Download as pdf or txt
Download as pdf or txt
You are on page 1of 17

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/287800146

Design methods of horizontal axis wind turbine rotor blades

Article  in  International Journal of Industrial Electronics and Drives · January 2015


DOI: 10.1504/IJIED.2015.072789

CITATIONS READS

7 251

Some of the authors of this publication are also working on these related projects:

wind turbine View project

All content following this page was uploaded by Yasser El-Okda on 30 July 2021.

The user has requested enhancement of the downloaded file.


Int. J. Industrial Electronics and Drives, Vol. 2, No. 3, 2015 135

Design methods of horizontal axis wind turbine rotor


blades

Yasser Mohamed El-Okda


Mechanical Engineering Department,
Higher Colleges of Technology,
Abu Dhabi Men’s College,
P. O. Box 25035, Abu Dhabi, UAE
Email: yelokda@vt.edu
Email: yelokda@hct.ac.ae

Abstract: Horizontal axis wind turbines (HAWTs) development and utilisation is growing
rapidly. No such a unique methodology in literature describes the design of HAWT rotor blades.
Each researcher usually adopts one of the design methods known to his best knowledge. In this
paper, we outline the HAWT design analysis using the blade element momentum (BEM) theory.
A total of ten different methods for the calculation of the chord length distribution along the rotor
blade, with and without considering wake rotation, are presented. The performance of various
methods for the correction of the axial induction factor is also presented. An important outcome
of this work is the elimination of the discrepancy encountered when trying to select among the
methods of blade design as we found that a few of the published methods are actually different
faces of the same coin and they lead to exactly the same final output yet through a different
procedure. Several methods based on BEM analysis for optimum chord length and flow angle are
found to produce very close blade geometry along the outer half of the blade length. The
inclusion of drag in the calculation of the optimal chord length is found to be insignificant.

Keywords: wind energy; wind turbine; high speed wind turbine; blade element momentum
theory; horizontal axis wind turbine; HAWT; optimum blade design.

Reference to this paper should be made as follows: El-Okda, Y.M. (2015) ‘Design methods of
horizontal axis wind turbine rotor blades’, Int. J. Industrial Electronics and Drives, Vol. 2,
No. 3, pp.135–150.

Biographical notes: Yasser Mohamed El-Okda received his BS from Zagazig University, Egypt
in 1994. He earned his MS and PhD from Virginia Tech in 2002 and 2005 respectively. From
2005 till present, he has served as a faculty member in full or temporary basis in Zagazig
University (ZU), Egypt, Misr University for Science and Technology (MUST), Sultan Qaboos
University (SQU), Oman and Abu Dhabi Men’s College (ADMC), UAE. Currently, he is on
sabbatical leave from Zagazig University to work at Higher Colleges of technology, ADMC, in
UAE. His areas of interest are experimental and computational fluid mechanics and
aerodynamics, flow around low-rise building and wind energy.

1 Introduction high. Therefore, wind turbines must be designed for


maximum efficiency in order to extract the largest amount
With the increasing energy demand in the world, renewable
of energy from wind and to make the generation of
energy production has become significant in providing
electricity from wind power economically feasible. In
alternate sources of energy and in reducing greenhouse
addition, theoretically, the maximum energy that can be
effect. Clean energy is also achieved through renewable
extracted from wind is no more than 59.3% of the energy
energy sources such as hydropower, solar energy and wind
content in the wind according to Betz who is one of the
energy. Among these sources comes wind energy especially
pioneers in this field.
in electric power generation industry (Uddin and
Wind turbines are classified as vertical axis and
Abera, 2009; Bayoumi, 2014a, 2014b; Yousefi-Talouki and
horizontal axis turbines and high speed and low speed
Ramezanpour, 2014; Mosaad and Salem, 2014).
turbines. High and low speeds characterise how fast the
Wind turbines of different arrangements and
speed of rotation relative to the wind speed. Low speed
configurations are used to extract energy from the wind.
wind turbines are usually used in applications such as
The rotor is the main component in a wind turbine that is
pumping water and grinding. It is found that wind turbines
responsible for extracting such energy and converting it to
running at high speeds are proven to be more efficient than
useful mechanical energy. While wind energy is almost
those running at low speeds. Hence, high speed horizontal
free, the initial cost of a wind turbine, however, is rather
axis wind turbines (HAWTs) are most suitable for electric

Copyright © 2015 Inderscience Enterprises Ltd.


136 Y.M. El-Okda

power generation. This emphasises the need for efficient u is the axial velocity at the rotor and U∞ is incoming flow
turbines. The wind turbine rotor is the main component of velocity. The performance parameters of HAWTs are the
the machine that is actually responsible for the extraction of power coefficient CP, the trust coefficient CT and tip speed
energy from the wind. High speed wind turbines rotor ratio λ. CP and CT obtained by the actuator disk method are,
blades are developed by aerodynamic analysis of the blades.
Several methods for obtaining optimum or highly efficient CP = 4a (1 − a) 2 (2)
blade design are found in literature. No such a unique CT = 4a(1 − a) (3)
method or procedure is known to be the best. No
comparisons between these methods are found in literature The optimum value of Cp can be reached by setting the
as well. In this paper, we try to provide an account of the differentiation of Cp to zero
methods used in wind turbine blade design and compare the
blade geometry produced by several methods. dCP
= 4(1 − a)(1 − 3a ) = 0 (4)
This paper starts by introducing the analytical methods da
used in the analysis of wind turbine performance mainly the
which has an optimum value of 16/27 at a = 1/3 giving the
actuator disc model and the blade element momentum
well-known Betz limit. It is important to note here that the
(BEM) theory. Comparison of the correction methods for
actuator disc model does not consider wake rotation which
thrust coefficient is investigated since it is an integral part of
actually occurs as the flow exits the rotor.
this analysis. The methods used in obtaining optimum wind
turbine blade geometry are presented. Comparisons between
the resulting geometries are conducted with and without
3 Momentum theory with wake rotation
neglecting drag. Finally, highlights on the design of wind
turbine blade are presented. Conservation of angular momentum in an annular stream
tube, as shown in Figure 2, is considered in the momentum
theory analysis. The thrust on a differential element
2 Actuator disk model obtained from the axial momentum theory is given by,
The actuator disk model for HAWT rotor considers the dT = 2 ρu ( u − U ∞ ) dA = 4πρU ∞2 a (1 − a)rdr (5)
change of momentum in the axial direction. As shown in
Figure 1, as the free stream flow approaches the turbine, it Introducing wake rotation and employing the angular
decelerates and undergoes reduction in velocity. momentum theory, the thrust force on the element is given
by
Figure 1 Actuator disk model
dT = 4πρΩ 2 a ′ (1 + a ′ ) r 3 dr (6)

where r is the radius of the differential element of thickness


dr. a ′ is the tangential induction, or interference, factor
defined as
ω
a′ = (7)

Ω is the rotor angular speed. ω is the wake angular velocity.
Figure 2 shows that there is no flow rotation at the inlet
to the blade and the flow exhibits rotation with speed ω in
the wake. Hence, the average tangential flow velocity that
the blade experiences is calculated from the average of inlet
and exit conditions as
1
Ωr + ωr = Ωr (1 − a ′ ) (8)
2
The torque on the differential element is
dQ = 4πρU ∞ Ωa ′(1 − a)r 3 dr (9)
Source: Gundtoft (2009)
The reduction in axial velocity is characterised by the axial and the power on a differential element is given by
induction factor a which is defined as 1 8
dP = ΩdQ = ρAU ∞3 2 a ′(1 − a) λ3r dλr (10)
U −u 2 λ
a= ∞ (1)
U∞ λr and λ are the local speed ratio and the tip speed ratio
respectively defined as,
Design methods of horizontal axis wind turbine rotor blades 137

Ωr the blades and φ is the flow angle, the angle of relative wind
λr =
U∞ which is the angle at which the airfoil sees the flow.
(11)
ΩR
λ= Figure 3 The blade element model (see online version
U∞ for colours)
where R is the radius of the rotor.

Figure 2 Rotating annular stream tube analysis (see online


version for colours)

Source: Ingram (2011)

5 BEM theory
The momentum theory analysis considers the conservation
of angular momentum in an annular stream tube. Hence, a
rotational component to the fluid velocity or wake rotation
is added unlike the actuator disk analysis which only
includes the axial velocity component. Hence, no
formulation for the torque is obtained. In the meantime, the
Source: Ingram (2011) momentum theory does not consider the effects of the main
geometrical parameters of the rotor, namely the chord and
twist distributions along the blade as well as the number of
4 Blade element theory blades. The blade element theory considers the aerodynamic
forces on an element of the blade and produce formulation
The blade element theory involves dividing the blade into a for the thrust and power corresponding to those produced by
number of segregated elements of finite length as shown in the momentum theory. The blade element theory and the
Figure 3. Each element is assigned an airfoil with a certain momentum theory are then combined to determine the wind
chord length c and twist angle θ with the rotor plane. Only turbine blade performance via the BEM theory.
the aerodynamic characteristics of the airfoil, mainly lift and The BEM theory is quite simple yet it gives
drag, as well as the chord and twist are considered in the considerable accurate results. However, due to its
analysis. Each of the blade elements experiences a slightly assumptions, it cannot model some aerodynamic effects on
different rotational speed (Ωr) and different twist angles as wind turbine rotor. Corrections are usually needed for
well as different chord, c, than its neighbouring elements. effects such as high induction factors on heavily loaded
The forces on the blade elements are solely determined by rotors, blade hub and tip losses and skewed rotor when the
the lift and drag coefficients. There is no aerodynamic rotor axis is at an angle with the wind direction.
interaction between different blade elements assumed. Equating equations (5) and (13) yields,
From the analysis of the blade aerodynamic forces,
shown in Figure 4, and the velocity triangle, Figure 5, we a cos φ ⎡ ⎛ CD ⎞ ⎤
can obtain relations for the torque and thrust respectively as 1− a
= σCL 1+
4sin 2 φ ⎢⎣ ⎜⎝ CL ⎟ tan φ ⎥ (14)
⎠ ⎦
follows (Kulunk, 2011; Song, 2012; Tenguria et al., 2011),
Equating equations (37) and (12),
U ∞2 (1 − a) 2
dQ = σπρ ( CL sin φ − CD cos φ ) r 2 dr (12) a′ σCL ⎡ ⎛ CD ⎞ ⎤
sin 2 φ =
1 − a 4 λr sin φ ⎢1 − ⎜ C ⎟ cot φ ⎥ (15)
⎣ ⎝ L ⎠ ⎦
U ∞2 (1 − a) 2
dT = σπρ ( CL cos φ + CD sin φ ) rdr (13) The ratio CL/CD is called the glide ratio and is usually
sin 2 φ
chosen to be maximum for the design angle of attack so as
CL and CD are the lift and drag coefficients of the airfoil that the maximum torque is obtained from the blade element.
is assigned to the blade element under consideration. Hence, CL is usually much larger than CD and CD may be set
Bc to zero in equations (14) and (15).
σ= is the solidity of the element. B is the number of From the velocity triangle in Figure 5
2πr
138 Y.M. El-Okda

1− a Rearranging equations (14) and (15) and solving for a and


tan φ = (16)
a ′,
(1 + a ′ ) λr

Combining equations (15) and (16) and rearranging 1


a= (21)
4sin 2 φ
a′ σCL ⎡ ⎛ CD ⎞ ⎤ 1+
=
1 ∓ a ′ 4 cos φ ⎢1 − ⎜ C ⎟ cot φ ⎥ (17) σCn
⎣ ⎝ L ⎠ ⎦
1
Setting CD to zero since the glide ratio CD/CL is usually a′ = (22)
4sin φ cos φ
much less than 1.0, equations (14), (15) and (17) can be −1
σCt
written as,
a cos φ The normal and tangential force coefficients are given by
= σCL (18)
1− a 4sin 2 φ Cn = CL cos φ + CD sin φ

a′ σCL Ct = CL sin φ − CD cos φ


= (19)
1 − a 4 λr sin φ

a′ σCL
= (20)
1 + a 4 cos φ

Figure 4 Aerodynamic forces on a blade element

FN

Blade FT
rotation

c
θ

α
φ

Urel

Note: φ is the flow angle, θ is the twist angle and α is the angle of attack.

Figure 5 Velocity triangle for the flow over a blade element showing the relative velocity of the flow with respect to the blade element
Blade
a′Ωr rotation

φ aU
w Ωr Ω r(1 + a′)
φ φ
U

U(1 – a)
Urel
Urel
Design methods of horizontal axis wind turbine rotor blades 139

Setting CD equals zero, equations (21) and (22) yield, n


8Δλr
1
CP = ∑
i =1
λ2
F sin 2 φ ( cos φ − λr sin φ )
a= (23) (32)
4sin 2 φ
( sin φ + λr cos φ ) ⎛⎜1 − D cot φ ⎞⎟ λ2r
1+ C
σCL cos φ ⎝ CL ⎠
1 where
a′ = (24)
4 cos φ
−1 Δλr = λr (i ) − λr (i −1)
σCL
The effect of shed vortices on the flow over the blade
From equations (18) and (19), we can prove that
outboard and towards the tip is not accounted for by the
a = a ′λr cot φ which can also be obtained from the velocity
BEM theory. A reduction in the pressure under the blade
triangle in Figure 5. Using this later relation for a in results as the air flows around the tip as shown in Figure 6.
conjunction with equations (23) and (24), we can obtain the This effect reduces the circulation of air which leads to
lift coefficient as a function of local speed ratio, flow angle reduction of torque and performance known as tip loss.
and element solidity
Figure 6 Tip loss flow diagram
4sin φ ( cos φ − λr sin φ )
CL = (25)
σ ( sin φ + λr cos φ )

The output power from each element is calculated by


dP = ΩdQ (26)
R
P= ∫
rh
ΩdQ (27)

where rh is the hub radius. Then,


R

CP =
P
=
∫ rh
ΩdQ
(28)
Pwind 1
ρπU ∞3 R 2
2
Using equation (12) and integrating for the whole blade
Source: McCosker (2012)
λ
8 ⎡ C ⎤
CP = 2
λ ∫
λh
λ3r a ′(1 − a) ⎢1 − D cot φ ⎥ dλr
⎣ DL ⎦
(29) The optimum blade is found when circulation distribution
remains uniform along its length according to Glauert
theory (Cencelli, 2006). Hence, vorticity may only be shed
Equation (29) can be discretised to perform numerical at the root and the tip. According to Prandtl, optimum rotor
calculations. If the total length from hub to tip is divided sheds a helical vortex sheet that moves as a rigid body
into n equal elements (Song, 2012), then while it is convected away from rotor. The tip-loss factor
8 R − rh
n models these vortex sheets as a series of parallel planes
∑ Fλ a′(1 − a) ⎛⎜⎝1 − C ⎞
CD
CP = 3
r cot φ ⎟ (30) with uniform spacing. Prandtl developed a method to
λ nR i =1 L ⎠ approximate the radial flow effect near the blade tip which
where F is loss factor to be introduced shortly. is sufficiently accurate for high tip speed ratios for turbines
An another form of the equation is given by Manwell with two or more blades (McCosker, 2012):
et al. (2009), 2 ⎛ B (1 − r / R ) ⎞
Ftip = cos −1 exp ⎜ − ⎟ (33)
8
n π ⎝ 2(r / R ) sin φ ⎠
CP =
λn ∑ F sin φ ( cos φ − λ sin φ )
i =1
2
r
The tip loss factor decreases towards zero at the tip. In
(31)
⎛ CD ⎞ 2 addition, the hub loss factor is given by,
⎜ sin φ + λr cos φ − C cot φ ⎟ λr
⎝ L ⎠ 2 ⎛ B ( r − rhub ) ⎞
Fhub = cos −1 exp ⎜ − ⎟ (34)
It should be noted that if the elements are not of equal π ⎝ 2r sin φ ⎠
length, the power coefficient can be numerically estimated The total loss factor is thus calculated by
using equation (32).
F = Ftip .Fhub (35)
140 Y.M. El-Okda

The formulas of a and a ′ including loss correction and with Figure 7 Relationship between axial induction, flow state and
drag term omitted, thrust of a rotor

1
a= (36)
4 F sin 2 φ
1+
σCL cos φ

1
a′ = (37)
4 F cos φ
−1
σCL
The equation of CL including loss factor is given by
4 F sin φ ( cos φ − λr sin φ )
CL = (38)
σ ( sin φ + λr cos φ )

It should be noted that the flow angle φ can be calculated


from
⎛ a ′λr ⎞ Source: Spera (2009)
φ = tan −1 ⎜ ⎟ (39)
⎝ a ⎠ Empirical relationships between CT and the axial induction
This relation can be proven by equating the thrust force in factor are often used in this region to correct CT and a where
equations (5), (6) and using equation (16). Equation (39) the momentum theory is invalid. These corrections are
constitutes an alternative form for the flow angle φ. Its embedded and iterated within the iterative routine that is
existence justifies the fact that the resultant induced usually employed to calculate the induction factors a and
velocity, w, is normal to the resultant relative velocity, Urel, a ′. Below is an account of the common methods used to
that the airfoil sees as shown in Figure 5 as long as the local correct for the axial induction factor,
angle of attack is below stall. 1 Glauert method:
Peters and Modarres (2013) showed that the resultant
induced velocity, w, shown in Figure 5, acts along the same Glauert method is employed for a > 0.4 or equivalently
line of the lift force L but in opposite direction. Hence, only CT > 0.96 (Glauert, 1926; Eggleston and Stoddard,
one equation for the axial induced velocity is required that 1987). Glauert’s empirical equation is given by
is obtained from the axial momentum analysis. The angular 1⎡
momentum equation is thus rendered redundant. With their a= 0.143 + 0.0203 − 0.6427 ( 0.889 − CT ) ⎤⎦ (40)
F⎣
theory, they were able to obtain a closed form formulation
for the optimum power coefficient. F is Prandtl’s tip loss factor and CT is calculated from
Fσ (1 − a) 2 Cn
CT = (41)
6 Axial induction factor correction sin 2 φ

From the assumptions of the momentum theory, the far Cn is calculated from equation (23).
wake velocity, Vw, is given by
2 Buhl method:
Vw = U ∞ (1 − 2a) The following relation for the thrust coefficient for
When a reaches 0.5, the wake velocity is thus zero, which is CT > 0.96 is given by Buhl (2005) and Marshall and
a condition where the streamlines no longer exist (Stoddard, Buhl (2005),
1976). The momentum theory assumptions have thus been 8 ⎛ 40 ⎞ ⎛ 50 ⎞
violated and its solution at this state becomes invalid. CT = + ⎜ 4F − ⎟ a + ⎜ − 4F ⎟ a2 (42)
9 ⎝ 9 ⎠ ⎝ 9 ⎠
Figure 7 shows that the momentum theory fails to predict
the thrust coefficient for values of a above 0.5 (Manwell Buhl’s equation for a (Molenaar, 2003),
et al., 2009; Stoddard, 1976). Measured data show that
thrust continue to increase for as large as CT = 2.0 as a 18 F − 20 − 3 CT (50 − 36 F ) + 12 F (3F − 4)
a= (43)
increases above 0.5. 36 F − 50
The BEM theory predicts the values for the axial
Molenaar (2003) also pointed out that equation (29)
induction factor throughout the windmill state, i.e., for
only gives reliable results for axial induction factor
a < 0.4. At high winds, a turbulent wake state occurs which
values between 0.0 and 0.4. For axial induction factors
is characterised by turbulence and recirculation behind the
greater than 0.4 the BEM theory does not yield reliable
rotor.
results. Glauert’s correction for determination of the
Design methods of horizontal axis wind turbine rotor blades 141

axial induction factor is valid only for F = 1. If the λ


8 ⎡ CD ⎤
losses at the tip of the blade are taken into account
(F < 1), Buhl and Wilson-Walker corrections give
CP = 2
λ ∫ Fλ a′(1 − a) ⎢⎣1 − D
λh
3
r
L
cot φ ⎥ dλr

(47)

reliable results. In addition, Buhl and Wilson-Walker


corrections overcome the numerical instability
associated with the implementation of Glauert 7 3D correction of Snel et al.
correction in conjunction with the presence of tip
losses. Figure 8 shows the performance of different An increase in the lift coefficient results as an effect of
correction methods for values of loss factor, F, of 1.0, rotation. This comes as a result of the phenomenon of stall
0.9, 0.7, and 0.5. It is shown in Figure 8 that Glauert delay which is also known as centrifugal pumping or
correction does not preserve continuity for values of F rotational augmentation. Snel, Houwink, and Bosschers
less than 1.0. derived what is called ‘3D correction’ method that corrects
the lift coefficient to account for rotational effects (Snel and
3 Wilson-Walker method: Schepers, 1995; Lindenburg, 2003, 2004; Liu and Janajreh,
Above a value of a = 0.2, Wilson-Walker method may 2012).
be used (Spera, 2009; Song, 2012) 2 2
⎛ Ω ⎞ ⎛c⎞
Cl , rot = Cl ,2 D + 3.1⎜ r ⎟ ⎜ ⎟ ( Cl , pot − Cl ,2 D ) (48)
⎧ 4a(1 − a ) F a ≤ ac ⎝ Vrel ⎠ ⎝ r ⎠
CT = ⎨ (44)
⎩4 ( ac + (1 − 2ac ) a ) F a > ac
2
or equivalently,
1⎡ 2
a= 2 + K (1 − 2ac ) ⎛ λ2 ⎞ ⎛ c ⎞
2 ⎢⎣ Cl ,rot = Cl ,2 D + 3.1 ⎜ r 2 ⎟ ⎜ ⎟ ( Cl , pot − Cl ,2 D ) (49)
(45) ⎝ 1 + λr ⎠ ⎝ r ⎠

( K (1 − 2ac ) + 2 )
2
− + 4 ( Kac2 − 1) ⎥⎦
where Cl,rot is the corrected lift coefficient to account for
where rotational effects, Cl,2D is the lift coefficient of non-rotating
airfoil and Cl,pot is the potential lift coefficient for the
4 F sin 2 φ non-rotating airfoil which is calculated by
K=
σCn
Cl , pot = 2π sin (α − α 0 )
4 Spera’s method:
and α0 is the angle of attack when lift coefficient is zero.
Another simple empirical relation that covers the full This correction has to be applied for locations up to 80%
range of a is given by Spera (2009). Spera gave an radius and for angles-of-attack up to 30°. For larger
empirical equation for a as a function of the thrust angles-of-attack the correction on the lift coefficient was
coefficient for the entire range that is given by, reduced linearly to zero at a 50° angle of-attack
(Lindenburg, 2004).
1
a= ( 0.27CT + 0.1CT3 ) (46) Liu and Janajreh (2012) showed that this model predicts
F well the power coefficient compared to experimental data
Other formulae for the correction factor are also given when applied in conjunction with the tip loss correction. A
by Anderson (1980), Garrad Hassan (Bossanyi, 1997) comparison of the performance of several stall delay models
and Johnson (1980) and can be found in Molenaar can be found in Breton (2008).
(2003). In addition to correction for rotational effects, there is
also the skewed rotor corrections. Liu and Janajreh (2012),
It can be inferred that Spera correction seems to better and Lindenburg (2003) can be referred to for more details.
match the experimental results than the other methods.
However, no results are reported in literature for the
application of Spera correction in the analysis of 8 Chord length distribution
turbine blade performance. Most of the work done used
Buhl to a great extent then Glauert and Wilson-Walker The chord length affects the performance of the wind
corrections. Detailed account of BEMT and correction turbine blade as well as the blade structural stiffness. The
factors can be found in Aagaard Madsen et al. (2012), maximum airfoil thickness is given in terms of percentage
Shen et al. (2005), Hansen and Johansen (2004), of the chord length. Several researchers have used several
Doessing (2009), Bak et al. (2006), Montgomerie formulas for calculating chord distribution (Hansen, 2008;
(2004), and Moriarty and Hansen (2005). Hau, 2006; Tenguria et al., 2010, 2011; Ingram, 2011).
A formula based on Betz ideal blade for
It should be noted that the power coefficient formula aerodynamically optimum chord distribution is given by
has to be modified to account for the loss factor as Hau (2006), and Schubel and Crossley (2012),
shown in equation (47),
142 Y.M. El-Okda

2πr 8 U ∞ where CLD is the design lift coefficient.


C (r )opt = (50)
B 9CL U rel A simpler form for the chord distribution without
considering wake rotation is given in Manwell et al. (2009),
where U∞ is the design flow velocity, and U rel = U ∞2 + Vw2
8πr ⎛ sin φ ⎞
which is the local resultant air velocity. Vw is the local c= ⎜ 3λ ⎟ (53)
angular speed. BCl ⎝ r ⎠
Another form of the above equation and a formula for For rotors with tip speed ratios of six to nine utilising airfoil
the flow angle φ based on the optimum value of the axial sections where drag and tip losses are negligible, Betz’s
induction factor a = 1/3 is given by Gundtoft (2009) which momentum theory gives a good approximation for the chord
are referred to as Betz formulas, distribution. In case of low tip speeds, high drag airfoil
2 sections and blade sections around the hub, the accuracy of
φ = tan −1 (51) this method is questionable.
3 λr
Blades designed in such a way for optimum power
and production yield larger chord and twist angle as one gets
closer to the blade root (Manwell et al., 2009). Although
16πR this design provides an insight into how the desired blade
c= (52)
1 should be, blade cost is high and its manufacturing is
9 BCLD
4 difficult with this method. In this case, different design
λ λ2r +
9 approach may be considered close to the blade root.

Figure 8 Axial induction factor correction methods for different values of loss coefficient (see online version for colours)
Design methods of horizontal axis wind turbine rotor blades 143

Including wake rotation, Gundtoft (2009) after Schmitz cos(2 x) = 1 − 2sin 2 ( x)


derived equations for the optimal flow angle and the chord
length distribution starting with momentum analysis, It can be shown that Schmitz equation, equation (55), for
optimal chord distribution is exactly the same as equation
2 1 (60).
φ = tan −1 (54)
3 λr Equation (59) is the same as equation (54). Schmitz
equations for optimum flow angle and optimum chord
and length gained significant attention in literature and was used
16πr 2 ⎛ 1 −1 1 ⎞ by many authors. These equations are based on ideal blade
c= sin ⎜ tan (55)
BCL ⎝3 λr ⎟⎠ shape derived with wake rotation or swirl but with neglected
drag and zero tip losses.
Equation (55) can be rewritten as In his book Wind Energy Explained, Manwell et al.
(2009) derived a formula for ideal rotor without wake
16πr 2 ⎛ φ ⎞ rotation given by equation (61) below.
c= sin ⎜ ⎟
BCL ⎝ 2⎠
8πr sin φ
Using equations (21) and (25) and inserting them into c= (61)
3BCL λr
equation (29) for the power coefficient (Duran, 2005;
Manwell et al., 2009) and after some algebra, we get Hansen (2008) reported the following relation for the
λ
optimum chord distribution which is a function of the
8 optimum values for a and a ′
CP =
λ2 ∫ sin φ ( cos φ − λ sin φ )( sin φ + λ cos φ )
2
r r
λh (56) 8Raλr sin 2 φ
C (r )opt = (62)
⎡ CD ⎤ (1 − a) BCn λ
⎢1 − D cot φ ⎥ dλr
⎣ L ⎦
which can be reduced to
Taking the partial derivative with respect to the flow angle
8ar sin 2 φ
while neglecting drag and equating it with zero, we can get C (r )opt = (63)
an optimum value for the flow angle φ and the chord c. (1 − a) BCn

∂ where a is the optimum value for the axial induction factor.


⎡sin 2 φ ( cos φ − λr sin φ )( sin φ + λr cos φ ) ⎤⎦ = 0 (57)
∂φ ⎣ This equation is identically equation (21). Description of
how a can be calculated is provided later in this section.
This yields, Where

sin φ (2 cos φ − 1) Cn = CL cos φ + CD sin φ


λr = (58)
(1 − cos φ )(2 cos φ + 1) In his book, Innovation in Wind Turbine Design, Jamieson
Expanding the numerator and the denominator and using the (2011) arrives at a definition for an optimum chord, where
following trigonometric relations successively, Cld is the design lift coefficient of the foil (usually lift at
which peak lift/drag ratio occurs).
sin 2 x = 2sin x cos x
16πR 2
cos 2 x = 2 cos 2 x − 1 c(r ) = (64)
9 BCld λ2 r
⎛ x+ y⎞ ⎛ x− y⎞
sin x − sin y = 2 cos ⎜ ⎟ sin ⎜ ⎟
⎝ 2 ⎠ ⎝ 2 ⎠ This formula for the chord length is used in conjunction
with the optimum flow angle given by equation (59).
⎛ x+ y⎞ ⎛ y−x⎞
cos x − cos y = 2sin ⎜ ⎟ sin ⎜ ⎟ Tenguria et al. (2010) calculated the chord length
⎝ 2 ⎠ ⎝ 2 ⎠ distribution with the same equation taken after Ryu and Kim
which reveals, (2004).
Equation (64) was used by Sohn et al. (n.d.) in the
2 ⎛1⎞ design of a 750 kW variable speed operation on the pitch
φ = tan −1 ⎜ ⎟ (59)
3 ⎝ λr ⎠ controlled direct drive wind turbine generator. The chord
length computed with this formula is optimum for
and aerodynamic performance, but not practical for fabrication
8πr and connection to the rotor hub as the blade becomes very
c= (1 − cos φ ) (60) wide at the root. They adopted a compromised chord
BCL
between aerodynamic performance and fabrication easiness
Considering the following trigonometric function, in which the inner part (up to 50%) is modified as a straight
line instead of a hyperbola as shown in Figure 9.
144 Y.M. El-Okda

Figure 9 Chord along blade of KBP-750D (see online version where the 1/3 factor represents the value for the axial
for colours) induction factor at the optimised performance based on Betz
theory. The 0.8 in equation (68) refers to the 80% point, i.e.,
at r = 0.8 R.
The blade design with equation (68) gives results very
close to that presented in Figure 9. It should be noted that
this design not only simplifies the plan-form but also
removes a lot of material close to the root. Burton et al.
(2001) showed that the required span-wise variation of lift
coefficient, Cl, for optimal operation is determined by
8 1
Cl =
9 Bcλ ⎛ 1 ⎞2 ⎡ 2 ⎤
2

⎜ 1 − ⎟ + λ2r ⎢1 + 2 ⎥
2π ⎝ 3 ⎠ ⎣ 9 λr ⎦
Source: Sohn et al. (n.d.) Raju (2011) and Edon (2007) used equation (70) to
Based on Glauert’s optimum conditions, Maalawi (2011) calculate the chord distribution. The formula was given
arrived at an analytical formula for the theoretical optimum without a proof.
chord distribution, R ⎛ 5.957 3.1 0.5433 0.2917 ⎞
c= 1.868 + − 2 + − (70)
8πrF sin φ BCL λ ⎜⎝ λr λr λ3r λ4r ⎟⎠
c= (65)
⎛ λr + tan φ ⎞ It should be noted that in most of the methods followed to
BCL ⎜ − CD / C L ⎟
⎝ 1 − λr tan φ ⎠ obtain an optimum blade geometry, the effect of drag is
usually ignored since it complicates the derivation
This formula includes tip losses, F, and drag is not omitted.
significantly. Burton et al. (2001) showed that omitting drag
Apparently, equation (65) is derived in the same way as
from the derivation of optimum geometric parameters has
equation (38) but without omitting drag. Kulunk (2011)
little effect on geometry of the blade. However, drag should
gave the same formula, equation (66), but with neglecting
not be omitted from torque and power calculations.
drag. This formula was also reported and used by several
authors (Kulunka and Yilmazb, 2009; Rathore and Ahmed,
2012).
9 Discussion
8πrF sin φopt ( cos φopt − λr sin φopt ) Figure 1 shows the flow angle distributions calculated by
c(r ) = (66)
BCld ( sin φopt + λr cos φopt ) Betz and Schmitz formulas for a three bladed rotor running
at tip speed ratio of 5. It can be seen that the two curves are
The optimum flow angle φopt is obtained from equation (59) matching until r/R = 0.4. Schmitz method seems to produce
and Cld, the design lift coefficient, is chosen at the slightly smaller flow angle than that produced by Betz
maximum Cl/Cd ratio. method as we approach the rotor hub which is usually
Ingram (2011) used a simplified version for the chord designed with a radius of about 20% of the whole rotor
distribution, equation (67), based on ideal blade theory in radius as indicated in the figure. However, there is relatively
the design of HAWT accompanied with equation (59) for large difference between the flow angle calculated by
flow angle calculation. Schmitz method and Burton method for the blade inboard.
8πr cos φ Figure 11 shows the chord distribution obtained via five
c(r ) = (67) different methods. It is apparent in the figure that equation
3Bλr
(70) resulted in about 70% larger chord at the tip compared
A direct method to calculate linear distribution of the chord to the other methods. This seems to be a large departure
and the twist angle is given by Burton et al. (2001) in from the other methods presented in the figure that are
equations (68) and (69), derived considering optimum solutions. It also produced
chord of large size at the blade root. Although it is
c 8 ⎛ λ ⎞ 2π consistent for the outboard with the other methods presented
= ⎜ 2− (68)
R 9 λ0.8 ⎝ λ0.8 ⎟⎠ Cl λB in the figure, equation (64) seems to produce very large
chords towards the blade root which complicates
⎛⎛ 1⎞ ⎞ manufacturing and raises the cost. Betz and Schmitz
⎜ ⎜1 − 3 ⎟ 1 ⎟
−1 ⎜ ⎝ ⎠ ⎟ methods produced similar results until r/R = 0.4. Schmitz
φ = tan (69)
⎜ 1+ 2 λr ⎟ and Burton methods resulted in appealing chord sizes at the
⎜ 3 λ2 ⎟ hub while maintaining optimality for the blade inboard.
⎝ r ⎠
Apparently, Schmitz formula outperforms the other
methods. This is because it is not an approximation like
Design methods of horizontal axis wind turbine rotor blades 145

Burton formula, yet it is derived based on optimum theory for a and on maximising the power output based on
including wake rotation. Additionally, it resulted in equation (29) for the power which is a function of a and a ′.
moderate chord length at the blade root. That renders it as a Following Hansen’s (2008) book, the following derivation
good solution for performance and manufacturability. On for optimum a and a ′ can be outlined.
the other hand, Burton formula not only simplifies the Using
plan-form but also removes a lot of material close to the
root and reduces manufacturing complexity (Burton et al., 1− a
tan φ = (71)
2001). (1 − a ′ ) λr

Figure 10 Flow angle calculated by Betz and Schmitz formulas and


for B = 3, λ = 5, α = 7° and CLD = 1.0 (see online a′
version for colours) tan φ = λr (72)
a
we obtain,
λ2r a ′ (1 + a ′ ) = a(1 − a ) (73)

Differentiating equation (29) w.r.t. a and equating the


differential with zero for maximum CP,
dCP
=0
da
yields,
da ′ a′
= (74)
da 1 − a
Differentiating equation (73) w.r.t. a yields,

(1 + 2a ′ ) da λ2r = 1 − 2a (75)
da
Figure 11 Comparison of chord distribution calculated by Combining equations (74) and (75) and using equation (73),
different methods for B = 3, λ = 5, α = 7° and we obtain an optimum relation between a and a ′,
CLD = 1.0 (see online version for colours)
1 − 3a
a′ = (76)
4a − 1
Combining equations (73) and (76) and eliminating a ′
yields a unique relation between a and λr which can be
solved for a,

(1 − a)
λr = (4a − 1) (77)
(1 − 3a)

Dixon and Hall (2010) reached the same equation in his


analysis for optimum rotor on his optimisation analysis
based on the maximisation of for the power calculated via
equation (29). Equation (77) can be rearranged to produce a
polynomial equation for the optimum value of a,

16a 3 − 24a 2 + ( 9 − 3 λ2r ) a − 1 + λ2r = 0 (78)

It can be seen from Figures 10 and 11 that the optimised Now we are able to arrive at values for a, a ′ given λr.
formulas based on BEM analysis produce almost identical The following procedure is used to reach optimum blade
flow angle and chord distribution for the blade outboard. geometry for a certain radius with a given local speed ratio
This is about 50% of the blade length which is actually the λr:
portion of the blade that incorporates most of the turbine
power production. 1 Evaluating the optimum value for a by solving equation
Hansen (2008) proposed a procedure based on equation (78) which is a cubic equation. Only the root between
(63) which is essentially a rearrangement of equation (21) 1/4 and 1/3 is to be selected. This range is inferred from
equation (77).
146 Y.M. El-Okda

2 Evaluating the optimum value for a ′ using equation Figure 14 Comparison of chord distribution calculated by
(76). Schmitz method and equation (66) using Schmitz
optimum flow angle [equation (59)] for B = 3, λ = 5,
3 Using equation (71), we now can evaluate the optimum α = 7° and CLD = 1.0
flow angle.
4 Substituting by a and φ into equation (63) to finally
reach an optimum value for the chord length.
It is very interesting to find that the optimum flow angle and
the optimum chord length obtained by Hansen’s procedure
are identical to those obtained by Schmitz formulas
[equations (59) and (60)]. Figures 12 and 13 show
comparisons of Schmitz and Hansen methods with respect
to the chord distribution and the flow angle distribution
respectively. However, Schmitz method is much simpler
and straight forward to implement than Hansen’s method.

Figure 12 Comparison of chord distribution calculated by


Schmitz method and Hansen method [equation (63)],
for B = 3, λ = 5, α = 7° and CLD = 1.0

Figure 15 Comparison of chord distribution calculated by


Schmitz method and equation (65) using Schmitz
optimum flow angle for B = 3, λ = 5, α = 7° and
CLD = 1.0, and CL/CD = 10

Figure 13 Comparison of flow angle distribution calculated by


Schmitz method and Hansen method, for B = 3, λ = 5,
α = 7° and CLD = 1.0

Dixon and Hall (2010) performed similar analysis to that


done by Hansen that leads to the same results; however,
their algorithm for the calculation of the optimum geometry
is a bit convoluted.
Another interesting finding is that by substituting with
optimum flow angle obtained via Schmitz method or
Hansen’s method into equation (66) we reach identical
results for the chord length to those obtained by Schmitz
and Hansen. Figure 14 shows a comparison of the chord
length distribution obtained by equation (66) with that
obtained by Schmitz method.
The inclusion of drag in the calculation of the chord
seems to be totally insignificant even for very low values of
glide ratio CL/CD. Figure 15 shows a comparison of Schmitz
Design methods of horizontal axis wind turbine rotor blades 147

method with Maalawi’s formula [equation (65)] at a glide speed ratio of 6 to 7 results in the best operational
ratio of as low as 10 which is one order of magnitude less performance. Additionally, for much higher tip speed ratio,
than actual values encountered in practice. the rotor appears to the wind as a large flat disk which
subjects the turbine to excessively larger stresses which
could lead to catastrophic failure in addition to blocking air
10 Rotor efficiency from passing through the turbine.
It is favourable to express the performance of the rotor Table 1 Suggested blade number, B, for different tip speed
along a scale of unity. Considering propulsive efficiency, ratios λ
marine engineers call it the Froude efficiency after William
Froude (1810–1879) who first used it. It is defined as the λ B
ratio of useful power output to the rate of energy input 1 8–24
(Tenguria et al., 2011). In the same manner, the wind 2 6–12
turbine rotor efficiency can be expressed in terms of Froude
3 3–6
efficiency ηFroude which is defined as ratio of the power
extracted to the available power. 4 3–4
>4 1–3
Power extracted Power
ηFroude = = Source: Manwell et al. (2009)
Power available 1 ρU 3 AC
∞ pmax
2 Ragheb and Ragheb (2011) showed that a theoretically
optimum tip speed ratio is approximately estimated as
Cp Cp λopt = 2π/B, where B is the number of rotor blades. They
ηFroude = = (79)
C pmax 16 / 27 added that with proper airfoil design, the optimal tip speed
ratio is around 25% to 30% of this estimated value. This
renders an optimum value for three-bladed rotor to be in the
11 Design methodology range of 5.24–5.45.
All necessary input parameters, U∞, R, λ and B, should
The aerodynamic design of rotor blades means the be set initially. According to flow conditions, more
determination of the geometric parameters, essentially the specifically roughness producing contaminants, species or
chord and twist angle distributions along the blade whose particles, as well as the type of operation, pitch controlled or
power, length and tip-speed ratio are already determined a stall regulated, suitable airfoil or airfoils are to be chosen for
priori to a great extent during the design process. the design. It should be noted that airfoils with thick nose
First, we decide upon design variables or make rough are less sensitive to surface roughness. However, they may
estimation of them using the formula for the power exhibit low lift coefficient in case of small wind turbine
coefficient CP and tip-speed ratio λ. inboard due to separation at low Reynolds numbers. The
P choice of the design angle of attack may be done as follows.
CP = The one that should be tried first is choosing the angle of
1
ρU ∞3 πR 2 attack corresponding to the maximum glide ratio (CL/CD).
2
For stall regulated wind turbines, the design angle of attack
ΩR may be around 80% of the stall angle of attack especially as
λ=
U∞ the tip is approached. The effect of Reynolds number for
both clean and dirty conditions as well as for different
From the design load, power required and angular speed relative flow velocity at different radii of the blade may be
requirement can be figured out. Assuming an initial considered here. Some data on smooth and rough airfoils
reasonable value for CP, and with the knowledge of site- can be found in Abbott and Von Doenhoff (1959). In his
specific wind speed, an initial estimate of the turbine radius thesis, Cencelli (2006), he interpolated data for CL and CD
R can be calculated. Now a reasonable value for the tip from data obtained via XFOIL through an interface
speed ratio and the number of blades are assigned as shown MATLAB code. Two XFOIL-MATLAB interface codes
in Table 1 after Manwell et al. (2009). can be found on mathworks.com website. A blade can be
For aesthetic considerations, the number of blades of divided into several sections each with a suitable airfoil.
three is usually employed for tip speed ratios of 4 or more. The last part of the blade close to the tip is the least power
Theoretically, the performance of the turbine improves as producing section during normal or rated operation. Hence,
the tip speed ratio is increased. If a rotor rotates too slowly, for stall regulated wind turbines, it is a good practice to
it allows too much wind to pass through undisturbed, and design at maximum glide ratio for about 80% of the blade
thus does not extract as much as energy as it could. length while maintaining the rest of the blade near stall to
However, as the speed of rotation increases overly, a blade help slow down the turbine at high winds. This strategy
will fall in the turbulent wake of the other blade(s) leading maximises the output from the rotor for the design wind
to significant degradation of the turbine performance. speed. Adding to that, for structural stiffness, thick airfoils
Hence, for high speed wind turbines, it is found that a tip are needed at the inboard of the blade.
148 Y.M. El-Okda

After the above discussion and comparison of the the flow angle rather than an optimisation of the chord
methods for calculating the flow angle and the chord length, length. The optimum design for maximum power reveals
it seems that Schmitz formulas are best candidates in this that the optimum chord distribution is dependent on the
respect. An alternative way to consider is to use a optimised flow angle at a certain radius. The local optimum
combination of design methods such as the work done by flow angle depends on the local tip speed ratio. The optimal
Sohn et al. (n.d,) shown in Figure 9. They used a formula chord length can be evaluated via equations that results
for optimum chord for the outboard where most of the from the BEM analysis that were obtained before any
power is extracted. On the other hand, they used linear optimisation analysis. It was found that most of the
chord distribution for the rest of the blade to reduce the size optimised formulas based on BEM analysis for the flow
of the blade at the root. An almost similar approach is to use angle and chord distributions are almost identical for the
equation (68), Burton method, to obtain an optimal linear blade outboard, about 50% of the blade length. Schmitz
distribution of the chord along the span of the blade. method for optimum flow angle and chord distribution is
It should be noted that optimal blade design is found to be the simplest and most straight forward result
influenced by the mode of operation of the turbine, that is from BEM optimisation analysis. The blade geometry
fixed rotational speed or variable rotational speed, and the obtained from Schmitz method is not only an optimum one,
wind distribution at the intended site. Hence, such factor but it produces a good compromise between performance
should be taken into account for more efficient design. and stiffness at the inboard of the blade.
The generated lift by an airfoil section is a function of Axial induction factor for optimum design assumes
the angle of attack. The inflow angle of the air stream is values between 1/4 and 1/3. However, these values are for
dependent on the rotational speed at a specified radius and optimum performance and vary considerably for off-design
the wind speed. The angle of twist required is dependent on performance.
tip speed ratio and the desired airfoil angle of attack. The The performance of several methods for the correction
angle of attack is obtained from the airfoil lift curve of the axial induction factor is investigated under different
corresponding to the selected lift, usually maximum lift or values of the loss factor, F. Out of these methods, Buhl, and
maximum CL/CD. The angle of twist, the angle between the Wilson-Walker seem to perform well and have proven to be
airfoil chord line and the rotor plane, is thus calculated working.
based on the angle of attack α and the flow angle φ,
Figure 4
References
θ = φ −α (80)
Aagaard Madsen, H., Riziotis, V., Zahle, F., Hansen, M.O.L.,
Assuming the same angle of attack to be prescribed for each Snel, H., Grasso, F., Larsen, T.J., Politis, E. and
airfoil section along the blade, i.e., neglecting the Rasmussen, F. (2012) ‘Blade element momentum modeling
dependence of the angle of attack on Reynolds number, the of inflow with shear in comparison with advanced model
results’, Wind Energy, Vol. 15, No. 1, pp.63–81.
smaller the flow angle implies less twist of the blade
(Cuerva, 2009). Referring to Figure 10, it can be seen that Abbott, I. and Von Doenhoff, A. (1959) Theory of Wing Sections,
Dover Publications Inc, New York.
Schmitz method is expected to produce larger twist towards
the root of the blade than that produced by Burton method. Anderson, M. (1980) ‘A vortex-wake analysis of a horizontal axis
wind turbine and a comparison with modified blade element
Large twist is required at the inboard to produce the theory’, in Proceedings of 3rd International Symposium on
necessary starting torque yet it is more complex in Wind Energy Systems, BHRA Fluid Engineering,
manufacturing. Copenhagen, pp.357–374.
For large-scale wind turbines, where the airfoil runs at Bak, C., Johansen, J. and Andersen, P.B. (2006) ‘Three
high Reynolds number conditions, the angle of attack at dimensional corrections of airfoil characteristics based on
maximum glide ratio is considered constant. However, for pressure distributions’, in EWEC2006, Athens, Greece.
small wind turbines, the local Reynolds number may Bayoumi, E.H.E. (2014a) ‘Matrix converter for static synchronous
become small at the inboard such that the characteristic series compensator using cooperative bacteria foraging
curve is no longer the same at all locations along the span of optimization’, Int. J. of Industrial Electronics and Drives.
the blade. In this case, the Reynolds number dependence of Bayoumi, E.H.E. (2014b) ‘Stator resistance estimator for direct
the lift coefficient and the glide ratio should be considered. torque control of permanent magnet synchronous motor drive
systems using multi-resolution analysis wavelet’, Int. J. of
Industrial Electronics and Drives, Vol. 1, No. 3, pp.191–201.
Bossanyi, E. (1997) Bladed for Windows – Theory Manual,
12 Conclusions Garrad Hassan & Partners Limited, No. 3, Document
In the paper, the BEM theory is investigated. A detailed No. 282/BR/009, Bristol, England, 54p.
BEM analysis for HAWTs is outlined. A comprehensive Breton, S-P. (2008) Study of the Stall Delay Phenomenon and of
review of different methods used to calculate the chord Wind Turbine Blade Dynamics Using Numerical Approaches
length distribution for high-speed HAWT is presented. The and NREL’s Wind Tunnel Tests, Doctoral thesis, Norwegian
University of Science and Technology, June.
optimisation of the blade geometry based on BEM theory is
investigated and is found to be primarily an optimisation of
Design methods of horizontal axis wind turbine rotor blades 149

Buhl, M.L.J. (2005) A New Empirical Relationship between Thrust Liu, S. and Janajreh, I. (2012) ‘Development and application of an
Coefficient and Induction Factor for the Turbulent Windmill improved blade element momentum method model on
State, NREL, USA. horizontal axis wind turbines’, International Journal of
Burton, T., Sharp, D., Jenkins, N. and Bossanyi, E. (2001) Wind Energy and Environmental Engineering, Vol. 3, No. 1,
Energy Handbook, John Wiley & Sons, New York. R 10.1186/2251-6832-3-30 [online] http://www.journal-ijeee.
com/content/3/1/30.
Cencelli, N.A. (2006) Aerodynamic Optimization of Small-Scale
Wind Turbine Blade for Low Wind Speed Conditions, Master Maalawi, K. (2011) edited by D.I. Al-Bahadly, Special Issues on
thesis, The University of Stellenbosch. Design Optimization of Wind Turbine Structures, InTech,
ISBN: 978-953-307-221-0 [online] http://www.intechopen.
Cuerva, A. (2009) Reynolds Number Implications on the com/books/wind-turbines/special-issues-on-design-
Determination of Wind Turbine Optimum Rotors, European optimization-of-wind-turbinestructures (accessed
Wind Energy Association, 16–19 March 2009. December 2013).
Dixon, S.L. and Hall, C.A. (2010) Fluid Mechanics and Manwell, J.F., McGowan, J.G. and Rogers, A.L. (2009) Wind
Thermodynamics of Turbomachinery, 6th ed., Elsevier, Energy Explained, Theory, Design and Application, 2nd ed.,
New York. Wiley, London.
Doessing, M. (2009) ‘A detailed investigation of the corrected Marshall, L. and Buhl, J. (2005) A New Empirical Relationship
BEM method and the potentialfor improving blade design’, in between Thrust Coefficient and Induction Factor for the
EWEC2009. Turbulent Windmill State, Technical Report, August,
Duran, S. (2005) Computer-Aided Design of Horizontal-Axis Wind NREL/TP-500-36834.
Turbine Blades, Master thesis, Middle East Technical McCosker, J. (2012) Design and Optimization of a Small Wind
University. Turbine, Master thesis, Rensselaer Polytechnic Institute.
Edon, M. (2007) Meter Wind Turbine Blade Design, June, Molenaar, D. (2003) Cost-effective Design and Operation of
Internship Report, Folkecenter for Renewable Energy. Variable Speed Wind Turbines, DUP Science, Delft.
Eggleston, D. and Stoddard, F. (1987) Wind Turbine Engineering Montgomerie, B. (2004) ‘Methods for root effects, tip effects and
Design, Van Nostrand Reinhold, New York, USA. extending the angle of attack range to 180 degree with
Glauert, H. (1926) The Analysis of Experimental Results in application to aerodynamics for blades on wind turbines and
Windmill Brake and Vortex Ring States of an Airscrew, propellers’, in FOIR1305, FOI, Swedish Defence Research
Reports and Memoranda, No. 1026, Aeronautical Research Agency.
Committee, London. Moriarty, P.J. and Hansen, A.C. (2005) AeroDyn Theory Manual,
Gundtoft, S. (2009) Wind Turbines, University of Aarhus, National Renewable Energy Laboratory, Golden, Colorado,
Denmark. USA.
Hansen, M.O. (2008) Aerodynamics of Wind Turbines, 2nd ed., Mosaad, M.I. and Salem, F. (2014) ‘Adaptive voltage regulation of
Earthscan, UK and USA. self excited induction generator using FACTS controllers’,
Hansen, M.O.L. and Johansen, J. (2004) ‘Tip studies using CFD Int. J. of Industrial Electronics and Drives, Vol. 1, No. 4,
and comparison with tip loss models’, Wind Energy, Vol. 7, pp.219–226.
No. 4, pp.343–356 Peters, D.A. and Modarres, R. (2013) ‘A compact closed form
Hau, E. (2006) Wind Turbines, Fundamentals, Technologies, solution for the optimum ideal wind turbines’, Wind Energy
Application, and Economics, Springer, New York. Journal, Vol. 17, No. 4, pp.589–603, DOI: 10.1002/we.1592.
Ingram, G. (2011) Wind Turbine Blade Analysis using the Blade Ragheb, M. and Ragheb, A.M. (2011) ‘Wind turbine theory – the
Element Momentum Method, 18 October [online] http://www. Betz equation and optimal rotor tip speed ratio’, in
dur.ac.uk/g.l.ingram/download/wind_turbine_design.pdf Carriveau, D.R. (Ed.): Fundamental and Advanced Topics in
(accessed October 2013). Wind Power, InTech [online] http://www.intechopen.com/
books/fundamental-and-advanced-topicsin-in-ind-power/wind
Jamieson, P. (2011) Innovation in Wind Technology, Wiley,
-turbines-theory-the-betz-equation-and-optimal-rotor-tip-
London.
speed-ratio (accessed September 2013).
Johnson, W. (1980) Helicopter Theory, Princeton University Press,
Raju, B.K. (2011) Design Optimization of a Wind Turbine Blade,
Princeton, NJ, USA.
Master thesis, The University of Texas at Arlington.
Kulunk, E. (2011) edited by D.R. Carriveau, Aerodynamics of
Rathore, A.S. and Ahmed, S. (2012) ‘Aerodynamic analyses of
Wind Turbines, Fundamental and Advanced Topics in
horizontal axis wind turbine by different blade airfoil
Wind Power [online] http://www.intechopen.com/books/
using computer program’, IOSR Journal of Engineering
fundamental-and-advanced-topics-in-wind-power/
(IOSRJEN), Vol. 2, No. 1, pp.118–123.
aerodynamics-of-windturbines (accessed October 2013).
Ryu, J. and Kim, D.H. (2004) ‘Blade design of a 360 KW direct
Kulunka, E. and Yilmazb, N. (2009) ‘HAWT rotor design and
drive wind turbine generator system’, Proceeding of ACCM4,
performance analysis’, Proceedings of the ASME 2009
Sydney.
3rd International Conference of Energy Sustainability,
San Francisco, California, USA, 19–23 July. Schubel, P.J. and Crossley, R.J. (2012) ‘Wind turbine blade
design’, Energies, Vol. 5, No. 9, pp.3425–3449.
Lindenburg, C. (2003) Investigation into Rotor Blade
Aerodynamics, Analysis of the Stationary Measurements on Shen, W.Z. et al. (2005) ‘Tip loss corrections for wind turbine
the UAE phase-VI Rotor in the NASA-Ames Wind Tunnel, computations’, Wind Energy, Vol. 8, No. 4, pp.457–475.
ECN [online] ftp://nrg-nl.com/pub/www/library/report/2003/ Snel, H. and Schepers, J. (1995) Joint Investigation of Dynamic
c03025.pdf (accessed December 2013). Inflow Effects and Implementation of an Engineering Method,
Lindenburg, C. (2004) ‘Modelling of rotational augmentation ECN, Petten.
based on engineering considerations and measurements’,
European Wind Energy Conference, London, pp.22–25.
150 Y.M. El-Okda

Sohn, Y.U., Chun, C.H., Kim, Y.C., Chung, C.W., Kim, Y.H., Tenguria, N., Mittal, N. and Ahmed, S. (2010) ‘Investigation of
Han, K.S. and Kim, D.H. (n.d.) Blade Design of a 750 kW blade performance of horizontal axis wind turbine based on
Direct-drive Wind Turbine Generator System, UNISON blade element momentum theory (BEMT) using NACA
Industrial Co., Ltd., 330-882, Jangsan-ri, Sooshin-myun airfoils’, International Journal of Engineering, Science and
Cheonan-si, Chung-nam, Korea R. Schmidt, M. Rees, Technology, Vol. 2, No. 12, pp.25–35.
Aerodyne Energy system GmbH, Provianthausstr. 9, D-24768 Tenguria, N., Mittal, N. and Ahmed, S. (2011) ‘Evaluation of
Rendsburg, Germany, pp.1–7 [online] http://www.unison.co. performance of horizontal axis wind turbine blades based on
kr/Include/Download.aspx?FILENM=Cyber_PR%5CTech optimal rotor theory’, Journal of Urban and Environmental
_Data%5CWWEC2004%20Blade%20design.pdf (accessed Engineering JUEE, Vol. 5, No. 1, pp.15–23.
October 2013).
Uddin, M.N. and Abera, F. (2009) ‘Efficiency optimisation based
Song, Q. (2012) Design, Fabrication, and Testing of a New Small speed control of IPMSM drive’, Int. J. of Industrial
Wind Turbine Blade, Master thesis, The University of Electronics and Drives, Vol. 1, No. 1, pp.34–41.
Guelph.
Yousefi-Talouki, M. and Ramezanpour, P. (2014) ‘Direct active
Spera, D.A. (2009) Wind Turbine Technology, Fundamental and reactive power control of 3-level NPC-fed DFIG
Concepts of Wind Turbine Engineering, 2nd ed., ASME, considering power ripple minimization’, Int. J. of Industrial
New York. Electronics and Drives, Vol. 1, No. 3, pp.174–190.
Stoddard, F.S. (1976) Discussion of Momentum Theory for
Windmills, Wind Energy Center Reports, University of
Massachusetts – Amherst.

View publication stats

You might also like