Small scale vertical axis wind turbine design

SMALL-SCALE VERTICAL AXIS WIND TURBINE DESIGN
Javier Castillo
Bachelor’s Thesis
December 2011
Degree program in Aeronautical Engineering
Tampereen ammattikorkeakoulu
Tampere University of Applied Sciences

ABSTRACT
Tampereen ammatikorkeakoulu
Tampere University of Aplied Sciences
Degree Programme in Aeronautical Engineering
CASTILLO, JAVIER: Small-scale vertical axis wind turbine design
Bachelor’s thesis 54 pages, appendixes 15 pages
December 2011
The thesis focuses on the design of a small vertical axis wind turbine rotor with
solid wood as a construction material. The aerodynamic analysis is performed
implementing a momentum based model on a mathematical computer pro-
gram.
A three bladed wind turbine is proposed as candidate for further prototype test-
ing after evaluating the effect of several parameters in turbine efficiency, torque
and acceleration.
The results obtained indicate that wood is a suitable material for rotor construc-
tion and a further development of the computer algorithm is needed in order to
improve the flow conditions simulation.
Key words: Wind turbine, giromill, vertical axis, small wind turbine

3
Table of Contents
1 INTRODUCTION.......................................................................................................................... 5
2 PRELIMINARY BUSINESS PLAN................................................................................................... 5
2.1 Product definition............................................................................................................... 6
2.2 Demand analysis ................................................................................................................. 6
2.3 Demand projection ............................................................................................................. 8
2.4 Offer analysis ...................................................................................................................... 8
3 WIND TURBINE TYPES.............................................................................................................. 11
3.1 Rotor axis orientation ....................................................................................................... 11
3.2 Lift or Drag type ................................................................................................................ 12
3.3 Power control.................................................................................................................... 12
3.4 Rotor speed: constant or variable .................................................................................... 13
3.5 Tower structure ................................................................................................................ 14
4 AERODYNAMIC ANALYSIS ........................................................................................................ 14
4.1 Wind turbine design parameters...................................................................................... 14
4.2 Blade performance initial estimations.............................................................................. 18
4.3 Aerodynamic model.......................................................................................................... 19
4.4 Model limitations.............................................................................................................. 28
4.5 Algorithm implementation ............................................................................................... 29
4.6. Algorithm validation ........................................................................................................ 30
5. ROTOR DESIGN........................................................................................................................ 33
5.1 Starting parameters .......................................................................................................... 33
5.2 Airfoil selection ................................................................................................................. 34
5.3 Design airspeed................................................................................................................. 35
5.4 Rotor dimensions.............................................................................................................. 37
5.5 Solidity............................................................................................................................... 39
5.6 Initial angle of attack......................................................................................................... 42
6. ACCELERATION ANALYSIS ....................................................................................................... 43
7. STRUCTURAL ANALYSIS........................................................................................................... 45
7.1 Rotor configuration........................................................................................................... 45
7.2 Allowable stress and selected materials........................................................................... 46
7.3 Load analysis..................................................................................................................... 49
8. FINAL DESIGN AND OPERATION ............................................................................................. 49
9. CONCLUSIONS......................................................................................................................... 52
10. BIBLIOGRAPHY ...................................................................................................................... 53

4
APPENDIXES................................................................................................................................ 55
SYMBOLS................................................................................................................................. 55
GLOSSARY................................................................................................................................ 56
WORLD ENERGY TRENDS........................................................................................................ 57
COMPETENCE ANALYSIS ......................................................................................................... 58
MATLAB ALGORITHM.............................................................................................................. 60
NACA0021 AIRFOIL DIMENSIONS ........................................................................................... 67
WOOD PROPERTIES AND STRUCTURAL STRENGTHS.............................................................. 68

5
1 INTRODUCTION
Wind power energy is getting more shares in the total energy production every
year, with wind turbines growing bigger and bigger at the rhythm of technology
does.
While we cannot expect nowadays a totally renewal energy supply in Europe,
(some estimations say that it is possible for 2030) there are places where ener-
gy grid has not even arrived and has not any plans for short or midterm time.
As we charge our mobile phone in any socket, there are people that has to illu-
minate their homes with kerosene lamps, students that cannot study when the
daylight extinct, medicines that cannot be stored in a freezer… There are other
people that are connected to the grid but still have energy cuts.
For these people, it seems to be cheaper to have a grid connection than to have
an independent energy source, reasons for that are that in an isolated installa-
tion the energy has to be chemically stored in batteries, with limited lifetime that
makes renewal costs comparable to those of a grid connection.
Furthermore, there are places which will not have electricity access, these plac-
es coincide with the poorest and isolated rural areas of developing countries,
and these people cannot afford the cost of a wind turbine, nor small nor bigger.
Maybe a free turbine design, that anyone with common workshop materials
could build, would make a difference in the living conditions of this people. The
purpose of this project is to provide an alternative to these people.
2 PRELIMINARY BUSINESS PLAN
A preliminary analysis of the situation of wind turbines used for similar applica-
tions is done analyzing the current manufacturers and market situation.

6
2.1 Product definition
The product is a vertical axis small-scale wind turbine, corresponding to the mi-
cro-generation classification of wind turbines, which is less than 1 kW. The aim
of the project is to make an affordable turbine, made locally with available mate-
rials in developing countries.
There are two different poles in micro-generation markets:
- On one hand we have manufacturers that had invested in a certification
program, use ultimate technologies (i.e. electronic controllers or carbon
fiber blades) and have a product oriented to developed country consum-
er.
- On the other hand there are some non-governmental organizations or in-
dividuals that build a model from wood, they also construct the generator,
but there is no certification process.
The product is intended to be somewhere in the middle of this situation, having
raw materials like wood and without a very sophisticated construction process,
the idea is to make it constructible with usual and non electrically driven work-
shop tools.
The IEC 61400 Wind turbine generator systems standards are the reference
used for the certification of wind turbines. The IEC61400-2 focuses in design
requirements for small wind turbines but they are not applicable to vertical-axis
wind turbines (Wood, 2011).
Technical standards should be considered in the design in order to ensure safe-
ty, reliability and durability of the wind turbine, but standards for vertical-axis
wind turbines have not been developed and a complete certification should be
done by accredited institutions, which is expensive (Cace, et al., 2007).
2.2 Demand analysis
1.6 billion People, 80% of them living in Sub-Saharan Africa and South Asia, do
not have access to electricity. This population uses kerosene and batteries for

7
their households and diesel generators for their businesses (Legros, Havet,
Bruce, & Bonjour, 2009).
Access to energy closely correlates with poverty: insufficient access to electrici-
ty or other energy sources also means that health services, access to clean wa-
ter and sanitation, and education all suffer (Cheikhrouhou, 2011).
With scarcity of resources, efforts on grid extension are focused to serve large
population cities while rural remote areas are considered to have too much
electrification costs to serve few people.
According to (winafrique, 2011):
- A small wind turbine production facility will cost less than a photovoltaic
factory of the same capacity.
- Wind turbines are very difficult to develop but not to manufacture.
- Local manufacturing, under license or through a joint venture, is neces-
sary if widespread utilization is to be pursued.
A small wind turbine design with very low regular maintenance, available spare
parts and with local users trained could meet the requirements needed for a
long operation in developing countries.
The following figure shows the geographical distribution of the areas that could
need the product.
Figure 1. Energy access percentage in developing countries (Legros, Havet, Bruce, & Bonjour, 2009)

8
2.3 Demand projection
Developing countries have potential demand for small wind systems because
they normally do not have electricity grid in rural areas. People need financial
assistance to buy small wind systems but government assistance is directed
nowadays to subsidizing extension of the grid and installing diesel generators
(American Wind Energy Association, 2002).
World electricity generating capacity is projected to increase a 46 percent by
2035, being the non-OECD countries the ones with a major contribution in total
capacity increase (71%). The figures on appendix 2, made with extracted data
from (U.S. Energy Information Administration, 2011), show the projected in-
crease in wind-powered installed generating capacity by 2035, which will be
2,75 times the actual generating capacity and will represent a 7.37% in the
world total generating capacity (nowadays represents the 3,95%).
The use of batteries and subsequent charging is becoming popular as a solu-
tion to the lack of electricity in small rural areas. Households in Sri Lanka use
automobile batteries to power radio and TV sets. The number of such batteries
is claimed to be 300,000. However, recharging the battery takes both time and
money (Dunnett).
2.4 Offer analysis
For the present offer analysis, a small wind turbine database has been used
(allsmallwindturbines.com, 2011) to compare those models that were more simi-
lar to the original idea, which is a vertical axis wind turbine with less than 1 kW
of rated power.
From this database we can extract that from the 558 models of wind turbines
listed, the 20% was of vertical axis type (VAWT from now on) and the other part
was of horizontal axis type (HAWT).

9
As price was not always available, some contacts with the wind turbine manu-
facturers have been made. A comparison between some wind turbines is in
competence analysis table (see appendix 4).
Another issue was the stated power, there are some studies made in Nether-
lands and UK that show quite difference between what is stated in the turbine
performances and what is really giving (Encraft, 2009).
Manufacturer’s characteristics
Approximately 250 companies in the world manufacture, or plan to manufac-
ture, small wind turbines. Of these (American Wind Energy Association, 2002):
• 95 (36%) are based in the United States
• At least 47 (12 U.S.) have begun to sell commercially
• 99% have fewer than 100 employees
• They are spread per 26 different countries
• The vast majority are in start-up phases and roughly half the world mar-
ket share is held by fewer than 10 U.S. manufacturers.
The Largest Manufacturers in 2009 were:
Table 1. Largest manufacturers considering sales in power installed (American Wind Energy Association, 2002)
Company Country kW Sold Worldwide
Southwest Windpower U.S. 11700
Northern Power Systems U.S. 9200
Proven Energy U.K. (Scotland) 3700
Wind Energy Solutions Netherlands 3700
Bergey WindPower Co. U.S. 2100
From this five, Bergey Windpower models are being used for rural areas electri-
fication projects in developing countries, some of their projects are:
- Installation in 1998 of three 1.5 kW and one 7.5 kW turbine at a cyclone
shelter in Charduani, Bangladesh

10
- Installation in 1995 of a 1.5 kW turbine for water pumping to Assu, a re-
mote village of 500 people in Brazil.
- One problem, which is common in these types of pioneering projects, is
that technical support is based 600 km away in a big city. With only a few
systems in the Assu area it is not economically feasible to maintain a
support structure closer to the projects, particularly since support is sel-
dom needed. The emphasis, therefore, is placed on making the systems
as reliable and maintenance-free as possible.
- Installation on 2005 of a 7.5 kW turbine for powering a cell phone provid-
er station, this resulted in collaboration with winafrique which supplies its
products.
- Installation on 1990 of a 10 kW for irrigation purposes in Timbuktu, Mali.
- Installation on 1992 of a 1.5 kW turbine for small-scale irrigation in Oe-
sao, on the island of Timor in East Indonesia.
Bergey Windpower turbines are not perceived as a competitor but as a model to
follow; its prices are still unaffordable for the vast majority of people living in re-
mote areas. Their models, with few moving parts and very well proven and de-
signed had achieved a reputation of dependable and reliable in these applica-
tions.
Another existing offer is that provided by some organizations like V3power in
the UK who make handmade wind turbines courses along Europe, they don’t
seem to have a direct contact with developing countries or NGOs. They use a
model designed by Hugh Piggot, who has been involved in projects in Peru and
Sri Lanka among others and has written a “recipe book” to build a wind turbine
with common workshop tools and materials.
V3power also sell Hugh Piggot’s model prices ranging 1293 $ the 1.2 m diame-
ter turbine to the 2716 $ the 4.2 m rotor, power rating not mentioned. These
prices, approximately half value of Bergey Wind Turbines make them more af-
fordable.
This idea of making a wind turbine construable in developing countries without
advanced tools and skills coincides with the idea of the project. However it
seems to be usual to found cracked or faulty wood blades after storms.

11
To sum up, the position of the project is intended to be between the reliable,
well-proven market turbines and the DIY (Do It Yourself) not standardized mod-
els. The intention of the project is to have a viable alternative to those models.
3 WIND TURBINE TYPES
Wind turbines can be classified in a first approximation according to its rotor
axis orientation and the type of aerodynamic forces used to take energy from
wind. There are several other features like power rating, dimensions, number of
blades, power control, etc. that are discussed further along the design process
and can also be used to classify the turbines in more specific categories.
3.1 Rotor axis orientation
The major classification of wind turbines is related to the rotating axis position in
respect to the wind; care should be taken to avoid confusion with the plane of
rotation:
• Horizontal Axis Wind Turbines (HAWT): the rotational axis of this turbine
must be oriented parallel to the wind in order to produce power. Numer-
ous sources claim a major efficiency per same swept area and the major-
ity of wind turbines are of this type.
• Vertical Axis Wind Turbines (VAWT): the rotational axis is perpendicular
to the wind direction or the mounting surface. The main advantage is that
the generator is on ground level so they are more accessible and they
don’t need a yaw system. Because of its proximity to ground, wind
speeds available are lower.

12
One interesting advantage of VAWTs is that blades can have a constant shape
along their length and, unlike HAWTs, there is no need in twisting the blade as
every section of the blade is subjected to the same wind speed. This allows an
easier design, fabrication and replication of the blade which can influence in a
cost reduction and is one of the main reasons to design the wind turbine with
this rotor configuration.
3.2 Lift or Drag type
There are two ways of extracting the energy from the wind depending on the
main aerodynamic forces used:
• The drag type takes less energy from the wind but has a higher torque
and is used for mechanical applications as pumping water. The most rep-
resentative model of drag-type VAWTs is the Savonius.
• The lift type uses an aerodynamic airfoil to create a lift force, they can
move quicker than the wind flow. This kind of windmills is used for the
generation of electricity. The most representative model of a lift-type
VAWT is the Darrieus turbine; its blades have a troposkien shape which
is appropriate for standing high centrifugal forces.
The design idea is to make a lift type turbine, with straight blades instead of
curved. This kind of device is also called giromill and its power coefficient can
be higher than the maximum possible efficiency of a drag type turbine, like Sa-
vonius (Claessens, 2006).
3.3 Power control
The ability to control the power output as well as stand worst meteorological
conditions can be achieved passively with stall control or active with pitch con-
trol; the majority of small wind turbines use stall control in addition to some
blocking device, the NACA44 series airfoils are considered particularly useful
for this type of control (Jha, 2011, p. 263) but they will not be used for the cur-

13
rent design because there is not available data about its behavior depending on
the angle of attack.
The Stall control is a passive control which provides overspeed protection, the
generator is used to electrically control the turbine rotational speed, when the
rated conditions are surpassed (i.e. in a windy storm).
This kind of control mechanism should be the one appropriate for a simple, reli-
able and low maintenance design because it doesn’t add any moving part to the
structure nor its associated bearings, which can affect the final cost of the tur-
bine.
The Variable pitch control is achieved by varying actively the pitch (orientation)
of the blade in such a way it can adapt to the current airflow speed to take the
maximum energy from the wind, its efficiency is increased compared to a fixed
pitch windmill and it increases the maximum power that can be extracted from
wind. Its main drawback is the need to have a control system to allow these
pitch movements. For complex models its worth to have some electronics taking
care of that purpose but it is not the case of our design.
3.4 Rotor speed: constant or variable
A constant rotor speed maintains the same rotational speed while the windmill
is generating energy; they don’t need power electronics to adapt to grid fre-
quency which makes them cheaper; one of their drawbacks is that their opti-
mum efficiency is only achieved at the design airspeed. A stall-regulated wind
turbine falls into this category as it maintains constant rpm once the rated rota-
tional speed is achieved.
A variable speed rotor tries to achieve the optimum rotational speed for each
wind speed, maintaining constant the optimum tip speed ratio will ensure opti-
mum efficiency at different airspeeds (the concept of tip speed ratio will be de-
veloped in section 4.1.3). From a structural point of view, letting the rotor to

14
change its speed reduces the load supported by the wind turbine in presence of
gusts or sudden starts. (Talayero & Telmo, 2008).
3.5 Tower structure
There are two main types of structures, the tube type, which requires low main-
tenance but higher acquisition costs or the lattice type which is cheaper to build
but requires more maintenance. The chosen structure should be related with
what is easier and more available to achieve in the community or area of instal-
lation.
4 AERODYNAMIC ANALYSIS
In this section, the rotor design parameters are described as well as the model
used to calculate its aerodynamic performance. The model limitations are ex-
posed and the computer algorithm and its validation are presented.
4.1 Wind turbine design parameters
The wind turbine parameters considered in the design process are:
• Swept area
• Power and power coefficient
• Tip speed ratio
• Blade chord
• Number of blades
• Solidity
• Initial angle of attack

15
4.1.1 Swept area
The swept area is the section of air that encloses the turbine in its movement,
the shape of the swept area depends on the rotor configuration, this way the
swept area of an HAWT is circular shaped while for a straight-bladed vertical
axis wind turbine the swept area has a rectangular shape and is calculated us-
ing:
ܵ = 2ܴ ‫ܮ‬ [4.1]
where S is the swept area [m2
], R is the rotor radius [m], and L is the blade
length [m].
The swept area limits the volume of air passing by the turbine. The rotor con-
verts the energy contained in the wind in rotational movement so as bigger the
area, bigger power output in the same wind conditions.
4.1.2 Power and power coefficient
The power available from wind for a vertical axis wind turbine can be found from
the following formula:
ܲ‫ݓ‬ =
1
2
ρ ܵ ܸ௢
ଷ [4.2]
where Vo is the velocity of the wind [m/s] and ρ is the air density [kg/m3
], the
reference density used its standard sea level value (1.225 kg/m^3 at 15ºC), for
other values the source (Aerospaceweb.org, 2005) can be consulted. Note that
available power is dependent on the cube of the airspeed.
The power the turbine takes from wind is calculated using the power coefficient:
‫ܥ‬௣ =
‫݀݁ݎݑݐ݌ܽܥ‬ ݉݁ܿℎ݈ܽ݊݅ܿܽ ‫ݎ݁ݓ݋݌‬ ܾ‫ݕ‬ ܾ݈ܽ݀݁‫ݏ‬
‫݈ܾ݈݁ܽ݅ܽݒܣ‬ ‫ݎ݁ݓ݋݌‬ ݅݊ ‫݀݊݅ݓ‬
[4.3]
Cp value represents the part of the total available power that is actually taken
from wind, which can be understood as its efficiency.

16
There is a theoretical limit in the efficiency of a wind turbine determined by the
deceleration the wind suffers when going across the turbine. For HAWT, the
limit is 19/27 (59.3%) and is called Lanchester-Betz limit (Tong, 2010, p. 22).
For VAWT, the limit is 16/25 (64%) (Paraschivoiu I. , 2002, p. 148).
These limits come from the actuator disk momentum theory which assumes
steady, inviscid and without swirl flow. Making an analysis of data from market
small VAWT (see Table 7 from appendix 4), the value of maximum power coef-
ficient has been found to be usually ranging between 0.15 and 0.22.
This power coefficient only considers the mechanical energy converted directly
from wind energy; it does not consider the mechanical-into-electrical energy
conversion, which involves other parameters like the generator efficiency.
4.1.3 Tip Speed Ratio
The power coefficient is strongly dependent on tip speed ratio, defined as the
ratio between the tangential speed at blade tip and the actual wind speed.
ܴܶܵ =
ܶܽ݊݃݁݊‫݈ܽ݅ݐ‬ ‫݀݁݁݌ݏ‬ ܽ‫ݐ‬ ‫ݐ‬ℎ݁ ܾ݈ܽ݀݁ ‫݌݅ݐ‬
‫݈ܽݑݐܿܣ‬ ‫݀݊݅ݓ‬ ‫݀݁݁݌ݏ‬
=
ܴ ߱
ܸ‫݋‬
[4.4]
where ω is the angular speed [rad/s], R the rotor radius [m] and Vo the ambient
wind speed [m/s]. Each rotor design has an optimal tip speed ratio at which the
maximum power extraction is achieved. This optimal TSR presents a variation
depending on ambient wind speed as will be analyzed on section 5.3.
4.1.4 Blade chord
The chord is the length between leading edge and trailing edge of the blade
profile. The blade thickness and shape is determined by the airfoil used, in this

17
case it will be a NACA airfoil, where the blade curvature and maximum thick-
ness are defined as percentage of the chord.
4.1.5 Number of blades
The number of blades has a direct effect in the smoothness of rotor operation
as they can compensate cycled aerodynamic loads. For easiness of building,
four and three blades have been contemplated.
As will be illustrated in section 4.3.1, the calculations used to evaluate the pow-
er coefficient of the turbine do not consider the wake turbulence effects of the
blade, which affect the performance of adjacent blades. The effect of the num-
ber of blades will be analyzed further on section 5.5.
4.1.6 Solidity
The solidity σ is defined as the ratio between the total blade area and the pro-
jected turbine area (Tullis, Fiedler, McLaren, & Ziada). It is an important non-
dimensional parameter which affects self-starting capabilities and for straight
bladed VAWTs is calculated with (Claessens, 2006):
σ =
N c
R
[4.5]
where N is the number of blades, c is the blade chord, L is the blade length and
S is the swept area, it is considered that each blade sweeps the area twice.
This formula is not applicable for HAWT as they have different shape of swept
area.
Solidity determines when the assumptions of the momentum models are appli-
cable, and only when using high σ ≥ 0.4 a self starting turbine is achieved
(Tong, 2010).

18
4.1.7 Initial angle of attack
The initial angle of attack is the angle the blade has regarding its trajectory,
considering negative the angle that locates the blade’s leading edge inside the
circumference described by the blade path. The effect of the initial angle of at-
tack in overall performance will be discussed further on section 5.6.
4.2 Blade performance initial estimations
An initial estimation had to be made to provide guidance in the design process,
in order to do that, the extracted data from similar market competitors (see Ta-
ble 8 from appendix 4) has been analyzed in order to find out reasonable val-
ues.
With all this data, an estimation of the required dimensions as well as a rota-
tional speed has been done, as it is not possible to approximate the average
wind speed of the whole developing countries, the average wind speed of Tam-
pere has been chosen as this is the location where the possible prototype tests
would be conducted. This value has been estimated to be 6 m/s looking at the
data provided by (Windfinder, 2011).
Also, it is interesting to have a low value of design wind speed as it will ensure
more availability of electrical energy supply with averaged and frequent air-
speeds. The turbines found on the market are rated by their maximum power
output; this could give false expectative on the final user as the rated wind con-
ditions can be higher than the average wind conditions.
Table 2 shows what were the initial estimated performance of the wind turbine,
the power coefficient and tip speed ratios had been estimated analyzing the
data taken from similar market wind turbines and making an average.

19
Table 2. Power performance and rotational speed initial estimations for the design wind speed
Power Performance initial estimation
Design Parameters Calculated Parameters
Rotor radius (m) 1 Swept area (m2
) 4
Blade Lenght (m) 2 Solidity 0,8
Blade Chord (m) 0,2
Power Coefficient 0,17 Rated blade speed (m/s) 12
Tip speed ratio 2 Actual Rotational speed (rad/s) 12
Number of blades 4 Actual Rotational speed (rpm) 114,59
Air constants
Air density (Kg/m^3) 1,225
Wind speed (m/s) 6 Power Available from wind (W) 529,2
Conversion factors Power output (W) 89,964
rad/s --> rpm 9,54
4.3 Aerodynamic model
In order to model the performance of a vertical-axis wind turbine there are four
main approaches (Cooper, 2010):
- Momentum models
- Vortex models
- Local circulation models
- Viscous models – Computational Fluid Dynamics (CFD)
Each model has its advantages and drawbacks, the main advantage of momen-
tum models is that their computer time needed is said to be much less than for
any other approach (Spera, 2009).
Computational fluid dynamics was considered in first place but the time needed
for mesh generation in three dimensional analyses was supposed to be too
large for the purposes of the present thesis.
The model chosen for the aerodynamic analysis is the double-multiple stream-
tube with variable interference factor, often abbreviated DMSV, which is en-

20
closed in the momentum models category and is based in the conservation of
momentum principle, which can be derived from the Newton’s second law of
motion. It has been used successfully to predict overall torque and thrust loads
on Darrieus rotors.
Single Streamtube Multiple Streamtube
Double Streamtube Double-Multiple Streamtube
Figure 2. Streamtube models used for vertical-axis wind turbines analyisis, adapted from Spera, 2009, p. 330.
This analytical model was developed by I. Paraschivoiu for determining aerody-
namic blade loads and rotor performance on the Darrieus-type wind turbine with
straight and curved blades (Paraschivoiu I. , 2002).
The model uses the actuator disk theory, two times, one for the upstream and
one for downstream part of the rotor. This theory represents the turbine as a
disc which creates a discontinuity of pressure in the stream tube of air flowing
through it. The pressure discontinuity causes a deceleration of the wind speed,
which results in an induced velocity.

21
4.3.1 Aerodynamic model explanation
The swept volume of the rotor is divided into adjacent, aerodynamically inde-
pendent, streamtubes; each one is identified by his middle θ angle, which is
defined as the angle between the direction of the free stream velocity and the
position of the streamtube in the rotor (see Figure 3).
The analysis of the flow conditions is made on each streamtube using a combi-
nation of the momentum and blade element theories, the former uses the con-
servation of the angular and linear momentum principle and the later divides the
blade in N elements and analyzes the forces on the blades (lift and drag) as a
function of blade shape. Further details on blade element momentum theory
can be found on Manwell, McGowan and Rogers, 2009, p. 117-121.
It is assumed that the wind velocity experiments a deceleration near the rotor, if
we represent the front and rear part of the turbine by two disks in series, the
velocity will be decelerated two times, one for the upstream and the other for
the downstream.
Unlike for the analysis of a troposkien shaped Darrieus turbine, it has been as-
sumed no vertical variation of the induced velocity as a straight vertical blade is
subjected to the same flow velocity along its length.

22
Figure 3. Division of the rotor swept area into streamtubes, definition of velocities and angles.
According to Paraschivoiu (2002) the induced velocities decrease in the axial
streamtube direction in the following manner:
The induced velocity in the upstream part of the rotor is:
ܸ‫ݑ‬ = Vo ܽ‫ݑ‬ [4.6]
where Vu is the upstream induced velocity, Vo is the free stream air velocity
and au is the upstream interference factor, which is less than 1 as the induced
velocity is less than the ambient velocity.
In the middle plane between the upstream and downstream there is an equili-
brium-induced velocity Ve:
ܸ݁ = Vo (2 ܽ‫ݑ‬ − 1) [4.7]
Finally, for the downstream part of the rotor, the corresponding induced velocity
is:
ܸ݀ = ܸ݁ ܽ݀ [4.8]
where Vd is the downstream induced velocity and ad is the downstream interfe-
rence factor which is smaller than the upstream interference factor.
Knowing the induced velocities all around the blade trajectory will allow us to
calculate the lift and drag forces associated to every blade position and then,

23
these forces can be related to the torque and power coefficient produced by the
wind turbine.
The resultant air velocity that the blade sees is dependent on the induced veloc-
ity and the local tip speed ratio:
ܹ‫ݑ‬ = ඥܸ‫ݑ‬ଶ ሾ(ܴܶܵ − sinଶ ߠ)ଶ + (cosଶ ߠ)ሿ [4.9]
where Wu is the resultant air velocity and TSR is the local tip speed ratio de-
fined as:
ܴܶܵ = ܴ
߱
ܸ‫ݑ‬
[4.10]
where R is the rotor radius and ω is the angular speed.
The resultant air velocity is used then to determine the local Reynolds number
of the blade.
ܴ݁௕ =
ܹ‫ݑ‬ c
K୴
[4.11]
Where Reb is the local Reynolds number, c is the blade chord and Kv is the ki-
nematic viscosity of the air, which has a reference value of 1.4607e-5 [m2
/s] for
an air temperature of 15ºC (Aerospaceweb.org, 2005).
The equation for the angle of attack of the blade has been taken from
Paraschivoiu (2002), the result has been demonstrated in order to ensure what
is considered the 0º of the azimuth angle (θ) and also what is considered a
negative angle of attack.
ߙ = arcsin ቆ
cos ߠ cos ߙ଴ − (ܺ − sin ߠ) sin ߙ଴
ඥሾ(ܴܶܵ − sinଶ ߠ)ଶ + (cosଶ ߠ)ሿ
ቇ [4.12]
where α is the angle of attack which is defined as the angle between the blade
chord and the resultant air velocity vector.
The previous equation allows the introduction of an initial angle of attack α0 dif-
ferent from 0º. It considers positive angles of attack those which have the rela-
tive wind speed vector outside the rotor as in Figure 3, which occurs in the
whole upstream half of the rotor.

24
Two different more intuitive formulas can be used to calculate the angle of at-
tack at each blade position.
ߙ௥௜௚௛௧௛௔௟௙ = arctan ൬
ܸ‫ݑ‬ cos ߠ
߱ ܴ + ܸ‫ݑ‬ sin ߠ
൰ [4.13]
ߙ௟௘௙௧௛௔௟௙ = arctan ൬
ܸ‫ݑ‬ cos ߠ
߱ ܴ − ܸ‫ݑ‬ sin ߠ
൰ [4.14]
The local Reynolds number and the angle of attack are then used to find the
corresponding lift and drag coefficients using double interpolation (i.e. one in-
terpolation for the Reynolds number and another for the angle of attack). The
data tables had been extracted from Sheldahl and Klimas (1981).
The variation of lift and drag coefficients depending on Reynolds number is only
patent while α<30º (see Figure 4), then all the curves converge. This is because
the data is a joint between experimental data until 30º and computed data with
an airfoil synthesizer code.
Figure 4. Lift coefficient variation with Reynolds number
0 20 40 60 80 100 120 140 160 180
-1
-0.5
0
0.5
1
1.5
Lift coefficient depending on angle of attack and Reynolds number
Angle of attack [deg]
Liftcoefficient
Reynolds = 1e4
Reynolds = 2e4
Reynolds = 4e4
Reynolds = 8e4
Reynolds = 16e4
Reynolds = 36e4
Reynolds = 70e4
Reynolds = 100e4
Reynolds = 200e4
Reynolds = 500e4

25
The existing step between the two types of data has been considered a problem
for a smoother simulated torque behavior, but a close analysis at the Reynolds
number involved in the calculations for the design wind speed of 6 m/s has
shown that the Reynolds ranges between 16·104
(in green) and 36·104
(in red)
which have a smoother transition.
Furthermore, the range of angle of attack variation becomes smaller as the tip
speed ratio increases; this can be appreciated in the following simulation of the
angle of attack variation of a blade in a full revolution:
Figure 5. Angle of attack variation in a blade revolution for different tip speed ratios.
For the purpose of calculating the torque produced by the blade, the lift and
drag forces are split in a blade trajectory tangential component and a normal
component (radial).
The normal and tangential coefficients can be calculated using:
‫݊ܥ‬ = ‫݈ܥ‬ cos ߙ + ‫݀ܥ‬ sin ߙ [4.15]
‫ݐܥ‬ = ‫݈ܥ‬ sin ߙ − ‫݀ܥ‬ cos ߙ [4.16]
where Cn and Ct are the normal and tangential coefficients respectively and Cl
and Cd are the lift and drag coefficients.
-100 -50 0 50 100 150 200 250 300
-30
-20
-10
0
10
20
30
Blade position (theta angle) [º]
Angleofattack[º]
TSR=2
TSR=4
TSR=6

26
According to Paraschivoiu, using the blade element theory and the momentum
equation, the upwind flow conditions can be characterized by fup:
݂‫݌ݑ‬ =
N ܿ
8 π R
න |sec ߠ|
గ
ଶ
ି
గ
ଶ
(‫݊ܥ‬ cos ߠ − ‫ݐܥ‬ sin ߠ) ݀ߠ [4.17]
where N is the number of blades.
Then the interference factor can be obtained by:
ܽ‫ݑ‬ =
ߨ
݂‫݌ݑ‬ + ߨ
[4.18]
For a given rotor geometry, rotational speed and free stream velocity, a first set
of calculations from equations 4.6 to 4.18 is made using au=1, then with the
new value of au the procedure is repeated until the initial and final au value are
similar. This is repeated for each streamtube position and once the upstream
induced velocities have been calculated, the downstream half is calculated us-
ing the same set of formulas, interchanging the upstream induced velocity Vu
by the downstream induced velocity Vd (see equation 4.8).
4.3.2 Rotor torque and performance
The tangential and normal forces as function of the azimuth angle θ are calcu-
lated the same way as lift and drag forces in an airfoil:
‫݊ܨ‬ (ߠ) =
1
2
ߩ ܿ ‫ܮ‬ ܹଶ
‫݊ܥ‬ [4.19]
‫ݐܨ‬ (ߠ) =
1
2
ߩ ܿ ‫ܮ‬ ܹଶ
‫ݐܥ‬ [4.20]
where Fn and Ft are the normal and tangential force respectively, ρ is the air
density [kg/m3
], c is the blade chord [m], L is the blade length [m] and W is the
relative wind speed.

27
The Torque produced by a blade is calculated using a combination between the
Lift and Torque formulas.
ܳ (ߠ) =
1
2
ߩ ܿ ܴ ‫ܮ‬ ‫ݐܥ‬ ܹଶ [4.21]
The average torque produced by the upstream half of the rotor (N/2 blades) is
estimated averaging the contributions of each streamtube:
ܳ௔௩ =
ܰ
2 ߨ
න ܳ(ߠ) ݀ߠ
గ
ଶ
ି
గ
ଶ
[4.22]
In order to work with non-dimensional magnitudes the average torque coeffi-
cient Cqav is calculated by:
‫ݍܥ‬௔௩ =
ܳ௔௩
1
2
ߩ ܸ‫݋‬ଶ ܵ ܴ
[4.23]
And finally the upstream half power coefficient Cpu is calculated with:
‫ݑ݌ܥ‬ = ‫ݑݍܥ‬௔௩ ܺ‫ݐ‬ [4.24]
where Xt is the rotor tip speed ratio which is calculated by:
ܺ‫ݐ‬ =
ܴ ߱
ܸ‫݋‬
[4.25]
The average torque and power coefficient for the downstream half of the rotor
are calculated using the same set of formulas from 4.21 to 4.24 using its cor-
responding values of tangential coefficient and induced velocities.
The total power coefficient of the rotor is calculated then adding the two contri-
butions:
‫ݐ݌ܥ‬ = ‫݀݌ܥ‬ + ‫ݑ݌ܥ‬ [4.26]
where Cpt and Cpd are the total and downstream power coefficients respective-
ly.

28
4.4 Model limitations
4.4.1 Streamtube Expansion Model
The double-multiple streamtube model does not account for expansion of the
fluid as it losses velocity. A more realistic model considers the expansion of the
streamtube affecting thus the interference factor calculation.
The significance of the streamtube expansion effects is greater as the tip speed
ratio increases, being the difference between the power coefficients calculated
with the streamtube expansion effect and without it less than 4% beyond a tip
speed ratio of 8. The consideration of the expansion effects result in lower loads
on the downwind half of the turbine (Paraschivoiu I. , 2002, págs. 194-199).
As the tip speed ratio is designed to be below 3, the streamtube expansion ef-
fects will not be included in the calculation as its contribution to the accuracy of
the results is not enough relevant.
4.4.2 Dynamic stall
Dynamic stall is an unsteady aerodynamic effect that occurs at low tip speed
ratios, due to quick and big variations of the angle of attack a vortex originates
on the leading edge of the airfoil; this vortex produces higher lift, drag and pitch-
ing moment coefficients than those produced in static stall.
The MDSV code can be modified to include a model to compute these dynamic
effects at low speed ratios. A description of the phenomenon as well as the dy-
namic-stall models that can be used can be found at (Paraschivoiu I. , 2002).

4.5 Algorithm implementation
4.5.1 Algorithm explanation
A Matlab algorithm, included in
above described process as a reference to calculate the aerodynamic loads,
torque and power coefficient.
The algorithm has as inputs the ro
ty and a set of tables containing the airfoil
on Reynolds number and angle of attack
tube it computes with
locity, Reynolds number, angle of attack and tangential and normal forces.
The resulting torque and normal load (radial load) are used then for the accel
ration and structural analysis respectively and the torque and power coefficient
are obtained by integration of all the blade positions.
implementation
4.5.1 Algorithm explanation
A Matlab algorithm, included in appendix 5 has been developed using the
torque and power coefficient.
The algorithm has as inputs the rotor design parameters, the ambient air veloc
ty and a set of tables containing the airfoil lift and drag coefficients depending
on Reynolds number and angle of attack (see figure below)
tube it computes with iteration process the interference factor, relative wind v
eynolds number, angle of attack and tangential and normal forces.
Figure 6. Algorithm schematic diagram
The resulting torque and normal load (radial load) are used then for the accel
and structural analysis respectively and the torque and power coefficient
are obtained by integration of all the blade positions.
29
has been developed using the
tor design parameters, the ambient air veloci-
lift and drag coefficients depending
(see figure below). For each stream-
ce factor, relative wind ve-
eynolds number, angle of attack and tangential and normal forces.
The resulting torque and normal load (radial load) are used then for the accele-
and structural analysis respectively and the torque and power coefficient

30
4.5.2 Number of streamtubes used
For the whole process, 36 streamtubes had been used, which means evaluating
the wind conditions at blade positions in 5º increments. This number is the
same used by the author of the procedure for presenting the results of his mod-
el (Paraschivoiu I. , 2002, p. 164).
A bigger number of streamtubes has been tested to evaluate the air conditions
in smaller increments but the resulting power coefficient and torque doesn’t
present any variation.
4.6. Algorithm validation
The Matlab Algorithm has been validated using the known results of one of the
aerodynamic prediction models developed by I. Paraschivoiu. This model called
CARDAAV considers the interference factor varying in function of the azimuth
angle and also can consider secondary effects.
It should be noted that the reference data was taken reading a plot because the
numeric results were unavailable, furthermore the code version CARDAAV
which has been used to compute the reference plot consider secondary effects
like the rotating tower and the presence of struts, which makes a difference in
maximum power coefficient of 6% (Paraschivoiu I. , 2002, p. 185), the devel-
oped algorithm gives a difference in maximum Cp of 4.7%.
A comparison between the qualitative accuracy of the algorithm compared with
the reference turbine give us a good approximation of the turbine performance
at TSRs lower than 3.5, which will be the range of our turbine (see figure 7).

31
The reference turbine is found on (Paraschivoiu, Trifu, & Saeed, 2009) and its
parameters are:
Table 3. Rotor parameters for the 7kW reference turbine (Paraschivoiu, Trifu, & Saeed, 2009)
Parameter Value
Rotor diameter 6.0 m
Rotor height 6.0 m
Blade length 6.0 m
Blade chord (constant) 0.2 m
Blade airfoil NACA0015
Number of blades 2
Rotor ground clearance 3 m
Figure 7. Comparison between own code results and the CARDAAV code, data taken from (Paraschivoiu, Trifu, &
Saeed, 2009).
The own version of the aerodynamic model takes into account a varying interfe-
rence factor in function of the azimuth angle but doesn’t consider the following
effects that should improve the accuracy of the model:
• Presence of struts and mast.
• Vertical variation of the freestream velocity.
• Expansion of the streamtubes.
• Dynamic stall effects.
1 2 3 4 5 6 7 8 9 10
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
TSR
Cpt
Own code version
Reference plot

32
A second validation with the results found from computational fluid dynamics
and experimental analysis of a small wind turbine has been done in order to see
the possible inaccuracy of the model for our design.
The second reference turbine is analyzed on (Raciti Castelli;Englaro;& Benini,
2011) and its parameters are:
Table 4. Rotor parameters for the second reference turbine (Raciti Castelli, Englaro, & Benini, 2011)
Parameter Value
Rotor diameter 1.03 m
Blade length 1.4564 m
Blade chord (constant) 0.0858 m
Blade airfoil NACA 0021
Number of blades 3
Figure 8. Comparison between CFD, DMSV and experimental results, data taken from (Raciti Castelli, Englaro, &
Benini, 2011)
We can see that the algorithm is not accurate for small wind turbines at small tip
speed ratios, this can be, as noted in (Claessens, 2006), because the momen-
tum based models are limited to solidities below 0.2, for bigger solidities the
assumptions of the momentum model are not valid (e.g. homogeneous, incom-
pressible, steady state flow; no frictional drag or non-rotating wake).
Despite these results, the momentum based model is still considered a useful
tool in the design process as it gives quicker and qualitative results than other
1.5 2 2.5 3 3.5
-0.1
0
0.1
0.2
0.3
0.4
Power Coefficient depending on tip speed ratio
TSR
Cpt
Modelled with DMSV
Experimental results
Modelled with CFD

33
methods. Different computational options, like CFD, are too time consuming to
be considered for use within the scope of this thesis.
5. ROTOR DESIGN
5.1 Starting parameters
A numerical optimization process of the design parameters have been done in
order to select which could be the proper parameters for the intended applica-
tion. The optimization process has been carried out focusing on the effect each
parameter has in the power coefficient and torque produced; the results are
presented in plots.
In order to speed up the optimization process, some of the parameters have to
be fixed for our purposes. The fixed parameters will be the design airspeed and
the blade swept area.
The usual power coefficient values shown at the comparison with the compe-
tence turbines are between 0.15 and 0.2, so a swept area between 4 and 5.2
m2
is needed to have 100W at 6 m/s. The selected swept area will be kept to 4
m2
.
The design airspeed is kept at 6 m/s, because the purpose of the application is
to take profit of the average wind speed rather than higher speeds, which are
not so frequent.

34
The starting parameters and conditions in which the wind turbine has been
tested are shown in table 5:
Table 5. Starting values of the turbine parameters for the optimization process.
Symbol Parameter Starting value
Vo Airspeed 6 m/s
TSR Tip speed ratio 2
R Rotor radius 1 m
c Blade chord 0.3 m
L Blade length 2 m
N Number of blades 4
αo Initial angle of attack 0º
5.2 Airfoil selection
The airfoil has been selected considering the availability of airfoil data for angles
of attack between -30 and 30º and the final thickness of the blade which is as-
sociated with its ability to withstand the loads.
The available airfoils with enough data to conduct the study of the aerodynamic
loads and behavior of the wind turbine were available in a document issued by
Sandia National Laboratories (Sheldahl & Klimas, 1981), this document pro-
vides the lift and drag coefficients depending on Reynolds number and angle of
attack ranging from 0 to 180º for several NACA airfoils.
The selected airfoil is the NACA0021, which aerodynamics characteristics were
determined using an airfoil property synthesizer code. This profile is one of the
thickest available (21% of chord) and when compared with NACA0015 (thick-
ness is 15% of chord) it can be seen that the self-starting behaviour is improved
with thicker airfoils (see Figure 9).

35
Figure 9. Performance comparison between two different airfoils.
The actual shape of the NACA0021 airfoil can be found at (Abbott & Von
Doenhoff, 1959, p. 335); Table 9 from appendix 6 shows the ordinates in per-
cent of chord.
The airfoil selection can be done using an airfoil simulation program like RFOIL,
which according to (Claessens, 2006) can have problems to accurately predict
the airfoil behavior at Reynolds numbers lower than 1·106
. Other alternative to
wind tunnel testing are the two-dimensional computational fluid dynamics simu-
lations.
5.3 Design airspeed
A modification in the free stream velocity from 0 to 15 m/s varies the power
coefficient significantly, beyond this airspeed the power coefficient remains con-
stant at each tip speed ratio as shown in Figure 10.
1 1.5 2 2.5 3 3.5 4
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
TSR
Cpt
NACA0015
NACA0021

36
Figure 10. Power coefficient depending on ambient airspeed (Vo) and tip speed ratio (TSR).
The torque behavior increases notably due to its quadratic dependence on air-
speed.
Figure 11. Torque depending on ambient airspeed (Vo) and Tip Speed Ratio (TSR).
0 0.5 1 1.5 2 2.5 3 3.5 4
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
Power Coefficient depending on Vo and TSR
TSR
Cp
Vo=3m/s
Vo=6m/s
Vo=9m/s
Vo=12m/s
Vo=15m/s
Vo=18m/s
Vo=21m/s
0 0.5 1 1.5 2 2.5 3 3.5 4
-50
0
50
100
150
200
Torque depending on Vo and TSR
TSR
Torque[N·m]
Vo=3m/s
Vo=6m/s
Vo=9m/s
Vo=12m/s
Vo=15m/s
Vo=18m/s
Vo=21m/s

37
The target performance is 100 W at 6 m/s, which should correspond to the rated
wind speed, and therefore the maximum rpm of the turbine should be the cor-
responding to the optimum tip speed ratio of the turbine. But, in order to take
advantage also of higher wind speeds, the rated wind speed will be set to 9 m/s
but maintaining the initial performance target at 6 m/s.
The usual rated wind speed value ranges from 11.5 to 15 m/s, but a lower value
may produce more energy overall because it will be more efficient at wind
speeds between cut-in and rated.
5.4 Rotor dimensions
The blade length and rotor radius have a major contribution in the torque beha-
vior of the turbine as can be deduced from the torque equation. In general as
bigger these parameters, bigger the torque produced. These parameters are
involved also in the solidity calculation. The solidity becomes an important pa-
rameter when scaling down or up wind turbines and also determines the appli-
cability of the momentum model.
The radius and blade variation analysis is done maintaining constant swept
area. It can be seen in Figure 12 that an increase in rotor radius leads to greater
maximum power coefficients, but they are achieved at greater tip speed ratios,
so the little the radius, the less tip speed ratio is necessary to work at maximum
power coefficient.

38
Figure 12. Power coefficient dependent on rotor radius (R) and tip speed ratio (TSR) maintaining constant Swept
area.
The power coefficient is not affected by a sole variation of the blade length. This
is deduced if the equation for calculating the power coefficient is developed and
simplified (Equation 4.24), in this case the upwind power coefficient is illu-
strated, but the dependences are equal on the downstream half. We can see
blade length is present both in the numerator and denominator of the equation.
‫ܥ‬௣௨ =
ܰ ܿ ߱ ‫׬‬ ‫ݐܥ‬ ܹଶ
݀ߠ
గ
ଶ
ି
గ
ଶ
4 ߨ ܸ‫݋‬ଷ
[5.1]
The rotor radius R is simplified also from the equation but it is still implicit in the
calculation of the relative velocity W, so we can state that the rotor radius have
influence both in torque (see figure 13) and power coefficient.
Nevertheless, the torque is affected by the blade length the same way the tan-
gential force coefficient is dependent on the blade surface. Thus the blade
length only affects torque and its optimization has to be done considering other
factors as the swept area, which determines the power available from wind.
The blade tip losses are not contemplated in the model, in fact, one of the initial
assumptions of the model is that there is no interaction between streamtubes,
0.5 1 1.5 2 2.5 3 3.5
0
0.05
0.1
0.15
0.2
0.25
TSR
Cp
R=0.5m
R=0.75m
R=1m
R=1.25m
R=1.5m
R=1.75m
R=2m

39
so in a real model, an increase on blade length should make the blade more
efficient.
The optimization needs the input of a fourth factor, the maximum load due to
both aerodynamic loads and rotational speed, as the radius will be the one de-
termining them.
Figure 13. Torque dependent on rotor radius (R) and tip speed ratio (TSR) maintaining constant Swept area.
5.5 Solidity
Solidity is dependent on blade chord, number of blades and rotor radius; as ra-
dius effect has been already analyzed for constant swept area, the chord will be
analyzed independently.
An increase in blade chord advances the TSR at which the maximum power
coefficient is achieved, as bigger the chord, smaller the tip speed ratio. In order
to reduce the centrifugal force, a big chord may be more effective than a lighter
blade.
0 0.5 1 1.5 2 2.5 3 3.5 4
-5
0
5
10
15
20
25
30
35
40
TSR
Torque[N·m]
R=0.5m
R=0.75m
R=1m
R=1.25m
R=1.5m
R=1.75m
R=2m

40
Figure 14. Power coefficient dependent on blade chord (c) and tip speed ratio (TSR).
A bigger chord advances also the point of maximum torque, blades with smaller
chords need a bigger tip speed ratio to develop maximum torque and it can af-
fect the self-starting capabilities.
We can anticipate looking at figures 14 and 15 that a carefully decision about
the chord length affects self-starting behavior, optimum TSR and maximum
power coefficient, so the maximum allowable loads will determine the suitable
chord for our model.
Figure 15. Torque dependent on blade chord (c) and tip speed ratio (TSR).
1 1.5 2 2.5 3 3.5 4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
TSR
Cp
c=0.1m
c=0.15m
c=0.2m
c=0.25m
c=0.3m
c=0.35m
c=0.4m
1 1.5 2 2.5 3 3.5 4
-8
-6
-4
-2
0
2
4
6
8
10
12
TSR
Torque[N·m]
c=0.1m
c=0.15m
c=0.2m
c=0.25m
c=0.3m
c=0.35m
c=0.4m

41
A better self-starting behavior will allow a smaller cut-in wind speed but will limit
the maximum achievable power coefficient; the chord will be selected between
0.25 or 0.3 m depending on their acceleration behavior and structural limits.
The other parameter affecting solidity is the number of blades, keeping in mind
the possible interference between blades and also a low solidity value, only
three and four-bladed designs will be analyzed. Furthermore, the algorithm
doesn’t account for the wake turbulence that is created behind each blade and
can seriously affect the adjacent blade’s lift and drag forces so a big number of
blades may give too optimistic results.
Figures 16 and 17 compare the effect of blade number on the power coefficient
and average torque. As can be deduced from equations 4.17 and 4.18, a bigger
number of blades lead to a bigger deceleration of the air, and this effect is
greater than the increase in torque that can be produced having more blades.
Figure 16. Number of blades effect in power coefficient at several rotational speeds, the two models had blade
chords of 0.28 m, NACA0021 airfoils and Vo is 9 m/s.
50 100 150 200 250 300 350 400
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
rotational speed [rpm]
Cp
N=4 blades
N=3 blades

42
Figure 17. Number of blades effect in average torque at several rotational speeds, the two models had blade
chords of 0.28 m, NACA0021 airfoils and Vo is 9 m/s.
In this case, three blades are more efficient than four for the same rotational
speed; the final decision will be made considering the acceleration behavior,
which will provide a better view of the self-starting characteristics.
5.6 Initial angle of attack
A positive initial angle of attack broadens the range of angular speed operation
and a negative one shortens it (see figure 18); this is interesting when fixing the
maximum rpm. Furthermore the torque is influenced the same way resulting in a
lower maximum power coefficient and torque for negative angles of attack.
The initial angle of attack for our design will be set at 0 degrees as the advan-
tages of a different angle of attack are, according to this model, only evident at
higher tip speed ratios than the intended for the model.
50 100 150 200 250 300 350 400
-20
-10
0
10
20
30
40
Averagetorque[Nm]

43
Figure 18. Power coefficient variation depending on the initial angle of attack.
6. ACCELERATION ANALYSIS
An acceleration analysis has been performed in order to ensure the wind tur-
bine will achieve the desired rpm. The first step is to make an estimation of the
inertial momentum.
The rotor has been divided in N blades with two beams each, one on the bottom
and one on the top. Approximatting the form of the blade to a cuboid and using
the Steiner teorem, we have that the inertial moments are:
‫ܫ‬௕௟௔ௗ௘ =
1
12
݉௕௟௔ௗ௘(ܿଶ
+ (0.21 ܿ)ଶ) + ݉௕௟௔ௗ௘ ܴଶ [6.1]
‫ܫ‬௕௘௔௠ =
1
12
݉௕௘௔௠(‫ݐ݀݅ݓ‬ℎଶ
+ ܴଶ) + ݉௕௟௔ௗ௘ ൬
ܴ
2
൰
ଶ
[6.2]
where I stands for Inertia moment [kg m2
] and m for mass [kg]. In order to calcu-
late the mass of the wind turbine, a wood density of 500 kg/m3
is used.
1 1.5 2 2.5 3 3.5 4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
TSR
Cp
Ao=-3º
Ao=0º
Ao=3º

44
Multiplying the inertial momentum by the number of blades we have an approx-
imate value of the total inertial momentum and we can calculate the angular
acceleration:
ߙ =
ܳ
‫ܫ‬
[6.3]
where α is the angular acceleration [rad/s2
], Q the torque [Nm] and I the rotor
inertia moment [kg m2
].
A qualitative acceleration analysis has been done (note that blades are approx-
imated to cuboids) between a three-bladed rotor and a four-bladed rotor at a
constant airspeed of 25 m/s in order to see if the diminution of inertia moment
and different blade arrangement can improve the acceleration characteristics.
Figure 19. Acceleration analysis of a three and a four-bladed design, at 25 m/s wind speed and chord=0.24.
The four-bladed design has better behavior at low rpm, but the three bladed
rotors present a much quicker response, which can be further improved using a
lighter material and can be very useful to take profit from brief gusts. The four-
bladed rotor will not be further considered in the design.
0 0.5 1 1.5 2 2.5 3 3.5 4
0
100
200
300
400
500
600
700
800
900
1000
time (minutes)
rotationalspeed(rpm)
N=4 blades
N=3 blades
Structural limit

45
7. STRUCTURAL ANALYSIS
A structural analysis will provide the limits of operation of the turbine, in terms of
rotational speed, in order to prevent structural failure. The analysis is done as
follows:
• A configuration of the wind structure is proposed.
• The maximum bending moment, normal and shear loads on a blade are
analyzed.
• Consulting tables of mechanical properties of materials, the ultimate
stress modules are found.
• The allowable stresses and loads are calculated using a safety factor.
• The moment of inertia of the blade shape is found using a CAD program.
• The numeric values of the allowable loads are calculated.
• A check is done to see if the turbine is able to accelerate to a point where
the structural allowable limits can be surpassed.
7.1 Rotor configuration
Three different rotor configurations (see figure 20) have been tested and are
presented in the following figure, they have been numbered from left to right in
order to refer them further in the analysis.
The first configuration attaches each blade by two points, which correspond to
the Bessel points and minimize the maximum bending moment. These points
are located at 20.7% of the blade length from each blade tip; a procedure for
obtaining these points by static analysis can be found at (Saranpää, 2010).

46
Figure 20. Rotor configurations analyzed.
The second and third configuration are equivalent in terms of maximum bending
strength and shear strength, so the third rotor configuration is discarded as the
second has lower exposed area per same strength, which will reduce the result-
ing drag associated with central mast and struts.
7.2 Allowable stress and selected materials
The allowable stress is defined as:
‫݈ܾ݁ܽݓ݋݈݈ܣ‬ ‫ݏݏ݁ݎݐݏ‬ =
ܻ݈݅݁݀ ‫ݎ݋‬ ܷ݈‫݁ݐܽ݉݅ݐ‬ ‫ݐ݃݊݁ݎݐݏ‬ℎ
‫ݎ݋ݐܿܽܨ‬ ‫݂݋‬ ‫ݕݐ݂݁ܽݏ‬
[7.1]
with a factor of safety bigger than 1 if the structure has to avoid failure. For brit-
tle materials such as concrete or materials without a defined yield strength, like
wood, the ultimate strength instead the yield stress is used.
The ultimate strength is dependent on the material used; in this design, the
chosen material is wood for blades, beams and mast. The type of wood has to
be determined according to the local availability in each region.

47
General guidelines for choosing an appropriate wood are:
• Lightweight softwood is the most suitable and heartwood is more appro-
priate than sapwood (Piggot, 2001).
• It should be free from knots. Knots introduce discontinuities which lower
the mechanical properties, having a greater effect in axial tension than in
axial compression so they should be orientated in order to stand com-
pression loads (Forest Products Laboratory, 2010, p. 5.28).
• The grain should run parallel as possible to the longitudinal axis.
Wood selection has been done regardless of wood availability in the country of
destination; three different types of wood had been considered in order to have
wider range of manufacturing possibilities. Final dimensions will then be func-
tion of the type of wood used for the construction of the model. The strength
data of these species can be found on appendix 7.
• Sitka spruce (Picea sitchensis) has been by far the most important wood
for aircraft construction and, although it may not be readily available, its
properties can be taken as reference in order to find technically similar
woods.
• Balsa (Ochroma pyramidale) wood is widely distributed throughout tropi-
cal America from southern Mexico to southern Brazil and Bolivia, it has
been included in the analysis because is the lightest and softest of all
woods on the market (Forest Products Laboratory, 2010, p. 36).
• Obeche (Triplochiton scleroxylon), original from west-central African area
is a hardwood type wood that can be found in some retail shops so it has
been included for a possible prototype candidate.
In order to determine which factor of safety should be applicable, this data has
been considered:
• Compression parallel to grain has an average coefficient of variation of
18%.
• Shear parallel to grain has an average coefficient of variation of 14%.
As an average value, a factor of safety of 1.5 will be considered in first place in
order to absorb the possible inaccuracies in the load calculation, the variability
of the material used or the deterioration due to atmospheric conditions.

48
Once the allowable stress is calculated, the allowable load is calculated with
(Gere, 2004):
‫݀ܽ݋݈_݈ܾ݁ܽݓ݋݈݈ܣ‬ = ‫ܽ݁ݎܣ‬ ‫ݏݏ݁ݎݐݏ_݈ܾ݁ܽݓ݋݈݈ܣ‬ [7.2]
Figure 21 presents the maximum allowable loads for bending moment as a
function of the blade material and rotor configuration.
Figure 21. Maximum allowable bending load depending on chord length, blade material and rotor configuration.
We can appreciate a big difference in rotor strength between rotor configuration
one and the others, second and third configurations are not further considered
in the analysis as the limitation in maximum load relates directly with the maxi-
mum tip speed ratio and thus power coefficient.
Wood blades structural strength are limited by the bending moment strength
rather than their shear stress as can be appreciated in maximum strength calcu-
lations from Table 11 in appendix 7. The values of blade maximum allowable
bending load for the first rotor configuration using different woods are found on
Table 13 from appendix 7.
0.12 0.14 0.16 0.18 0.2 0.22 0.24 0.26 0.28 0.3 0.32
0
10
20
30
40
50
60
70
chord length [m]
maximumallowablebendingload[kN]
Sitka spruce (conf.1)
Sitka spruce (conf.2&3)
Balsa (conf.1)
Balsa (conf.2&3)
Obeche (conf.1)
Obeche (conf.2&3)

49
7.3 Load analysis
The loads analyzed for the blade dimensioning are divided into centrifugal and
aerodynamic loads. The centrifugal force is dependent on both radius and the
square of the rotational speed.
‫ܨ‬௖ = ݉ ܴ ߱ଶ [7.3]
where Fc is the centrifugal force [N], m is the blade mass [kg], R is the rotor ra-
dius [m] and ω the rotational speed [rad/s].
The normal force Fn is the aerodynamic load, an analysis of the resultant of the
normal force and centrifugal force is made for one blade in a revolution. As they
have equal direction, their resultant is:
‫ܨ‬௥ = ‫ܨ‬௖ + ‫ܨ‬௡ [7.4]
Knowing the maximum allowable load for a given blade chord and a given ma-
terial, the maximum rpm for each rotor construction can be found. The aerody-
namic load is considered to be the load produced when the turbine achieves its
cut-out speed, which has been deduced from the average value found on ap-
pendix 4 and corresponds to 25 m/s.
The maximum allowable loads, as well as blade mass and maximum rpm are
found on tables, we can establish the structural limit of the wind turbine at 580
rpm and 25 m/s wind speed regardless of blade material considering a blade
chord of 0.28 m.
8. FINAL DESIGN AND OPERATION
With all the previous considerations, keeping in mind that a prototype must be
built in order to validate the algorithm results and possibly make further refine-
ments, the candidate for a prototype is presented on table 6

50
Table 6. Final rotor design parameters
Symbol Parameter Value
Vci Cut-in windspeed 4 m/s
Vrs Rated windspeed 9 m/s
Vco Cut-out windspeed 25 m/s
TSR Optimum tip speed ratio 2.8
Cp
Maximum power coeffi-
cient
0.45
R Rotor radius 1 m
c Blade chord 0.28 m
L Blade length 2 m
N Number of blades 3
αo Initial angle of attack 0º
It’s power curve has been predicted graphically according to the procedure
stated at Manwell, McGowan and Rogers (2009, p. 336-340) assuming a gene-
rator efficiency of 0.85; the procedure consists in matching the power output
from the rotor as a function of wind and rotational speed (see next figure) and
with their data make a Power output versus wind speed plot.
Figure 22. Power output at several rotational and wind speeds. A generator efficiency of 0.85 is assumed.
50 100 150 200 250 300 350 400 450 500
0
200
400
600
800
1000
1200
1400
1600
1800
Rotational speed [rpm]
Poweroutput(W)
3 m/s
6 m/s
9 m/s
12 m/s
170
rpm
240
rpm

51
The rotor design gives (theoretically) 690 W at rated wind speed and the power
obtained at 6 m/s at the same rotational speed is almost 0. Nevertheless, the
power prediction curve should be made with field test data, as the generator
specifications, algorithm inaccuracies and control system used play an impor-
tant role on the shape of the power curve.
The rotor torque curves at different rotational and wind speeds are presented on
next figure.
Figure 23. Average torque curves for several wind and rotational speeds.
A mechanism to mechanically stop the wind turbine before the structural limit is
surpassed has to be designed. In HAWT models, the over speed control is
achieved turning away the frontal area of the turbine from the wind direction or
changing the blade pitch (angle of attack) in order to stall the blades. These
mechanisms may present spring-loaded devices, and although a fixed pitch was
considered for the initial design, perhaps a furling mechanism of the rotor
blades when certain centrifugal force is achieved can be considered in the de-
sign.
50 100 150 200 250 300 350
0
10
20
30
40
50
Averagetorque[Nm]
6
m/s
10
m/s
12
m/s
8
m/s
4
m/s

52
9. CONCLUSIONS
Wood is a suitable material for blade construction in small-scale wind turbines
and the construction of an affordable wind turbine with common workshop tools
can be contemplated.
A three-bladed design is more efficient than a four-bladed rotor; a low solidity
wind turbine may present self-starting problems as rotor efficiency is poor at low
tip speed ratios.
There is an optimum turbine rotational speed for each ambient wind speed at
which the maximum efficiency is achieved. The energy production of a fixed-
pitch wind turbine can be improved adjusting the rated airspeed to the place of
installation average wind conditions in order to reach its maximum efficiency.
Larger radius turbines are more efficient than small turbines at same rotational
speed as the tangential airspeed increase leads to smaller angles of attack,
bigger Reynolds numbers and thus bigger blade lift coefficients.
Experimental tests are needed in order to refine the design and validate the
aerodynamic model used.

53
10. BIBLIOGRAPHY
Abbott, I. H., & Von Doenhoff, A. E. (1959). Theory of Wing Sections Including a Summary of
Airfoil Data. New York: Dover Publications, Inc.
Aerospaceweb.org. (2005). Retrieved 10 29, 2011, from
http://www.aerospaceweb.org/design/scripts/atmosphere/
allsmallwindturbines.com. (2011). Retrieved September 2011, from allsmallwindturbines.com:
http://www.allsmallwindturbines.com/
American Wind Energy Association. (2002). The U.S. Small Wind Turbine Industry Roadmap.
Retrieved September 2011, from American Wind Energy Association:
http://www.awea.org/learnabout/smallwind/upload/US_Turbine_RoadMap.pdf
Blackwell, B. F., & Reis, G. E. (1974). Blade Shape for a Troposkien Type of Vertical-Axis Wind
Turbine. Sandia Laboratories.
Brahimi, M. T., Allet, A., & Paraschivoiu, I. (1995). Aerodynamic Analysis Models for Vertical-
Axis Wind Turbines. International Journal of Rotating Machinery , Vol.2 No. 1 pp. 15-21.
Cace, J., ter Horst, E., Syngellakis, K., Maíte, N., Clement, P., Heppener, R., et al. (2007). Urban
windturbines, guidelines for small wind turbines in the built environment. Retrieved September
2011, from Urbanwind.org: http://www.urban-
wind.org/pdf/SMALL_WIND_TURBINES_GUIDE_final.pdf
Claessens, M. C. (2006). The Design and Testing of Airfoils for Application in Small Vertical Axis
Wind Turbines. Delft: Delft University of Technology-Faculty of Aerospace Engineering.
Cooper, P. (2010). Development and analysis of vertical-axis wind turbines. In W. Tong, Wind
power generation and wind turbine design (pp. 289-296). WIT Press.
Cheikhrouhou, H. (2011). Energy and Power. Retrieved September 2011, from African
Development Bank Group: http://www.afdb.org/en/topics-and-sectors/sectors/energy-
power/
Dunnett, S. (n.d.). Small wind energy systems for battery charging. Retrieved September 2011,
from Practical Action:
http://practicalaction.org/docs/energy/wind_energy_battery_charging.pdf
Encraft. (2009). Encraft Warwick Wind Trials Project. Retrieved September 2011, from
http://www.warwickwindtrials.org.uk/resources/Warwick+Wind+Trials+Final+Report+.pdf
Forest Products Laboratory. (2010). Wood handbook, wood as an engineering material.
Madison: United States Department of Agriculture.
Gere, J. M. (2004). Mechanics of materials. Brooks/Cole-Thomson Learning.
Jha, A. R. (2011). Wind Turbine Technology. New York: CRC Press.
Legros, G., Havet, I., Bruce, N., & Bonjour, S. (2009). The Energy Acces Situation in Developing
Countries. New York: UNDP and World Health Organization.

54
Manwell, J. F., McGowan, J. C., & Rogers, A. L. (2009). Wind Energy Explained - Theory, Design
and Application. John Wiley & Sons Ltd.
Paraschivoiu, I. (2002). Wind Turbine Design With Emphasis on Darrieus Concept. Presses
internationales Polytechnique.
Paraschivoiu, I., Trifu, O., & Saeed, F. (2009). H-Darrieus Wind Turbine with Blade Pitch Control.
International Journal of Rotating Machinery , Article ID 505343.
Piggot, H. (2001). Windpower workshop. Centre for Alternative Technology.
Raciti Castelli, M., Englaro, A., & Benini, E. (2011). The Darrieus wind turbine: Proposal for a
new performance prediction model based on CFD. Elsevier Energy , 4919-4935.
Saranpää, O. (2010). Painerungon osien kuljetuskehikoiden analysointi ja kehittäminen.
Tampere: Opinnäytetyö (Thesis).
Sheldahl, R. E., & Klimas, P. C. (1981). Aerodynamic Characteristics of Seven Symmetrical Airfoil
Section Trough 180-Degree Angle of Attack for Use in Aerodynamic Analysis of Vertical Axis
Wind Turbines. Albuquerque: Sandia National Laboratories.
Spera, D. (2009). Wind Turbine Technology, fundamental concepts of wind turbine engineering.
ASME Press.
Talayero, & Telmo. (2008). Energía eólica. Zaragoza: Prensas universitarias de zaragoza.
Tong, W. (2010). Fundamentals of wind energy. In W. Tong, Wind power generation and wind
turbine design (p. 112). WIT Press.
Tullis, S., Fiedler, A., McLaren, K., & Ziada, S. Medium-solidity Vertical Axis Wind Turbines for
use in Urban Environments. Hamilton, Canada.
U.S. Energy Information Administration. (2011, March). U.S. Department of Energy. Retrieved
November 2011, from
http://www.eia.gov/oiaf/aeo/tablebrowser/#release=IEO2011&subject=1-IEO2011&table=2-
IEO2011&region=0-0&cases=Reference-0504a_1630
Wahl, M. (2007). Designing an H-rotor type Wind Turbine for Operation on Amundsen-Scott
South Pole Station. Uppsala: Uppsala Universitet.
winafrique. (2011). Retrieved September 2011, from winafrique.com:
http://www.winafrique.com/index.php?option=com_content&view=article&id=92&Itemid=14
5
Windfinder. (2011). Retrieved September 2011, from Windfinder.com:
http://www.windfinder.com/windstats/windstatistic_pyhajarvi_tampere.htm
Wood, D. (2011). Small Wind Turbines – Analysis, Design, and Application. Springer-Verlag.

55
APPENDIXES
SYMBOLS APPENDIX 1
ad, au downstream and upstream interference factors
c blade chord
Cd drag coefficient
Cl lift coefficient
Cn normal coefficient
Cp, Cpd, Cpu power coefficient, downstream and upstream
Cqav average torque coefficient
Ct tangential coefficient
Fc centrifugal force
Fn normal force
Fr resultant force
Ft tangential force
fup function characterizing upwind flow conditions
Gb basic wood specific gravity
GMC wood specific gravity at MC moisture content
I inertia moment
Kv air kinematic viscoity
L blade length
MC moisture content
MCft wood fiber saturation point
N number of blades
Pw power available from wind
Q, Qav torque, average torque
R rotor radius
Reb local blade's Reynolds number
S swept area
SMC
wood volumetric percent shrinkage at MC moisture con-
tent
So wood volumetric percent shrinkage from green to ovendry
TSR tip speed ratio
W, Wu resultant air velocity
Vd downstream induced velocity
Ve equilibrium-induced velocity
Vo ambient airspeed
Vu upstream induced velocity
Xt rotor tip speed ratio
α angle of attack/angular acceleration
αo initial angle of attack
θ angle between Vo and the position of the blade in the rotor
ρ air density
ρMC wood density at x moisture content
ρw water density at 4°C
σ solidity
ω rotor angular speed

56
GLOSSARY APPENDIX 2
Compressive strength parallel to grain: Maximum stress sustained by a
compression parallel-to-grain specimen having a ratio of length to least dimen-
sion of less than 11 (Definition taken from (Forest Products Laboratory, 2010)).
Cut-in windspeed: The minimum windspeed at which the wind turbine starts
producing energy.
Cut-out windspeed: The maximum windspeed at which the wind turbine stops
operating, mainly for safety reasons.
NACA-series Airfoil: an airfoil which section and aerodynamic properties are
available in reference books like Abbott & Von Doenhoff, 1959, these airfoil data
are the result of the research conducted at the National Advisory Commite for
Aeronautics (NACA).
Rated windspeed: windspeed at which the rated power is produced, this value
defines the shape of the power curve.
Shear strength parallel to grain: Ability to resist internal slipping of one part
upon another along the grain. (Definition taken from (Forest Products
Laboratory, 2010).
Troposkien shape: is a word derived from the Greek, means turning rope and
is used to describe the shape assumed by a cable if its ends are attached to
two points and it is spun at constant angular velocity (Blackwell & Reis, 1974).
This shape is used to minimize the bending stress in the blades of Darrieus tur-
bines.

57
WORLD ENERGY TRENDS APPENDIX 3
Figure 24. World wind-powered generating capacity trend (U.S. Energy Information Administration, 2011)
Figure 25. World total installed generating capacity trend (U.S. Energy Information Administration, 2011)
0
100
200
300
400
500
600
2005
2007
2009
2011
2013
2015
2017
2019
2021
2023
2025
2027
2029
2031
2033
2035
Wind-poweredgeneratingcapacity
(GW)
Years
OECD
Non-OECD
World
0
1000
2000
3000
4000
5000
6000
7000
8000
2005
2007
2009
2011
2013
2015
2017
2019
2021
2023
2025
2027
2029
2031
2033
2035
Totalinstalledgeneratingcapacity(GW)
Years
OECD
Non-OECD
World

58
COMPETENCE ANALYSIS APPENDIX 4: 1(2)
Table 7. Market turbines analysis (1)
Product Manufacturer
Rated
Power
(W)
Rated
Wind
Speed
(m/s)
Rated
rpm N
Price
€
Turby 2500 14
120/
400 3 11466
Windspire Windspire Energy, US 1200 10,7 3 11354
P-1000 Shanghai Aeolus Windpower 1000 12 5 2826
Easy Vertical Ropatec 1000 12 240 3
FL1KWVAWT Flexienergy 1000 10 55
Windterra Eco 1200
Windterra systems
1000 11
270
max 3
VENCO-Twister 1000 T VENCO Power GmbH 1000 12 3
Mini Vertical Ropatec 750 14 180 3
Aeolos-V 600w Aeolos Wind Turbine 600 10
P-500
Shanghai Aeolus Wind-
power 500 13 5 1680
FSW Gyro.5 Four Seasons Windpower 500 10,72 5 1949
Seahawk 500 Wepower 500 12,5 608 6
FLEX500VAWT Flexienergy 500 13
Aeolos-V 300w Aeolos Wind Turbine 300 10
VENCO-Twister 300 T VENCO Power GmbH 300 14 3
Wind Smile 200W Wind Smile 200 13 4
ELY-100 Qingdao 100 10 200 5

59
COMPETENCE ANALYSIS APPENDIX 4: 2(2)
Table 8. Market turbine analysis (2)
Product
Blade
Lenght
(m)
Rotor
radius
(m)
Swept
Area
(m^2)
Total
Height
(m)
Min
Wind
(m/s)
Max
Wind
(m/s)
Survival
Wind
(m/s) Calc. Cp
Turby 2,6 1,9 9,88 4 14 55 0,15
Windspire 5,79 0,61 7,06 9,1 3,8 47 0,23
P-1000 2 0,9 3,60 5,5 4 25 50 0,26
Easy Vertical 1,9 0,95 3,61 6,8 3 53 0,26
FL1KWVAWT 1,65 1,25 4,13 8 3 25 40 0,40
Windterra Eco 1200 2,66 1,125 5,99 3 53 0,20
VENCO-Twister 1000
T 1,9 0,95 3,61 3,5 20 50 0,26
Mini Vertical 1,5 1,5 4,50 3 53 0,10
Aeolos-V 600w 1,6 0,65 2,08 2 50 0,47
P-500 1,05 0,68 1,43 5,5 4 25 45 0,26
FSW Gyro.5 2,21 0,68 3,01 5,49 1,96 24,6 40,2 0,22
Seahawk 500 1,2 0,381 0,91 1,35 54 0,46
FLEX500VAWT 1 0,6 1,20 4,5 0,31
FSW Gyro.3 1,29 0,68 1,75 5,49 1,96 24,6 40,2 0,29
Aeolos-V 300w 1,4 0,6 1,68 2 50 0,29
VENCO-Twister 300 T 1 0,5 1,00 3,5 25 50 0,18
FSW Gyro.2 0,89 0,68 1,21 5,49 1,96 24,6 40,2 0,28
Wind Smile 200W 1 0,5 1,00 1,4 2,5 60 0,15
ELY-100 0,8 0,8 1,28 6 1,5 25 50 0,13

60
MATLAB ALGORITHM APPENDIX 5: 1(7)
%This function calculates the power coefficient and average torque%This function calculates the power coefficient and average torque%This function calculates the power coefficient and average torque%This function calculates the power coefficient and average torque
%for a vertical axis wind turbine according to the procedure stated at:%for a vertical axis wind turbine according to the procedure stated at:%for a vertical axis wind turbine according to the procedure stated at:%for a vertical axis wind turbine according to the procedure stated at:
%"Double%"Double%"Double%"Double----Multiple Streamtube modelMultiple Streamtube modelMultiple Streamtube modelMultiple Streamtube model for Darrieus wind turbines,for Darrieus wind turbines,for Darrieus wind turbines,for Darrieus wind turbines,
%I.Paraschivoiu", the equations have been adapted to a straight bladed%I.Paraschivoiu", the equations have been adapted to a straight bladed%I.Paraschivoiu", the equations have been adapted to a straight bladed%I.Paraschivoiu", the equations have been adapted to a straight bladed
%wind turbine when necessary.%wind turbine when necessary.%wind turbine when necessary.%wind turbine when necessary.
%function [Cpt, av_T] = f_Cp_and_av_T2 (Vo, w, R, N, c, L, Ao, NACA)%function [Cpt, av_T] = f_Cp_and_av_T2 (Vo, w, R, N, c, L, Ao, NACA)%function [Cpt, av_T] = f_Cp_and_av_T2 (Vo, w, R, N, c, L, Ao, NACA)%function [Cpt, av_T] = f_Cp_and_av_T2 (Vo, w, R, N, c, L, Ao, NACA)
%%INPUT PARAMETERS%%INPUT PARAMETERS%%INPUT PARAMETERS%%INPUT PARAMETERS
%Vo = ambient air speed (m/s).%Vo = ambient air speed (m/s).%Vo = ambient air speed (m/s).%Vo = ambient air speed (m/s).
%w = r%w = r%w = r%w = rotor angular speed (rad/s).otor angular speed (rad/s).otor angular speed (rad/s).otor angular speed (rad/s).
%R = rotor radius (m).%R = rotor radius (m).%R = rotor radius (m).%R = rotor radius (m).
%N = number of blades.%N = number of blades.%N = number of blades.%N = number of blades.
%c = blade chord (m).%c = blade chord (m).%c = blade chord (m).%c = blade chord (m).
%L = blade length (m).%L = blade length (m).%L = blade length (m).%L = blade length (m).
%Ao = initial angle of attack (rad).%Ao = initial angle of attack (rad).%Ao = initial angle of attack (rad).%Ao = initial angle of attack (rad).
%NACA = airfoil used (15 or 21).%NACA = airfoil used (15 or 21).%NACA = airfoil used (15 or 21).%NACA = airfoil used (15 or 21).
%%OUTPUT PARAMETERS%%OUTPUT PARAMETERS%%OUTPUT PARAMETERS%%OUTPUT PARAMETERS
%Cpt = power coefficient.%Cpt = power coefficient.%Cpt = power coefficient.%Cpt = power coefficient.
%av_T = average torq%av_T = average torq%av_T = average torq%av_T = average torque (Nm).ue (Nm).ue (Nm).ue (Nm).
%%CONSTANTS FOR STANDARD AIR CONDITIONS AND SEA LEVEL%%CONSTANTS FOR STANDARD AIR CONDITIONS AND SEA LEVEL%%CONSTANTS FOR STANDARD AIR CONDITIONS AND SEA LEVEL%%CONSTANTS FOR STANDARD AIR CONDITIONS AND SEA LEVEL
kv = 1.4607ekv = 1.4607ekv = 1.4607ekv = 1.4607e----5; % Kinematic viscosity at 15ºC [m^2/s].5; % Kinematic viscosity at 15ºC [m^2/s].5; % Kinematic viscosity at 15ºC [m^2/s].5; % Kinematic viscosity at 15ºC [m^2/s].
rho = 1.225; % air density (Standard density at sea level)[kg/m^3]rho = 1.225; % air density (Standard density at sea level)[kg/m^3]rho = 1.225; % air density (Standard density at sea level)[kg/m^3]rho = 1.225; % air density (Standard density at sea level)[kg/m^3]
%%OTHER PARAMETERS%%OTHER PARAMETERS%%OTHER PARAMETERS%%OTHER PARAMETERS
n=36; % number of streamtubes (180/5)n=36; % number of streamtubes (180/5)n=36; % number of streamtubes (180/5)n=36; % number of streamtubes (180/5) divides one half of the rotor in 5ºdivides one half of the rotor in 5ºdivides one half of the rotor in 5ºdivides one half of the rotor in 5º
% increments.% increments.% increments.% increments.
% theta = vector containing the angles between streamtube and local radius% theta = vector containing the angles between streamtube and local radius% theta = vector containing the angles between streamtube and local radius% theta = vector containing the angles between streamtube and local radius
% to the rotor axis.% to the rotor axis.% to the rotor axis.% to the rotor axis.
thetau = linspace (thetau = linspace (thetau = linspace (thetau = linspace (----89*pi/180, 89*pi/180, n); % upstream angles (rad).89*pi/180, 89*pi/180, n); % upstream angles (rad).89*pi/180, 89*pi/180, n); % upstream angles (rad).89*pi/180, 89*pi/180, n); % upstream angles (rad).
thetad = linspace (91*pi/180,thetad = linspace (91*pi/180,thetad = linspace (91*pi/180,thetad = linspace (91*pi/180, 269*pi/180, n); % downstream angles (rad).269*pi/180, n); % downstream angles (rad).269*pi/180, n); % downstream angles (rad).269*pi/180, n); % downstream angles (rad).
Xt = w*R/Vo; %Tip speed ratioXt = w*R/Vo; %Tip speed ratioXt = w*R/Vo; %Tip speed ratioXt = w*R/Vo; %Tip speed ratio

61
S = 2*L*R; %Swept area [m^2]S = 2*L*R; %Swept area [m^2]S = 2*L*R; %Swept area [m^2]S = 2*L*R; %Swept area [m^2]
%%%%%%%%------------------------------------------------------------ UPSTREAM CALCULATIONUPSTREAM CALCULATIONUPSTREAM CALCULATIONUPSTREAM CALCULATION ------------------------------------------------------------------------------------
%%Get the corresponding au value for each s%%Get the corresponding au value for each s%%Get the corresponding au value for each s%%Get the corresponding au value for each streamtube (each theta value)treamtube (each theta value)treamtube (each theta value)treamtube (each theta value)
i = 0; %initialize the counter valuei = 0; %initialize the counter valuei = 0; %initialize the counter valuei = 0; %initialize the counter value
while (i~=n)while (i~=n)while (i~=n)while (i~=n)
i = i+1;i = i+1;i = i+1;i = i+1;
au = 1.01; % velocity induction factor upstream.au = 1.01; % velocity induction factor upstream.au = 1.01; % velocity induction factor upstream.au = 1.01; % velocity induction factor upstream.
newau = 1; % initialize, au must be different from newaunewau = 1; % initialize, au must be different from newaunewau = 1; % initialize, au must be different from newaunewau = 1; % initialize, au must be different from newau
while ((auwhile ((auwhile ((auwhile ((au----newau)>1enewau)>1enewau)>1enewau)>1e----3) %Iterative process to find au3) %Iterative process to find au3) %Iterative process to find au3) %Iterative process to find au
au = newau;au = newau;au = newau;au = newau;
Vu = Vo*(au); %Vu =velocity upstream of wind turbine cylinderVu = Vo*(au); %Vu =velocity upstream of wind turbine cylinderVu = Vo*(au); %Vu =velocity upstream of wind turbine cylinderVu = Vo*(au); %Vu =velocity upstream of wind turbine cylinder
X = R*w/Vu; %Local Tip speed ratioX = R*w/Vu; %Local Tip speed ratioX = R*w/Vu; %Local Tip speed ratioX = R*w/Vu; %Local Tip speed ratio
% Wu=local resultant air velocity% Wu=local resultant air velocity% Wu=local resultant air velocity% Wu=local resultant air velocity
WuWuWuWu = sqrt ( Vu^2*( (X= sqrt ( Vu^2*( (X= sqrt ( Vu^2*( (X= sqrt ( Vu^2*( (X ---- sin (thetau(i)))^2 + (cos (thetau(i)))^2));sin (thetau(i)))^2 + (cos (thetau(i)))^2));sin (thetau(i)))^2 + (cos (thetau(i)))^2));sin (thetau(i)))^2 + (cos (thetau(i)))^2));
Reb = Wu*c/kv; %Blade Reynolds numberReb = Wu*c/kv; %Blade Reynolds numberReb = Wu*c/kv; %Blade Reynolds numberReb = Wu*c/kv; %Blade Reynolds number
% the values from airfoil lift and drag coefficients depending on% the values from airfoil lift and drag coefficients depending on% the values from airfoil lift and drag coefficients depending on% the values from airfoil lift and drag coefficients depending on
% the angle of attack are interpolated for the given Reynolds num% the angle of attack are interpolated for the given Reynolds num% the angle of attack are interpolated for the given Reynolds num% the angle of attack are interpolated for the given Reynolds num----
% ber% ber% ber% ber
if (NACA==15)if (NACA==15)if (NACA==15)if (NACA==15)
[A1, Cl1, Cd1] = NACAfinder15 (Reb); %NACA0015[A1, Cl1, Cd1] = NACAfinder15 (Reb); %NACA0015[A1, Cl1, Cd1] = NACAfinder15 (Reb); %NACA0015[A1, Cl1, Cd1] = NACAfinder15 (Reb); %NACA0015
elseelseelseelse
[A1, Cl1, Cd1] = NACAfinder21 (Reb); %NACA0021[A1, Cl1, Cd1] = NACAfinder21 (Reb); %NACA0021[A1, Cl1, Cd1] = NACAfinder21 (Reb); %NACA0021[A1, Cl1, Cd1] = NACAfinder21 (Reb); %NACA0021
endendendend
costh = cos(thetau(i));costh = cos(thetau(i));costh = cos(thetau(i));costh = cos(thetau(i));
cosao = cos(Ao);cosao = cos(Ao);cosao = cos(Ao);cosao = cos(Ao);
sinth = sin(thetau(i));sinth = sin(thetau(i));sinth = sin(thetau(i));sinth = sin(thetau(i));
sinao = sin(Ao);sinao = sin(Ao);sinao = sin(Ao);sinao = sin(Ao);
A = asin((costh*cosaoA = asin((costh*cosaoA = asin((costh*cosaoA = asin((costh*cosao----(X(X(X(X----sinth)*sinao)/sqrt((Xsinth)*sinao)/sqrt((Xsinth)*sinao)/sqrt((Xsinth)*sinao)/sqrt((X----sinth)^2+(costh^2)));sinth)^2+(costh^2)));sinth)^2+(costh^2)));sinth)^2+(costh^2)));

62
% if angle% if angle% if angle% if angle of attack is negative the sign is changed for interpolaof attack is negative the sign is changed for interpolaof attack is negative the sign is changed for interpolaof attack is negative the sign is changed for interpola----
% ting in the data tables and then the lift coefficient sign is% ting in the data tables and then the lift coefficient sign is% ting in the data tables and then the lift coefficient sign is% ting in the data tables and then the lift coefficient sign is
% changed.% changed.% changed.% changed.
neg = 0;neg = 0;neg = 0;neg = 0;
if (sign(A)==if (sign(A)==if (sign(A)==if (sign(A)==----1)1)1)1)
neg = 1;neg = 1;neg = 1;neg = 1;
endendendend
A = abs(A*180/pi); %ConvA = abs(A*180/pi); %ConvA = abs(A*180/pi); %ConvA = abs(A*180/pi); %Conversion to degrees to match tableersion to degrees to match tableersion to degrees to match tableersion to degrees to match table
Cl = interp1 (A1, Cl1, A, 'linear');Cl = interp1 (A1, Cl1, A, 'linear');Cl = interp1 (A1, Cl1, A, 'linear');Cl = interp1 (A1, Cl1, A, 'linear');
Cd = interp1 (A1, Cd1, A, 'linear');Cd = interp1 (A1, Cd1, A, 'linear');Cd = interp1 (A1, Cd1, A, 'linear');Cd = interp1 (A1, Cd1, A, 'linear');
if (neg==1)if (neg==1)if (neg==1)if (neg==1)
A =A =A =A = ----1*A; %Restablish the original value for plots1*A; %Restablish the original value for plots1*A; %Restablish the original value for plots1*A; %Restablish the original value for plots
Cl =Cl =Cl =Cl =----1*Cl;1*Cl;1*Cl;1*Cl;
Cd =1*Cd;Cd =1*Cd;Cd =1*Cd;Cd =1*Cd;
endendendend
% Cn = normal coefficient, % Ct = tangential coefficient% Cn = normal coefficient, % Ct = tangential coefficient% Cn = normal coefficient, % Ct = tangential coefficient% Cn = normal coefficient, % Ct = tangential coefficient
Cnu = Cl*cosd (A) + Cd*sind (A); %note that A is in degreesCnu = Cl*cosd (A) + Cd*sind (A); %note that A is in degreesCnu = Cl*cosd (A) + Cd*sind (A); %note that A is in degreesCnu = Cl*cosd (A) + Cd*sind (A); %note that A is in degrees
Ctu = Cl*sind (A)Ctu = Cl*sind (A)Ctu = Cl*sind (A)Ctu = Cl*sind (A) ---- Cd*cosd (A);Cd*cosd (A);Cd*cosd (A);Cd*cosd (A);
% fup = function to find interfer% fup = function to find interfer% fup = function to find interfer% fup = function to find interference factor (au).ence factor (au).ence factor (au).ence factor (au).
g=@(thetau) (abs(sec (thetau)).*(Cnu.*cos(thetau)g=@(thetau) (abs(sec (thetau)).*(Cnu.*cos(thetau)g=@(thetau) (abs(sec (thetau)).*(Cnu.*cos(thetau)g=@(thetau) (abs(sec (thetau)).*(Cnu.*cos(thetau)----
Ctu.*sin(thetau)).*(Wu./Vu).^2);Ctu.*sin(thetau)).*(Wu./Vu).^2);Ctu.*sin(thetau)).*(Wu./Vu).^2);Ctu.*sin(thetau)).*(Wu./Vu).^2);
y = quadl (g,y = quadl (g,y = quadl (g,y = quadl (g, ----89*pi/180, 89*pi/180);89*pi/180, 89*pi/180);89*pi/180, 89*pi/180);89*pi/180, 89*pi/180);
fup = N*c*y/(8*pi*R);fup = N*c*y/(8*pi*R);fup = N*c*y/(8*pi*R);fup = N*c*y/(8*pi*R);
%[newau] = solve ('fup*newau = pi*(1%[newau] = solve ('fup*newau = pi*(1%[newau] = solve ('fup*newau = pi*(1%[newau] = solve ('fup*newau = pi*(1----newau)')newau)')newau)')newau)')
%syms ne%syms ne%syms ne%syms newauwauwauwau
%S1 = solve ('fup*newau*R = pi*(1%S1 = solve ('fup*newau*R = pi*(1%S1 = solve ('fup*newau*R = pi*(1%S1 = solve ('fup*newau*R = pi*(1----newau)*r(k)', 'newau')newau)*r(k)', 'newau')newau)*r(k)', 'newau')newau)*r(k)', 'newau')
newau = pi/(fup+pi); % new interference factor value for the nextnewau = pi/(fup+pi); % new interference factor value for the nextnewau = pi/(fup+pi); % new interference factor value for the nextnewau = pi/(fup+pi); % new interference factor value for the next
% iteration process.% iteration process.% iteration process.% iteration process.
endendendend
end of while Au....

63
Auvector (i) = A; % Store angle of attack valueAuvector (i) = A; % Store angle of attack valueAuvector (i) = A; % Store angle of attack valueAuvector (i) = A; % Store angle of attack value
auvector (i) = newau; % Store au value in a vectorauvector (i) = newau; % Store au value in a vectorauvector (i) = newau; % Store au value in a vectorauvector (i) = newau; % Store au value in a vector
Fnu (i) = (c*L/S)*Cnu*(Wu/Vo)^2; % normal force coefficientFnu (i) = (c*L/S)*Cnu*(Wu/Vo)^2; % normal force coefficientFnu (i) = (c*L/S)*Cnu*(Wu/Vo)^2; % normal force coefficientFnu (i) = (c*L/S)*Cnu*(Wu/Vo)^2; % normal force coefficient
Ftu (i) = (c*L/S)*Ctu*(Wu/Vo)^2; % tangential force coefficientFtu (i) = (c*L/S)*Ctu*(Wu/Vo)^2; % tangential force coefficientFtu (i) = (c*L/S)*Ctu*(Wu/Vo)^2; % tangential force coefficientFtu (i) = (c*L/S)*Ctu*(Wu/Vo)^2; % tangential force coefficient
Tup (i) = 0.5*rho*c*R*L*Ctu*Wu^2; % torque produced in when the bladeTup (i) = 0.5*rho*c*R*L*Ctu*Wu^2; % torque produced in when the bladeTup (i) = 0.5*rho*c*R*L*Ctu*Wu^2; % torque produced in when the bladeTup (i) = 0.5*rho*c*R*L*Ctu*Wu^2; % torque produced in when the blade
% crosses this streamtube% crosses this streamtube% crosses this streamtube% crosses this streamtube
endendendend
% Average upstream torque% Average upstream torque% Average upstream torque% Average upstream torque
ts2 = f_trapezoidal_integration_s (thetau, Tup)ts2 = f_trapezoidal_integration_s (thetau, Tup)ts2 = f_trapezoidal_integration_s (thetau, Tup)ts2 = f_trapezoidal_integration_s (thetau, Tup)
av_Tup = N*(ts2)/(2*pi) %upstreamav_Tup = N*(ts2)/(2*pi) %upstreamav_Tup = N*(ts2)/(2*pi) %upstreamav_Tup = N*(ts2)/(2*pi) %upstream average torqueaverage torqueaverage torqueaverage torque
% Average torque coefficient% Average torque coefficient% Average torque coefficient% Average torque coefficient
av_Cqu = av_Tup/(0.5*rho*S*R*Vo^2)av_Cqu = av_Tup/(0.5*rho*S*R*Vo^2)av_Cqu = av_Tup/(0.5*rho*S*R*Vo^2)av_Cqu = av_Tup/(0.5*rho*S*R*Vo^2)
Cpu = av_Cqu*Xt; %upstream power coefficientCpu = av_Cqu*Xt; %upstream power coefficientCpu = av_Cqu*Xt; %upstream power coefficientCpu = av_Cqu*Xt; %upstream power coefficient
%%%%%%%%------------------------------------------------------------ DOWNSTREAM CALCULATIONDOWNSTREAM CALCULATIONDOWNSTREAM CALCULATIONDOWNSTREAM CALCULATION ------------------------------------------------------------------------------------
j = n+1;j = n+1;j = n+1;j = n+1;
flag =0;flag =0;flag =0;flag =0;
i = 0; %i = 0; %i = 0; %i = 0; %initialize the counter valueinitialize the counter valueinitialize the counter valueinitialize the counter value
while (j~=1)while (j~=1)while (j~=1)while (j~=1)
j = jj = jj = jj = j----1;1;1;1;
i = i+1; % interference factor downstream.i = i+1; % interference factor downstream.i = i+1; % interference factor downstream.i = i+1; % interference factor downstream.
ad = 1.01; % velocity induction factor upstream.ad = 1.01; % velocity induction factor upstream.ad = 1.01; % velocity induction factor upstream.ad = 1.01; % velocity induction factor upstream.
newad = auvector(j); % initialize, au must be different from newaunewad = auvector(j); % initialize, au must be different from newaunewad = auvector(j); % initialize, au must be different from newaunewad = auvector(j); % initialize, au must be different from newau
while ((adwhile ((adwhile ((adwhile ((ad----newad)>1enewad)>1enewad)>1enewad)>1e----3) %Iterative process to find ad3) %Iterative process to find ad3) %Iterative process to find ad3) %Iterative process to find ad
ad = newad;ad = newad;ad = newad;ad = newad;
Ve = Vo*((2*auvector(j))Ve = Vo*((2*auvector(j))Ve = Vo*((2*auvector(j))Ve = Vo*((2*auvector(j))----1); %Ve = air velocity inside cylinder1); %Ve = air velocity inside cylinder1); %Ve = air velocity inside cylinder1); %Ve = air velocity inside cylinder
Vd = Ve*ad; %Vd = air velocity downstream of cylinderVd = Ve*ad; %Vd = air velocity downstream of cylinderVd = Ve*ad; %Vd = air velocity downstream of cylinderVd = Ve*ad; %Vd = air velocity downstream of cylinder
X = R*w/Vd; %LocaX = R*w/Vd; %LocaX = R*w/Vd; %LocaX = R*w/Vd; %Local Tip speed ratiol Tip speed ratiol Tip speed ratiol Tip speed ratio
% Wd=local resultant air velocity% Wd=local resultant air velocity% Wd=local resultant air velocity% Wd=local resultant air velocity
end of while n~=1

64
Wd = sqrt ( Vd^2*( (XWd = sqrt ( Vd^2*( (XWd = sqrt ( Vd^2*( (XWd = sqrt ( Vd^2*( (X ---- sin (thetad(i)))^2 + (cos (thetad(i)))^2));sin (thetad(i)))^2 + (cos (thetad(i)))^2));sin (thetad(i)))^2 + (cos (thetad(i)))^2));sin (thetad(i)))^2 + (cos (thetad(i)))^2));
Reb = Wd*c/kv; %Reynolds number of the bladeReb = Wd*c/kv; %Reynolds number of the bladeReb = Wd*c/kv; %Reynolds number of the bladeReb = Wd*c/kv; %Reynolds number of the blade
if (NACA==15)if (NACA==15)if (NACA==15)if (NACA==15)
[A1, Cl1, Cd1] = NACAfinder15 (Reb);[A1, Cl1, Cd1] = NACAfinder15 (Reb);[A1, Cl1, Cd1] = NACAfinder15 (Reb);[A1, Cl1, Cd1] = NACAfinder15 (Reb);
elseelseelseelse
[A1, Cl1, Cd1] = NACAfinder21 (Reb);[A1, Cl1, Cd1] = NACAfinder21 (Reb);[A1, Cl1, Cd1] = NACAfinder21 (Reb);[A1, Cl1, Cd1] = NACAfinder21 (Reb);
endendendend
costh = cos(thetad(i));costh = cos(thetad(i));costh = cos(thetad(i));costh = cos(thetad(i));
cosao = cos(Ao);cosao = cos(Ao);cosao = cos(Ao);cosao = cos(Ao);
sinth = sin(thetad(i));sinth = sin(thetad(i));sinth = sin(thetad(i));sinth = sin(thetad(i));
sinao = sin(Ao);sinao = sin(Ao);sinao = sin(Ao);sinao = sin(Ao);
AAAA = asin((costh*cosao= asin((costh*cosao= asin((costh*cosao= asin((costh*cosao----(X(X(X(X----sinth)*sinao)/sqrt((Xsinth)*sinao)/sqrt((Xsinth)*sinao)/sqrt((Xsinth)*sinao)/sqrt((X----sinth)^2+(costh^2)));sinth)^2+(costh^2)));sinth)^2+(costh^2)));sinth)^2+(costh^2)));
neg = 0;neg = 0;neg = 0;neg = 0;
if (sign(A)==if (sign(A)==if (sign(A)==if (sign(A)==----1)1)1)1)
neg = 1;neg = 1;neg = 1;neg = 1;
endendendend
A = abs(A*180/pi); %Conversion to degrees to match tableA = abs(A*180/pi); %Conversion to degrees to match tableA = abs(A*180/pi); %Conversion to degrees to match tableA = abs(A*180/pi); %Conversion to degrees to match table
Cl = interp1 (A1, Cl1, A, 'linear');Cl = interp1 (A1, Cl1, A, 'linear');Cl = interp1 (A1, Cl1, A, 'linear');Cl = interp1 (A1, Cl1, A, 'linear');
Cd = interp1 (A1, Cd1, A, 'linear');Cd = interp1 (A1, Cd1, A, 'linear');Cd = interp1 (A1, Cd1, A, 'linear');Cd = interp1 (A1, Cd1, A, 'linear');
if (neg==1)if (neg==1)if (neg==1)if (neg==1)
A =A =A =A = ----1*A; %Restablish the original value for plots1*A; %Restablish the original value for plots1*A; %Restablish the original value for plots1*A; %Restablish the original value for plots
Cl =Cl =Cl =Cl =----1*Cl;1*Cl;1*Cl;1*Cl;
Cd =1*Cd;Cd =1*Cd;Cd =1*Cd;Cd =1*Cd;
endendendend
% Cn = normal coefficient, % Ct = tangential coeff% Cn = normal coefficient, % Ct = tangential coeff% Cn = normal coefficient, % Ct = tangential coeff% Cn = normal coefficient, % Ct = tangential coefficienticienticienticient
Cnd = Cl*cosd (A) + Cd*sind (A); %note that A is in degreesCnd = Cl*cosd (A) + Cd*sind (A); %note that A is in degreesCnd = Cl*cosd (A) + Cd*sind (A); %note that A is in degreesCnd = Cl*cosd (A) + Cd*sind (A); %note that A is in degrees
Ctd = Cl*sind (A)Ctd = Cl*sind (A)Ctd = Cl*sind (A)Ctd = Cl*sind (A) ---- Cd*cosd (A);Cd*cosd (A);Cd*cosd (A);Cd*cosd (A);
g=@(thetad) (abs(sec (thetad)).*(Cnd.*cos(thetad)g=@(thetad) (abs(sec (thetad)).*(Cnd.*cos(thetad)g=@(thetad) (abs(sec (thetad)).*(Cnd.*cos(thetad)g=@(thetad) (abs(sec (thetad)).*(Cnd.*cos(thetad)----
Ctd.*sin(thetad)).*(Wd./Vd).^2);Ctd.*sin(thetad)).*(Wd./Vd).^2);Ctd.*sin(thetad)).*(Wd./Vd).^2);Ctd.*sin(thetad)).*(Wd./Vd).^2);
y = quadl (g, 91*pi/180, 269*pi/180);y = quadl (g, 91*pi/180, 269*pi/180);y = quadl (g, 91*pi/180, 269*pi/180);y = quadl (g, 91*pi/180, 269*pi/180);
fdw = N*c*y/(8*pi*R);fdw = N*c*y/(8*pi*R);fdw = N*c*y/(8*pi*R);fdw = N*c*y/(8*pi*R);
%[newau] = solve ('fup*newau = pi*(1%[newau] = solve ('fup*newau = pi*(1%[newau] = solve ('fup*newau = pi*(1%[newau] = solve ('fup*newau = pi*(1----newau)')newau)')newau)')newau)')

65
%syms newau%syms newau%syms newau%syms newau
%S1 = solve ('fup*newau*R = pi*(1%S1 = solve ('fup*newau*R = pi*(1%S1 = solve ('fup*newau*R = pi*(1%S1 = solve ('fup*newau*R = pi*(1----newau)*r(k)', 'newau')newau)*r(k)', 'newau')newau)*r(k)', 'newau')newau)*r(k)', 'newau')
if (flag ==0)if (flag ==0)if (flag ==0)if (flag ==0)
newad = pi/(fdw+newad = pi/(fdw+newad = pi/(fdw+newad = pi/(fdw+pi);pi);pi);pi);
endendendend
%if the iteration process does not converge, the value of the%if the iteration process does not converge, the value of the%if the iteration process does not converge, the value of the%if the iteration process does not converge, the value of the
%interference factor upstream from the same streamtube is taken%interference factor upstream from the same streamtube is taken%interference factor upstream from the same streamtube is taken%interference factor upstream from the same streamtube is taken
if (newad<0.01)if (newad<0.01)if (newad<0.01)if (newad<0.01)
warning('newad<warning('newad<warning('newad<warning('newad<0.01 at theta = %d and A = %d', (thetad(i)*180/pi),0.01 at theta = %d and A = %d', (thetad(i)*180/pi),0.01 at theta = %d and A = %d', (thetad(i)*180/pi),0.01 at theta = %d and A = %d', (thetad(i)*180/pi),
A);A);A);A);
if (i>1)if (i>1)if (i>1)if (i>1)
newad = advector(inewad = advector(inewad = advector(inewad = advector(i----1);1);1);1);
elseelseelseelse
newad = auvector (i);newad = auvector (i);newad = auvector (i);newad = auvector (i);
endendendend
flag = 1;flag = 1;flag = 1;flag = 1;
endendendend
endendendend
Advector (i) = A;Advector (i) = A;Advector (i) = A;Advector (i) = A;
advector (i) = newad; %Store ad value in a vectoradvector (i) = newad; %Store ad value in a vectoradvector (i) = newad; %Store ad value in a vectoradvector (i) = newad; %Store ad value in a vector
% Force coefficients and torque calculation% Force coefficients and torque calculation% Force coefficients and torque calculation% Force coefficients and torque calculation
Fnd (i) = (c*L/S)*Cnd*(Wd/Vo)^2;Fnd (i) = (c*L/S)*Cnd*(Wd/Vo)^2;Fnd (i) = (c*L/S)*Cnd*(Wd/Vo)^2;Fnd (i) = (c*L/S)*Cnd*(Wd/Vo)^2;
Ftd (i) = (c*L/S)*Ctd*(Wd/Vo)^2;Ftd (i) = (c*L/S)*Ctd*(Wd/Vo)^2;Ftd (i) = (c*L/S)*Ctd*(Wd/Vo)^2;Ftd (i) = (c*L/S)*Ctd*(Wd/Vo)^2;
Tdw (i) = 0.5*rho*c*R*L*Ctd*Wd^2;Tdw (i) = 0.5*rho*c*R*L*Ctd*Wd^2;Tdw (i) = 0.5*rho*c*R*L*Ctd*Wd^2;Tdw (i) = 0.5*rho*c*R*L*Ctd*Wd^2;
endendendend
% Average upstream torque% Average upstream torque% Average upstream torque% Average upstream torque
ts4 = fts4 = fts4 = fts4 = f_trapezoidal_integration_s (thetad, Tdw)_trapezoidal_integration_s (thetad, Tdw)_trapezoidal_integration_s (thetad, Tdw)_trapezoidal_integration_s (thetad, Tdw)
av_Tdw = N*(ts4)/(2*pi) % upstream average torqueav_Tdw = N*(ts4)/(2*pi) % upstream average torqueav_Tdw = N*(ts4)/(2*pi) % upstream average torqueav_Tdw = N*(ts4)/(2*pi) % upstream average torque
% Average torque coefficient% Average torque coefficient% Average torque coefficient% Average torque coefficient
av_Cqd = av_Tdw/(0.5*rho*S*R*Vo^2)av_Cqd = av_Tdw/(0.5*rho*S*R*Vo^2)av_Cqd = av_Tdw/(0.5*rho*S*R*Vo^2)av_Cqd = av_Tdw/(0.5*rho*S*R*Vo^2)
Cpd = av_Cqd*Xt; % downstream power coefficientCpd = av_Cqd*Xt; % downstream power coefficientCpd = av_Cqd*Xt; % downstream power coefficientCpd = av_Cqd*Xt; % downstream power coefficient
end of while Add-new
end of while n~=1

66
Cpt = Cpd+Cpu % Total power coefficientCpt = Cpd+Cpu % Total power coefficientCpt = Cpd+Cpu % Total power coefficientCpt = Cpd+Cpu % Total power coefficient
av_T = av_Tup + av_Tdw % Total average torque [Nm]av_T = av_Tup + av_Tdw % Total average torque [Nm]av_T = av_Tup + av_Tdw % Total average torque [Nm]av_T = av_Tup + av_Tdw % Total average torque [Nm]

67
NACA0021 AIRFOIL DIMENSIONS APPENDIX 6
Table 9. NACA0021 airfoil ordinates and measures.
Ordinates chord 28 cm
x(%c) y(%c) x y
0 0,000 0 0
1,25 3,315 0,35 0,9282
2,5 4,576 0,7 1,2813
5 6,221 1,4 1,7419
7,5 7,350 2,1 2,058
10 8,195 2,8 2,2946
15 9,354 4,2 2,6191
20 10,040 5,6 2,8112
25 10,397 7 2,9112
30 10,504 8,4 2,9411
40 10,156 11,2 2,8437
50 9,265 14 2,5942
60 7,986 16,8 2,2361
70 6,412 19,6 1,7954
80 4,591 22,4 1,2855
90 2,534 25,2 0,7095
95 1,412 26,6 0,3954
100 0,221 28 0,0619
L.E. radius (%c) 4,85

68
WOOD PROPERTIES AND STRUCTURAL STRENGTHS APPENDIX 7
Table 10. Wood properties (Forest Products Laboratory, 2010)
Timber type
Sitka Spruce
(Picea sit-
chensis)
Balsa (Och-
roma py-
ramidale)
Obeche
(Triplochiton
scleroxylon)
Basic Specific gravity 0.40 0.16 0.3
Volumetric Shrinkage from green to
ovendry moisture content [%]
11.5 10.8 9.2
Calculated density at 19% moisture
content [kg/m3]
496.95 198.25 369.46
Compression parallel to grain [kPa] 38700 14900 27100
Shear parallel to grain [kPa] 7900 2100 6800
Average parallel to grain tensile
strength [kPa]
59300
Modulus of Elasticity [MPa] 10800 3400 5900
Modulus of rupture [kPa] 70000 21600 51000
The wood density (included on table 9) varies in function of moisture content,
having the specific gravity of the wood and the moisture content MC, the density
is calculated using the procedure stated at (Forest Products Laboratory, 2010):
ߩெ஼ = ߩ௪ ‫ܩ‬ெ஼(1 +
‫ܥܯ‬
100
) [1.1]
where ρw is the water density at 4°C which is considered to be (1000 kg/m3
), x
is the moisture content in % and Gx is the specific gravity at the given moisture
content, which is calculated using:
‫ܩ‬ெ஼ = ‫ܩ‬௕/(1 −
ܵெ஼
100
) [1.2]
where Gb is the basic specific gravity and Sx is calculated using:
ܵெ஼ = ܵ௢ ቆ1 −
‫ܥܯ‬
‫ܥܯ‬௙௧
ቇ [1.3]
where MCft is the fiber saturation point taken as 30% when unknown and So is
the wood volumetric percent shrinkage from the green condition to ovendry
which depends on wood type and its value is found at (Forest Products
Laboratory, 2010, pág. 4.8), the moisture content will be assumed to be 19%
which is the value of moisture content at which wood is considered dry by the
American Softwood Lumber Standard (Forest Products Laboratory, 2010, pág.
6.18).

69
Table 11. Maximum bending moment and shear stress for the three rotor configurations
Maximum Bending moment Maximum Shear Stress
First ‫ݍ‬
(‫ܮ‬ଵ)ଶ
2
‫ݍ‬ ‫ܮ‬ଵ
Second ‫ݍ‬
‫ܮ‬ଶ
8
‫ݍ‬
‫ܮ‬
2
Third ‫ݍ‬
‫ܮ‬ଶ
8
‫ݍ‬
‫ܮ‬
2
Table 12. Airfoil geometric properties for use in the structural calculations.
Blade airfoil geometric properties used to predict blade stresses
Chord length [m] 0,16 0,2 0,24 0,28 0,32
Total Area [mm]^2: 3682,3 5753,59 8285,17 11277,04 14729,19
Centroid X [mm]: 67,25 84,06 100,87 117,69 134,5
Centroid Y [mm]: 0 0 0 0 0
Area Moment of
Inertia in X (lxx)
[mm^4]: 239204,84 583996,18 1210974,48 2243479,73 3827277,37
Area Moment of
Inertia in Y (lyy)
[mm^4]: 5198450,12 12691528,62 26317153,75 48755776,35 83175201,98
Upper half blade airfoil geometric properties
Area [mm]^2: 1841,15 2876,79 4142,58 5638,52 7364,59
Centroid X [mm]: 67,25 84,06 100,87 117,69 134,5
Centroid Y [mm]: 6,75 8,44 10,13 11,82 13,51
Area Moment of
Inertia in (lxx)
[mm^4]: 35597,2 86907,23 180210,83 333862,81 569555,21
Area Moment of
Inertia in (lyy)
[mm^4]: 2599225,06 6345764,31 13158576,88 24377888,18 41587600,99
Table 13. Rotor blade limits depending on construction material.
Sitka spruce Obeche Balsa
Maximum allowable
bending load [kN]
45.95 32.17 17.69
Blade mass [kg] 11.2 8.33 4.47
Maximum revolutions
[rpm]
604.15 582.85 583.47

Small scale vertical axis wind turbine design

More Related Content

What's hot (8)

Similar to Small scale vertical axis wind turbine design (20)

More from Phuong Dx (20)

Recently uploaded (20)

Small scale vertical axis wind turbine design