Laín Nieto Gómez 2009/2010 Curse

Contents

1.-Introduction…………………………………………………………………2

1.1.-History of wind energy........................................................................2

1.1.1.-Ancient times and farming use of the wind energy, also
pumping of water...................................................................................................2
1.1.2.- Modern times and electrical use of the wind energy.............3
1.1.2.2.-The era of inventions...........................................................3
1.1.2.3.-Scoping the utility size wind turbines to generate
electricity.....................................................................................................3
1.1.2.4.-The invention of the three more developed techniques.......4

1.2.-Types of wind turbines.........................................................................6

1.2.1.-Horizontal axis wind turbines.................................................6
1.1.2.1.-The number of blades influencing the initial torque and
explanation about the solidity......................................................................6
1.2.1.2.-Classification of the HAWT turbines in terms of the
direct/indirect contact of the wings with the air .........................................7
1.2.2.-Vertical axis wind turbines.....................................................8
1.2.3.-Darrieus rotor..........................................................................9
1.2.4.-Savonius rotor.........................................................................9
1.2.5.-Musgrove rotor......................................................................10

2.-Aerodynamic calculus...................................................................................11

2.-Power energy basics and aero dynamical calculus...............................11

2.1.-NACA airfoils..........................................................................11
2.2.-Reynolds dimensionless number calculations and variations of
drag-lift coefficients..................................................................................12
2.3.-NACA report 586, tables, basics and explanations..................14
2.4.-Relations between the coefficients and the forces, momentum
coefficient..................................................................................................18
2.5.-Outputpower and its relation with the spinning velocity.........20

2.6.-Step by step calculus of one section´s lift and drag force……23

3.-Mechanical calculus

3.1.-Axis calculation respect to DIN 743..................................................33

3.2.-Bearings selection..............................................................................34
3.2.1.-6206-RS1 (axis support one) selection justification.............34

Wind Turbine design and modelling 1

Laín Nieto Gómez 2009/2010 Curse

3.2.2.-Bearings concerning the vertical axis

spinning.....................................................................................................38
3.3.-Comprobation of tabs according to DIN 6885...................................46
3.4.-Axis break dimensioning...................................................................48

4.-Electrical calculus……………………………………………………..........58

4.1.-Generator selection............................................................................58
4.2.-Batteries selection..............................................................................61
4.3.-Transformation of the electrical output power to utility parameters of
tension voltage and frequency.............................................................................63

5.-Conclusions....................................................................................................64

6.-Bibliography..................................................................................................67

Annexes:

Draftings……………………………………………….................................[-]

Wind Turbine design and modelling 2

Laín Nieto Gómez 2009/2010 Curse

1.-Introduction:

1.1.-History of wind energy

Human efforts to harness wind for energy date back to the ancient times, when
he used sails to propel ships and boats. Later, wind energy served the mankind
by energising his grain grinding mills and water pumps. During its
transformation from these crude and heavy devices to today’s efficient and
sophisticated machines, the technology went through various phases of
development.

1.1.1. - Ancient times and farming use of the wind energy, also pumping of
water.

There is disagreement on the origin of the concept of using wind for mechanical
power. Some believe that the concept originated in ancient Babylonia. The
Babylonian emperor Hammurabi planned to use wind power for his ambitious
irrigation project during seventeenth century B.C. Others argue that the birth
place of wind mills is India. In Arthasastra, a classic work in Sanskrit written by
Kautiliya during 4th century B.C., references are seen on lifting water with
contrivances operated by wind. However, there are no records to prove that
these concepts got transformed to real hardware.

The earliest documented design of wind mill dates back to 200 B.C. The
Persians used wind mills for grinding grains during this period. Those were
vertical axis machines having sails made with bundles of reeds or wood. The
grinding stone was attached to the vertical shaft. The sails were attached to the
central shaft using horizontal struts. The size of the sails was decided by the
materials used for its fabrication, usually 5 m long and 9 m tall. By the 13th
century, grain grinding mills were popular in most of Europe. The French
adopted this technology by 1105 A.D. and the English by 1191 A.D. In contrast
with the vertical axis Persian design, European mills had horizontal axis. These
post mills were built with beautiful structures. The tower was circular or
polygonal in cross-section and constructed in wood or brick. The rotor was
manually oriented to the wind by adjusting the tail. The mill was protected
against high winds by turning the rotor out of the wind or removing the canvas
covering the rotor.

The Dutch, with renowned designer Jan Adriaenszoon, were the pioneers in
making these mills. They made many improvements in the design and invented
several types of mills. Examples are the tjasker and smock mills. The rotors
were made with crude airfoil profile to improve the efficiency. Apart from grain
grinding, wind mills were employed to drain marshy lands in Holland. These
wind mills reached America by mid-1700, through the Dutch settlers.
Wind Turbine design and modelling 3
Laín Nieto Gómez 2009/2010 Curse

This is followed by the water pumping wind mill, which is still considered as
one of the most successful application of wind power. The so-called American
multi bladed wind turbine appeared in the wind energy history by the mid-1800.
Relatively smaller rotors, ranging from one to several meters in diameter, were
used for this application. The primary motive was to pump water from a few
meters below the surface for agricultural uses. These water pumpers, with its
metallic blades and better engineering design, offered good field performance.

1.1.2.- Modern times and electrical use of the wind energy:

1.1.2.1.- The era of inventions:

Over six million of such units were installed in US alone, between 1850 and
1930.The era of wind electric generators began close to 1900’s. The first
modern wind turbine, specifically designed for electricity generation, was
constructed in Denmark in 1890. It supplied electricity to the rural areas.

During the same period, a large wind electric generator having 17 m ‘picket
fence’ rotor was built in Cleveland, Ohio. For the first time, a speed-up gear box
was introduced in the design. This system operated for 20 years generating its
rated power of 12 kW. More systematic methods were adopted for the
engineering design of turbines during this period. With low-solidity rotors and
aerodynamically designed blades, these systems could give impressive field
performance. By 1910, several hundreds of such machines were supplying
electrical power to the villages in Denmark. By about1925, wind electric
generators became commercially available in the American market. Similarly,
two and three bladed propeller turbines ranging from 0.2 to 3 kW in capacity
were available for charging batteries.

1.1.2.3.- Scoping the utility size wind turbines to generate electricity

Turbines with bigger capacity were also developed during this period. The first
utility-scale system was installed in Russia in 1931. A 100 kW turbine was
installed on the Caspian sea shore, which worked for two years and generated
about 20,000 kW.h electricity. Experimental wind plants were subsequently
constructed in other countries like United States, Denmark, France, Germany,
and Great Britain. A significant development in large-scale systems was the
1250 kW turbine fabricated by Palmer C. Putman. The turbine was put to use in
1941 at the Grandpa’s Knob, near Rutland, Vermont [8]. Its 53 m rotor was
mounted on a 34 m tall tower. This machine could achieve a constant rotor
speed by changing the blade pitch. The machine operated for 1100 hours during
the next five years, i.e., till the blades failed in 1945. The project is considered to
be a success as it could demonstrate the technical feasibility of large- scale

Wind Turbine design and modelling 4

Laín Nieto Gómez 2009/2010 Curse

wind-electric generation. Some interesting designs of wind turbine were

experimented during this period.

1.1.2.4.-The invention of the three more developed techniques

Darrieus G.J.M, a French engineer, put forth the design of Darrieus turbine in
1920,which was patented in United Sates in 1931. In contrast with the popular
horizontal axis rotors, Darrieus turbines had narrow curved blades rotating about
its vertical axis. During the same period, Julius D. Madaras invented a turbine
working on Magnus effect. Magnus effect is basically derived from the force on
a spinning cylinder placed in a stream of air. Another significant development at
this time was the Savonius rotor in Finland, invented by S.J. Savonius. This
rotor was made with two halves of a cylinder split longitudinally and arranged
radially on a vertical shaft. The transverse cross-section of the rotor resembled
the letter ‘S’ [10]. The rotor was driven by the difference in drag forces acting
on its concave and convex halves, facing the wind. Intensive research on the
behavior of wind turbines occurred during 1950's.
The concept of high tip speed ratio-low solidity turbines got introduced during
this period. For example, light-weight constant-speed rotors were developed in
Germany in 1968. They had fibre glass blades attached to simple hollow towers
supported by guy ropes. The largest of this breed was of 15 m diameter with a
rated output of 100 kW.

In the later years, cheaper and more reliable electricity, generated from fossil
fuel based plants became available. When the electricity generated from wind
cost 12 to 30 cents/kWh in 1940, the same generated from other sources was
available at 3 to 6 cents/kWh [7]. Cost of electricity from fossil fuels further
declined below 3 cents/kWh by 1970. Fossil fuels were available in plenty at a
relatively cheaper rate at that time. Several nuclear power projects were also
embarked on, believing that it would be the ultimate source for the future energy
needs. Thus, the interest in wind energy declined gradually, especially by 1970.

The oil crisis in 1973, however, forced the scientists, engineers and policy
makers to have a second thought on the fossil fuel dependence. They realised
that political tampering can restrict the availability and escalate the cost of fossil
fuels. Moreover, it was realised that the fossil fuel reserve would be exhausted
one day or the other. Nuclear power was unacceptable to many, due to safety
reasons. These factors caused the revival of interest in wind energy. Research on
resource analysis, hardware development, and cost reduction techniques were
intensified. United States entrusted its National Aeronautics and Space
Administration (NASA) with the development of6 large wind turbines. As a
result, a series of horizontal axis turbines named MOD-0,MOD-1, MOD-2 and
MOD-5 were developed . These projects were stopped by mid-1980’s due to
various reasons. During the same period, scientists at Sandia Laboratories
Wind Turbine design and modelling 5
Laín Nieto Gómez 2009/2010 Curse

focussed their research on the design and development of the Darrieus turbine.
They fabricated several models of the Darrieus machine in different sizes during
1980’s.
Research and development on wind energy are seen intensified in the later years.
A few innovative concepts like the vortex turbine, diffuser augmented design,
Musgrove rotor etc. were also proposed during that time. Prototypes of these
turbines were constructed and tested. However, only the horizontal axis
propeller design could emerge successfully on a commercial scale.

1.2.-Types of wind turbines

1.2.1.-Horizontal axis wind turbines

Horizontal axis wind turbines (HAWT) have their axis of rotation horizontal to
the ground and almost parallel to the wind stream (Fig. 2.3). Most of the
commercial wind turbines fall under this category. Horizontal axis machines
have some distinct advantages such as low cut-in wind speed and easy furling.
In general, they show relatively high power coefficient. However, the generator
and gearbox of these turbines are to be placed over the tower which makes its
design more complex and expensive. Another disadvantage is the need for the
tail or yaw drive to orient the turbine towards wind.

1.2.1.1.-The number of blades influencing the initial torque and explanation

about the solidity=Sblades / Sswept

Depending on the number of blades, horizontal axis wind turbines are further
classified as single bladed, two bladed, three bladed and multi bladed, as shown
in Fig. 1.4. Single bladed turbines are cheaper due to savings on blade materials.
The drag losses are also minimum for these turbines. However, to balance the
blade, a counter weight has to be placed opposite to the hub. Single bladed
designs are not very popular due to problems in balancing and visual
acceptability. Two bladed rotors also have these drawbacks, but to a lesser
extent. Most of the present commercial turbines used for electricity generation
have three blades.

Wind Turbine design and modelling 6

Laín Nieto Gómez 2009/2010 Curse

Fig.1.4

LeftSingle bladed, two bladed, three bladed and multi bladed turbines
Classification of wind turbines,
RightUpwind and downwind turbines

They are more stable as the aerodynamic loading will be relatively uniform.
Machines with more number of blades (6, 8, 12, 18 or even more) are also
available. The ratio between the actual blade area to the swept area of a rotor is
termed as the solidity. Hence, multi-bladed rotors are also called high solidity
rotors. These rotors can start easily as more rotor area interacts with the wind
initially. Some low solidity designs may require external starting.

Now consider two rotors, both of the same diameter, but different in number of
blades; say one with 3 blades and the other with 12 blades. Which will produce
more power at the same wind velocity? As the rotor swept area and velocity are
the same, theoretically both the rotors should produce the same power. However
aerodynamic losses are more for the rotor with more number of blades. Hence,
for the same rotor size and wind velocity, we can expect more power from the
three bladed rotor.

Then why do we need turbines with more blades? Some applications like water
pumping require high starting torque. For such systems, the torque required for
starting goes up to 3-4 times the running torque. Starting torque increases with
the solidity. Hence to develop high starting torque, water pumping wind mills
are made with multi bladed rotors.

1.2.1.2.-Classification of the HAWT turbines in terms of the direct/indirect

contact of the wings with the air
Wind Turbine design and modelling 7
Laín Nieto Gómez 2009/2010 Curse

Based on the direction of receiving the wind, HAWT can be classified as

upwind and down wind turbines as shown in Fig. 1.5. Upwind turbines have
their rotors facing the wind directly. As the wind stream passes the rotor first,
they do not have the problem of tower shadow. However, yaw mechanism is
essential for such designs to keep the rotor always facing the wind. On the other
hand, downwind machines are more flexible and may not require a yaw
mechanism. But, as the rotors are placed at the lee side of the tower, there may
be uneven loading on the blades as it passes through the shadow of the tower.

1.2.2.- Vertical axis wind turbines

The axis of rotation of vertical axis wind turbine (VAWT) is vertical to the
ground
and almost perpendicular to the wind direction as seen from Fig.1. 4. The
VAWT

Fig.1.5.-Darrieus wind turbine

can receive wind from any direction. Hence complicated yaw devices can be
eliminated. The generator and the gearbox of such systems can be housed at the
ground level, which makes the tower design simple and more economical.
Moreover the maintenance of these turbines can be done at the ground level. For
these systems, pitch control is not required when used for synchronous
applications. The major disadvantage of some VAWT is that they are usually
not self starting. Additional mechanisms may be required to ‘push’ and start the
turbine, once it is stopped. As the rotor completes its rotation, the blades have to
pass through aerodynamically dead zones which will result in lowering the
system efficiency.

There are chances that the blades may run at dangerously high speeds causing
the system to fail, if not controlled properly. Further, guy wires are required to
Wind Turbine design and modelling 8
Laín Nieto Gómez 2009/2010 Curse

support the tower structure which may pose practical difficulties. Features of
some major vertical axis designs are discussed below.

1.2.3.-Darrieus rotor

Darrieus rotor, named after its inventor Georges Jeans Darrieus, works due to
the lift force generated from a set of airfoils (Fig. 2.6). In the original design the
blades are shaped like egg beaters or troposkein (turning rope) and are under
pure tension while in operation. This typical blade configuration helps in
minimizing the bending stress experienced by the blades. There are several
variations in the Darrieus design of which some are with straight vertical blades,
usually called Giromills

Fig.1.7 A low cost Savonius wind turbine with rotors arranged 90º out of
phase

Darrieus rotor usually works at high tip speed ratio which makes it attractive for
wind electric generators. However, they are not self-starting and require external
‘excitation’ to cut-in. Moreover, the rotor produces peak torque only twice per
revolution.

1.2.4.-Savonius rotor

The Savonius wind turbine, invented by S.J. Savonius, is a vertical axis machine
consisting of two half cylindrical (or elliptical) blades arranged in ‘S’ shape
(Fig.1.7). Convex side of one of the half cylinder and the concave side of the
other are facing the wind at a time as shown in Fig. 1.8. The basic driving force
of Savonius rotor is drag. The drag coefficient of a concave surface is more than
the convex surface. Hence, the half cylinder with concave side facing the wind
will experience more drag force than the other cylinder, thus forcing the rotor to

Wind Turbine design and modelling 9

Laín Nieto Gómez 2009/2010 Curse

rotate. Sometimes two or more rotors fixed one over the other at 90O offset may
be used to smoothen the torque fluctuations during rotation.

Another way to improve the performance is to attach deflector augmenters with

the rotor. The augmenter shades the convex half facing the wind and directs the
flow to the concave half thus enhancing the performance. Being drag machines,
Savonius rotors have relatively lower power coefficient. However, some
experimental rotors have shown power coefficient up to 35 percent. These rotors
have high solidity and thus high starting torque

Fig. 1.8.-Principle of Savonius rotor

They work at low tip speed ratios, with the maximum of about 1. They are very
simple in construction-even can be made from oil barrels cut in two halves
lengthwise.
Hence they are preferred for high torque-low speed applications like water
pumping.

1.2.5.-Musgrove rotor

Musgrove rotor was developed by a research team under Prof. Musgrove at the
Reading University, UK. It is basically a vertical axis lift machine having ‘H’
shaped blades and a central shaft (Fig. 1.9). At high wind speeds the rotor
feathers and turn about a horizontal point due to the centrifugal force.

This eliminates the risk of higher aerodynamic forces on the blades and the
structure. Based on driving aerodynamic force, wind turbines are classified as
lift and drag machines. Turbines that work predominantly by the lift force are
called the lift machines and that by the drag force are called drag turbines. It is
always advantageous to utilize the lift force to run the turbine.

Wind Turbine design and modelling 10

Laín Nieto Gómez 2009/2010 Curse

Wind turbines are available in various sizes ranging from a fraction of kW to

several MW. Based on the size, the turbines may be classified as small (< 25
kW) medium (25-100 kW), large (100-1000 kW) and very large (>1000 kW)
machines.

Fig.1.9.-Principle of Musgrove turbine

Wind Turbine design and modelling 11

Laín Nieto Gómez 2009/2010 Curse

2.-Power energy basics and aero dynamical calculus.

A wind flow moving with the c velocity carries within it a kinematical energy.
This Ec depends on the fluid velocity principally but also is function of the
density. The energy which a wind turbine can take from the moving air is:

1
Ec =C p× × A×ρ×v 3
2
[2.0]

The coefficient Cp is integrated into the formula to express that it is impossible

to get profit of the whole kinematical energy inside the fluid.

2.1.-NACA airfoils:

The NACA airfoils are airfoil shapes for aircraft wings developed by the
National Advisory Committee for Aeronautics (NACA). The shape of the
NACA airfoils is described using a series of digits following the word "NACA."
The parameters in the numerical code can be entered into equations to precisely
generate the cross-section of the airfoil and calculate its properties.
In our case the airfoil provided by the company has the NACA 4412 profile
section. This kind of profiles has a different geometry depending on the number
of profile and the Chord (the maximum length of the section). The Klein wind
energy company let us choose between the kind of section we want but in a
concrete size and in a concrete geometry.

Fig. 2.1: Geometry of an airfoil section, its chord length and other
dimensions

The velocity and the angle of attack as the image shows depend on the two
components of the velocity. The first one is the air speed, which depends on the
height of the wind turbine and the atmosphere, or the surrounding barriers
erected. The angle of attack depends on the disposition of the wings and the

Wind Turbine design and modelling 12

Laín Nieto Gómez 2009/2010 Curse

relative to wing velocity of the air taking on account the spin of the wings
around the axis (namely the relative spin velocity). This compounded velocity is
the one we use to search in the abacus of the NACA report taking into account
the angle of attack.

2.2.-Reynolds dimensionless number calculations and variations of drag-lift

coefficients

Once we know the resultant of the velocities we must look in the NACA report
586 about “Airfoil section characteristics as affected by variations on the
Reynolds number”.
First of all we must define Reynolds dimensionless number:

c×L
Re=
υ
[2.1]

Where (in the International System of Units):

c: The absolute velocity (vector addition of the relative velocity and the air
velocity) [m/s]
L: The length of the Chord [m]
υ : Kinematical viscosity of the fluid in this case the air [m²/s]

The tables provided by the NACA institute need firstly the angle of attack that
can be obtained from the geometrical characteristics and angles of the blades
that the company provided us and the velocity conditions of the air. Then the
Reynolds number where we enter the compound velocity and the Length of that
concrete section.

So with the c (compound velocity of the air considering that the wings do not
spin so there are the u= ω× r component and the v velocity of the air), L (length
of the chord of the wing) and ν the kinematical viscosity of the air (supposing
it is constant), we can calculate each of the Reynolds numbers for each sections
that will make us to choose between the different curves of the abacus for that
chosen NACA blade section type.
The type of the airfoil chose for this blade is the following:

Wind Turbine design and modelling 13

Laín Nieto Gómez 2009/2010 Curse

Fig.2.2.- NACA 4412 airfoil section appearance

The information provided by the company is the following:

Fig. 2.3: Sizes and parameters of the wings provided by Heyde Windtechnik
Deutschland

Last table´s (Fig 2.3) type of blade is the 1,28/8r Wi one. The most important
characteristics are the following ones rotor diameter is of 2,55m and at 580 rpm
and with a 10m/s wind velocity the nominal power output is of 1400W.

In the 7th page of the 586 NACA report we can find dealing with “Airfoil
section characteristics as affected by variations on the Reynolds number”. We
can see for the selected section type NACA 4412 the different lift coefficients
C L , drag coefficients Cd and moment coefficients from the aerodynamic
C
centre ma.c . for each Reynolds Number.

Now it is going to be explained the use and the information inside the NACA
report. The graphics contain the following information. In the first abacus we
can see the profile geometry, instead of giving us the polynomial that describes
the upper and the down camber they give us graphically the information. Also in
this graphic we can find the position of the aerodynamic centre point. That is the
point where the resultant force of pressure is applied theoretically. The way they
provide us with a.c. point is the Y (vertical distance) and the X (horizontal
distance) from a point in the chord that is at a quarter from the attack point.

Wind Turbine design and modelling 14

Laín Nieto Gómez 2009/2010 Curse

2.3.-NACA report 586, tables, basics and explanations

Fig. 2.4: Geometry of NACA 4412 airfoil and the placement of the pressure
centre.
The numerical data of the first graphic gives us the upper and the lower
coordinates of the section in function of the chord coordinate.

Wind Turbine design and modelling 15

Laín Nieto Gómez 2009/2010 Curse

Fig. 2.5: Lift coefficients for different Reynolds numbers

Wind Turbine design and modelling 16

Laín Nieto Gómez 2009/2010 Curse

Fig. 2.5: Drag and Moment coefficients for different Reynolds numbers

Wind Turbine design and modelling 17

Laín Nieto Gómez 2009/2010 Curse

As we can see in the excel page the variation of the Reynolds number for this
cases is not so huge and it is in between 164.000 and 331.000, that is the reason
for choosing the line with squares as points. With that line we have built up a
polynomial that fits the data and that with a mathematical formula provides us
with the lift coefficient in function of the angle of attack.

It has been developed taken points from the graphics and a polynomial of 3th
degree has been calculated that best fits the data. We can see in the table bellow
that for the Reynolds Number range within which we move there are very
similar values for the lift coefficient. That is the reason for approximately
choosing a first degree polynomial, because are more or less linear and very
close one to each other.

Lift coefficient VS Angle of attack

1.6
1.4 f(x) = 0.16 x
R² = 0.97
1.2
1

CL 0.8
0.6
0.4
0.2
0
1 2 3 4 5 6 7 8 9 10 11
Angle of attack, α

CL = 0,1565*alpha [2.2]

Fig. 2.6: Lift coefficient VS angle of attack

The Pearson coefficient that appears below the deductible by excels polynomial
gives a good measure of what has been the polynomial approximation of degree
1. The approximation level is so good that we can use it safely. Once we have
the attack angle and the lift coefficient in the second graphic we can get the
moment coefficient and the drag coefficients.

For the drag coefficient we have made the same operation, but as the curve’s
slope changes like a four degree one the polynomial is different:

Cd = 0,0905α4 - 0,0513α3 - 0,0968α2 + 0,0769α [2.3]

Wind Turbine design and modelling 18

Laín Nieto Gómez 2009/2010 Curse

Drag coefficient VS angle of attack

0.09
0.08 f(x) = 0.09 x⁴ − 0.05 x³ − 0.1 x² + 0.08 x
0.07 R² = 0.98

0.06
0.05
Cd 0.04

0.03
0.02
0.01
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Angle of attack

Fig 2.7: Drag coefficient VS angle of attack

2.4.-Relations between the coefficients and the forces, momentum

coefficient

These coefficients were obtained from the previous abacus used within the
following formulas in order to obtain the forces of lift and drag:

ρ
F L=C L× ×v 2 ∞×L×1
2
[2.4]

ρ
F D=C D× ×v2 ∞×L×1
2
[2.5]

For these formulas we have supposed that the length of the piece of wing which
we are calculating measures 1m. Afterwards when we are calculating the
integral along the length of the blade we will multiply times a number which is
less than the unit.

The supposed density and nominal velocity of the wind are the following:

ρ : Density
v ∞ : Actual velocity of the air
L : It is a variable depending of which is the diameter

Wind Turbine design and modelling 19

Laín Nieto Gómez 2009/2010 Curse

From these two equations we can obtain Lift Force and Drag Force. The units
are corresponding with Newton in the IS (International System of units).
Geometrically the direction of the drag force is parallel to the c velocity vector
and goes through the following point obtained from a moment equation at a
distance d from the point of attack.

ρ
F L×d=C M × ×v 2∞ ×( L×1 )×L
2
[2.6]
So finally the distance from the border of attack to the point of application of the
drag and the lift forces is “d” that we can express like this:

ρ
C M × ×v 2∞ ×L2
2
d=
FL [2.7]

In the Fig 2.6 we can see the forces and the distances where in the next image
we see a diagram of the lift and drag forces and the angles they form. The point
of application of both forces is on the line just at a distance d of the contact point
of the air with the profile.

The direction of the drag force vector is the same as the absolute velocity of the
air (supposing that the system of reference is spinning with the wing at a
velocity equal to ω. The sense of the vector is the same as the projection of the
vector velocity. Lift force is perpendicular to the drag force and the sense is
directing to the convex part of the profile as showed in the picture.

Wind Turbine design and modelling 20

Laín Nieto Gómez 2009/2010 Curse

Fig 2.8: Lift and drag forces representation in a wing

The resulting vector addition of those forces gives us another force that will be
split into horizontal and vertical forces. At this point there are two forces; the
vertical provides us power for the windmill. The second one must be withstand
by the bearings.

For each section must be calculated the torque produced by the vertical
component which is proportional to the distance from the section to the rotary
axis. With this information for each section and considering the different chord
lengths and the different sizes of the profile we can calculate the output torque.

2.5.-Outputpower and its relation with the spinning velocity

We have spoken about the different chord sections and so on and once we have
the value of the total torque in the axis, with the velocity of spin we have the
output power which is the desirable parameter to maximize. For that we should
vary the spin velocity parameter and get the different output powers for different
rpm.

This way where obtained the following graphics that show us for a fixed
velocity of the air which is the output power varying the spin velocity. This
graphic show us that there are some points of maximum output power for each
velocity of the air.

Wind Turbine design and modelling 21

Laín Nieto Gómez 2009/2010 Curse

These curves represent the different values that our wind turbine can provide us
with for each spin velocity. We also have obtained the different third degree
polinomic that best suit the points calculated.

Output power VS spin velocity

2500
2250
2000
1750
1500
1250power P [W]
Output
1000
750
500
250
0
200 400 600 800 1000 1200 1400 1600
6m/s 6 m/s
8 m/s
Spin velocity
Polynomial (8 m/s)
n [rpm]
10 m/s Polynomial (10 m/s)

Fig. 2.9: Outputpower [W] VS spinning velocity [rpm], depending on the

parameter of the wind velocity

Each of the adjusted curves has an equation as follows:

P(v=14m/s)= 2E-06n3 - 0.0094n2 + 13.002n - 3373.9 R2 = 0.9929 [2.8]

P(v=12m/s)= -2E-07n3 - 0.0021n2 + 3.7719n - 696.61R2 = 0.9853 [2.9]

P(v=10m/s)= -2E-07n3 - 0.0021n2 + 3.7719n - 696.61R2 = 0.9853 [2.10]

P(v=8m/s)= -8E-07n3 - 0.0004n2 + 1.5462n - 165.29R2 = 0.993 [2.11]

P(v=6m/s)= -0.0016n2 + 1.5437n - 173.54R2 = 0.9865 [2.12]

The maximum outputpower points have the following table’s coordinates:

Wind speed 6m/s 8m/s 10m/s 12m/s 14m/s

n[rpm] 515 685 825 895 1035
P[W] 191,8864 448,9806 871,97 1376,6285 2231,011

Wind Turbine design and modelling 22

Laín Nieto Gómez 2009/2010 Curse

Table 2.1: Maximum outputpower values in both P[W] and n[rpm]

coordinates

This all leads us to the conclusion that the power curve needed by the generator
must cross the maximum points of the second table. Also there is a polynomial
of second degree which suit the data obtained before with a Pearson correlation
factor of 0.99, which is quite good.

Generator Output Power VS n [rpm]

2500

2000 f(x) = 0 x² − 1.9 x

R² = 0.99
1500 rpm VS Max
power
Output Power P [W] output
1000

500

0
400 500 600 700 800 900 10001100
n [rpm]

Fig. 2.10: Generator outputpower VS n[rpm], for an ideal operation

between the wing’s power optimum points and the generator

This graphic indicates different output powers deppending on the revolutions per
minute with which the generator must carry out.

About the geometrical specifications of the air turbine wings:

The data provided by the company heyde-windtechnik is the following one.

Firstly they provide the geometry of the wings. That we can resume in the
following table 2.2:

Wind Turbine design and modelling 23

Laín Nieto Gómez 2009/2010 Curse

Geometrical description of the airfoil Width

Length of the section
Section’s Chord length starting in each
Radius [m] spin θ [deg] [m] measure[m]
1 0,1 36,1 0,1435 0,025
2 0,15 28,7 0,1375 0,05
3 0,2 23,3 0,135 0,075
4 0,3 16,1 0,115 0,1
5 0,4 11,8 0,106 0,1
6 0,5 9,13 0,098 0,1
7 0,6 7,2 0,0835 0,1
8 0,7 5,8 0,0715 0,1
9 0,8 4,7 0,0595 0,1
10 0,9 3,9 0,0475 0,1
11 1 3,2 0,0355 0,1
12 1,1 2,7 0,0235 0,05

Table 2.2: Geometry of the wings given by Heyde Windtechnik

The first column is the number of the section, the second one is the radius of the
section. The third one is the section´s spin angle that we have called β. This β
angle is the angle in between the chord and the plane of spin of the machine.
The other important angle is the one we have named with θ. This angle is the
one between the Plane of spin and the absolut velocity so called “c”. This angle
is the one we have used to split both drag and lift forces into vertical and
horizontal forces. In the following picture all those parameters are drawed.

2.6.-Step by step calculus of one section´s lift force and drag force

We will choose the data provided by “Heyde Windtechnik” company for the
blade´s propierties explained before.
1)We fix the density ρ and the kinematic viscosity ν of the air as:
kg
ρ=1 . 18
m3
[2.13]

Wind Turbine design and modelling 24

Laín Nieto Gómez 2009/2010 Curse

2
m
υ=0 . 00001
s
[2.14]

2) The nominal operating point of the Wind Turbine is the following:

n=580r . p . m
[2.15]

2π rad
ω= ×580=60. 74
60 s
[2.16]

m
v =10
s
[2.17]

m
u=ω×r=60 .74×0,6=36 , 44
s
[2.18]

This was the section number 7 we have choosen this section because the
information given by the company was not for this section but we have
supposed the section’s length of chord used in the following calculus supposing
that this length will vary linearly. We have drawn in a graphic the different
chord length and a line was almost crossing them so we will we will use this
assumption for for the calculus. So for the given section we have L=0.0835m.

Chord length VS radius

0.16
0.14
0.12 f(x) = − 0.12 x + 0.16
R² = 0.95
0.1
Radius [m] 0.08 Linear ()
0.06
0.04
0.02
0
0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55
Chord length [m]

Wind Turbine design and modelling 25

Laín Nieto Gómez 2009/2010 Curse

Fig. 2.11:Chord length decrement with respect to the radius of the section

We establish like in some other books is explained that the approaching speed of
wind to the border of attack is two thirds of the v ∞ in the sorrounding area.

2
v real = ×v ∞
3
[2.19]

Spinnig axis
u
c
Fig. 2.12:

According to the Fig. 2.12 we can calculate the absolute velocity as:

m
|⃗c|= √ u2 +v 2 =37 . 05
s
[2.20]
The angle etween the ⃗c and the plane of spin is the following

u
β=arctan ()
v
=10. 37 °

[2.21]
With the module of the vector velocity relative to the airfoil we calculate the
Reynolds number for each case, in this section number 7 we have:

c×L 37 . 05×0 .0835

Re= = =309 . 345[−]
υ 0 . 00001
[2.22]

Once calculated for all the other sections, we can note that all the Reynolds
5
numbers for the different sections remain between 1÷4×10 . That is the reason
for approximatig the lift and the drag coefficient with the lines of the squares
and the triangles. Both of this curves are very close, so we pick up the higher
one and approximate it with the exlplained polinomials. This following ones are
the polinomials required :

C L=0.0917α 0+ 0.4611
[2.23]
Wind Turbine design and modelling 26
Laín Nieto Gómez 2009/2010 Curse

C D=0. 2421C 4 - 0.5263C 3L+ 0 . 4181C 2L - 0 . 1458CL + 0 . 0311

L
[2.24]

α
Where 0 changes from one section to each other, it is the subtraction of the
angle of the c velocity to the angle θ of the position of the section.

The “θ” angle is the spin angle of each section distanced at a radius from the
axis is provided by the manufacturer, the second one “β” is dependent on the
ω and on the wind velocity “c”. And finally the difference between β and θ is
α as follows:

α =β−θ
[2.25]

Section's Spin angle θ with the radius

40

35

30

25
Angle θ [°]

20 Reihe1

15

10

0
0 0.2 0.4 0.6 0.8 1 1.2
Radius [m]

Fig. 2.13:Section’s spinning angle θ with the radius

Wind Turbine design and modelling 27

Laín Nieto Gómez 2009/2010 Curse

Angle between plane of spin "β" and "c" wind absolute

velocity
60

50

40
Angle β [°]

30 Reihe1

20

10

0
0 0.2 0.4 0.6 0.8 1 1.2
Radius [m]

Fig. 2.14: Angle between plane of spin β and c wind absolute velocity

Angle of attack α=β-θ [°] VS Radius [m]

14

12

10
Angle of attack α [°]

8
Reihe1
6

0
0 0.2 0.4 0.6 0.8 1 1.2
Radius [m]

Fig. 2.15: Angle of attack for each of the sections corresponding with one
radius

Wind Turbine design and modelling 28

Laín Nieto Gómez 2009/2010 Curse

The lift coefficient varies from 0,7 to 1,25 for the nominal parameters. And for
this section is the following one:

C L=0. 0917 α 0 + 0 . 4611|α =3 ,17 =0.7515

0
[2.26]

With this coefficient, we are able to find out the value of the drag one:

C D=0. 2421C 4 - 0.5263C3L+ 0 . 4181C2L - 0. 1458CL + 0 . 0311|C L= 0 .75150205=0. 0115

L
[2.27]

Fig. 2.16: Parts of the turbine and geometrical disposition

F total
FL Fu

Ω Fi Fd

θ
Ωr
c

Fig. 2.17: Forces and angles used for each section’s forces

Wind Turbine design and modelling 29

Laín Nieto Gómez 2009/2010 Curse

So the drag and lift forces are descrived as follows for section number 7:

ρ 0 . 00001
F L=C L× ×v 2 ∞×L×R l=0 ,7515× ×37 . 04722×0 . 0835×0 .0835=5 .0814N
2 2
ρ 2∞ 0 . 00001
F D=C D× ×v ×L×Rl=0. 0115× ×37 . 04722 ×0 . 0835×0 . 0835=0 . 0778N
2 2

[2.28];[2.29]

With the forces and the angles as drawn in the Fig.2.16 we can obtain the
parallel force to the plane of spin and the perpendicular one.
We are going to divide both forces in horizontal and vertical directions. First one
is useful so will call it
Fu , the other one will be called Fi .

Fu =F L ×sin( β)−F D×cos( β)=5.08×sin(10.37° )−0,078×cos(10.37° )=0.8379N

[2.30]

Fi =F L×cos( β)+F D ×sin( β)=5.08×cos(10.37)+0.078×sin(10.37° )=5.0124N

[2.31]

The useful force has been so called because for each of the sections there is a
momentum created. The integration or sum of the momentums gives us the total
momentum. Momentum times spin velocity is equal to power in [W]. And as we
have 3 wings the total output power is three times that.

Before calculating the momentum which gives power to the system, we must
know for each secction, in this case the number 7 the point of application of the
two forces, or the so called a.c. (aerodynamical centre). For that we will use the
previously explained formula:

ρ 0 . 00001
C M × ×v 2∞ ×L2 0. 1× ×102 ×0 .0835 2
2 2
d= = =0 . 01111108
FL 8 . 0814

[2.32]

This is the distance from the attack point where the force is actuating. It would
be useful for terms of torque along the axis of the blade which produce torsion.
But as the wings are not calculated in this work we suppose that for the worst
conditions they will support the forces.

Wind Turbine design and modelling 30

Laín Nieto Gómez 2009/2010 Curse

Momentum per each section. Considering that the radial length of the wings is
the one in between twoo secctions and its centre, as the Fig. 2.16 shows.
i=12
M=3× ∑ F u ×r i =756 . 4188 [ Nm ]
i=1 i

[2.33]

dM=Fui x rui

Fig. 2.16: Momentum sum along the diameter of the turbine

The output power is the momentum times the angular velocity:

P=M ×ω=756 . 4188×60 .7375=756 . 4187782 [W ]

[2.34]

3)The variation of the parameters “ v ∞ ” and “n”:

In this part we have developed a graphic with the output power obtained with
variations in n[ r . p.m.] . Each of the graphics below has different velocity
variations.

Wind Turbine design and modelling 31

Laín Nieto Gómez 2009/2010 Curse

Output power VS spin velocity

2500
2250
2000
1750
1500
1250power P [W]
Output
1000
750
500
250
0
200 400 600 800 1000 1200 1400 1600
6m/s 6 m/s
8 m/s Spin(8velocity
Polynomial m/s) n [rpm]
10 m/s Polynomial (10 m/s)
12 m/s Polynomial (12 m/s)

Fig.2.17 : Pointing of the best outputpower to each of the velocities

The following ones are the correlations more suitable to the different
outputpowers for each of the velocities of the wind:

P(n)= 2E-06n3 - 0.0094n2 + 13.002n - 3373.9 R2 = 0.9929

P(n)= -2E-07n3 - 0.0021n2 + 3.7719n - 696.61R2 = 0.9853

P(n)= -2E-07n3 - 0.0021n2 + 3.7719n - 696.61R2 = 0.9853

P(n)= -8E-07n3 - 0.0004n2 + 1.5462n - 165.29R2 = 0.993

P(n)= -0.0016n2 + 1.5437n - 173.54R2 = 0.9865

Signaled points are the maximum output power for each of the velocities of the
wind. This points must be the same for the output power of the generator to
obtain the maximum power for each wind velocity.

Wind Turbine design and modelling 32

Laín Nieto Gómez 2009/2010 Curse

3.-Mechanical calculus

3.1.-Axis calculation respect to DIN 743

There had been calculated two different hipotetical charges, one of them is
related to the aerodinamical forces the other one comes from the own weight of
the mechanical parts.

The forces coming from the wings are the following ones

NA=156 N NB=156 N

VA=80 N VB=20 N

50 mm 160mm

Fig. 3.1: Flexional stress on the axis

This charges create the following flexion tensions:

50 mm 160mm

Fig. 3.2: Flexional stress on the axis

It is when the maschine slows down when we can observe a maximum torque
along the axis. The value is constant and the distribution goes from the head of

Wind Turbine design and modelling 33

Laín Nieto Gómez 2009/2010 Curse

the wings till the point where the axis is installed. Due to the torque it can be
observed a torsional stress in the axis which distribution along it is the
following:

T=20 Nm

130mm

Fig. 3.3: Torque and torsional stress along the axis

The diameter of the axis must be strengthful enough to support both stresses of
the torque and the vertical forces. As the DIN 743 asserts we can dimensionate
the axis with the following formula:

16 T 3 16 ∙ 20000
√
d min = 3
π τ max√=
π ∙ 150
=8.79 mm → d=30 mm [3.1]

The flexion stress is related to the diameter of the axis as follows:

M f max y max 1.8 ∙ 0.15 N

σ= = =0.679 [3.2]
I ln mm 2
−8
4 ∙ 10

This value for the flexional stress is negligible for the resistant stress of the
material. By the way, it has been choosen an special iron for the windturbine
cosntruction, wich’s resistance is:

N
σ max adm=800 [3.3]
mm2

This is a DIN 1.4401 semisoft steel usually utilized to develop this kind of
mechanical elements.

3.2.-Bearings selection

3.2.1.-6206-RS1 (axis support one) selection justification

The turbine plays two different rotational movements. The first one takes part
around the horizontal axis. While the axis is rotating and boosting the generator
coupling there are some forces coming out that must be abide by our bearings.

Wind Turbine design and modelling 34

Laín Nieto Gómez 2009/2010 Curse

The opposite reactions of the axis turn out to be the clouts supported by the
bearings. Remembering the Fig 3.1 when the resultant forces are isolated then
those are the forces wich must be supported by the bearings.

A B

Fig. 3.3: Forces affecting A and B bearings (aerodinamical forces)

There have been calculated both horizontal and vertical forces beeing this ones
the results:

Horizontal Load [N] Vertical Load [N]

Bearing A 156 80
Bearing B 156 20

Table 3.1: Aerodynamical loads in each bearing

For the selection of the bearing there have been choosen the SKF company. For
reasons of simplicity both of them will be the same. The web page of SKF
provides us with a calculation program wich results for the selected bearing are
the following:

Wind Turbine design and modelling 35

Laín Nieto Gómez 2009/2010 Curse

Fig. 3.4: Equivalent bearing loads and basic rating life.

It has not been taken into account the fact that the wind turbine will be facing
gusty winds, which in fact is providing the axis and the whole structure with P ( t )
variable charges pending the day.

That is the reason why it must be used the so called Miner's rule or the
Palmgren-Miner linear damage hypothesis, which states that where there are k
different stress magnitudes in a spectrum, Si (1 ≤ i ≤ k), each contributing ni(Si)
cycles, then if Ni(Si) is the number of cycles to failure of a constant stress
reversal Si, failure occurs when:

[3.4]

Where C is experimentally found to be between 0, 7 and 2, 2. Usually for design

purposes, C is assumed to be 1. This calculus gets us into air wind frequency
calculus and estatistical estimations which are not the purpose of this thesis.

On the other hand it is up to the hill to clear the patch between the axis an the
bearings. At this point we must look at the following tables taken from spanish

Wind Turbine design and modelling 36

Laín Nieto Gómez 2009/2010 Curse

ISO rules which contribute to a hint about how to choose between the different
hole adjustment and the shaft adjustment.

Fig. 3.5.a: Guide for radial bearing adjustment, support housing

adjustment

Fig. 3.5.b: Guide for radial bearing adjustment, shaft tolerance

Wind Turbine design and modelling 37

Laín Nieto Gómez 2009/2010 Curse

If a constant wind velocity facing the turbine is supposed then the loads are
constant in value but not in direction. This is because the axis is spinning around
while the structure, if the wind direction keeps the same, is fixed. Thus is the
second choice in Fig.3.5a H7 for the hole and k5 for the axis.

3.2.2.-Bearings concerning the vertical axis spinning

The wind turbine must be positioned in front of the air orthogonally in order to
obtain the most possible quantity of energy from the position where it stands.
Therefore it must rotate around a vertical axis.

There has been performed a mechanical construction for letting the machine
swivel around this vertical axis but also pass on the forces coming from
aerodynamic interactions and the self weight of the construction above.
In the following image this construction can be shown.

Fig 3.6: Mechanical construction for vertical turn around, 3D and

separated construction

The following lay outs show how is the construction in between the UPN and
the rotatory axis:

Wind Turbine design and modelling 38

Laín Nieto Gómez 2009/2010 Curse

N
M
V

R v2

R v1
O

RN

Fig 3.7: Mechanical construction for vertical turn around, section and
model

We assume that the radial bearings do not absorb any axial force and that the
whole vertical force is absorbed by the axial bearing. This is the momentum will
be absorbed also by the radial bearings. This forces can be obtained from the
static planar mechanics:

∑ M oi=0; M +225 V −225 R v2 −150 R v 1=0 [3.5]

i

∑ V i =0 ; V −R v2 −R v 1=0 [3.6]
i

∑ N i=0 ; N−R N =0 [3.7]

i

This equations have only three unknown terms, R v2 , R v 1and R N . M, V and N are
calculated as follows.

The “N” force is the resultant of the complete weight of the mechanical
construction minus the vertical component of the air impulse.

In the other hand V is caused by the horizontal force coming solely from the
aerodynamic forces.

The whole construction weight supposing a homogeneous density of the

different iron alloys of 7850 kg /m3is 184,509kg, we must take in account the
weight of the wings that more or less are 2kg each wich together can be like
190kg more or less rounding up is about 200kg mass. This in term of force
would be like N= 1000N.

Wind Turbine design and modelling 39

Laín Nieto Gómez 2009/2010 Curse

The V force comes from the aerodynamic drag and lift forces. Therefore the
decomposition has been yet made. Resulting for the nominal wind velocity at his
optimum point of output power in V=54.04N.
The momentum is generated by two forces, one V and the other one the weight.
Firstly the gravity center must be traced back.

Fig. 3.8: Catia Screenshot of the centre of gravity and the coorodinates of it
according to the assembly coordinate system

Once known the centre of gravity, then the moment can be calculated as follows:

80 N 20 N

156 N
156 N

M
V
1900 N

Wind Turbine design and modelling 40

Laín Nieto Gómez 2009/2010 Curse

Fig.3.9: Reduction of the system of forces to the upper point of the pillar
bearing junktor

We can calculate the loads in the top part of the metallic piece numbered as 1
that in the following document will be called „pillar bearing junktor“ with the
same equations of static equilibrium.

M =2 ∙156 ∙ 228+1900 ∙ 61−80 ∙256−20 ∙ 100=165 N ∙m [3.8]

N=1900+ 100=2000 N [3.9]

V =156 +156=312 N [3.10]

These values provide us with the following results for the written equilibrium
equations [3.5], [3.6] and [3.7]:

165+225 ∙ 312−225 ∙ R v2
R v1= [3.11]
150

R v1=312−Rv 2 [3.12]

165+ 225∙ 312−225 ∙ R v 2

312−R v 2= → R v 2=75 N
150

R N =N =2000 N [3.13]

R v1=237 N

R v2=75 N

R N =2000 N

With this data in hand we can check if the three different bearings chosen are the
good ones for this shipping. There have been used for three bearing
comprobation again the SKF providers of mechanical components program.
The results obtained are as follows:

There have been choosen two different ones, for the upper bearing it has been
choosen 62307-2RS1, and for the lower one it has been used the RLS 12 (these
are the product names that can be selected from the SKF catalog).

Wind Turbine design and modelling 41

Laín Nieto Gómez 2009/2010 Curse

Fig 3.10: RLS 12 bearing corresponding to the lower radial one in the
costruction

Fig 3.10: 62307-2RS1 bearing corresponding to the upper radial one in the
costruction

Wind Turbine design and modelling 42

Laín Nieto Gómez 2009/2010 Curse

Fig 3.11: 89413 TN bearing corresponding to the axial one in the

costruction

The constuction with those three bearings is explained in the previous Fig.3.7
pag. 23. So the checkout of those three bearings named:

-62307-2RS1

Fig 3.11: 62307-2RS1 bearing life calculus

Wind Turbine design and modelling 43

Laín Nieto Gómez 2009/2010 Curse

-RLS 12

Fig 3.12: RLS 12 bearing life calculus

Wind Turbine design and modelling 44

Laín Nieto Gómez 2009/2010 Curse

-89413 TN

Fig 3.13: 89413 TN bearing life calculus

Wind Turbine design and modelling 45

Laín Nieto Gómez 2009/2010 Curse

3.3.-Comprobation of tabs according to DIN 6885

The mechanical transmission of power between the axis and the generator is
worked up at two points in the mechanism by tabs. The first one „1“ transmits
wind turbine obtained energy from the wind along the wings into mechanical
torque spinning arround the transmission axis.On the other hand there is a
second coupling „2“ between the opposite axis part and the generator (See Fig
3.14).

As the Fig 3.14 shows it has been designed with 1 tab per connexion.In order to
be able to transmit the not negligible amount of 1.4 kW the following calculated
neccesary length of the tabs has been rounded up.

Fig. 3.15: Two mechanical couplings developed with tabs

Firstly we must make sure that for the worst possible consitions of work this
mechanical conexion will be accesible to support them. This conditions are fixed
by the axis slow down terms. As in the brake mechanical calculus chapter is
stated it has been fixed the maximum time expected from the brake to slow
down from the maximum velocity of swiveling to cero in 5seconds. More time
could have been dangerous for the sorrounding area due to the possibility that
the blades get broken by their roots and injury people or cause damages in the
nearby goods.

For slowing down as in [3.11] equation is obtained, a braking torque of 165 Nm

is required . That is the limiting condition to design the „1“ possition tabs wich
will be absorving the torque coming from the air and the kinematic energy
stored along the axis.

For the comprobation and dimmensioning of the tabs it has been used the DIN
6885 and the DIN 6892 norm. First of it provides us with the forms of the tabs
and the seccond one with the .
Wind Turbine design and modelling 46
Laín Nieto Gómez 2009/2010 Curse

There are two different dimensioning systems that involve firstly the shear stress
and then the compression strees but as the τ maxinvolved in calculus is much more
small than the compression axial stress σ comp .max must be said that before
calculating both of them the more restrictive one is τ max.

First of all the force F u must be calculated. This is the force wich once put at a
distance of radius, can perform the torque of transmission. Namely:

T transmission 165
F u= = =1333,33 N [3.11]
d /2 30/2

This whole load must be held by one tab, wich must have the following length
„l“:

Firstly the compression tension suffered by the tab is the following one in the
upper part of its keyway.

Fu
S=
(h−t 1 )∙ l t ∙ i∙ k

[3.12]
Where

S: Compression stress suffered by the tab/or the resistance of the material when
the l t value is cleared
F u:Tangential force along the tab
l t :length of the tab
t 1:distance from the axis cavity of the tag till the end of it once inserted
i:Number of tabs
k :Coefficient of impairment

The unknown variable is the l t wich must be as follows cleared from the
equation. This is going to be the design parameter. Once substituted the values
in the equation:

Fu 1333.33
lt = = =12mm
( h−t 1 ) ∙ i∙ k ∙ S ( 5−3.1 ) ∙ 1 ∙1 ∙60

[3.13]

This length has been calculated the value of the resistance in Mpa of 60 N ⁄ mm 2.
This resistance always must be less than the other two materials. The reason for
Wind Turbine design and modelling 47
Laín Nieto Gómez 2009/2010 Curse

this is that when the mechanisim brokes up to broke along the most cheapest
part of three of them.

The normalized sizes of the tabs are the following ones for the choosen type:

bxh t1 Clearance Tighten Needed screw

8x5 3.1+02 2+0.1 1.4+0.1 M3x8 (DIN
84)

Table 3.2: Geometrical characteristics of the selecctioned tab according to

the norm DIN 6885

The other tab number“2“ is designed as well as the other because it is passing
down the same torque and consequently it will have the same size.

3.4.-Axis break dimensioning

It has been installed for security reasons a brake which is able to back away the
wind mill. Its design can be showed in the following image Fig. [3.16]. It

Fig.3.16: Selected geometrie and design for the brake, disc brake.

There is a mechanical junktion between the following parts of the wind turbine
costrucction: Wings, torque transmission axis, tabs, bearings, and generator
rotor. This is why when time to dimensioning the security brake it must be clear
the moments of inertia involved in the calculus of the rapid deceleration of the
axis.

According with the last paragraph comment there are three different inertia
moments wich ares as follows: generator, axis coupligs construction, and the
wings.

I TOTAL=I wings + I axis+coupligs + I generator rotor [3.14]

Wind Turbine design and modelling 48

Laín Nieto Gómez 2009/2010 Curse

Each of them has been calculated in different ways. The most simple calculus is
the one of the axis and the couplings one because in CATIA we can obtain it
directly. The following Fig [3.17] shows us the value of it:

Fig. 3.17: Catia screenshot involving direct measures of inertia

I_(axis+coupligs)=0.005kg∙m^2 [3.15]

This one is automatically generated by CATIA.

For the other two calculations the method has been less exact done. As the
manufacturer does not provide any information about the inertia of the rotor it
has been searched in the web one sincronous generator of 1kw wich inertia is as
follows:

I generator , 1 KW =0.0075 kg ∙ m 2 [3.16]

That was the generator inertia for one generator of less output power, so that we
must say that it must have more or less 1.4 times more inertia as our turbine
generator will be of this output power. So the final inertia for the generator is:

I generator , 1,4 KW =1.4 ∙0.0075=0.0105 kg ∙ m 2 [3.17]

Wind Turbine design and modelling 49

Laín Nieto Gómez 2009/2010 Curse

The wings are aproximatedly as a beam and it is known that the beams inertia
moment in respect to one axis crossing his butt is dependent on the length, the
mass as follows.

1
I wings =3 ∙ I wing=3 ∙ ∙ M ∙ L2=M ∙ L2 =2∙ 1.2752 kg ∙ m2 [3.18]
3

Where:

L: Length of the beam or wing in this case, wich’s value is half of the diameter
M: total mass of the wing, wich has been estimated in 2kg

All this three inertias put toghether make a total inertia of 3.25 Kg m^2, wich is
the data tied to be calculated. This helped we can calculate helped by the
kinematic formulas and the relation between inertia moment and torque the
angular acceleration neccesary for stoppig the whole assembly in a short period
of time. It has been estimated a time of t=5s, within the wings must have
stopped spinning. Calculus are as follow:

First of all we must propose the kinematics formulas for a circular decelerated
movemet, with constant deceleration:

1
θ=θ0 +ω ∙ t− ∙α ∙ t 2 [3.19]
2

ω=ω 0−α ∙ t [3.20]

α =constant ? [3.21]

Its also known the relation between the torque and the innertia:

T =I total ∙ α [3.22]

With those equations [3.19]-[3.22] we open up the value of the angular

acceleration and finally the value of the necessary torque wich will be provided
by the brake M . In last equations the variables have the following values:

Where:

θ :Angle of spin measured in [rad]

θ0 : initial angle of spin related to t 0wich in the beginning of swiveling is taken as

0 to make calculus more simple. So θ0 =θ ( t 0 ) =θ ( t=0 )=0

Wind Turbine design and modelling 50

Laín Nieto Gómez 2009/2010 Curse

ω :Angular spinning velocity measured in [rad / s]

t :Final value of time, as said before ot has been choosen a final time of t=5s
because it is estimated that dynamic problems will not appear in such a little
time. Is the time wich the brake needs to slow down from 125 rad/s (the
maximum spinning velocity) to 0 rad/s

α : This is the angular deceleration, it is said deceleration because the rpm are
decreasing, and that is the reason to it be negative in the equation

Fig. 3.18: Negative angular aceleration of the brake

ω 0 : Initial spinning angular velocity wich as it has been stated is 125 rad/s

T :Torque needed from the break to reduce the velocity of the machine. Wich
really is the second unknown number of four of the equations. It will be
obtained in second term after the obtention of the α .

I total :It is the total junktion momemt of inertia measured aling the turning down
axis. It is measured in [kg ∙ m2]

The past equations are solved as:

1
θ=0+125 rad /s ∙5 s− ∙ α rad /s2 ∙5 s2 [3.23]
2

0 rad / s=125rad /s−αrad /s 2 ∙ 5 s [3.24]

125
α= =50 rad /s 2 [3.25]
5

θ=0 rad [3.26]

T =3.25 kg ∙ m 2 ∙ 50 rad /s 2=165 N ∙ m [3.26]

So this is the essential parameter of the brake the torque neccesary to slow down
the velocity from the most possible one 1200 rpm= 125 rad/s to cero.

Once calculated this term it must be calculated the normal force wich the beam
must be making against the disc of the brake. This is like that because as we see
Wind Turbine design and modelling 51
Laín Nieto Gómez 2009/2010 Curse

in the following Fig3.19 the fuctionament of the brake consists on a long „U“
form beam wich is deformated by an extensor until the swiveling velocity of the
axis is too high and can cause damages. After this the extensioner turn around its
axis 90º and activates automatically the brake wich as it was said before will
expend like 5s to stop with security the axis.
brake slow down torque
Wing torque

brake acctioner force

Fig. 3.19 work and functionament of the mechanical security brake.

As the Fig. 3.20 shows there is a circle delimitated by a line wich is the area
where the beam is making an estimated costant pressure into the disc. In both
sides. In the following screenshot Fig.3.20 it can be seen wich is the size of the
area of contact between the beam and the disk. It really simple to measure it
without no table or integral just with CATIA.

Fig. 3.20: Screenshot of CATIA drawn semicircle delimitated by a line.

Wind Turbine design and modelling 52

Laín Nieto Gómez 2009/2010 Curse

Fig. 3.21: Screenshot of the measured area in CATIA of the Fig 3.20

So, as it is surely known the fricction coefficient is the constant between the
normal force and the Fricction force. As the surface of contact between the beam
and the disk is symetric, then we can consider both the normal and friccition
forces just getting of the centre of gravity from the crossed by lines surface in
the Fig. 3.22.

This center of gravity is centered in the axis of the disc at a distance from the
centre of gravity of the disk of 57.33 mm as in the Fig 3.21 can be seen.

Once known the distance where both normal and fricction forces actuate, then it
is possible to calculate provided the torque the value of the normal needed. It
can be calculated with the following equations:

Wind Turbine design and modelling 53

Laín Nieto Gómez 2009/2010 Curse

F r ∙ d 1=T brake [3.27]

T brake 165 000 N ∙ mm

F r= = =2878 N [3.28]
57.33 mm 57.33 mm

First of all we must resolve the needed friction force, wich is related to the
distance obtained before and to the torque. It is 2878 N as the equation [3.28]
states.

The relationship between the normal force and the fricction force is the
following one:
F r=μ ∙ N [3.29 ]
Where :

F r : Is the fricction force

μ : Is the coefficient of friction wich really depend on the two materials wich do
get friction in between. This coefficient is between two Iron pieces arround 0.15.
But the beam will be provided with some speacil materials wich resist the
abrassive action and wich give an upper coefficient of friction. This gives us a
better (hihger) F r.

N : Normal force actuating against the disk.

F r 2878 N
N= = =1.44 KN [3.29]
μ 2

This equation needs the value of this coefficient that will be taken as 2 because
the material is an specific one with good propierties of heat dissipation, non
wear propierties and so on.

Now the unknown parameter to design is the inertia moment and furtherly the
secction area of the beam wich produces teh slowing down because of the
friction of the axis.

It has been made an aproximation where the beam is considered as in the

following Fig 3.22 as a beam with a constant load whose value is the normal

Wind Turbine design and modelling 54

Laín Nieto Gómez 2009/2010 Curse

divided times the area of the circular surface part in contact with the rotatory
disk.

Fig. 3.22: Table with the deformation necessary in the beam so that it
pushes the neccesary load against the disk.

Once it has been calculated the c constant linear load the equation in the
Fig.3.22 can be used. It gives us a relationship between the flexion resistance,
loads, the positioning of them and the displacement of the most furtherly placed
point of the beam.

The following [3.40] equation has as unknown parameters EI, wich is the
rigidness of the beam. Young module is E=210 000 Pa. From this value we can
get the two dymensions of the beam seccition.

Wind Turbine design and modelling 55

Laín Nieto Gómez 2009/2010 Curse

One of them is really the calculated one, the inertia involved on the flexion of
the beam wich is drawn in the Fig. 3.23 as I y− y. This value must get out of
Equation [3.40]

x-x

y-y

Fig. 3.23: Inertia axis of the section of the beam involves in the calculus of
the folowing equation [3.40]

q ∙ c ∙ a2 2
c 2
δ=
3 ∙ E∙ I[ ( ( ))
l∙ 3+
c
a
−a∙(1+ ()
a ]
) [3.40]

Where:

δ : Displacement of the extremal point of the beam wich will be fixed by the
designer in order to only have an unknown parameter in the [3.40]
equation. It has been fixed to 5mm just not ot be very stiff the beam.
[mm]

q : The linear constant load suffered by the beam. As we only know the normal
force we can divide it by the area calculated with CATIA of the contact
surface wich is AT =0.03 m m2 [mm2 ]

a : as the diagram in the Fig.2.23 shows, it is the distance from the ligated point
of the beam till the centre of the longitudinal linear load.[m]

l : Is the total length of the beam [m]

Wind Turbine design and modelling 56

Laín Nieto Gómez 2009/2010 Curse

c : Is the half of the total length along the load is placed [m]

If the moment of inertia is got from this equation [3.40] it changes to:

E the Young coefficient wich is the value of the slope in the σ −ϵ diagram in the
N
linear zone of it. It is measured in
mm 2 [ ]
q ∙ c ∙ a2 2
c 2
I=
3∙ E ∙ δ[ ( ( ))
l ∙ 3+
c
a
−a ∙(1+ ()
a ]
) [3.40]

Substituting the values of the geometrical and force parameters:

( 20/0.03 ) ∙58 ∙ 732 2 2

I= 5
3∙ 2.1 ∙10 ∙5 [ (
230 ∙ 3+
58
73 ( ) ) ( ( ) )]→[3.41]
−73∙ 1+
58
73

I =46839m m4

This is the required inertia from the secction to be 5mm less thick as the disk
and to provoke on it a torque of 165 Nm wich makes the whole construction to
slow dwn in 5s from 125 rad/s to 0rad/s and to stop.

The secction could have been designed in plenty of different ahapes but a
rectangle is one of the simplest ines, that is why it has been choosen. The Fig.
3.24 shows the two axis where it works.

There has been fixed the „b“ dymension to 20mm so that „h“ dymension must
be obtained from the formula [3.42] wich gives us the relationship between the
moment of inertia and the „b“ and „h“ dymensions.

1
I= b h3 [3.42]
12

b=20 mm

I ∙ 12 3 46839m m4 ∙12
h=
√(
3
b
= ) √(
20 mm
=42 mm )

Wind Turbine design and modelling 57

Laín Nieto Gómez 2009/2010 Curse

4.-Electrical calculus

4.1.-Generator selection:

As in the chapter 2 has been explained there is a graphic wich give the output
power of the wings against the spinning velocity of the generator. This is the so
called graphic:

Output power VS s pin veloc ity

2500
IDEAL GENERATOR CURVE
y = 2E -06x 3 - 0.0094x 2 + 13.002x - 3373.9
2250 R 2 = 0.9929

2000
6m/s
8 m/s
1750 y = 3E -06x 3 - 0.0101x 2 + 10.605x - 2255.3 10 m/s
R 2 = 0.9986
Output power P [W]

12 m/s
14 m/s
1500

1250

1000

750
y = -8E -07x 3 - 0.0004x 2 + 1.5462x - 165.29
R 2 = 0.993 y = -2E -07x 3 - 0.0021x 2 + 3.7719x - 696.61
500
R 2 = 0.9853
y = -0.0016x 2 + 1.5437x - 173.54
250 R 2 = 0.9865

0
0 200 400 600 800 1000 1200 1400 1600

S pin veloc ity n [rpm ]

Fig: 4.1: Swivelling velocity against the mechanical outputpower and the
output generated electricpower supperposed, Ideal generator.

The ideal generator would be that which passes along the point of maximum
output power produced. When the producction of this windturbine would be
massive then it would be justified to develop an special generator for this
characteristics of the machine.

Reality is the Fig 4.2 generator, it has been produced by „Maurer

Electromaschinen“ electrical machines constructor this machine has its
characteristics in the Fig 4.3 wich the most important ones for us are the
nominal power 500W, the tipe of generator trifasic altern current 50=Hz
frequency generator with variable intensity generation depending on the rpm of
spinning.

Wind Turbine design and modelling 58

Laín Nieto Gómez 2009/2010 Curse

Fig. 4.2: Measures and 3D reality appearance of the Generator.

In figure 4.1 it can be seen different points of work to the turbine, wich really
depend on the power provided by the wind for each of output powers there is a
rotation velocity wich is marked by the generator.

The Fig. 4.2 shows how the generator reacts giving one output power to
each of the velocities

It has been choosen the generator from Maurer company it fits perfectly the
generator curves needed because really although it is supposed to provide with

Wind Turbine design and modelling 59

Laín Nieto Gómez 2009/2010 Curse

1.4 KW with 10m/s wind velocity, this will not be the average velocity of the
place where it is placed.

The generator will be moving in terms of outputpower allways beyond the green
line of 10m/s, and in average in the 6m/s one. In case that the wind velocity
increases dangerously, the generator wont be ablo to catch the whole power, and
that is why :

T −T ' =T extra =α ∙ I [4.1]

Where

T : the wing’s provided torque measured in Nm

T ' :Torque absorved by the generator wich is then converted into electricity in
the rotor of the generator. Measured in Nm

T extra :The difference between them, T and T‘ what makes the rotor gettin
accelerated till it would break down or till the infinite if the generator was
unbreakable. Measured in Nm

α : Angular acceleration, rotor junktion with all his inertia is accelerated at a

constant rate (taking into account that the velocity of the wind keeps like the
rad
starting one) of α s

I :Inertia of the whole construcction measured in kg m2

When the angular velocity achieves 1200 rpm or 125 rad/s, then the security
brake starts to decelerate.

The normal voltage generated by the turbine will be at 500 rpm, that as the Fig
4.3 shows provides us with more or less 40V of tension between borns.

Wind Turbine design and modelling 60

Laín Nieto Gómez 2009/2010 Curse

Fig 4.3: Rotary speed [rpm] VS Voltage in [V]

And current [A] vs rpm

4.2.-Batteries selection

This energy can be collected by four batteries of 12V serial connected. This
makes a total of 48V, which is enough for this construcction. The batteries
seleccted are those ones of the Fig 4.4 and the Table 4.1 wich shows its
Electrical characteristics.

Wind Turbine design and modelling 61

Laín Nieto Gómez 2009/2010 Curse

Fig4.4: batterries used to keep the energy roduced by the turbine

Table 4.1: Ah values prizes and voltages of different tipes of bateris, the
ones between red lines are the selected ones.

As it has been said in the second pharagraph the generator characteristics show
that the generator is a triphasic generator with constant magnest inside 8 pair of
poles. The number of poles gives the relation between the spinning velocity and
the frecuaency of the generated electricity:

n ∙ p=60 ∙ f [4.2]

As the wind turbine is situated in Europe the frequency is 50Hz, but the
generator will provide us with a variable frequency because the spinning
velocity will be changins at the same time as the wind velocity does. The
Frequency in fact will be:

f =0.13 ∙ n[4.3]

Wind Turbine design and modelling 62

Laín Nieto Gómez 2009/2010 Curse

Really for each velocity of the air and this concrete turbine there is a direct
relation among the wind velocity and the rpm which is spinning at.

As the frequency is variable this does not work for the purpose of the turbine
wich really is feeding a house electricity requirements and works at a voltage
level of 240V trifasic voltage at a constant frequency.

4.3.-Transformation of the electrical output power to utility parameters of

tension voltage and frequency

The use of this windturbine is what gives us the neccesity of a power electronics
intermediate exchanger. It is needed a AC/DC transformer, then a set of batteries
that will work at 12V and finally an inversor wich provides us with a constant
feeding of 50Hz electricity at 240V.

The machine it will not be able to feed all the electrical requirements, and still it
must be taken electricity from the facility. But it allows to be less independent
from the net, and to save an important amounth of money along the year.

The following Fig4.5 shows us the electrical circuit to feed the house:

Fig 4.5: The electrical construction of AC/DC+batteries+DC/AC

Wind Turbine design and modelling 63

Laín Nieto Gómez 2009/2010 Curse

6.-Conclussions:

1)The small wind turbines are as this report shows a good option for house
electricity feeding, the invests in this machines can be within 10 years payback
amortized.

It is not possible to feed all the energetical requirements of one modern house by
means of this technology, but it is possible to get not only a natural benafit from
the installation of this turbines but also economical.

2)The wind energy itself appears in a lot of countries where the installation of a
net can be very expensive. This utilities also can appear as a dangerous weapon
against the animals and the enviroment.

Agreed to this it is very profitable to install this kind of technology in

emplacements as farms, etc where installations for nets are very expensive.

3)There is a factor generally aplicated to the productors of energy:

In physics, energy economics and ecological energetics, EROEI (energy

returned on energy invested), ERoEI, or EROI (energy return on investment), is
the ratio of the amount of usable energy acquired from a particular energy
resource to the amount of energy expended to obtain that energy resource. When
the EROEI of a resource is equal to or lower than 1, that energy source becomes
an "energy sink", and can no longer be used as a primary source of energy.

This is very important to the decission makin, just because this factor when less
than 1 clears that is not the well worth to try to invest societies generated energy
into it, cause if every source of energy was like that then it would lead to a
general lost of energy resources.

In the case of the wind energy Currently (2006) the EROEI in North America
and Europe is about 20:1 which has driven its adoption.

By this way theoretically it would be possible to achieve a 100% wind energy

production within a long period of time. But this is not real because not all the
locations are energetically giving the same energetic payback.

Wind Turbine design and modelling 64

Laín Nieto Gómez 2009/2010 Curse

In the case of the small energy it is said that this kind of installations can take a
20 years period o life having finished they payback period of time in 10 years.
That provides us with a economical and energetically profit.

Wind Turbine design and modelling 65

Laín Nieto Gómez 2009/2010 Curse

7.-Bibliography:

Books and PDFS:

[1]Numerische Strömungssimulation von horizontalachsen windturbinen,

Florian J. Kronschnabl, TECHNISCHE UNIVERSITÄT MÜNCHEN
INSTITUT FÜR ENERGIETECHNIK MW7
LEHRSTUHL FÜR FLUIDMECHANIK

[2]Ajustes de rodamientos, “Fundación Universidad de Atacama“Especialidad

Mecánica Automotriz.

[3
[1]DIN 743: Wellen Durchbiegung, Querkraft, Momente Festigkeit,
Eigenfrequenzen

[2]DIN 6885 Keyway and keyway details

[3]DIN 1.4401 Stainless steel norms.Resistance of the materials

Wind Turbine design and modelling 66

Laín Nieto Gómez 2009/2010 Curse

Wind Turbine design and modelling 67

Laín Nieto Gómez 2009/2010 Curse

6.-Draftings, guide of draftings:

Further contact or questions over: [email protected]

Wind Turbine design and modelling 68