turbine operating in steady conditions. This theory is very useful to wind turbine design
because the effect of varying design parameters such as number of blades, blade pitch, blade radius, lift
and drag coefficients, freestream velocity, and finally the angular velocity of the wind turbine all will affect
the power output. By finding this torque distribution across each blade, the total power can be predicted
for a given angular velocity of the wind turbine. With this approach, the final result is usually presented
as a non-dimensional power coefficient, CP , as a function of the tip speed ratio, . While assuming wind
turbine blades are rigid and that wind speeds and aerodynamics are steady may be sufficient for a first order
design, optimizing wind turbines to extract more kinetic energy from the wind should include the real world
unsteadiness of this natural resource, and for inclusiveness, blade aeroelasticity. It is the focus of this project
to extend BEM theory to the temporal domain through the incorporation of unsteady thin airfoil theory. In
this way, the effect of wind gusts and unsteadiness in the freestream on the performance of a wind turbine
can be investigated. This will better quantify the actual power and energy production and true optimal
operating conditions of wind turbines functioning in realistic, unsteady wind conditions.
A broad range of analysis techniques exist in the study of wind turbine aerodynamics. Leishman details
three main areas of analysis which are often hard to all include in one model: unsteady wake modeling,
blade section unsteady aerodynamics, and dynamic stall1 . Within the area of wake modeling, Carpenter
& Freidovitch2 , have developed the dynamic inflow approach which includes the apparent mass term of an
impermeable solid rotor disc, but the analogy of impermeability is not rigorous. In addition the model
assumes C is equal to a constant, which is clearly not the case in unsteady thin airfoil theory. So there
L

are a few drawbacks to the dynamic inflow model.


The next area of analysis detailed by Leishman is blade section unsteady aerodynamics. This area is
very applicable to being modeled in BEM theory because the work of Theodorsen, Wagner, von Karman
and Sears, and Kussner, all provide analytical solutions to lift in the time domain. A harmonic plunging and
pitching of thin airfoils from the work of Theodorsen3 is certainly relevant to wind turbines. Experiments
have shown common structural dynamics of wind turbine blades oscillating on the order of Hertz with order
centimeter amplitudes4 . In this way the Theodorsen function is well suited to be incorporated into BEM
theory since the lift coefficient appears explicitly in the BEM equations. Rather than assume a constant
lift coefficient, it can be varied in time if it was known how the wind turbine blade section was pitching or
plunging due to structural flexure or, for the purposes of this paper, known from experiment. The work of
Wagner5 in step changes in angle of attack using the indicial method with the Duhamels integral is not as
well suited for incorporation into BEM theory because the lift coefficient must be convolved with all previous
times putting the unknowns (axial and tangential flow induction factors) within the convolution integral.
However accounting for fluctuating wind speeds seems significant as blade flexure. The work of von Karman
and Sears6 is useful in that they provide an analytical solution for unsteady lift of an airfoil in a transverse
gust, which is the main component seen by a rotating airfoil, which is actually a longitudinal wave to the
freestream velocity as seen from the inertial frame because the wind turbine blades are nearly orthogonal to
1 Leishman, J. G., Challenges in Modeling the Unsteady Aerodynamics of Wind Turbines, 40th AIAA Aerospace Sciences

Meeting, Reno, NV, 14-17 January 2002.


2 Carpenter, P. J., and Fridovich, B., Effect of A Rapid Blade- Pitch Increase on the Thrust and Induced-Velocity Response

of a Full-Scale Helicopter Rotor, NACA TN 3044, Nov., 1953.


3 Theodorsen, T., General Theory of Aerodynamic Instability and the Mechanism of Flutter, NACA TR 496, 1935.
4 Burton, T., Sharpe, D., Jenkins, N., Bossanyi, E., Wind Energy Handbook, J. W. Wiley, NY, 2001, pp. 225271.
5 Wagner, H., Uber die Entstehung des dynamischen Auftriebes von Tragflugeln, Zeitschrift fur Angewandte Mathematik

und Mechanick, Vol. 5, No. 1, 1925.


6 von Karmn, T., Sears, W.R., Airfoil Theory for Non- Uniform Motion, Journal of the Aeronautical Sciences, Vol. 5,

No. 10, 1938, pp. 379390.

1
the freestream. In this way the effect of freestream turbulence intensity and turbulence wave number can
be again incorporated directly into the BEM theory via the lift coefficient, Cl . Finally the Kussner function
could provide useful information from sharp edged gusts, but the relevance to wind turbines is questionable
because in the boundary layer of the earth where most wind turbines operate, rarely does turbulence have
this well defined abruptness.
The final area relevant to wind turbine aerodynamics is dynamic stall to incorporate viscosity into an
otherwise inviscid, incompressible flow. When an airfoil dynamically exceeds the stall angle of attack,
viscosity is important, and hence the area of dynamic stall tries to quantify the significant differences in lift
and drag in this non-linear regime. The LeishmanBeddoes7 Model is the most comprehensive yet simple
model developed in the study of dynamic stall. Due to its complexity in implementation, dynamic stall will
be ignored for the present work.
To summarize, BEM theory provides the fundamental equations that must be solved at each instant in
time. These equations take into account the usual steady design parameters, but also include unsteady lift
coefficients. The effect of aeroelasticity can be explored through the variation of blade plunging frequency
and amplitude. In addition, the effect of turbulence intensity and gust frequency can also be explored to see
the effect on power output of a wind turbine, which is not possible in the original, steady BEM theory.

Formulation
Blade Element Momentum Theory
BEM Theory is straightforward from the conservation of linear and angular momentum. It assumes that
the change in momentum of air is solely due to pressure drop across the annulus swept by a small blade
element. An implicit assumption of BEM theory is that there is no interaction between flow along the blade
span and the adjacent span locations. This assumption is actually quite good8 . So we have a freestream
velocity U approaching the wind turbine as a stream-tube seen in Fig. 19 . This stream-tube diffuses and

Figure 1: Stream-tube and Rotor Annulus Velocities

slows down the incoming air by a percentage of the freestream flow, known as the axial flow induction factor
a. Also, the air gains angular momentum as it passes through the area swept by the rotor blades, known
as the rotor disc. The air gains angular velocity equal to ra0 , where is the angular velocity of the wind
turbine rotor, r is the blade radius coordinate, and a0 is the tangential flow induction factor. Because the
blades are spinning, the relative angle of attack that the blade sees () depends on the blade pitch and
the freestream velocity just before the blade, U (1 a). So the relative velocity that the wind turbine sees
2 (1 a)2 + 2 r 2 (1 + a0 )2 at an angle of attack = arcsin( U (1a) ) as seen in Fig. 210 .
p
is W = U W
This shows the total angle = arcsin( UW (1a)
). Now a control volume analysis is performed equating the
rate of change of linear momentum to the force applied to the control volume at the rotor disc. In addition
the rate of change of angular momentum is found for the same control volume being equated to the external
7 Leishman, J. G.
8 Lock,C. N. H., Experiments to Verify the Independence of the Elements of an Airscrew Blade, ACR R&M, No. 953.
9 Burton, T., pp. 42, 60.
10 Burton, T., pp. 61.

2
torque. The details of this analysis can be seen in Burton11 . But the results are the two non-linear system
of algebraic equations with the independent variables being the flow induction factor, a and a0 as seen in
Eqns. 12 which are the governing equations of BEM theory.

W2 c
N (Cl cos + Cd sin ) = 8(a(1 a) + (a0 )2 ) (1)
U2 R

W2 c
N (Cl sin Cd cos ) = 82 a0 (1 a) (2)
U2 R
In the two BEM equations, the design parameters are typically number of blades N , blade chord c, blade
radius R, and the lift and drag coefficients Cl and Cd all of which can be functions of span location = Rr ,
and the tip speed ratio = Ur
. The designer should specify these parameters and then find the solution for
a and a0 . Typically thin airfoil theory, or experimental data is used for the lift and drag coefficients. Thus
Cl can be rewritten in the form for thin airfoil theory as 2. And then replacing angle of attack in terms
of the independent variable a, Cl = 2(arcsin( UW (1a)
) ).
To find a total power output, first one should integrate the lift along the blade span to get the total
torque Q.
1  
1 2 3 2 0 W Nc 0
Q(, t) = U R 8a (1 a) Cd (1 + a ) d (3)
2 U R
0

Then multiplying by the angular velocity the power output for the wind turbine is known. In this way BEM
theory connects choice of design parameters with the total power that a wind turbine can produce over a
range of tip speed ratios.

Incorporating Unsteady Aerodynamics


Original to this work is the idea that at any instant in time, conservation of linear and angular momentum
should be satisfied. So let us borrow from the field of unsteady aerodynamics and incorporate those analytical
solutions of lift into the equations that govern BEM theory. Experimentally it has been shown that wind
turbine blades bend in harmonic motion around 2 Hz with an amplitude up to 20 cm at the tip12 . Of course
these vary based on the specific cross-section, materials, size of the wind turbine, and turbulence intensity
and frequency. Since this aeroelastic response has been observed it can be incorporated with the Theodorsen
function C() into the lift coefficient.
!
h c
Cl (t) = 2C() + + 2
h (4)
U 2U

Here, the harmonic plunging motion is described by h(t) = h0 eit which puts the oscillating frequency
and amplitude now into the BEM theory and thus can be varied as design parameter to see the effect of
total power output. The non-dimensional angular velocity is related to the plunging angular velocity by
c
= 2U
.
Just as the interaction of inertial, aerodynamic, and elastic forces produce blade harmonic blade plunging,
there is also inherent unsteadiness to the wind. Through the use of the Sears function, S(), one can also
incorporate the unsteady lift due to gusts into BEM theory. A transverse gust is described as v(t) = v0 eit .
 
v
Cl (t) = 2S() + (5)
U
At most tip speed ratios, the blade pitch is nearly orthogonal to the freestream. This means that longitudinal
wind gusts in the inertial frame are the same as transverse gusts as seen by the blade section. So again
substituting this lift function into BEM theory allows a wind turbine designer to see the temporal effects of
turbulence intensity and amplitude on the power output rather than just a constant tip speed ratio.
11 Burton, T., pp. 6064.
12 Burton, T., pp. 225271.

3
Numerical Method
To solve the system of two equations (Eqns. 12) with the two unknowns (a and a0 ), a MATLAB script was
written which is included in the appendix. Because of the non-linearity of this system of algebraic equations,
the fsolve function was used. The algorithm for this solver is known as the Trust-Region Dogleg method.
This trust-region method is more robust than a Newton iteration in the event of handling singularities. In
all cases the solution converged rapidly with guesses for the solution a and a0 to be between 0 and 1. These
equations and unknowns were found for multiple instants in time and tip speed ratios, which are treated as
the independent variables, while the other design parameters were held constant. Then the code was modified
to include various plunging amplitudes and frequencies for the Theodorsen lift coefficient. The code was
then again modified to include the effects of unsteady wind gusts of various amplitudes and frequencies for
the Sears lift coefficient.

Results
Effects of Harmonic Blade Flexure Theodorsen Function
It was decided that a typical wind turbine, similar to the one being built at the Whitefield Research facility
should be studied. This meant using an average wind speed of U = 8 m/sec. Three blades were used
N = 3. The blade radius chosen is 10 m for a swept area of 314 . A chord of 60 cm is common. BEM theory
has been well studied for the affect of varying these design parameters, so for the present work the focus will
be on a temporal solution. Let us first only allow the blades to harmonically plunge and hence Eqn. 4 will
be used for the lift coefficient.
Fig. 3 shows the solution to the power output as a function of tip speed ratio, , and time t in seconds.
The following four subfigures include the aeroelastic plunging of the blade with various amplitudes 2.5 cm, 5
cm, 10 cm, and 20 cm, all oscillating at a frequency of 1.78 Hz ( = 11.2 radsec , = 0.42). As can be seen in
Fig. 3, the power output with plunging has a definite period that is constant for all amplitudes as evidenced
by the repeating pattern that looks similar to an x-ray of human teeth. The frequency seen in the power is
0.22 Hz, or 8 times slower than the frequency of plunging. For low amplitude plunging the maximum power
is seen at higher tip speed ratios. For the higher amplitudes, the maximum shifts down to lower tip speed
ratios. The interesting results here is that greater plunging amplitude increases the power output because
the amplitude of harmonic motion appears in all three forms of unsteady lift: quasi-steady, apparent mass,
and wake lift. The trade-off is a tighter band on optimal power output, that is the power band is narrowed
for greater plunging amplitudes. Peak powers on the order of 30-50 kW are to be expected for a 10 m radius
wind turbine. Next the power output was averaged in time to better see the effect over a larger runtime.
Fig. 4 shows that there is an initial power loss in going from steady operation to unsteady plunging due
to blade flexure. This is due to the fact that at a reduced plunging frequency of = 0.42, the Theodorsen
function has a magnitude of about 0.54, thus reducing the lift coefficient nearly to 54% of its original value.
The narrower and slightly increasing peaks in power for higher amplitudes is also seen in this time averaged
power coefficient. Also a check that verifies the code is that no power curve exceeds the Betz limit of 0.593.
Also it can be seen that the power curve goes negative, which is physical as the current model only considers
one blade pitch angle of 3o .
Next, the effect plunging frequency had on power output in time was investigated. The amplitude was
held fixed at 10 cm, while the frequency of oscillation was varied: 0.89 Hz, 1.78 Hz, 3.56 Hz, and 5.34 Hz
(corresponding to an of 5.6, 11.2, 22.4, and 33.6 rad
sec and also a reduced frequency of 0.21, 0.42, 0.84, and
1.26 respectively). The effect of the blade plunging frequency is apparent in the power contours of Fig. 5. All
of the plunging frequency to power frequency ratios held constant at 8, although the primary frequency in the
smallest frequency case ( = 0.21) is harder to determine, but the period is likely 9 seconds to fit the same
plunging frequency to power frequency ratio of 8. The time average power coefficient was again found and
is shown in Fig. 6. As the oscillating frequency increases the power band gets narrower and the peak power
output increases a little, but never exceeds the Betz limit. Finally it was investigated how this unsteady
power production affected actual energy production. The time series of power was numerically integrated
at the optimal tip speed ratio, that is the peaks of the power coefficient curves. Fig. 7 shows that the total
energy capture of a wind turbine is nearly cut in half for all oscillating amplitudes and frequencies even when
they are operated at their optimum tip speed ratios. This means that deflection of a wind turbine blades
should be increased (to a safe level) to increase their efficiency but due to fatigue this is likely undesirable.
Rather, the resonant frequency of the plunging blade due to aeroelasticity should be increased such that the
blade plunges at the highest frequency possible. This will maximize energy capture.

4
Figure 2: Velocity and Force Vectors

Power Contour [kW] Amplitude = 0.025 Power Contour [kW] Amplitude = 0.050 Power Contour [kW]
10 50 10 50 10 50

9 45 9 45 9 45

8 40 8 40 8 40

7 35 7 35 7 35

6 30 6 30 6 30

25

25 5 25 5

4 20 4 20 4 20

3 15 3 15 3 15

2 10 2 10 2 10

1 5 1 5 1 5

0 0 0
0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
t [sec] t [sec] t [sec]

Amplitude = 0.10 Power Contour [kW] Amplitude = 0.20 Power Contour [kW]
10 50 10 50

9 45 9 45

8 40 8 40

7 35 7 35

6 30 6 30

25 5 25

4 20 4 20

3 15 3 15

2 10 2 10

1 5 1 5

0 0
0 2 4 6 8 10 0 2 4 6 8 10
t [sec] t [sec]

Figure 3: Effect of Plunging Amplitude on Power Contour as a Function of Time and Tip Speed Ratio

5
Effect of Plunging Amplitude on Power Output
0.5

0.45

0.4

0.35

0.3
CP

0.25

0.2

0.15
steady
h = .025ei(11.2)t
0.1
h = .05ei(11.2)t
h = .1ei(11.2)t
0.05 h = .2ei(11.2)t

0
0 1 2 3 4 5 6 7 8 9 10 11

Figure 4: Effect of Plunging Amplitude on Time Averaged Power Coefficient

Power Contour [kW] Freq = 0.21 Power Contour [kW] Freq = 0.42 Power Contour [kW]
10 50 10 50 10 50

9 45 9 45 9 45

8 40 8 40 8 40

7 35 7 35 7 35

6 30 6 30 6 30

25 5 25

25 5

4 20 4 20 4 20

3 15 3 15 3 15

2 10 2 10 2 10

1 5 1 5 1 5

0 0 0
0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
t [sec] t [sec] t [sec]

Freq = 0.84 Power Contour [kW] Freq = 1.26 Power Contour [kW]
10 50 10 50

9 45 9 45

8 40 8 40

7 35 7 35

6 30 6 30

25

25 5

4 20 4 20

3 15 3 15

2 10 2 10

1 5 1 5

0 0
0 2 4 6 8 10 0 2 4 6 8 10
t [sec] t [sec]

Figure 5: Effect of Plunging Frequency on Power Contour as a Function of Time and Tip Speed Ratio

6
Effect of Plunging Frequency on Power Output
0.5

0.45

0.4

0.35

0.3
CP

0.25

0.2

0.15
steady
h = .1ei(5.6)t
0.1 h = .1ei(11.2)t
h = .1ei(22.4)t
0.05 h = .1ei(33.6)t

0
0 1 2 3 4 5 6 7 8 9 10 11

Figure 6: Effect of Plunging Frequency on Time Averaged Power Coefficient

Effect of Plunging Amplitude on Energy Production at opt Effect of Plunging Frequency on Energy Production atopt
250 250
steady steady
h =.025ei(11.2)t h =.1ei(5.6)t
h =.05ei(11.2)t h =.1ei(11.2)t
200 h =.1ei(11.2)t 200 h =.1ei(22.4)t
h =.2ei(11.2)t h =.1ei(33.6)t
Energy [kW*hr]

Energy [kW*hr]

150 150

100 100

50 50

0 0
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
t [sec] t [sec]

Figure 7: Plunging Blades and Energy Production at Optimal Tip Speed Ratio

7
Effects of Harmonic Gusts Sears Function
Now the effects of power production for a wind turbine in the influence of wind gusts will be examined.
In the inertial frame let us consider a longitudinal gust, which is seen nearly as a transverse gust by the
blade section. In this sense the harmonic, transverse gust problem is applied to BEM theory with the Sears
function in the lift coefficient seen in Eqn. 5. The effect of gust intensity and gust frequency was found for
typical atmospheric conditions. First the effect of gust intensity was found for typical values of 5%, 10%,
15%, and 20% as seen in Fig. 8. Gust intensity reduces the optimum tip speed ratio. For example a gust

Power Contour [kW] Intensity = 5% Power Contour [kW] Intensity = 10% Power Contour [kW]
10 50 10 50 10 50

9 45 9 45 9 45

8 40 8 40 8 40

7 35 7 35 7 35

6 30 6 30 6 30

25


25 5

25 5

4 20 4 20 4 20

3 15 3 15 3 15

2 10 2 10 2 10

1 5 1 5 1 5

0 0 0
0 2 4 6 8 10 0 5 10 15 20 0 5 10 15 20
t [sec] t [sec] t [sec]

Intensity = 15% Power Contour [kW] Intensity = 20% Power Contour [kW]
10 50 10 50

9 45 9 45

8 40 8 40

7 35 7 35

6 30 6 30

25

25 5

4 20 4 20

3 15 3 15

2 10 2 10

1 5 1 5

0 0
0 5 10 15 20 0 5 10 15 20
t [sec] t [sec]

Figure 8: Effect of Gust Intensity on Power Contour as a Function of Time and Tip Speed Ratio

with intensity of 20%, CPmax occurs at a tip speed ratio 4, whereas for gusts of 5%, the operating speed
should be at = 5.5.
The time average power coefficients were again found as seen in Fig. 9. For low tip speed ratios the time
averaged power coefficients very closely match the steady solution. This is due to the fact that the frequency
of the gusts is on the order of 0.1 Hz or a period of 10 seconds. For this low of a frequency the Sears function
is nearly equal to unity.
The effect of gust frequency was examined. Burton attributes most wind gust frequencies to the order
of 0.1 sec13 . So a range of angular frequencies was selected as 0.31, 0.62, 1.26, and 2.51 rad ( = 0.012,
0.024, 0.047, and 0.094). The effect of changing the gust frequency is explicitly seen in the repeating pattern
of the power contour. In fact the frequency content of the power is identical to the gust frequency, unlike
the previous airfoil plunging analysis. This is due to the slow gust changes. Again, the time averaged power
coefficients were found. The interesting result seen in Fig. 11 is how closely the power output follows the
steady solution for wind gusts, although it does reduce the width of optimal performance slightly.
Finally the effect wind gusts have on actual energy extraction from this unsteady flow has been calculated
as before as a time integral of power operating at Cpmax . Fig. 12 shows that wind gusts have very little
effect on the energy a wind turbine can extract from the wind. This is due to the fact the frequencies of
wind gusts are very slow and hence the magnitude of Sears function is almost unity. Clearly this indicates
that a wind turbine designer should operate the wind turbine at an optimal tip speed ratio as prescribed by
blade flexure results and not due to the effects of wind gusts. This also shows that the time averaged power
coefficient results do not tell the whole story. Specifically, even though there are power coefficient peaks from
airfoil plunging above the steady solution, the time integration of this power yields a much smaller energy
than considering just the effects of wind gusts.
13 Burton, T., pp. 27.

8
Effect of Gust Intensity on Power Output
0.5

0.45

0.4

0.35

0.3
CP

0.25

0.2

0.15 steady
I = 5%, w = .57ei(0.63)t
0.1 I = 10%, w = 1.13ei(0.63)t
I = 15%, w = 1.70ei(0.63)t
0.05
I = 20%, w = 2.26ei(0.63)t
0
0 1 2 3 4 5 6 7 8 9 10 11

Figure 9: Effect of Gust Intensity on Time Averaged Power Coefficient

Power Contour [kW] Freq=0.012 Power Contour [kW] Freq=0.024 Power Contour [kW]
10 50 10 50 10 50

9 45 9 45 9 45

8 40 8 40 8 40

7 35 7 35 7 35

6 30 6 30 6 30

25 5 25

25 5

4 20 4 20 4 20

3 15 3 15 3 15

2 10 2 10 2 10

1 5 1 5 1 5

0 0 0
0 2 4 6 8 10 0 5 10 15 20 0 5 10 15 20
t [sec] t [sec] t [sec]

Freq=0.047 Power Contour [kW] Freq=0.094 Power Contour [kW]


10 50 10 50

9 45 9 45

8 40 8 40

7 35 7 35

6 30 6 30

25

25 5

4 20 4 20

3 15 3 15

2 10 2 10

1 5 1 5

0 0
0 5 10 15 20 0 5 10 15 20
t [sec] t [sec]

Figure 10: Effect of Gust Frequency on Power Contour as a Function of Time and Tip Speed Ratio

9
Effect of Gust Frequency on Power Output
0.5

0.45

0.4

0.35

0.3
CP

0.25

0.2

0.15 steady
i(0.31)t
w = 1.13e
0.1 w = 1.13ei(0.62)t
w = 1.13ei(1.26)t
0.05
w = 1.13ei(2.51)t
0
0 1 2 3 4 5 6 7 8 9 10 11

Figure 11: Effect of Gust Frequency on Time Averaged Power Coefficient

Effect of Gust Intensity on Energy Production at opt Effect of Gust Frequency on Energy Production at opt
250 250
steady steady
I =5%, w =.57ei(0.62)t w =1.13ei(0.31)t
I =10%, w =1.13ei(0.62)t w =1.13ei(0.62)t
200 I =15%, w =1.70ei(0.62)t 200 w =1.13ei(1.26)t
I =20%, w =2.26ei(0.62)t w =1.13ei(2.51)t
Energy [kW*hr]

Energy [kW*hr]

150 150

100 100

50 50

0 0
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
t [sec] t [sec]

Figure 12: Wind Gusts and Energy Production at Optimal Tip Speed Ratio

10
Conclusion
Unsteady aerodynamics have been incorporated into BEM theory with some success. Specifically, the effect
of plunging blade sections and wind gusts has made possible more accurate energy capture predictions in the
time domain, rather than steady BEM theory as seen in the power contour plots. The Theodorsen function
was used in the lift coefficient in consideration of a harmonically plunging blade section and the consideration
of increasing amplitude and frequency increased the rate of energy production, but also decreased the width
of peak power on the time averaged power coefficient curve. This means that strict control of tip speed ratio
must be maintained or even further loss in efficiency will occur. Also, the effect of blade flexure shows that
the optimum tip speed ratio is always lower than that predicted by BEM. In the second study, the effect
of wind gusts was examined. Here the lift coefficient in BEM theory was based on the Sears function to
incorporate sinusoidal wind gusts. In general, these gust frequencies are on the order of 10 times smaller
than blade plunging, and thus the effect was purely quasi-steady. Again the effect of higher frequency and
intensity gusts lowered the optimal tip speed ratio slightly and reduced the power band width. In comparing
the effects of blade flexure to wind gusts, it is clear that blade flexure significantly affects power production
and hence energy capture much more than wind gusts. This is due to the inherent lower frequency of wind
gusts. Also it should be noted that time averaged power coefficients can be misleading. In comparing Fig. 4
(Effect of Plunging Frequency on Time Averaged Power Coefficient) with Fig. 9 (Effect of Gust Frequency
on Time Averaged Power Coefficient) we see that the peak time average power coefficient for plunging is
greater than that for the peak time average power coefficient for gusts. But when one examines the time
integration of power, and looks at the total energy capture in time, the effect of gusts alone was minimal
compared to blade plunging whose energy capture was nearly halved.

11
References
Burton, T., Sharpe, D., Jenkins, N., Bossanyi, E., Wind Energy Handbook, J. W. Wiley, NY, 2001.

Carpenter, P. J., and Fridovich, B., Effect of A Rapid Blade- Pitch Increase on the Thrust and
Induced-Velocity Response of a Full-Scale Helicopter Rotor, NACA TN 3044, Nov. 1953.

Leishman, J. G., Challenges in Modeling the Unsteady Aerodynamics of Wind Turbines, 40th AIAA
Aerospace Sciences Meeting, Reno, NV, 14-17 January 2002.

Lock, C. N. H., Experiments to Verify the Independence of the Elements of an Airscrew Blade, ACR
R&M, No. 953.

Theodorsen, T., General Theory of Aerodynamic Instability and the Mechanism of Flutter, NACA TR
496, 1935.

von Karmn, T., Sears, W.R., Airfoil Theory for Non- Uniform Motion, Journal of the Aeronautical
Sciences, Vol. 5, No. 10, 1938, pp. 379390.

Wagner, H., Uber die Entstehung des dynamischen Auftriebes von Tragflugeln, Zeitschrift fur
Angewandte Mathematik und Mechanick, Vol. 5, No. 1, 1925.

12

You might also like