DEPARTMENT OF MECHANICAL

MKSSS'S CUMMINS COLLEGE OF ENGINEERING

LITERATURE REVIEW REPORT

ON
“Modelling and Design of a Wind Turbine Blade for
Structural Integrity Analysis”

PRESENTED BY
Ms. SUMAYYA LAYAKPAHA KASAR
F.Y. MECHANICAL DESIGN ENGINEERING
Roll No: PME2025101

UNDER THE GUIDANCE OF

PROF. PARAG CHAWARE
MKSSS's Cummins College of Engineering for Women

TABLE OF CONTENTS
ABSTRACT

1.0 Introduction ........................................................................................................ 1

1.1 Wind turbine challenges ................................................................................... 1

1.2 Goals of this project ......................................................................................... 2

2.0 Theory and background ....................................................................................... 3

2.1 Wind turbine fundamentals ............................................................................... 3

2.1.1 Classifications .......................................................................................... 3

2.1.2 The wind turbine rotor ............................................................................... 6

2.2 Aerodynamics ................................................................................................. 7

2.2.1 Airfoils .................................................................................................... 7

2.2.2 Power and torque .................................................................................... 10

2.3 Loads on a turbine blade ................................................................................. 12

2.3.1 Uniform and steady state air flow.............................................................. 12

2.3.2 Vertical wind shear ................................................................................. 13

2.3.3 Wind turbulence stochastic loads .............................................................. 13

2.3.4 Gravitational and inertial loads ................................................................. 14

2.4 Computational methods .................................................................................. 14

2.4.1 Finite element method (FEM) ................................................................... 14

2.4.2 Computational fluid dynamics (CFD) ........................................................ 15

2.5 Fatigue assessment analysis ............................................................................ 16

3.0 Methodology .................................................................................................... 17

3.1 Model design ................................................................................................ 17

3.1.1 Geometry ............................................................................................... 18

3.1.2 Mesh...................................................................................................... 20

3.2 Simulations ................................................................................................... 22

3.2.1 CFD ...................................................................................................... 22

3.2.2 FEM ...................................................................................................... 23

3.2.3 Fatigue assessment analysis ...................................................................... 24

MKSSS's Cummins College of Engineering for Women

4.0 Results and discussion ....................................................................................... 25

4.1 Fluid flow analysis ......................................................................................... 254.2 Static

structural analysis ................................................................................. 30

4.3 Modal analysis .............................................................................................. 31

4.4 Fatigue assessment analysis ............................................................................ 32

5.0 Conclusions ...................................................................................................... 34

5.1 Further work ................................................................................................. 34

6.0 References ....................................................................................................... 36

MKSSS's Cummins College of Engineering for Women

List of Figures
Figure 1: Comparison between HAWTs and VAWTs…………………………………………4

Figure 2: Comparison between triple, double, single, and multi bladed wind turbines………..4

Figure 3: Comparison between upwind and downwind horizontal axis wind turbines………..5

Figure 4: Components of a horizontal axis wind turbine………………………………………6

Figure 5: Diagram of wind turbine hub………………………………………………………...7

Figure 6: Geometric airfoil parameters of the NACA airfoil series……………………………8

Figure 7: Examples of airfoil types…………………………………………………………….9

Figure 8: Forces on a wind turbine airfoil…………………………………………………….10

Figure 9: Vertical wind shear…………………………………………………………………13

Figure 10: Mesh element example in two dimensions………………………………………..15

Figure 11: Blade skeleton (left) and blade body (right)……………………………………….19

Figure 12: Fluid body geometry………………………………………………………………19

Figure 13: 3D cross section of fluid mesh……………………………………………………20

Figure 14: Blade mesh………………………………………………………………………..21

Figure 15: Convergence of integral static pressure…………………………………………...25

Figure 16: Velocity vectors of wind turbine rotor……………………………………………26

Figure 17: Velocity streamlines of wind for full model………………………………………27

Figure 18: Velocity streamlines of wind with velocity vector of the wind turbine rotor…….27

Figure 19: Velocity streamlines of wind behind the blade……………………………………28

Figure 20: Pressure contours on the front of the blades………………………………………29

Figure 21: Pressure contours on the back of the blades………………………………………29

Figure 22: Deformed wind turbine blade……………………………………………………..30

Figure 23: First six mode shapes of the wind turbine blade…………………………………..32

Figure 24: ε-N curve for glass-epoxy composite material…………………………………….33

MKSSS's Cummins College of Engineering for Women

List of Tables
Table 1: Wind turbine blade geometry…………………………………………………..........18

Table 2: Mesh metrics………………………………………………………………………...21

Table 3: Properties of air……………………………………………………………………...22

Table 4: Properties of fibre glass-epoxy composite material…………………………..…...…22

Table 5: Mass flow rate……………………………………………………………………….25

Table 6: Force and moment reaction vectors…………………………………………………30

Table 7: First six modes of the turbine blade……………………………………………..…..31

Table 8: Summary of maximum stress values………………………………………………..33

MKSSS's Cummins College of Engineering for Women

Symbol Description Units

A Area m2

𝐶𝐷 Drag coefficient -

𝐶𝐿 Lift coefficient -

𝐶𝑃 Power coefficient -
𝐶𝑇 Torque coefficient -
C Chord length m
𝐸𝑘 Kinetic energy J
E Young’s modulus Pa
F Force N
f Camber position m
G Shear modulus Pa
h Tower height m
m Mass kg
N Cycles to failure -
P Power W
p Momentum kgm/s
𝑃𝑇 Power output of W
𝑅𝐷
turbine
N
𝑅𝐿
Drag force
N
Lift force
R m
𝑆1
Radius
Pa
𝑆2
First principal stress
Pa
𝑆3
Middle principal
Pa
𝑆𝑢
stress
Pa
𝑆𝑣𝑚
Third principal stress
Pa
Ultimate stress
T Nm
Von Mises stress

Torque

𝑇𝑅 Torque of turbine Nm
MKSSS's Cummins College of Engineering for Women

V Velocity m/s

𝑥𝑓 Maximum camber m

α Angle of attack °

ε Strain -

λ Tip speed ratio -

ρ Density Kg/m3

σ Stress Pa

𝜎𝑎 Stress amplitude Pa

𝜎𝑒 Modified stress Pa
𝑞𝑣
amplitude
Kg/m·s
Viscosity
τ𝑟 s-1
Angular velocity
ω -
Poisson’s ratio
υ
MKSSS's Cummins College of Engineering for Women

ABSTRACT
The aim of this study was to further develop data on the structural integrity of wind turbine blades
by designing and modelling a wind turbine blade based upon current wind turbine blade
specifications and undergoing a fatigue analysis. The model uses available geometrical data from the
NREL 5MW turbine and the Vestas 117 4.2MW turbine in order to approximate a large scale, upwind,
horizontal, axis triple bladed wind turbine. The model uses fibre glass-epoxy composite as a material
property. CFD and FEM calculations were achieved in order to analyse its structural integrity. For a
rotor radius of 58.5m, the blade deformation was found at a maximum of 1.8874m at a maximum
operational wind-speed of 23m/s. A modal analysis was carried out and the natural frequency of the
blade was found to be 1.0236Hz. Finally, a fatigue analysis was carried out at the maximum
operational wind speed where it was found that the number of load cycles before failure, N ≈
100,000,000 for the fibre glass-epoxy composite wind turbine blade. More continued work is to be
done to carry out a mesh refinement analysis and to study a wider variety of parameters.
MKSSS's Cummins College of Engineering for Women

1.0 Introduction

We are currently living through a time where scientists and engineers alike are all racing climate
change in the hope to possibly halt its effects and prevent global disaster. Investors are bidding their
money on clean, renewable energy as the most effective method of doing so, by reducing the
demand for fossil fuels. Amongst these clean, renewable energies, the wind energy industry in
particular has exploded in the 21st century. According to some analysts, this was due to a surge in
the number of people concerned about climate change. [1] Wind power has since become the
fastest growing energy technology in the world. [2] Global onshore wind energy potential is
estimated to be approximately 119.5 petawatt hours per year, allowing for much further growth of
the industry. [3] This provides an excellent opportunity to expand on the wind energy industry and
implement this growth. Whilst doing so, it is just as important to continue building upon the
knowledge of wind turbines. This allows us to harness wind energy potential as efficiently and as
cleanly as possible.

1.1 Wind turbine challenges

Amongst the wind industry it is a commonly known problem that wind turbines require a great deal
of maintenance to keep them running in a safe and efficient manner. Blade failures in particular
exhibit one of the biggest numbers of structural incidents recorded. [4] This may include total blade
failures or breaks in the blade. Blade failure is a result of cracks in the blade which arise from fatigue
or material defects. [5, 6], Even minor blade damages may cause unbalance in the rotating body
which induces extra stress, speeding up the degradation process of the wind turbine. [7] Lightning
strikes and ice build-up are also commonly known causes of blade failures. [8] A much wider scope of
problems may incur in further areas of the wind turbine such as the power system, the auxiliary
system or fatigue in other structures such as the tower or the hub.

Blade damage requires extensive repair with a long down time. Down time leads to a total reduction
in renewable energy generation from a wind farm and thus a reduction in revenue for the wind farm.
Furthermore, the cost of the blades accounts for 15–20% of the total turbine cost. [9] In addition, the
replacement of wind turbine blades is not environmentally friendly, not only due to the down time in
renewable energy generation but due to the manufacturing process of a wind turbine blade.
Information provided by wind blade 202068608 2

manufacturers shows that each blade can weigh as much as 12 tonnes and that the manufacturing
process generates around 10–15% of waste. [10] In addition, 76.2% of waste produced in the
manufacturing process is inorganic material. [11]

1.2 Goals of this project

The goal of this project is to design a wind turbine blade based off current modern blade
technologies and analyse its structural integrity, specifically to conduct a fatigue assessment analysis.
Designing the wind turbine blade consists of using a computer aided design (CAD) with limited wind
turbine rotor specifications to model the more detailed geometry. Analysing the structural integrity
of the wind turbine blade will consist of carrying out a fluid flow analysis, a static structural analysis,
and a modal analysis. Decisions for the model must be carefully considered, such as whether it is a
MKSSS's Cummins College of Engineering for Women

partial or complete assembly, or if it includes solid and or surface bodies. Approximations and
boundary conditions must be carefully determined. The material properties must also be assigned to
various parts of the geometry and finally, realistic loads were assigned to the simulation.

Thus, by the end of this experiment, one will have developed a methodology for designing and
MKSSS's Cummins College of Engineering for Women

2.0 Theory and background

2.1 Wind turbine fundamentals

The fundamental role of a wind turbine is to convert kinetic energy contained in an air stream into
electrical power. This is done in via a turbine rotor/blade where the kinetic energy is first converted
into mechanical work. The rotor is then attached to a generator through a gearbox where the
mechanical work is converted into electrical power.

2.1.1 Classifications

Some wind turbine sites can contain over one hundred active wind turbines. These can be either
onshore or offshore wind farms. Offshore wind turbine farms tend to be faster growing due to the
larger area of space available and less planning restrictions. Offshore wind turbines also have the
benefit of being in open areas which allow for higher and more consistent wind speeds. [12] This
means they experience less turbulence and as a result are less prone to fatigue damage.
Maintenance and construction for offshore wind farms may be more expensive and this is the benefit
of onshore wind farms. They are also easier to connect to the national electricity grid with less losses
through cables.

Modern wind turbines can be classified depending on the rotational axis of the wind turbine blades.
Horizontal axis wind turbines (HAWTs) are far more common as they have a low cut-in wind speed.
Since they can be turned at an angle to the direction of incoming wind, they exhibit easy furling.
Vertical axis wind turbines (VAWTs) are a much older design of wind turbine, generally with a lower
energy efficiency. Another major problem of VAWTs is faster fatigue damage on the blades as a result
of the cyclic aerodynamic stress. [13] Although VAWTs are much less common they benefit by being
much cheaper to construct since the gearbox and generator can be situated on ground level. This is
shown in figure 1 as a side by side comparison.
MKSSS's Cummins College of Engineering for Women

Figure 1: Comparison between HAWTs and VAWTs [14]

HAWTs may be further classified by the number of wind turbine blades attached to the hub. They
could be single bladed, two bladed, three bladed or multi bladed. A single bladed HAWT is the least
widely used although they benefit by saving on materials. In order to balance the weight of the blade
there is usually a counterbalance attached to the hub on the opposing side of the blade. Two bladed
HAWTs may also save more materials than a three bladed HAWT and are more energy efficient than
a single bladed HAWT. They also require no need for a counterweight. A three bladed HAWT will
however capture more energy than either the single bladed or two bladed HAWT and it is for this
reason they are the most widely used. Multi bladed HAWTs are mostly used as windmill water
pumps. They are typically not used for generating electrical power. [15] See figure 2 for a side by side
comparison.

Figure 2: Comparison between triple, double, single, and multi bladed wind turbines. [16]
MKSSS's Cummins College of Engineering for Women

HAWTs can also be classified as upwind or downwind turbines. If the wind stream makes contact
with the rotor before the tower it is known as an upwind wind turbine. These have an advantage of
functioning in undistributed air flow. [12] The disadvantage of using an upwind wind turbine is the
need for a 'yaw' system to control the direction the wind turbine faces. This is needed to prevent the
turbine from turning downwind. Conversely, if the wind stream makes contact with the tower before
the rotor it is a downwind wind turbine. These have the advantage of not requiring a yaw system
since the wind will keep the rotor in the downwind configuration. These are however less efficient
due to a wake formed by the tower causing uneven air loads on the blades. This also leads to larger
fatigue damage on the blades. [13] See figure 3 for a side by side comparison.

Figure 3: Comparison between upwind and downwind horizontal axis wind turbines. [17]

Most wind turbines manufactured and produced are three bladed HAWTs. [18] The main
components

Figure 3: Comparison between upwind and downwind horizontal axis wind turbines. [17]

Most wind turbines manufactured and produced are three bladed HAWTs. [18] The main
components of such a wind turbine are shown in figure 4. This can be separated into the rotor, which
includes the rotor blades and the rotor hub; the nacelle which includes the gearbox, rotor break,
generator, and yaw mechanism; and the tower. [19]
MKSSS's Cummins College of Engineering for Women

Figure 4: Components of a horizontal axis wind turbine. [18]

2.1.2 The wind turbine rotor

The wind turbine rotor is the most essential component of a wind turbine as it is what extracts the
kinetic energy from an air stream. This energy gives the wind turbine rotor angular momentum which
can then drive an electrical generator. The rotor consists of the hub and the wind turbine blades.

The hub connects the wind turbine blades to each other and to the nacelle. By doing so it is able to
transfer loads in each of the blades to the drive train. The type of hub depends on forces and
moments which are transmitted to the rotor. See figure 5 for a diagram of the hub.
MKSSS's Cummins College of Engineering for Women

Figure 5: Diagram of wind turbine hub [20]

Wind turbine blades can be made from a vast array of materials including wood, steel, carbon fibre
and fibreglass. The properties of the blade depend strongly on the material used as well as the blade
dimensions, operating wind conditions and stresses that arise from bending on the hub. Large scale
wind turbine blades are typically constructed from glass-reinforced plastic (GRP), carbon fibre-
reinforced plastic (CFRP), aluminium and steel. [21] A composite is created in order to achieve long,
lightweight blades with high stiffness. For example, fibre glass wind turbine blades produce high
stiffness, flexural strength, and shear resistance characteristics under extreme thermal and
mechanical conditions. Some blades may also use high grade resins such as epoxy.

2.2 Aerodynamics

The aerodynamics of the wind turbine blades is the study of air forces on the wind turbine blades
when air passes through. The qualities of a wind turbine blade that determine the aerodynamic
interactions are the blade airfoil, twist, pitch, chord length and radius.

2.2.1 Airfoils

The airfoil of a wind turbine blade is simply a cross section of the blade, as shown in figure 6. It is the
shape of this cross section that can create the mechanical forces necessary to move the wind turbine
rotor due from the motion of air around the airfoil. The twist determines the angle at which the air
stream makes contact with the blade airfoil at a given radius, while the pitch determines the angle
the tip of the blade. The chord length determines the length of the airfoil at a given radius along the
blade. The radius simply refers to the swept radius of the whole wind turbine rotor.
MKSSS's Cummins College of Engineering for Women

There are many types of airfoils available when designing a wind turbine blade. Figure 6 shows a
diagram depicting a typical wind turbine airfoil from the National Advisory Committee for
Aeronautics (NACA) series of airofils, showing all the significant parameters such as chord length c,
the camber f and the position of maximum camber (xf).

Figure 6: Geometric airfoil parameters of the NACA airfoil series [18]

NACA airfoils are listed with a four digit code which contains data on the airfoil geometry and partly
on the aerodynamic properties. [18] The first digit contains information on the maximum camber to
chord ratio. The second digit is the camber position in tenths of the chord length. The third and fourth
digit are the maximum thickness to chord ratio in percentage. Airfoil coordinates are widely available
to download online, usually with a chord length equal to 1. Figure 7 shows some examples of varying
airfoil geometries
MKSSS's Cummins College of Engineering for Women

Figure 7: Examples of airfoil types [22]

Airfoils are designed for air to pass through the leading edge of the airfoil, as depicted on the left side
of figures 6 and 7. The air then passes over the upper and lower surfaces of the airfoil. Due to the
differences in shape between these two surfaces, the air will have to travel a different distance per
unit time depending on if it is passing the upper surface or the lower surface i.e. air particles will
move faster or slower on the upper or lower surfaces of the airfoil.

According to Bernoulli’s theorem, the difference in the speeds across the upper and lower sides of the
airfoil is created by the difference in pressures on these surfaces. These pressure differences create a
net force (R) on the blade. [23] These are characterised as the lift force or drag force. The lift force is
the component of the net force that is vertical with respect to the incoming air stream. The drag
force is the component of the net force which is horizontal to the incoming air stream. The lift force is
given by equation 1.

𝑅𝐿=𝐶𝐿12ρ𝐴𝑉2 (1)
Here 𝐶𝐿 is the lift coefficient. ρ𝐴𝑉2 represents the density of air, the projected airfoil area and the
velocity of the undisturbed air squared. The drag force is given by equation 2.

𝑅𝐷=𝐶𝐷12ρ𝐴𝑉2 (2)
Here 𝐶𝐷 is the drag coefficient. Figure 8 depicts a force diagram on an airfoil indicating lift and drag
forces.
MKSSS's Cummins College of Engineering for Women

Figure 8: Forces on a wind turbine airfoil [24]

Lift and drag coefficients are typically measured in experiments known as wind tunnel experiments.
Here, the lift and drag forces are measured by transducers located on the vertical and horizontal
planes. [25] These coefficients depend on the angle of attack of the airfoil and so the experiments are
usually done at a range of angles to generate a plot between the coefficient and the angle. This
angle of attack is described by the twist at a given radius of a wind turbine blade, and the pitch.

2.2.2 Power and torque

The kinetic energy 𝐸𝑘 of a given air stream can be described by the mass of air 𝑚 and the velocity
of the air stream 𝑉 using equation 3:

𝐸𝑘=1/2𝑚𝑉2 (3)
We can consider the volume of air to be described as a cylinder with the radius equal to the swept
radius of the wind turbine rotor. The energy per second can be given by equation 4:

𝐸𝑘=1/2
̇ 𝑚̇𝑉2 (4)

Here, 𝑚̇ is the change of mass per second which is given by equation 5:

𝑚̇=ρ𝐴𝑉 (5)

Here, 𝜌 is the density of air. 𝐴 is simply the cross sectional area of the cylindrical air stream,
𝐴=𝜋𝑟2. Since power is defined as the energy per unit time, we get equation 6 for the power 𝑃 of
the wind passing through the rotor: 202068608 11

𝑃=𝐸𝑘=1/2ρ
̇ 𝐴̇𝑉3 (6)

Since this is an ideal case, and it is not possible to extract all this energy, some of the kinetic energy is

transferred from the wind to the wind turbine rotor is defined by the power coefficient 𝐶𝑃. This
transferred to the rotor and the rest leaves the rotor by the wind. The amount of wind energy that is

describes the efficiency of the wind turbine rotor and is given by equation 7:

𝐶𝑃=2𝑃𝑇/ρ𝐴𝑉3 (7)

𝑃𝑇 is the power output of the wind turbine. There are many factors involved to determine the
power coefficient including the number of wind turbine blades, the airfoil, pitch and twist. The torque
MKSSS's Cummins College of Engineering for Women

of the wind turbine rotor may be determined by using equation 8 for the thrust force done by the
wind turbine rotor:

𝐹=1/2ρ𝐴𝑉2 (8)

From this we can now determine the torque to be shown in equation 9:

𝑇=1/2ρ𝐴𝑉2𝑟 (9)

Here 𝑟 is the radius of the wind turbine rotor. Since this is again the ideal case, the real value of

torque coefficient 𝐶𝑇. This is defined as the ratio between the real torque and the theoretical torque
torque will be less than this value. We can express this real value by using a term known as the

described in equation 9. This gives us the equation 10:

𝐶𝑇=2𝑇𝑅/ρ𝐴𝑉2𝑟 (10)
In equation 10, 𝑇 is the real torque of the wind turbine rotor.

One other value used to determine the efficiency of the interaction between the wind turbine rotor
and the stream of wind is defined as the ratio between the velocity of the tip of the wind turbine
rotor and the velocity of the stream of wind. This ratio is known as the tip speed ratio λ and is given
by equation 11:

λ=𝑟𝜔/𝑉 (11)
MKSSS's Cummins College of Engineering for Women

Here, 𝜔 is the angular velocity. This can also be expressed in terms if the torque coefficient and the
power coefficient using equation 12:

λ=𝐶𝑃/𝐶𝑇 (12)

These values based on aerodynamic theories are very useful when designing a wind turbine rotor.
The tip speed ratio for instance will depend on the function of the wind turbine. Wind pump rotors
require a low tip speed ratio due to the high starting torque. On the other hand, electricity
generating wind turbines require a high tip speed ratio due to the fast moving rotor. [25] The number
of blades a wind turbine uses directly affects the tip speed ratio.

2.3 Loads on a turbine blade

There are numerous sources for loads present on a wind turbine at any given time. Since it is these
loads that cause the stress required for fatigue failure to occur, we must study these loads in depth.
The most significant of these is the aerodynamic loads due to the wind. These may be considered
uniform, steady state air flow loads when the wind is in the direction of the turbine with a uniform
distribution of wind speed across the rotor. Typically loads are unsteady and spatially non-uniform
due to wind shear and cross winds, causing cyclic load changes on the rotor. There may also be,
gravitational loads, inertial loads due to the dead weight of the turning rotor blades and stochastic
loads caused by wind turbulence. [18] These loads can be determined by geographic and weather
conditions as well as relative locations of a turbine in a heavily populated wind farm. Finally, there
may be operational loads present at different operating states of the wind turbine. [25]

2.3.1 Uniform and steady state air flow

Assuming an even distribution of air across the rotor is a strong idealisation but it is useful for
practical purposes when calculating the mean loading over a long period of time. In this case the
aerodynamic loads present on a wind turbine blade can be described using equations 1 and 2 for the
lift and drag forces. These forces are largely determined by the effective wind speed, which varies
from the root of the blade to the tip when the rotor is in operation. This distribution of loads along
the blade also varies significantly from the start up wind speed to the shutdown wind speed, due to
the blades twist. The twist is designed for an optimal wind speed, and at higher wind speeds the flow
separates in the blade sections near the hub causing the change in load distribution. [18]
MKSSS's Cummins College of Engineering for Women

2.3.2 Vertical wind shear

In reality, wind flow velocity is non uniform across the wind turbine rotor at any given time. This
causes a much more complex variety of loads present. This however can be described by wind shear
allowing us to approximate a more accurate solution for loading. Wind shear describes the change in
wind velocity with respect to height. Wind typically moves much slower as it gets closer to the
ground due to the resistance of the surface, whereas at greater heights wind is much freer to move
quickly. This is depicted in figure 9.

Figure 9: Vertical wind shear [26]

Wind shear can be approximately described using the power law. This is detailed in the following
equation:

𝑉2/𝑉1=(ℎ2ℎ1)α (13)

Here, 𝑉1 and 𝑉2 represent the mean velocity at two given heights, ℎ1 and ℎ2. α is known as the
shear coefficient. Vertical wind shear causes a cyclic change in loading as the wind turbine rotates
around in a circle. This produces a peak loading every revolution when a wind turbine blade is
pointed vertically upwards and a minimum loading when the blade is pointed downwards to the
ground.

2.3.3 Wind turbulence stochastic loads

Incorporating wind turbulence into a loading calculation can prove to be challenging due to the
stochastic introduction of wind velocities. Wind normally moves in all three directions, longitudinal,
vertical, and lateral. [12] However, it is typical for most calculations to assume a one-dimensional
fluctuation in wind speed in the longitudinal direction due to the difficult
MKSSS's Cummins College of Engineering for Women

nature of incorporating a second dimension. Spatial distribution of wind speeds along the swept
rotor area in this longitudinal direction is much more significant as far as loading is concerned. [18]
Wind turbulence is characterised as high frequency fluctuations with occasional sporadic high
deviations in wind speed known as gusts, usually described by a gust factor. [27] While turbulence
can be very challenging to model and predict, it does contribute significantly to wind fatigue,
particularly to the rotor blades.

2.3.4 Gravitational and inertial loads

Gravitational loads for wind turbines become significant for large scale wind turbines. As the size of
the turbine increases, so does its mass. The structures weight becomes a considerable factor when
designing strong supports such as the tower or the hub.

Centrifugal loads are present while the wind turbine is in operation. These are not very significant in
larger turbines with lower rotational speeds. [25] Centrifugal loads can also be used to counter act
aerodynamic loads by a special trick known as coning. This is where the rotor blades are inclined at a
small angle downwind such that the centrifugal load opposes the upwind direction of the
aerodynamic loading. This can however be counter intuitive if the rotor is subject to strong gusts,
causing wind direction to be momentarily reversed.

Gyroscopic loads are present in upwind wind turbines while the rotor is being yawed into the
direction of the wind. If the rotor is yawed quickly these loads can be very significant but in most
cases the rate of yaw is relatively low.

2.4 Computational methods

2.4.1 Finite element method (FEM)

FEM is a powerful numerical method is a numerical operation that is used to find a solution to a wide
variety of engineering and physics problems. [28] These are typically problems involving stress
analysis, heat transfer, electromagnetism, and fluid flow. In engineering there are a wide variety of
complex problems for which we can’t achieve exact solutions. This may be due to complex governing
differential equations. For these difficult problems we resort to numerical methods such as FEM.

FEM involves an integral of the differential equation over the model domain, which is divided up into
individual elements connected by nodes. These elements are solved individually by defining
approximate interpolation for each element, which are then placed

together to provide a solution for the entire model. [29] The division of the model in to differing
elements is known as the mesh which can be shown in figure 10
MKSSS's Cummins College of Engineering for Women

Figure 10: Mesh element example in two dimensions [30]

Once the mesh is finalised, there are several methods that can be used by FEM to generate a
solution, such as direct formulation, minimum total potential energy formulation and weighted
residual formulations. In general, an approximate continuous function is assumed for an element to
represent its behaviour. Equations are then developed for every other element based on the solution.
Finally, the elements are assembled back together to present the entire problem. From here,
boundary conditions, initial conditions and loading may be applied to the problem to obtain solutions
from a set of algebraic equations. [28]

2.4.2 Computational fluid dynamics (CFD)

CFD is used to predict and model fluid flow or heat transfer in a system. CFD methods are used in
applications across the wind energy industry. As the wind energy continues to grow and potential
wind farm sites become more and more scare, complex CFD calculations will be required to advance
the industry further. [31]

CFD solves complex problems by using a using the model domain divided up into individual elements
known as the mesh. From there the physical and chemical phenomena need to be modelled, fluid
properties must be inputted the appropriate boundary conditions must be placed on cells that touch
the boundary. [32] CFD simulations can be done using FEM to solve differential equations, however it
is much more common to use the finite volume method (FVM) as the numerical method. FVM is a
numerical method based on transport equations governing fluid flow and heat transfer. These
governing equations are the continuity (conservation of mass, equation 14) and Navier-Stokes
(Conservation of momentum, equation 15) equations, and are defined as follows:
MKSSS's Cummins College of Engineering for Women

∂ρ/∂𝑡+∇⋅ρν𝑟⃗⃗⃗ =0 (14)

∇⋅(ρν𝑟⃗⃗⃗ ν𝑟⃗⃗⃗ )+ρ(2ω⃗⃗ ×ν𝑟⃗⃗⃗ +ω⃗⃗ ×ω⃗⃗ ×𝑟 )=−∇𝑝+∇⋅τ𝑟 (15)

Here, ν𝑟⃗⃗⃗ is the relative velocity and ω⃗⃗ is the angular velocity. 𝑝 is momentum and ρ is density. τ𝑟 is
the viscous stress term, 2ω⃗⃗ ×ν𝑟⃗⃗⃗ is the Coriolis effect term and ω⃗⃗ ×ω⃗⃗ ×𝑟 is the centripetal acceleration
term. [32] CFD also has turbulence modelling capabilities using a model known as the k-ω SST model,
which uses a modified Navier-Stokes equation known as the Reynolds averaged Navier-Stokes
equation. [32]

2.5 Fatigue assessment analysis

In material science fatigue is the weakening of a material caused by cyclic loading that results in
progressive and localised structure damage and the growth of cracks. If fatigue is considered a
possible failure mode in a component then the fatigue strength of the material, the operating stress
levels and the minimum cycles to failure must be considered as part of the design process.

Fatigue can be characterised as either being low cycle fatigue or high cycle fatigue depending on the
number of stress cycles required before bringing a body to failure. [33] Low cycle fatigue usually
occurs in less than 10,000 cycles whereas high cycle fatigue occurs thereafter.

The number of cycles to failure can be found for a material by using its S-N curve. This is a plot
between number of cycles to failure N and the allowable stress amplitude S of a cyclic stress. There
are currently numerous sources of available data for different material’s S-N curves, as well as a
methodology for approximating S-N curves. [34, 35]
MKSSS's Cummins College of Engineering for Women

3.0 Methodology

The methodology was split in to two large sections. Firstly, the model design was developed by
defining the geometry of the wind turbine blade and the mesh for both the blade and the air around
it. Time was spent to make sure this geometry was made as accurately as possible. Secondly, the
simulations of various loads were introduced to the model. Initially a fluid flow model was simulated
in order to achieve aerodynamic loads. In this model, a fluid flow, a static structural analysis and a
modal analysis are required. The model geometry was constructed on ANSYS DesignModeler and
would include just one wind turbine blade given the symmetry between the blades. Limits to the
number of mesh elements on ANSYS student version also limited the volume of air to be simulated.

3.1 Model design

A wind turbine model was chosen at the beginning of this experiment to guide the computer model
and act as a framework. Airfoil and twist coordinates were chosen from the NREL 5MW wind turbine
as these specifications are widely available online. [36] This a un upwind, three bladed HAWT. More
commercial wind turbines typically do not share airfoil and twist information as this would be
confidential. The NREL wind turbine was then further modified to be closer to a commercial turbine
model. Dimensions were modified according to available information regarding the Vestas V117 -
4.2MW model, changing the model maximum chord length and rotor radius. This model was chosen
as it is a very modern turbine designed to be used in strong wind conditions, like the NREL wind
turbine. [37] It is also currently being used for the current development of the Viking wind farm in the
Shetland Islands, United Kingdom. Details of the hybrid model developed component dimensions can
be found in table 1.
MKSSS's Cummins College of Engineering for Women

Table Radius (m) Twist (°) Chord length (m) Airfoil type

1 1.393 13.308 3.046 Cylinder

2 2.662 13.308 3.046

3 10.911 13.308 3.918 DU40 A17

4 14.718 11.480 4.000 DU35 A17

5 18.525 10.162 3.833

6 22.333 9.011 3.653 DU30 A17

7 26.140 7.795 3.445 DU25 A17

8 29.947 6.544 3.223

9 33.754 5.361 3.011 DU21 A17

10 37.562 4.188 2.800

11 41.369 3.125 2.588 NACA64 A17

12 45.176 2.319 2.377

13 48.983 1.526 2.165

14 52.156 0.863 1.989

15 54.694 0.37 1.794

16 57.232 0.106 1.220

17 58.501 0.106 1.220

For this model the pitch was kept to 0°and the tilt was kept to 0°.

3.1.1 Geometry

These coordinates were carefully inputted in to DesignModeler generating 17 different cross sections
of the wind turbine blade. With these cross sections defined, the skin/loft feature was then used to
generate a smooth surface over each of the elements and a body was created. This can be seen in
figure 11:
MKSSS's Cummins College of Engineering for Women

Two copies of this same blade were then generated. One of these will be used to model the mesh of the blade for FEA. A
varying thickness was added to the FEA geometry of this model as a linear function from 0.15m to 0.005m, from the base to
the tip of the blade. The second model will be used as a volume to be subtracted from a greater volume of air to be used
with CFD. The volume of air was designed for the CFD calculation with a length of 270m, an upwind radius 120m and a
downwind radius of 240m. It is designed as one third of a semi-circle as depicted in figure 12. This will then be extrapolated
to create all three wind turbine blades.

Figure 12:
MKSSS's Cummins College of Engineering for Women

3.1.2 Mesh

The fluid mesh was generated first since the results for the CFD calculations will be used for the FEM
calculations. This process started with a default 3D mesh generated automatically by ANSYS fluent
which gives a mesh with tetrahedral cells that are tailored to CFD calculations. The generated mesh
contains adequate inflation and sizing to run our calculation, but smaller element sizes may be used
to improve accuracy. Additional modifications to the mesh were made. ‘Match control’ was applied
such that the nodes of two surfaces with quadrilateral faces match up. This is important when
extrapolating the fluid geometry to 360°. The final mesh contains 42,378 nodes and 237,733
elements, which is within the limit of ANSYS student version. A cross section can be seen in figure 13.

Figure 13: 3D cross section of fluid mesh

Secondly, the mesh for the blade was created for FEM calculations. This requires much less elements
as it is treated as a hollow body with a small thickness. The process for meshing once again started
with the default mesh generated automatically by ANSYS mechanical tailored for FEM calculations. A
mapped face mesh was generated on the surface to create quadrilateral elements. The final mesh
contains 8,755 nodes and 8,507 elements, which is within the limit of ANSYS student version. A cross
section can be seen in figure 14
MKSSS's Cummins College of Engineering for Women

Figure 14: Blade mesh

To ensure good quality of the mesh produced a quick analysis of orthogonal quality and skewness
was done. These are values that range between 0 and 1. In general, it is advised to keep the
minimum orthogonal quality to be greater than 0.15 and the maximum skewness lower than 0.95.
[38] Details of the final mesh metrics that satisfied these conditions are given int table 2

Table 2: Mesh metrics

Mesh metric Blade Fluid

Minimum orthogonal 0.417 0.186

quality

Maximum orthogonal 1 0.993

quality

Average orthogonal 0.997 0.753

quality

Minimum skewness 0.00000874 0.000256

Maximum skewness 0.670 0.814

Average skewness 0.0412 0.246

3.2 Simulations
MKSSS's Cummins College of Engineering for Women

In order to carry out the simulations in our model we must first define material properties. The
material properties for air are simply defined with the following values in table 3.

Table 3: Properties of air

Property Value

Density (kg/m3) 1.225

Viscosity (kg/m·s) 1.789×10

The properties of the wind turbine blade were then chosen based upon a database found on wind
turbine blades composites. [39] For this analysis the blade design was to be made of a fibre glass-
epoxy composite material. These properties assume material is isotropic and are summarised in the
table 4.

Table 4: Properties of fibre glass-epoxy composite material

Property Value

Density (kg/m3) 2,600

Young’s modulus (GPa) 85

Poisson’s ratio 0.23

Shear modulus (GPa) 36

Ultimate tensile 2.05

strength (GPa)

Given the variable thickness of the wind turbine blade outlined in the geometry section, we can
compute the mass of the wind turbine blade to be 88,531kg.

3.2.1 CFD

CFD simulations were first carried out to determine the air flow and the loads these impose on the
wind turbine blade. The CFD model is set up by defining the governing equations and boundary
conditions within ANSYS fluent.

Settings were modified to carry out a steady simulation using the pressure based solver. Model
settings were modified to use the k-ω SST model. This simulation must be carried out in the rotated
frame of reference by enabling ‘frame rotating’ and inputting the blade rotational velocity to be -
1.80rad/s defined around the positive Z-axis. This is equal to approximately 17.5RPM which is the
upper limit of the rotor speed as specified by the Vestas 117 data sheet. [37]

Boundary conditions are then defined. The inlet surface of the fluid and the outer curved surface of
the fluid are given air velocities of -23m/s in the Z-direction. This is the re-cut in wind speed for the
specified wind turbine indicating it’s the highest stable operating wind speed. The turbulent intensity
is set to 5%. The turbulent viscosity ratio is set to 10. The outlet of the fluid is given a pressure
MKSSS's Cummins College of Engineering for Women

101,325Pa which is equal to 1atm. Next, the two match controlled surfaces are set as interfaces
connecting to each other.

Finally, the settings for the solution were set for high accuracy increasing the change for the solution
to converge. The pressure-velocity coupling scheme was set to ‘coupled’, the spatial discretization
pressure was set to ‘standard’, ‘pseudo transient’ and ‘high order term relaxation’ were turned on.
The Residuals for velocity and continuity were set to 10-6 to ensure the calculation doesn’t stop
before converging. The integral static pressure is set to be monitored so one can see in real time if the
solution is converging. The number of steps was set to 1,500, the settings were initialized, and the
calculation was run.

3.2.2 FEM

FEM calculations were carried out to complete a static structural analysis and a modal analysis
within ANSYS mechanical. Default static structural simulation settings were first loaded. The blade
was given a rigid node at the base to simulate the assumed effect of having the blade be rigidly
connected the hub. The blade was then given the same rotational velocity of 1.80 rad/s in order to
calculate centripetal loading. ‘Large deflection’ setting was turned on to simulate the stress stiffening
effect due to the fast spinning blade. The pressure loads from the CFD calculation were then simply
imported to ANSYS mechanical. The calculation was run and ANSYS will automatically perform the
calculation by forming the stiffness matrix for each element to get nodal displacements. From the
nodal displacements, ANSYS can work out equivalent stress, force reactions and moment reactions.

A modal analysis was also carried out using the same rigid node and rotational velocity to determine
the first six modes. The frequency for the first mode was then checked to determine if a harmonic
response analysis is required. If the time required to apply a load is greater than three times the
natural period, a dynamic response analyses is not required as the effects are negligible. [40]
MKSSS's Cummins College of Engineering for Women

3.2.3 Fatigue assessment analysis

The methodology used in carrying out a fatigue analysis is similar to the procedure used in the
European Pressure vessel code and the SAE design handbook. [41, 42] Areas of high stress are
located within the FEM model results file. Specifically, these include the maximum von Mises stress
(Svm), principal tensile stress (S1) and principle compressive stress (S3).

Using equation 16 the appropriate von Mises stress amplitude σa is then calculated based on the
actual von Mises stress found from the FEM results.

𝜎𝑎=𝑆𝑣𝑚/2 (16)

This equation is used when the load starts at zero, rises to a peak and falls back to a minimum stress
at zero as it adjusts the amplitude to cycle around the mean stress. If the mean stress is tensile we
must modify the amplitude further using equation 16.

𝜎𝑒𝑞𝑣=𝜎𝑎[𝑆𝑢/𝑆𝑢−(𝑆1+𝑆2+𝑆32)] (17)
Here, Su is the ultimate stress of the material. The term (𝑆1+𝑆2+𝑆32) is the mean stress using the
first, second and third principal stress. This is based on the Goodman diagram which assumes that a
compressive stress does not affect fatigue strength. [43] This modified von Mises stress is then used,
with the appropriate notched S-N curve to calculate predicted cycles to failure.
MKSSS's Cummins College of Engineering for Women

4.0 Results and discussion

In analysing the results, one must first determine if results have converged sufficiently. The first step
is to ensure mass is adequately balanced between the inlet, outer surface and the outlet. The mass
flow rate at each of these surfaces is summarised in table 5 below

Table 5: Mass flow rate

Fluid surface Value (kg/s)

Inlet surface 422,841.22

Outer surface 1,273,811.2

Outlet surface -1,696,652

Net -0.1128356

These were found to be adequately balanced indicating the model is good. The next step is to study
the convergence of the integral static pressure. This is shown in figure 15 as seen where the integral
static pressure converges at around -335,000Pa.

4.1 Fluid flow analysis

Results for CFD can be visualised and analysed using the program CFD-Post. Here we can visualise the
complete triple bladed wind turbine rotor. Figure 16 shows this with vector arrows representing the
tangential velocity of the wind turbine rotor varying along the radius. The legend shows a maximum
velocity of 105.3m/s.
MKSSS's Cummins College of Engineering for Women

Figure 16: Velocity vectors of wind turbine rotor

Next we can visualise the velocity streamlines of the wind as it passes through the wind turbine rotor.
These can be seen in figure 17 and 18. As we can see the velocity at the inlet is equal to the specified
23m/s. There is then a clear drop in velocity behind the wind, which is shown figure 19. This is due to
the wake behind the wind turbine blade. Finally, there is also an increase in velocity around the wake.
This can be described by a balance in momentum
MKSSS's Cummins College of Engineering for Women

Figure 17: Velocity streamlines of wind for full model

Figure 18: Velocity streamlines of wind with velocity vector of the wind turbine rotor
MKSSS's Cummins College of Engineering for Women

Figure 19: Velocity streamlines of wind behind the blade.

Next we can look at the pressure contours on the blade. These are shown in figure 20 and 21. Here
we can see that the maximum pressure on the front and back of the blade are different, and the back
surfaces of the blades experience a pressure that is much higher in magnitude. This pressure
difference creates the lift force required for the wind turbine to rotate, since there is a component of
this lift force that is in the plane of rotation. The component of lift that is parallel with the direction of
incoming wind is what creates deflection of the blade, as will be studied in the static structural
analysis.
MKSSS's Cummins College of Engineering for Women

Figure 20: Pressure contours on the front of the blades

Figure 21: Pressure contours on the back of the blades

MKSSS's Cummins College of Engineering for Women

4.2 Static structural analysis

The first result studied within the structural analysis is the total deformation of the wind turbine
blade. This is shown in figure 22 with the shadowed blade showing the original undeformed wind
turbine blade.

Figure 22: Deformed wind turbine blade

The deformation shown in figure 22 is to scale with the length of the blade. This data tells us the total
deformation under these conditions is a maximum displacement of 1.8874m at the tip of the blade,
in the direction of the wind. This is accurate to what happens in large wind turbines under high wind
speeds.

The force reaction and moment reaction calculated are given as vectors. This is summarised in table
6.

Table 6: Force and moment reaction vectors

Force reaction (N) Moment reaction (Nm)

X-axis 5,232,000 356,950

Y-axis -83,791 16,350,000

Z-axis 461,920 3,267,900

Total 5,253,000 16,677,000

MKSSS's Cummins College of Engineering for Women

In table 6 we can see that the largest component of the force reaction is in the X-axis. This is
attributed to the centripetal acceleration of the rotating blade. The majority of the reaction bending
moment is about the Y-axis due to how the blade is deforming.

blade to be 88,531kg and the centre of mass to be at 𝑌=−18.294, we can say 𝐹=−𝑚(𝜔𝑟)2𝑟 =
We can validate our results be hand calculating the radial force. Given the mass of the wind turbine

−(88,531)(−1.80)2(−18.294) = 5,247,459 𝑁

This value agrees with the ANSYS calculation with just a 3% difference.

4.3 Modal analysis

The first six mode frequencies from the modal analysis were found. These are summarised in table 7
along with the maximum deformation.

Table 7: First six modes of the turbine blade

Mode Frequency (Hz) Maximum deformation

(m)

1 1.0236 0.02487

2 2.3442 0.02407

3 2.6771 0.02266

4 4.9117 0.03448

5 7.3539 0.026434

6 7.8855 0.035795

It was found that the natural period of the structure is 0.9769s. This is much smaller than the period
of the turbine moving at 1.80rad/s, having a period of 3.49s. For this reason, it was considered
arbitrary to conduct a harmonic response analysis. Figure 23 shows the mode shapes corresponding
to each of the modes in table 7.
MKSSS's Cummins College of Engineering for Women

Mode 1 Mode 2

Mode 3 Mode 4

Mode 5 Mode 6

Figure 23: First six mode shapes of the wind turbine blade

The mode shapes shown represent the total deformation, with the maximum deformation shown in
red and the minimum deformation shown in blue.

4.4 Fatigue assessment analysis

The nodes of maximum von Mises stress, principal tensile stress and principal compressive stress
were found on the model and are summarised in table 8. The modified stress amplitude was also
calculated for each of these nodes using equation 17, using the ultimate tensile strength shown in
table 4 of 2.05 GPa.

From these equivalent values for the stress amplitude, we can see the maximum stress amplitude
occurs at node 12. To determine the number of stress cycles of this magnitude before failure we must
use an S-N curve for a glass-epoxy composite material. Figure 24 below shows an appropriate ε-N
curve for such a material, with a Young’s modulus of 29.6GPa. [44]
MKSSS's Cummins College of Engineering for Women

Table 8: Summary of maximum stress values

Node 𝑆𝑣𝑚 (Pa) 𝑆1 (Pa) 𝑆2 (Pa) 𝑆3 (Pa) 𝜎𝑒𝑞𝑣

(MPa)

12 5.1356×108 3.2376×108 514.5 -1.950×108 2.7879×108

7998 4.5163×108 4.4473×108 10,150 -4.6282×106 2.5296×108

3008 1.1014×108 1.3289×108 5.6512×107 6.6246×106 5.7835×107

Figure 24: ε-N curve for glass-epoxy composite material [44]

In order to study the curve, one must convert the equivalent stress amplitude calculated into a strain
percentage value using Young’s modulus using equation 18 as the relationship between stress and
strain.

𝐸 = 𝜎/𝜀 (18)
This gives a maximum equivalent strain at node 12 to be 0.32%. According to figure 24, this means
we can expect the number of load cycles before failure N ≈ 100,000,000. This is found to be a
relatively high value and may be attributed to the decision to use pure fibre glass-epoxy composite
material.
MKSSS's Cummins College of Engineering for Women

5.0 Conclusions

In conclusion, a wind turbine blade has been designed and modelled after current rotor blade
specifications, and the structural integrity of the blade was successfully analysed. Overall, it was
found that the experiment was quite effective.

The methodology for the blade design consisted of researching material specifications, geometry
specifications as well as wind turbine airfoils, then, using a CAD program to draw the blade in three
dimensions for use in simulations.

CFD calculations were performed using the CAD model. This proved successful as one was able to
visualise the wind speed variations due to the wake behind the blade as well as the pressure contour
across the blade due to the oncoming wind.

Further analysis was carried out on the structural integrity of the wind turbine blade. This included a
visualisation of the deformation of the blade due to the pressure contours. The force reaction and
moment reaction vectors were also calculated.

A modal analysis was carried out on the blade as well as a simple fatigue assessment analysis. The
number of cycles to failure for our given load was calculated and was found to be quite high,
indicating a long lifespan for the wind turbine blade.

While one has only scratched the surface on fatigue analysis and structural integrity, it was felt that
the simulations used, and the results found could be very useful in making design decisions for wind
turbine blades.

5.1 Further work

The information gathered in this report is limited to an analysis of just one wind turbine model, with
limited data available and limited computer power. The immediate next step for this model is to carry
out a mesh refinement analysis in both the CFD model and the FEM model. A finer mesh improves
accuracy of the model at the expense of computing power but is a necessary step when validating
the mesh. The ANSYS student version is limited to only 512,000 elements and nodes in a CFD model,
and 32,000 elements and nodes in a FEM model, so a full version of ANSYS is needed to show
convergence of the numerical solution with respect to the number of elements in the mesh.
MKSSS's Cummins College of Engineering for Women

While the purpose of this decision is to create a focus on modelling and designing for blade with
limited available data, the methods must be validated across a wider range of wind turbine models
and characteristics. Within the studied model, one might wish to study a wider variety of rated wind
speeds and rotational velocities. Blade material properties may be modified to represent CFRP, GRP,
steel, aluminium, or any composite material. The model geometry may be tested with varying radii,
twist, pitch, tilt and airfoil selection to achieve a more accurate representation of a given wind
turbine model. One may also choose to model the hub, nacelle, and tower for structural integrity.

Finally, a wider variety of analysis may be introduced to the model. These could include gravitational
loads, gyroscopic loads, and non-uniform loads. A transient analysis of the turbine may also be done
over a period of time to reflect potential changes in wind speed and turbulence, and how this might
impact fatigue.
MKSSS's Cummins College of Engineering for Women

6.0 References

[1] I. B. Vasi, Winds of change: The environmental movement and the global development of the
wind energy industry. Oxford University Press, 2011.

[2] M. Sathyajith and G. S. Philip, Advances in wind energy conversion technology. Springer Science
& Business Media, 2011.

[3] Y. Zhou, P. Luckow, S. J. Smith, and L. Clarke, “Evaluation of global onshore wind energy potential
and generation costs,” Environmental science & technology, vol. 46, no. 14, pp. 7857–7864, 2012.

[4] C. C. Ciang, J.-R. Lee, and H.-J. Bang, “Structural health monitoring for a wind turbine system: a
review of damage detection methods,” Measurement science and technology, vol. 19, no. 12, p.
122001, 2008.

[5] C. W. Kensche, “Fatigue of composites for wind turbines,” International journal of fatigue, vol.
28, no. 10, pp. 1363–1374, 2006.

[6] R. P. L. Nijssen, “Fatigue life prediction and strength degradation of wind turbine rotor blade
composites,” Contractor Report SAND2006-7810P, Sandia National Laboratories, Albuquerque, NM,
2006.

[7] P. Caselitz and J. Giebhardt, “Rotor condition monitoring for improved operational safety of
offshore wind energy converters,” J. Sol. Energy Eng., vol. 127, no. 2, pp. 253–261, 2005.

[8] B. Le and J. Andrews, “Modelling wind turbine degradation and maintenance,” Wind Energy, vol.
19, no. 4, pp. 571–591, 2016.

[9] B.-H. Han, D.-J. Yoon, Y.-H. Huh, and Y.-S. Lee, “Damage assessment of wind turbine blade under
static loading test using acoustic emission,” Journal of Intelligent Material Systems and Structures,
vol. 25, no. 5, pp. 621–630, 2014.

[10] P. C. Guerrero, S. D. Mancini, and C. M. Toubia, “Caracterização e reciclagem qumica via

pirólise de resduos da fabricação de pás eólicas,” Holos Environment, vol. 11, no. 2, pp. 147–157,
2011.

[11] P. S. Oliveira, M. L. P. Antunes, N. C. da Cruz, E. C. Rangel, A. R. G. de Azevedo, and S. F. Durrant,

“Use of waste collected from wind turbine blade production as an eco-friendly ingredient in mortars
for civil construction,” Journal of Cleaner Production, vol. 274, p. 122948, 2020.

[12] W. Tong, Wind power generation and wind turbine design. WIT press, 2010.

[13] J. G. McGowan and S. R. Connors, “Windpower: a turn of the century review,” Annual Review of
Energy and the Environment, vol. 25, no. 1, pp. 147–197, 2000.

[14] E. Kumara, N. Hettiarachchi, and R. Jayathilake, “Overview of the vertical axis wind turbines,”
Int. J. Sci. Res. Innov. Technol, vol. 4, pp. 56–67, 2017.

[15] J. Earnest and T. Wizelius, Wind power plants and project development. PHI Learning Pvt. Ltd.,
2011.
MKSSS's Cummins College of Engineering for Women

[16] K. WOLNIEWICZ, W. KUCZYNSKI, and A. ZAGUBIEN, “Evaluation of wind resources for horizontal
and vertical wind turbine,”

[17] M. Yassine Kebbati, “Design, modelling and control of a grid-connected hybrid pv-wind system
(case study of adrar),” Master’s thesis, 2018.

[18] E. Hau, Wind turbines: fundamentals, technologies, application, economics. Springer Science &
Business Media, 2013.

[19] J. F. Manwell, J. G. McGowan, and A. L. Rogers, Wind energy explained: theory, design and
application. John Wiley & Sons, 2010.

[20] E. L. Secco, A. Nagar, C. Deters, H. Wurdemann, H.-K. Lam, and K. Althoefer, “A neural network
clamping force model for bolt tightening of wind turbine hubs,” in 2015 IEEE International Conference
on Computer and Information Technology; Ubiquitous Computing and Communications; Dependable,
Autonomic and Se

[21] M. Hansen, Aerodynamics of wind turbines. Routledge, 2015.

[22] W. B. Tay and K. B. Lim, “Analysis of non-symmetrical flapping airfoils,” Acta Mechanica Sinica,
vol. 25, no. 4, pp. 433–450, 2009.

[23] N. F. Smith, “Bernoulli and newton in fluid mechanics,” The Physics Teacher, vol. 10, no. 8, pp.
451–455, 1972. 202068608 38

[24] D. C. Panigrahi and D. P. Mishra, “Cfd simulations for the selection of an appropriate blade
profile for improving energy efficiency in axial flow mine ventilation fans,” Journal of Sustainable
Mining, vol. 13, no. 1, pp. 15–21, 2014.

[25] S. Mathew, Wind energy: fundamentals, resource analysis and economics, vol. 1. Springer,
2006.

[26] G. Kavari, M. Tahani, and M. Mirhosseini, “Wind shear effect on aerodynamic performance and
energy production of horizontal axis wind turbines with developing blade element momentum
theory,” Journal of Cleaner Production, vol. 219, pp. 368–376, 2019.

[27] W. Frost, B. Long, and R. E. Turner, “Engineering handbook on the atmospheric environmental
guidelines for use in wind turbine generator development,” tech. rep., Tennessee Univ., Tullahoma
(USA). Space Inst.; National Aeronautics and …, 1978.

[28] S. Moaveni, Finite element analysis theory and application with ANSYS, 3/e. Pearson Education
India, 2011.

[29] A. Ismail-Zadeh and P. Tackley, Computational methods for geodynamics. Cambridge University
Press, 2010.

[30] H. Policarpo, Analytical, Numerical and Experimental Study of the Dynamical Response of
Helicoidal Springs and Periodic Bars, 2008. PhD thesis, MSc Thesis, Instituto Superior Técnico,
Universidade Técnica de Lisboa, Portugal.