Aerodynamic Analysis of a Novel Wind Turbine

for an Omni-flow Wind Energy System

Pei Ying

A thesis submitted to the University of Hertfordshire in partial fulfilment of the

requirements of the degree of Doctor of Philosophy

The programme of research was carried out in the School of Engineering &
Technology, University of Hertfordshire

April 2016
Abstract

The purpose of this research is to propose a novel wind turbine for an omni-flow wind energy

system and investigate its aerodynamic performance. The geometry of the novel wind turbine

is based upon the impulse turbine technology which has been successful in wave energy. In

this study, both numerical and experimental studies were conducted to evaluate the

aerodynamic features of this wind turbine. The numerical method was validated by a

comparison between numerical and experimental results over a range of tip speed ratios.

Results obtained from experiments and simulations indicate that the proposed wind turbine

can be employed. Additionally, on the basis of the analyses performed, this new wind turbine

has the potential for having a good startup feature, which means that this wind turbine can be

suitable for applications in an urban environment. As an important component, the stator of

this wind turbine can increase the passing flow velocity by 20%. Meanwhile, the passing flow

direction also can be optimised by the stator.

Aerodynamics of the wind turbine was analysed under the non-uniform flow condition,

because the flow is non-uniform inside the omni-flow wind energy system. It was found that

the maximum power coefficient of such a turbine under the non-uniform flow condition is

lower than that under the uniform flow condition. Due to the non-uniform flow, the blades

experience different flow velocities, and as a consequence, undergo different aerodynamic

loads during one operation cycle. Thus the generated torque and thrust on a blade are

subjected to frequent and periodical changes.

Influences of the geometrical parameters on the aerodynamic characteristics of this wind

turbine were investigated. From the initial study, it was found that changes of hub-to-tip

ratios, numbers of blades, aerofoils and numbers of guide vanes, can significantly affect the
I
power performance. Additionally, the wind turbine obtained high values of maximum torque

coefficients with changing geometrical parameters.

II
To my dear mother and father

谨此献给我亲爱的母亲和父亲
 Acknowledgements

Acknowledgements

Foremost, I wish to express my sincere gratitude to Dr. Yong Kang Chen, Dr. Yi Geng Xu

and Dr. Yuan Tian for their invaluable academic supervision. During this period, Dr. Yong

Kang Chen guided me with great encouragement and patience through this work.

I wish to thank the School technical staff. Many thanks must be given to the technicians of

both the Wind Tunnel Laboratory and the Workshop in the School. Sincere thanks go to

administrators in Science and Technology Research Institute.

Heartfelt thanks also go to my close friends, Aiyuan Yang, Zeng Shen, Yang Jiao, Xu Zhang

and Chuanli Zhao, who always stand by my side. Thanks go to my friends who always

encourage and trust me.

Last but not least, I must thank my family for their support and unconditional love, not just

through the past years of my study, but up until now.

III
 Publications

Publications

Journal Papers

[1] P. Ying, Y. K. Chen, and Y. G. Xu, "An aerodynamic analysis of a novel small wind

turbine based on impulse turbine principles," Renewable Energy, vol. 75, pp. 37-43, 2015.

[2] P. Ying, Y. K. Chen, Y. G. Xu, and Y. Tian, "Computational and experimental

investigations of an omni-flow wind turbine," Applied Energy, vol. 146, pp. 74-83, 2015.

Papers in Conference Proceedings

[3] P. Ying, Y. K. Chen, and Y. G. Xu, "Computational Modelling of Blades for a Novel

Wind Turbine, " Proc. of the International Conference on Modelling, Identification and

Control, Wuhan, China, 2012, pp. 878-882.

[4] Y. K. Chen, P. Ying, Y. G. Xu, and Y. Tian, "On the turbulent flow models in modelling

of omni-flow wind turbine, " Proc. of the International Conference on Next Generation Wind

Energy, Madrid, Spain, 2014.

IV
 Contents

Contents

Abstract………………………………………………………………….…………………….I

Acknowledgements ................................................................................................................III

Publications………………………………………………………………………………….IV

Contents………………………………………………………………………………………V

List of Figures ......................................................................................................................... IX

List of Tables ……………………………………………………………………………...XIII

Nomenclature ...................................................................................................................... XIV

Chapter 1 Introduction ......................................................................................................... 1

1.1 Necessity of a Novel Wind Turbine in Urban Areas ...................................... 1

1.2 Project Aim and Objectives ............................................................................ 3

1.3 Thesis Layout .................................................................................................. 4

Chapter 2 Overview of Existing Wind Turbines ................................................................ 6

2.1 Conventional Wind Turbines .......................................................................... 6

2.1.1 Horizontal Axis Wind Turbines ........................................................... 6

2.1.2 Vertical Axis Wind Turbines ................................................................ 8

2.1.3 Basic Aerodynamics of Aerofoils for Wind Turbines ........................ 10

2.2 Requirements and Development of Urban Wind Turbines ........................... 13

2.2.1 Wind Characteristics in Urban Areas ................................................. 14

2.2.2 Noise ................................................................................................... 16

2.2.3 Existing Urban Wind Turbines ........................................................... 17

2.2.3.1 Small-scale Conventional Wind Turbines ............................ 18

2.2.3.2 Small-scale Wind Turbines with Added Shrouds ................. 22

2.2.3.3 Re-designed Wind Turbines .................................................. 27

V
 Contents

2.2.3.4 Summary ............................................................................... 32

2.3 Omni-flow Wind Energy System .................................................................. 32

2.3.1 Geometry of Omni-flow Wind Energy System .................................. 33

2.3.2 Flow Characters inside Omni-flow Wind Energy System ................. 34

2.4 Impulse Turbines in Wave Energy Field ...................................................... 39

2.4.1 Flow Characteristics ........................................................................... 40

2.4.2 Geometrical Features of Impulse Turbines ........................................ 42

2.4.3 Performances of Impulse Turbines ..................................................... 45

2.5 Summary ....................................................................................................... 46

Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines.......................... 47

3.1 Computational Fluid Dynamics for Investigation of Wind Turbines ........... 47

3.1.1 Governing Equations .......................................................................... 47

3.1.2 Mesh Strategies for Wind Turbines .................................................... 49

3.1.3 Turbulence Models Used for Investigation of Wind Turbines ........... 51

3.1.4 Applications of CFD Method ............................................................. 54

3.2 Wind Tunnel Testing..................................................................................... 56

3.3 Summary ....................................................................................................... 59

Chapter 4 Proposed Configuration of a Novel Wind Turbine ........................................ 60

4.1 Introduction ................................................................................................... 60

4.2 Geometry of a Novel Wind Turbine ............................................................. 60

4.3 Unique Geometrical Features ........................................................................ 63

4.4 Summary ....................................................................................................... 64

Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine ........................................ 66

5.1 Test Rig in Wind Tunnel Testing .................................................................. 66

5.1.1 Wind Tunnel and Test Rig ................................................................. 66

VI
 Contents

5.1.2 Blockage Correction and Error Analysis ............................................ 70

5.2 CFD Modelling of the Proposed Wind Turbine ............................................ 73

5.2.1 Computational Domain and Boundary Conditions ............................ 74

5.2.2 Mesh Setup and Mesh Independence Test ......................................... 74

5.2.3 Physical Conditions ............................................................................ 78

5.3 Aerodynamic Analysis of the Proposed Wind Turbine ................................ 80

5.3.1 Validation of Numerical Results with Experimental Results ............. 81

5.3.2 Power Output and Power Coefficient ................................................. 87

5.3.3 Torque and Starting Wind Velocity ................................................... 91

5.4 Effects of the Stator on Passing Flow ........................................................... 92

5.4.1 Effects of the Stator on Flow Velocity ............................................... 93

5.4.2 Effects of the Stator on Flow Direction .............................................. 95

5.5 Summary ....................................................................................................... 99

Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow .... 100

6.1 Wind Turbine Model under Non-uniform Flow ......................................... 100

6.2 Power Coefficient and Torque Coefficient ................................................. 103

6.3 Thrust .......................................................................................................... 107

6.4 Summary ..................................................................................................... 110

Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind

Turbine…………………………………………...…………………………..112

7.1 Introduction ................................................................................................. 112

7.2 Effects of Hub-to-tip Ratios ........................................................................ 112

7.3 Influences of Blade Numbers ...................................................................... 117

7.4 Effects of Aerofoil Profiles ......................................................................... 122

7.5 Influences of Guide Vane Numbers ............................................................ 128

VII
 Contents

7.6 Summary ..................................................................................................... 132

Chapter 8 Discussion ......................................................................................................... 134

8.1 Comparison about Aerodynamics between the Proposed Wind Turbine

and Conventional Wind Turbines ................................................................ 134

8.1.1 Power Coefficient ............................................................................. 134

8.1.2 Torque Coefficient ............................................................................ 135

8.1.3 Thrust Coefficient ............................................................................. 135

8.2 Comparison about Aerodynamics of the Proposed Wind Turbine under

Uniform Flow and Non-uniform Flow ........................................................ 135

8.3 Effects of Geometrical Parameters on Aerodynamics ................................ 136

8.4 Approaches Adopted in this Project............................................................ 138

Chapter 9 Conclusions and Recommendations for Future Work ................................. 140

9.1 Conclusions ................................................................................................. 140

9.2 Recommendations for Future Work ............................................................ 142

9.2.1 Study of the Whole Energy System .................................................. 142

9.2.2 Optimisation of the Geometry .......................................................... 142

9.2.3 Study of the Wind Turbine with a Diffuser ...................................... 142

References …...……………………………………………………………………………..144

Appendix A Drawing of Guide Vane ................................................................................ 156

Appendix B Drawing of Stator Hub .................................................................................. 157

Appendix C Drawing of Rotor Hub .................................................................................. 158

Appendix D Coordinates of Three Aerofoils .................................................................... 159

Appendix E Flow Velocity and Pressure Change When Passing Guide Vanes ............ 162

Appendix F Flow Directions inside the Omni-flow System with and without a Stator 163

VIII
 List of Figures

List of Figures

Figure 1.1: Global wind power capacity in 2005 - 2014 [12].................................................... 2

Figure 1.2: Omni-flow wind energy system [15]....................................................................... 3

Figure 2.1: HAWTs: (a) schematic view; (b) offshore wind farm [20]. .................................... 7
Figure 2.2: Development of the rotor diameter and capability [21]. ......................................... 8
Figure 2.3: VAWTs: (a) Savonius turbine; (b) classical Darrieus turbine; (c) Darrieus turbine
with straight blades [24]. ......................................................................................... 9
Figure 2.4: An aerofoil: (a) profile [19]; (b) aerodynamic forces [31]. ................................... 11
Figure 2.5: Pressure coefficient distribution on the lower and upper surfaces of an aerofoil
[31]. ....................................................................................................................... 13
Figure 2.6: Wind velocity fields on different roof shapes [37]................................................ 15
Figure 2.7: Flow velocity field around a block building [40]. ................................................. 16
Figure 2.8: HAWTs placed multiply on a roof [58]. ............................................................... 21
Figure 2.9: Schematic views of the diffuser for HAWTs [47]. ............................................... 23
Figure 2.10: HAWT with the flanged diffuser: (a) schematic view; (b) prototype [66]. ........ 23
Figure 2.11: Ducted wind turbines: (a) original type [69]; (b) a bucket-shape duct with a
nozzle [68]. ......................................................................................................... 24
Figure 2.12: A shroud present by Ahmed [70]. ....................................................................... 25
Figure 2.13: An urban VAWT with guide vanes [9]. .............................................................. 26
Figure 2.14: Three types of shrouds for a Savonius wind turbine: (a) obstacle [71];
(b) curtain arrangement [72]; (c) guide-box tunnel [73]. ................................... 27
Figure 2.15: A turbine with a crossflow runner [74]. .............................................................. 28
Figure 2.16: Sistan wind turbine: (a) with guide vanes [76] ; (b) with a shroud [75]. ............ 29
Figure 2.17: Zephyr wind turbine: (a) 3D view; (b) schematic view [78]. .............................. 30
Figure 2.18: Crossflow wind turbine: (a) concept; (b) architectural integration with ridge
[79]...................................................................................................................... 30
Figure 2.19: VAWT with segments: (a) wireframe view; (b) 3D model [81]. ........................ 31
Figure 2.20: Illustrative view of the omni-flow wind energy system on the top of a building.
The background is a modified view of London by Google Earth. ..................... 33
Figure 2.21: The omni-flow wind energy system: (a) left view; (b) top view......................... 34
Figure 2.22: Four wind directions for the omni-flow wind energy system: (a) 0o; (b) 12o;
(c) 24o; (d) 36o. ................................................................................................... 35
Figure 2.23: Velocity distributions at the outlet of the omni-flow system under four wind
directions: (a) 0o; (b) 12o; (c) 24o; (d) 36o [15]. .................................................. 36
Figure 2.24: The maximum and average velocities at the outlet of the omni-flow system
under four wind directions [15]. ......................................................................... 37
Figure 2.25: Flow velocity vector field through the omni-flow system [15]........................... 38
Figure 2.26: Velocity vector field at the outlet under four wind directions: (a) 0o; (b) 12o;
(c) 24o; (d) 36o [15]. ............................................................................................ 38
Figure 2.27: Two types of OWCs: (a) vertical type; (b) horizontal type [84, 86]. .................. 40
Figure 2.28: Height change of the irregular wave in a short time inside an OWC [84]. ......... 41
Figure 2.29: 3D view of an impulse turbine [90]..................................................................... 42
Figure 2.30: Schematics of two types of impulse turbines: (a) IFGV; (b) ISGV [84]. ........... 43
Figure 2.31: Two profiles of guide vanes: (a) plate; (b) aerofoil [84]. .................................... 43
Figure 2.32: Two blade profiles: (a) simple; (b) elliptic. ......................................................... 44

IX
 List of Figures

Figure 2.33: Unidirectional impulse turbine: (a) asymmetric blade aerofoil; (b) upstream
plate guide vanes and blades [93]. ...................................................................... 45
Figure 2.34: Efficiency comparison of three impulse turbines [94]. ....................................... 46

Figure 3.1: Polyhedral cells on the proposed wind turbine model in this project. .................. 49
Figure 3.2: O-type cells for a blade aerofoil [101]. ................................................................. 50
Figure 3.3: The hybrid mesh strategy for an impulse turbine [103]. ....................................... 51
Figure 3.4: Measure system in wind tunnel testing [143]........................................................ 57
Figure 3.5: Wind turbines in wind tunnel tests [47, 143, 147]. ............................................... 58

Figure 4.1: The proposed wind turbine: (a) 3D view; (b) schematic view of the wind turbine
inside a chamber.................................................................................................... 61
Figure 4.2: Guide vane: (a) section view; (b) 3D view............................................................ 62
Figure 4.3: Schematic views of the blade: (a) aerofoil; (b) 3D view. ...................................... 63

Figure 5.1: The closed return wind tunnel used in this project. ............................................... 66
Figure 5.2: Wind turbine prototype: (a) stator inside a cylindrical chamber; (b) rotor with
blades. ................................................................................................................... 67
Figure 5.3: Schematic for the measurement system of the wind tunnel testing. ..................... 68
Figure 5.4: Test rig inside the wind tunnel test section. .......................................................... 68
Figure 5.5: Flow measurement: (a) tiny Pitot tube; (b) pressure meter ................................... 70
Figure 5.6: Schematic view of the computational domain setup in simulations. .................... 74
Figure 5.7: Torque values under different cell numbers. ......................................................... 75
Figure 5.8: Meshed model: (a) a section view in the flow direction; (b) an axial section view
of the wind turbine model. .................................................................................... 76
Figure 5.9: Thin layers near surfaces: (a) cell layers near the blade leading edge; (b) cell
layers near the guide vane edge. ........................................................................... 78
Figure 5.10: Positions of upstream interface and downstream interface in simulations. ........ 79
Figure 5.11: Comparison of power coefficient between numerical results and experimental
data at U = 8.2m/s. .............................................................................................. 82
Figure 5.12: Pressure distributions on the aerofoil predicted by four turbulence models at
r = 140 mm and λ = 0.76. ................................................................................... 83
Figure 5.13: Pressure coefficients on the blade aerofoil at r = 140 mm and λ = 0.76: (a) by
the standard k-𝜺 model; (b) by four turbulence models. ..................................... 85
Figure 5.14: Contributions of blade sections to the normal force at 𝜆 = 0.66 by the realisable
k-𝜀 model. ........................................................................................................... 86
Figure 5.15: Contributions of blade sections to torque at λ = 0.66 by the realisable k-𝜀
model. ................................................................................................................. 87
Figure 5.16: Distributions of experimental power coefficient of the proposed wind turbine
under different wind velocities. .......................................................................... 88
Figure 5.17: Distributions of experimental power output of the proposed wind turbine under
various wind velocities. ...................................................................................... 89
Figure 5.18: Pressure coefficient distributions on the blade section at r = 140 mm under the
flow speed of 8.2 m/s. ......................................................................................... 90
Figure 5.19: Distributions of experimental torque coefficient of the proposed wind turbine
under various wind velocities. ............................................................................ 91
Figure 5.20: Starting wind velocities of the proposed wind turbine in wind tunnel tests........ 92

X
 List of Figures

Figure 5.21: Stator model in simulations: (a) stator inside the cylindrical chamber; (b) testing
positions both in experiments and simulations. .................................................. 93
Figure 5.22: Comparison of wind velocity ratios between the experimental and numerical
results at the selected positions. .......................................................................... 94
Figure 5.23: Properties of the flow passing guide vanes: (a) flow passage between two guide
vanes; (b) velocity and pressure along the centre line at r = 100 mm. ............... 95
Figure 5.24: Study of flow direction: (a) computational domain; (b) the direction of free
stream in simulations [15]. ................................................................................. 96
Figure 5.25: Positions for recording flow directions: (a) the system model without a stator;
(b) the system model with a stator. ..................................................................... 96
Figure 5.26: Angle of the flow direction, φ. ............................................................................ 97
Figure 5.27: Angles of flow directions inside the omni-flow system without the stator. ........ 98
Figure 5.28: Angles of flow directions inside the omni-flow system with the stator. ............. 98

Figure 6.1: 20% non-uniform flow condition: (a) wind direction; (b) velocity distribution
at the outlet of the system. .................................................................................. 101
Figure 6.2: Wind turbine models: (a) for 20% non-uniform flow condition; (b) for 40%
non-uniform flow condition. ............................................................................... 101
Figure 6.3: 40% non-uniform flow condition: (a) wind direction; (b) velocity distribution
at the outlet of the system. .................................................................................. 101
Figure 6.4: Two flow regions caused by the cover ahead the wind turbine. ......................... 102
Figure 6.5: Distributions of power coefficient under two non-uniform flow conditions. ..... 104
Figure 6.6: Distributions of torque coefficient under two non-uniform flow conditions. ..... 104
Figure 6.7: Torque generated on every blade under the 20% non-uniform flow condition. . 106
Figure 6.8: Torque generated on every blade under the 40% non-uniform flow condition. . 106
Figure 6.9: Pressure coefficient distributions at r = 140 mm and λ = 0.3 under the 20%
non-uniform flow condition: (a) No. 6 blade; (b) No. 16 blade.......................... 107
Figure 6.10: Thrust on every blade under the 20% non-uniform flow condition. ................. 108
Figure 6.11: Thrust on every blade under the 40% non-uniform flow condition. ................. 109

Figure 7.1: Employed three hub-to-tip ratios and hub diameters in simulations: (a) 0.35;
(b) 0.45; (c) 0.6. .................................................................................................. 113
Figure 7.2: Distributions of torque per unit length on the blade with three hub-to-tip ratios
at λ = 0.66............................................................................................................ 114
Figure 7.3: Distributions of power coefficient of the proposed wind turbine with three
hub-to-tip ratios. .................................................................................................. 115
Figure 7.4: Distributions of torque coefficient of the proposed wind turbine with three
hub-to-tip ratios. .................................................................................................. 116
Figure 7.5: Distributions of thrust coefficient of the proposed wind turbine with three
hub-to-tip ratios. .................................................................................................. 117
Figure 7.6: Distributions of torque coefficient of the proposed wind turbine with different
blade numbers. .................................................................................................... 118
Figure 7.7: Distributions of pressure coefficient on the blade aerofoil at r = 140 mm: (a) the
wind turbine with 15 blades; (b) the wind turbine with 30 blades. ..................... 120
Figure 7.8: Distributions of torque coefficient of the proposed wind turbine with five blade
numbers. .............................................................................................................. 121
Figure 7.9: Distributions of thrust coefficient of the proposed wind turbine with five blade
numbers. .............................................................................................................. 122
XI
 List of Figures

Figure 7.10: Schematic view of three blade aerofoils employed. .......................................... 123
Figure 7.11: Distributions of power coefficient of the proposed wind turbine with three
aerofoils. ........................................................................................................... 124
Figure 7.12: Distributions of pressure coefficient on the blade aerofoil at r = 140 mm:
(a) Type 1; (b) Type 2; (c) Type 3. ................................................................... 126
Figure 7.13: Distributions of torque coefficient of the proposed wind turbine with three
aerofoils. ........................................................................................................... 127
Figure 7.14: Distributions of thrust coefficient of the proposed wind turbine with three
aerofoils. ........................................................................................................... 127
Figure 7.15: The stator model with four guide vane numbers: (a) 5; (b) 10; (c) 15; (d) 20. . 128
Figure 7.16: Distributions of power coefficient of the proposed wind turbine with four
guide vane numbers. ......................................................................................... 130
Figure 7.17: Distributions of torque coefficient of the proposed wind turbine with four
guide vane numbers. ......................................................................................... 131
Figure 7.18: Distributions of thrust coefficient of the proposed wind turbine with four
guide vane numbers. ......................................................................................... 132

XII
 List of Tables

List of Tables

Table 2.1: Efficiency of several current small-to-middle scale HAWTs [52]. ........................ 19

Table 5.1: Errors caused by the house bearing on power coefficient. ..................................... 71
Table 5.2: Errors caused by the house bearing on power coefficient under different flow
velocities. ................................................................................................................. 72
Table 5.3: Errors caused by gears on power coefficient. ......................................................... 72
Table 5.4: Cell size control on different surfaces. ................................................................... 76

Table 6.1: Values of Ucovered and Uuncovered under different rotational speeds. ....................... 103
Table 6.2: Predicted maximum thrust and frequency with different radii. ............................ 110

Table 7.1: Maximum speeds of the approaching flow with different guide vane numbers. .. 129
Table 7.2: Mass flow rates for the wind turbine with different guide vane numbers. ........... 129

Table 8.1: Key aerodynamic features of the proposed wind turbine with three hub-to-tip
ratios. ..................................................................................................................... 136
Table 8.2: Key aerodynamic features of the proposed wind turbine with five blades
numbers. ................................................................................................................ 137
Table 8.3: Key aerodynamic features of the proposed wind turbine with three aerofoils. .... 137
Table 8.4: Key aerodynamic features of the proposed wind turbine with four numbers of
guide vanes. ........................................................................................................... 138

XIII
 Nomenclature

Nomenclature

Latin Symbols

A projected area of aerofoil (m2) T torque (Nm)

A1, A2 entrance area of flow passage (m2) 𝑇𝑟 resistive torque (Nm)
B blade number u component of 𝑈 ⃗ in x direction
c chord of a blade aerofoil (m) (m/s)
′
𝐶𝐵 correction for blockage effect 𝑢 fluctuating components of u
𝐶𝐷 drag coefficient (m/s)
̅̅̅
𝐶𝑓 friction coefficient 𝑢̅ mean component of u (m/s)
𝐶𝐿 lift coefficient ut friction velocity (m/s)
𝐶𝑁 thrust coefficient 𝑈⃗ velocity vector (m/s)
𝐶𝑃 power coefficient Ucorrected corrected flow velocity in the
𝐶𝑃,𝑚𝑎𝑥 maximum power coefficient wind tunnel (m/s)
𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒 pressure coefficient Ucovered flow velocity at the entrance
𝐶𝑇 torque coefficient for a covered model (m/s)
D diameter of the chamber (m) Umeasured measured flow velocity in the
e error in experiments wind tunnel (m/s)
f frequency (Hz) Uuncovered flow velocity at the entrance
F aerodynamic force (N) (m/s)
F’ normal force (N) v component of 𝑈 ⃗ in y direction
𝐹𝐷 drag (N) (m/s)
′
𝐹𝐿 lift (N) 𝑣 fluctuating components of v
𝐹𝑁 thrust (N) (m/s)
𝐹𝑇 tangential force (N) 𝑣̅ mean component of v (m/s)
L length of the chamber (m) 𝑦 distance from the wall (m)
𝑚̇ mass flow rate (kg/s) 𝑦+ dimensionless distance from
N rotational speed (rpm) the wall
p pressure (N/m2) z position on z-axis (m)
po reference pressure (N/m2) w component of 𝑈 ⃗ in z direction
P power output (W) (m/s)
′
r local blade radius (m) 𝑤 fluctuating components of w
R blade radius (m) (m/s)
Re Reynolds number 𝑤
̅ mean component of w (m/s)
t time (s)

Greeks symbols

𝛼 angle of attack (o) δ the max height between the

𝛾 correction factor for adding a cover camber line and chord line (m)
XIV
 Nomenclature

η efficiency 𝜏𝑤 wall shear stress (N/ m2)

λ tip speed ratio φ angle between flow direction
υ kinematic viscosity (m2/s) and a vertical axis (°)
ρ fluid density (kg/m3) ω angular speed of the rotor
𝜌̅ mean fluid density (kg/ m3) (rad/s)
σ solidity

Acronyms

ABS acrylonitrile butadiene styrene LES large eddy simulation

BEM blade element momentum MRF moving reference frame
CFD computational fluid dynamics NREL National Renewable Energy
DNS direct numerical simulation Laboratory
HAWT horizontal axis wind turbine OWC oscillating water column
IFGV impulse turbine with fixed guide RANS Reynolds-averaged Navier
vanes Stokes equations
ISGV impulse turbine with self-pith- SST shear stress transport
controlled guide vanes VAWT vertical axis wind turbine

XV
 Chapter 1 Introduction

Chapter 1 Introduction

1.1 Necessity of a Novel Wind Turbine in Urban Areas

Due to the concerns about global warming, wind energy technology has been developed and

has become one of the major technologies in the renewable energy field. As shown in Figure

1.1, the capacity of global cumulative wind energy has increased continuously in the last

decade. In 2012, an employment of wind energy has avoided € 9.6 billion of fossil fuels costs,

which means less CO2 emissions [1]. In 2020, an estimated € 22 billion in fuel costs can be

avoided [1]. In Denmark, wind turbines have provided around 39% of total electricity

consumption in 2014 [2]. The European Wind Energy Association predicts that the wind

energy will account for 24.4% of the EU’s electricity demand in 2030 [3].

In recent years, technologies in wind turbines have matured. Wind turbines generally work

well in land based wind farms or in large offshore wind farms, where there is good quality

wind [4, 5]. However, wind farms are in remote areas far from residential areas. The

electricity produced from wind farms suffers from long-distance transmission through electric

grids with high losses and high costs. Due to the size, the modern large wind turbines need a

high cost in transportation and installation of components.

In the UK, approximately 50% of energy consumption is in buildings [6]. The UK

government has set a target of 80% carbon reduction by 2050 [7]. The wind turbine integrated

into building is one of emerging technologies that can reduce CO2 emission [8, 9]. Compared

with large wind turbines in wind farms, small wind turbines that function as distributed

electricity generators in urban areas can reduce the power losses and high costs in electricity

transmission over long distances. Because the wind speed on tops of buildings can be

1
 Chapter 1 Introduction

significantly higher than that near ground in the urban environment, several types of small

wind turbines have been presented for working on the roofs of tall buildings [10, 11].

MW
400,000
369,597

350,000
318,644

300,000 283,132

250,000 238,435

197,943
200,000
159,079
150,000
120,715
93,901
100,000
73,949
59,091
50,000

0
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Figure 1.1: Global wind power capacity in 2005 - 2014 [12].

An omni-flow wind energy system has been introduced for urban areas [13]. The term of

omni-flow can be defined as the flow of air in omni direction. The omni-flow energy system

can accept the omni-directional wind for a downstream wind turbine. Figure 1.2 shows a 3D

view of this wind energy system. The wind energy system can receive and effectively

transmit wind to a downstream wind turbine. The omni-flow wind energy system can cater

for a wide variety of wind conditions without a yaw mechanism. The aerodynamics of such a

wind energy system has been studied by Zhang et al. [14, 15]. Zhang [15] suggested that this

system could be used in the urban areas where wind speed was always low. However, it has

been found that the flow velocity distribution is not uniform inside the omni-flow wind

2
 Chapter 1 Introduction

energy system [14, 15]. Blades of a wind turbine have to operate under a non-uniform wind

condition, which can result in different aerodynamic loads during one operation cycle.

Figure 1.2: Omni-flow wind energy system [15].

Thin blades of current wind turbines are difficult to accommodate the non-uniform flow

condition. In this PhD project a new wind turbine is proposed for the omni-flow wind energy

system. The proposed wind turbine has a new geometry and no research has been carried out

on this new geometry [16]. Much of research about wind turbines is performed under the

uniform flow condition [17]. This PhD project studies the aerodynamics of the proposed wind

turbine under the non-uniform condition caused by the omni-flow system. Recently, the

employment of wind turbines in an urban environment has been receiving increased attention

in both academia and industry. This work will contribute to the use of wind turbines in the

urban areas, such as the omni-flow wind energy system.

1.2 Project Aim and Objectives

It has been reported that the omni-flow wind energy system can guide the omni wind to a

downstream wind turbine and has the potential to be employed in an urban environment [14,

15].

3
 Chapter 1 Introduction

This PhD project proposes a new wind turbine for the omni-flow wind energy system. The

project aims to investigate aerodynamics of the proposed wind turbine.

The project has the following objectives:

 A literature review about current small urban wind turbines, impulse turbines and the

omni-flow wind energy system.

 Study two common research methods: computational fluid dynamics (CFD)

simulation and wind tunnel testing.

 Presentation of the geometry for the proposed wind turbine.

 Test the power output performance and starting capability of the proposed wind

turbine in a wind tunnel and validate the numerical approach by a comparison

between numerical and experimental results.

 Investigate aerodynamics of the proposed wind turbine under the non-uniform flow

condition by the validated numerical approach.

 Study influences of geometrical parameters on the aerodynamic performances of the

proposed wind turbine.

1.3 Thesis Layout

There are nine chapters in this thesis. Brief summaries for Chapters 2-9 are shown as follows:

Chapter 2 contains the literature review about current wind turbines, small urban wind

turbines, wind characteristics in urban areas, aerodynamic features of the omni-flow wind

energy system and the impulse turbine technology.

Chapter 3 reviews two methods, CFD simulation and wind tunnel testing, for aerodynamic

investigation of the wind turbine.

4
 Chapter 1 Introduction

Chapter 4 presents the geometry of a novel wind turbine based upon the impulse turbine

technology.

Chapter 5 provides aerodynamic analyses of the proposed wind turbine by numerical and

experimental methods. The numerical approach is validated by the comparison between

numerical results and experiment data.

Chapter 6 addresses aerodynamic performance of the proposed wind turbine under the non-

uniform flow condition.

Chapter 7 presents influences of geometrical parameters, such as hub-to-tip ratios, numbers of

blades, aerofoils and numbers of guide vanes, on aerodynamics of the proposed wind turbine.

Chapter 8 discusses aerodynamics of the proposed wind turbine by comparing with

conventional wind turbines. Performances of the wind turbine under uniform and non-

uniform flow conditions and influences of geometrical parameteries are also discussed.

Chapter 9 lists the highlighted foundings of this project. Future work is also suggested.

5
 Chapter 2 Overview of Existing Wind Turbines

Chapter 2 Overview of Existing Wind Turbines

This chapter introduces features of conventional wind turbines and wind characters founded

in an urban environment. Aerodynamics of the omni-flow wind energy system is addressed.

Technologies about impulse turbines are also reviewed.

2.1 Conventional Wind Turbines

Conventional wind turbines can be categorised into two types by the axial direction of a wind

turbine: horizontal axis wind turbines (HAWTs) and vertical axis wind turbines (VAWTs).

Characteristics, advantages and disadvantages of HAWTs and VAWTs are reviewed in this

section.

2.1.1 Horizontal Axis Wind Turbines

A HAWT is a wind turbine whose rotor axis is horizontal and parallel to the wind direction.

Employing HAWTs for electricity can be traced back to the late nineteenth century in the

USA [18]. The HAWT has been the most common type of wind turbines [18, 19]. A classic

structure of a general HAWT is shown in Figure 2.1(a). A high tower is used to place a

nacelle and a rotor with long blades. The generator, gear box, brake system and yaw

mechanism are placed inside the nacelle. The rotor always consists of two or three blades and

a supporting hub. The rotor of a HAWT can be oriented upwind and downwind of the tower.

Aerodynamic forces generated on blades rotate the rotor and the generator to produce

electricity. In order to obtain a good efficiency, a yaw mechanism is needed to adjust the rotor

axis against the wind direction instantly [18]. Recently, HAWTs are always gathered in wild

and offshore wind farms, where there is good quality wind, as shown in Figure 2.1(b), and

produce electricity for grids.

6
 Chapter 2 Overview of Existing Wind Turbines

The power coefficient is an important dimensionless term for all wind turbines to estimate the

capability of power generation. The power coefficient is defined as a ratio between the

generated power from the wind turbine and the available power from wind [19]. For a modern

large HAWT, the peak power coefficient is in a range of 0.4-0.5 [18].

(a) (b)

Figure 2.1: HAWTs: (a) schematic view; (b) offshore wind farm [20].

To achieve high power output, long blade radii and high towers are required [19]. Over the

last 25 years, the largest rotor diameter of commercial wind turbines has increased from 20 m

to 140 m with improving the capability from 50 kW to 7 MW, as shown in Figure 2.2 [21].

But the large rotor diameter causes difficulty and high cost in manufacturing. The installation

and maintenance of key components, such as blades, the generator and the control system, on

the top of a high tower are also difficult and expensive, especially for offshore applications.

7
 Chapter 2 Overview of Existing Wind Turbines

Figure 2.2: Development of the rotor diameter and capability [21].

2.1.2 Vertical Axis Wind Turbines

A VAWT is a wind turbine whose rotor axis is perpendicular to the wind direction and the

ground. VAWTs can be divided into two types, the lift driven device and the drag driven

device. Savonius wind turbine is a classical drag driven VAWT that was presented by S.J.

Savonius in 1920s [22]. A classical structure of a Savonius turbine is shown in Figure 2.3(a).

It consists of “S-shape” blades fixed on a vertical axis shaft. The maximum power coefficient

of a Savonius wind turbine is approximately 0.2 [23]. The Savonius wind turbine has good

starting capability, which is an advantage for working under low velocity wind [24].

The Darrieus wind turbine is a lift driven VAWT that was originally patented by G.J.

Darrieus in 1931 [25]. There are two major types of Darrieus wind turbines: the wind turbine

with curved “edge beater” blades and the wind turbine with straight blades, as shown in

Figures 2.3(b, c). The Darrieus wind turbine can achieve a peak power coefficient of

8
 Chapter 2 Overview of Existing Wind Turbines

approximately 0.3 [26]. However, the Darrieus wind turbine has no ability to start up by itself

[24].

(a) (b) (c)

Figure 2.3: VAWTs: (a) Savonius turbine; (b) classical Darrieus turbine; (c) Darrieus
turbine with straight blades [24].

Although VAWTs were firstly used in wind energy utilisation, the interest in VAWTs was

lost due to the low power coefficient in the mid-twentieth century [19]. Recently, the research

about VAWTs has become popular since it is found that VAWTs are more suitable for low

velocity wind and turbulent wind when compared with HAWTs [23]. VAWTs have a number

of substantial advantages over HAWTs [24, 27]:

 There is no need of a yaw mechanism because the rotor axis of a VAWT is

perpendicular to wind directions at all times.

 The noise level of a VAWT is typically lower than that of a HAWT since the tangential

speed at the blade tip of a VAWT is always smaller than that of a HAWT.

 Due to simpler profiles of blades, the manufacturing cost of blades of a VAWT can be

lower than that of a HAWT.

9
 Chapter 2 Overview of Existing Wind Turbines

 VAWTs have an advantage on safe operation during a gust due to the stalling.

Additionally, key components, such as the gear box, generator system and brake system, can

be installed near to the ground. Thus VAWTs are easier to be accessed for installation and

maintenance than HAWTs.

However, VAWTs also have disadvantages compared with HAWTs [24]:

 The rotor of a VAWT is not always symmetric to wind direction and not all blades can

face to wind. The blade generates negative or zero torque so that the overall torque on

the rotor cannot be large enough to rotate the rotor at starting.

 The weight and stress are concentrated on the bottom bearing, which has a risk in

structure.

 There is a periodic change of aerodynamic forces on a blade in one revolution, which

may affect the structure service life.

2.1.3 Basic Aerodynamics of Aerofoils for Wind Turbines

The blade aerofoil affects performances of a wind turbine, such as power output and starting

capability. Many wind turbine blades use existing aerofoils from aircrafts, such as the

National Advisory Committee for Aeronautics (NACA) aerofoil family. Several special

aerofoil families for wind turbines have been presented for more requirements in the wind

turbine industry, such as the National Renewable Energy Laboratory (NREL) [28], Delft

University [29], FFA [18] and Risø [30] aerofoil families.

A schematic view of an aerofoil is shown in Figure 2.4(a). A straight line connecting the

leading edge and the trailing edge is the chord with a length of c. The angle between the wind

direction and the chord line is called angle of attack, α. A camber line is the mid-line between

10
 Chapter 2 Overview of Existing Wind Turbines

the upper and lower surfaces. The maximum height between the camber line and the chord

δ
line is presented as δ and the quantity 𝑐 ×100% is called the percentage camber of an aerofoil.

The percentage camber of an aerofoil is always in the range of 0% - 5% [18]. The percentage

camber of a symmetrical aerofoil is 0%.

(a)

(b)

Figure 2.4: An aerofoil: (a) profile [19]; (b) aerodynamic forces [31].

An aerofoil can provide an aerodynamic force to rotate the blade. The aerodynamic force on

an aerofoil is defined as follows:

𝐹
𝐶𝐹 = 1 (2.1)
𝜌𝐴𝑈 2
2

where F is the aerodynamic force, 𝜌 is the air density, A is the projected area of the aerofoil

and U is the velocity of free stream wind. There are two types of force coefficients of an

11
 Chapter 2 Overview of Existing Wind Turbines

aerofoil: lift coefficient and drag coefficient. Lift, FL, is the force component on the aerofoil

perpendicular to the relative approaching wind. Drag, FD, is the force component parallel to

the direction of relative approaching wind. The aerodynamic force diagram on an aerofoil is

shown in Figure 2.4(b). Lift coefficient and drag coefficient are defined by [18]:

𝐹𝐿
Lift coefficient: 𝐶𝐿 = 1 (2.2)
𝜌𝐴𝑈 2
2

𝐹𝐷
Drag coefficient: 𝐶𝐷 = 1 (2.3)
𝜌𝐴𝑈 2
2

Values of lift and drag depend on the pressure distribution on the aerofoil. Level of the

pressure can be estimated by a dimensionless term, pressure coefficient, which is defined as

[31],

𝑝−𝑝𝑜
𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒 = 1 (2.4)
𝜌𝑈 2
2

where 𝑝 is the pressure on an element of the aerofoil surface and 𝑝𝑜 is the reference pressure.

Lift coefficient and drag coefficient can be expressed as [31]

1 𝑥 𝑍 ⁄𝐶 𝑥
𝐶𝐿 = 𝑐𝑜𝑠𝛼 ∫0 (𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑙𝑜𝑤𝑒𝑟 − 𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑢𝑝𝑝𝑒𝑟 )𝑑 ( 𝑐 ) − 𝑠𝑖𝑛𝛼 ∫𝑍 2⁄𝐶 ∆𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑥 𝑑( 𝑐 ) (2.5)
1

1 𝑥 𝑍 ⁄𝐶 𝑥
𝐶𝐷 = 𝑠𝑖𝑛𝛼 ∫0 (𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑙𝑜𝑤𝑒𝑟 − 𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑢𝑝𝑝𝑒𝑟 )𝑑 (𝑐 ) + 𝑐𝑜𝑠𝛼 ∫𝑍 2⁄𝐶 ∆𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑥 𝑑( 𝑐 ) (2.6)
1

where 𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑙𝑜𝑤𝑒𝑟 and 𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑢𝑝𝑝𝑒𝑟 represent the pressure coefficient on the lower

surface and upper surface respectively, (𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑙𝑜𝑤𝑒𝑟 − 𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑢𝑝𝑝𝑒𝑟 ) is the pressure

coefficient difference between the lower surface and upper surface, ∆𝐶𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒,𝑥 is pressure

coefficient difference in the x-axis, c is the blade chord and α is the angle of attack, z1 and z2

denote the lowest and highest positions on the aerofoil in the z-axis as shown in Figure 2.4(b).
12
 Chapter 2 Overview of Existing Wind Turbines

Figure 2.5 shows the pressure coefficient distribution on an aerofoil. The top curve represents

pressure coefficient on the upper surface of an aerofoil and the bottom curve means pressure

coefficient on the lower surface. The area enclosed by two pressure coefficient curves

represents the pressure coefficient difference. Equation 2.5 shows that a high lift coefficient

requires a large pressure difference that means a large area enclosed by the two pressure

coefficient curves.

Upper surface

Lower surface

Figure 2.5: Pressure coefficient distribution on the lower and upper surfaces of an
aerofoil [31].

2.2 Requirements and Development of Urban Wind Turbines

Wind turbines can be placed in urban areas to provide renewable energy directly for buildings.

Employing wind turbines in urban areas is limited by complex wind conditions and wind

turbine technologies. Wind characteristics, noise level requirement and development of

current urban wind turbines are reviewed in this section.

13
 Chapter 2 Overview of Existing Wind Turbines

2.2.1 Wind Characteristics in Urban Areas

Wind is air in motion and generated by atmospheric pressure difference. In urban areas, wind

is influenced by various buildings, which results in the low velocity, frequently changed

direction and high turbulence intensity [32]. These characteristics cause complexities in

efficient utilisation of wind energy.

Wind in urban areas always has low velocities due to influences of various buildings. Drew et

al. [33] estimated the variability of annual mean wind speeds for typically installing current

wind turbines in the Greater London, UK. They found that the wind speed across the Greater

London was very low and the wind speed exceeded 4 m/s only at 27% of the neighbourhoods

in the city [33]. Millward-Hopkins et al. [34] predicted that the long-term mean wind speed in

the urban area of Leeds, UK, and found that the wind speed was mostly in a range of 1 m/s to

6 m/s at a height of 3m above a mean building height. It was also reported that there were

2000 – 9500 locations to building-mount wind turbines when the minimum assuming wind

speed was 4 m/s and only 10 – 80 locations when the assuming wind speed was larger than 6

m/s in Leeds [34]. In order to produce an efficient power output in urban areas, a wind turbine

should start up and capture energy under low velocity wind. Therefore, good starting

capability is important for an urban wind turbine.

Balduzzi et al. [35] pointed out that the potential of wind energy was dependent on building

heights, widths and distances between buildings. If a wind turbine is located in a wrong place

in urban areas, it is possible that the power output decreases greatly even wind has high

velocity [35, 36]. Investigations about influences of building heights have shown that a

chosen building for installing a wind turbine should be higher than surround buildings [37]. In

addition to the heights of buildings, the potential of wind energy is also strongly dependant on

14
 Chapter 2 Overview of Existing Wind Turbines

the shape profiles of building roofs. Flow characteristics around three roof profiles, namely:

pitched roof, pyramidal roof and flat roof, were compared by Led et al. [38]. It was reported

that a wind turbine mounted on a flat roof had more potential on power generation. A suitable

roof has an advantage of accelerating wind speeds so more wind energy can be captured.

Abohela et al. [37] studied wind features around six types of roofs as shown in Figure 2.6.

They found that wind on the vaulted roof had the highest velocity and 56.1% more electricity

could be achieved.

Figure 2.6: Wind velocity fields on different roof shapes [37].

The frequently changed wind direction and high flow turbulence intensity are unavoidable in

urban areas. The reason for such complex wind is vortexes around buildings. Figure 2.7

shows the generated vortexes around a block building. It can be seen that there are vortexes

on the roof and behind the building. Other presented results show that there can be five

vortexes around a building when wind passing [37]. Kotb et al. [39] theoretically investigated

power output performances of a HAWT under the flow with a swirl. It was reported that the

power coefficient under the flow with a swirl was lower by 40% than that under a steady flow

[39]. Therefore, an urban wind turbine has to suit the turbulent wind. As mentioned in Section

15
 Chapter 2 Overview of Existing Wind Turbines

2.1.2, VAWTs have advantages on operation with turbulent flows in urban areas when

compared with HAWTs.

Figure 2.7: Flow velocity field around a block building [40].

2.2.2 Noise

In addition to operation under complex flow conditions, urban wind turbines have to meet

requirements about noise. Many countries have noise regulations for public areas and

residential areas. However, noise from wind turbines is unavoidable due to aerodynamic

interactions of blades, mechanical components and flow passing holes or slits [41]. It was

reported that noise from wind turbines could be annoying and associated with some health

effects [42]. Compared with other community noise sources, perception and annoyance

caused by wind turbine noise are higher at the same sound level due to visual interference and

noise characteristics [43]. Noise from small wind turbines in urban areas has been realised as

one of the most important sitting constraints.

Sources of noise from wind turbines can be divided into two types: (a) mechanical noise

generated from interaction of components, such as gearboxes and generators; (b)

16
 Chapter 2 Overview of Existing Wind Turbines

aerodynamic noise produced by wind passing blades [44]. Mechanical noise has been almost

solved, thus reducing the aerodynamic noise level is a topic of considerable research in the

wind energy field [44].

Since aerodynamic noise generation is very sensitive to the rotational speed, an efficient way

to decrease aerodynamic noise is to reduce the rotational speed [41, 44]. Vick et al. [45]

tested the noise level of a 3m-diameter three-blade HAWT under different rotational speeds.

They found that the wind turbine achieved the maximum noise level of approximately 90 dB

at the maximum rotational speed of 900 rpm. By reducing the rotational speed to 600 rpm, the

measured noise level was less than 60 dB.

In addition to the good performance under complex wind conditions and the low working

noise level, some other features needed for an urban wind turbine can be considered as

follows [46]:

 Safe operation

 Minimised cost in maintenance

 Simple, rugged design and appearance

2.2.3 Existing Urban Wind Turbines

Wind turbines that function as distributed electricity generators can provide electricity

directly for buildings, which can reduce the power losses in electricity transmission over long

distances and cost on grids. The utilisation of wind turbines in urban areas is also an efficient

method to reduce the building carbon emission. Wind turbines can be divided into four types

by the rotor diameter [47]: i) micro-scale (rotor diameter < 0.1 m); ii) small-scale (0.1 m <

rotor diameter < 1 m); iii) middle-scale (1 m <rotor diameter < 5 m) and iv) large-scale (rotor

diameter > 5 m). Micro-scale wind turbines cannot generate enough electricity to satisfy the
17
 Chapter 2 Overview of Existing Wind Turbines

electric requirement of buildings. Both middle-scale and large-scale wind turbines have high

requirements for space and installation of such a wind turbine may influence structures of

buildings. Therefore, small-scale wind turbines are suitable to be used for buildings. Small-

scale wind turbines can be categorised into three types: small-scale conventional wind

turbines; small-scale conventional wind turbines with added shrouds; re-designed small-scale

wind turbines. Three types of urban wind turbines are reviewed in following subsections.

2.2.3.1 Small-scale Conventional Wind Turbines

A small-scale wind turbine has lower efficiency than a large-scale wind turbine due to low

Reynolds numbers (Re < 500,000) [48]. With low Reynolds numbers, laminar separation and

laminar separation bubbles are generated on the blade of a small wind turbine [49]. The

laminar separation and laminar separation bubbles affect the aerodynamics of blades and

result in the poor power generation and start-up capability [50]. Statistics show that 97% of

all micro/small wind turbines are HAWTs in the UK [51]. Kishore and Priya [52] summarised

performances of fourteen small HAWTs as shown in Table 2.1. It can be found that the

minimum and maximum values of the overall efficiency are 12% and 26%.

18
 Chapter 2 Overview of Existing Wind Turbines

Table 2.1: Efficiency of several current small-to-middle scale HAWTs [52].

Number Description Rotor diameter Rated wind speed Overall efficiency

(cm) (m/s)

1 NE-100S 120 10 14%

2 NE-200S 130 11 18%

3 NE-300S 130 13 17%

4 Energy Ball V100 110 10 17%

5 Micro Wind Turbine 23.4 2-7 18% *

6 TAOS 600 180 12.5 20%

7 Chinook 200 100 12.5 21%

8 EP-220 137 12.5 12%

9 Bornay 600 200 11 23%

10 AC 120 120 9 24%

11 AC 240 165 9 25%

12 µF500 50 12 25%

13 Alladin 400 130 12.5 25%

14 Superwind 350 120 12.5 26%

*power coefficient

There are several methods to enhance the low efficiency of a small-scale wind turbine, such

as aerofoils designed for low Reynolds numbers and optimised blades.

The blade aerofoils working under low Reynolds numbers have been developed. Singh et al.

[48] designed an AF300 aerofoil and studied its aerodynamic features by experimental and

numerical approaches under low Reynolds numbers. Compared with the other eight aerofoils,

19
 Chapter 2 Overview of Existing Wind Turbines

the optimised AF300 aerofoil has both a high lift coefficient and a high lift-to-drag ratio with

low Reynolds numbers. Moreover, a blade with this aerofoil can have low cut-in wind speed

and fast start-up [48]. Selig and McGranahan [53] studied six aerofoils of E387, FX63-137,

S822, S834, SD2030 and SH3055 for small HAWTs with a wind tunnel and investigated

aerodynamic characteristics of each aerofoil under the Reynolds number range from 100, 000

to 500,000. Gieuere et al. [54] compared fifteen aerofoils for an optimal operation with low

Reynolds numbers. They suggested that an aerofoil with a high lift-to-drag ratio was suitable

for variable-speed HAWTs, an aerofoil with a wide lift range was suitable for variable-pitch

HAWTs and an aerofoil for stall regulated HAWTs should be insensitive to roughness.

Henriques et al. [55] proposed a pressure-load inverse approach to design an aerofoil of small

wind turbines for the urban environment and found that an aerofoil designed by this approach

could reduce the leading edge suction peak, control the soft-stall behaviour and increase the

maximum lift.

In addition to development of low Reynolds number aerofoils, other geometrical parameters

of blades also can be optimized to suit the urban environment. Singh and Ahmed [49]

investigated influences of blade pitch angles of a two-blade small HAWT with an aerofoil of

AF300. They found that with an optimal pitch angle of 18o, the instantaneous cut-in wind

speed was reduced from 3.58 m/s to 2.34 m/s and the peak power coefficient was increased

from 0.15 to 0.255 under a wind speed of 6 m/s [49].

A start-up capability is important for wind turbines in urban areas [56]. Multiplying blades

can achieve a greater value of torque and improve the start-up behaviour of a wind turbine

[18]. Typically, a HAWT with five blades has the better start-up capability than that with two

blades [18]. Duquette and Visser [57] studied relationships between power coefficient and

20
 Chapter 2 Overview of Existing Wind Turbines

solidity of a HAWT. The solidity represents the blade number of a wind turbine and is

defined as total blade area divided by the rotor disc area [18]
𝐵 𝑐
𝜎 = 2𝜋 𝑅 (2.7)

where B is number of blades, c is the blade chord and R is the radius of blades. They found

that the wind turbine with a solidity of 15% – 25% had the highest maximum power

coefficient. The lower cut-in wind speed, the less blade erosion and lower noise were also

found as additional advantages [57].

Several small wind turbines can be installed on the top of a building as a wind farm, as shown

in Figure 2.8. This method increases the total generated power output greatly. Sixteen small

wind turbines were installed on a top of a building in Manchester, which may be one of the

first mini wind farms on a building in the UK [58]. The layout of wind turbines on a roof

should be considered because wakes behind wind turbines can diminish the power generation

of downstream wind turbines [59].

Figure 2.8: HAWTs placed multiply on a roof [58].

Howell et al. [27] studied a 0.6m-diameter Darrieus wind turbine with straight blades and the

wind turbine obtained the maximum power coefficient of 0.2. Pope et al. [60] predicted that a

Savonius VAWT with a rotor diameter of 2 m could achieve the maximum power coefficient
21
 Chapter 2 Overview of Existing Wind Turbines

of 0.18. Akwa et al. [61] reviewed the effects of geometrical parameters, such as end plates,

aspect ratios, overlap, number of buckets, buckets spacing and rotor shapes, on the

performance of a Savonius wind turbine. The maximum averaged power coefficient of

Savonius wind turbines is in a range of 0.05 - 0.3 [61]. Although the power coefficient of a

small VAWT is always lower than that of a small HAWT, VAWTs have advantages under

turbulent wind in an urban environment when compared with HAWTs [23, 27].

2.2.3.2 Small-scale Wind Turbines with Added Shrouds

Adding shrouds on small-scale wind turbines is a remarkable approach to accelerate the speed

of the passing wind so that a wind turbine can start up easily and produce more power output.

There are two popular types of shrouds: shrouds for small HAWTs, such as diffusers and

ducts; shrouds for small VAWTs, such as guide vanes.

A schematic view of general diffuser shrouds is shown in Figure 2.9. A wind turbine is placed

inside the shroud and near the flow inlet. Concentration of wind energy by a diffuser was

proposed around 1980 [62-64]. The diffuser can cause a larger flow pressure difference

between front and behind of a wind turbine to increase the flow speed [65]. Because the wind

power is proportional to the wind velocity cubed, a slight increment on the wind velocity can

provide a large increment of the wind power [18]. Kishore et al. [47] found that with an aid of

the diffuser as shown in Figure 2.9, the power output was increased by 40% - 60%, when

compared with a bare one.

22
 Chapter 2 Overview of Existing Wind Turbines

Figure 2.9: Schematic views of the diffuser for HAWTs [47].

(a) (b)

Figure 2.10: HAWT with the flanged diffuser: (a) schematic view; (b) prototype [66].

A flanged diffuser is an improved type. Figure 2.10 shows the geometry of a flanged diffuser.

The flange can produce a large flow separation behind the diffuser and attract more wind

entrancing [65]. Ohya et al. [66] found that the power output of a HAWT was improved 4-5

times when the flanged diffuser was used. Their results showed that the achieved maximum

power coefficient of 1.4 had exceeded the Betz limit of 0.59 [66]. However, it should be

noted that employing a diffuser has to add more weight and stresses on the tower of a HAWT.

In addition to the diffuser, a duct is another type of shrouds designed for HAWTs. Figure

2.11(a) shows the duct geometry. The duct can change wind direction from horizontal to
23
 Chapter 2 Overview of Existing Wind Turbines

vertical. Moreover, experiments have shown that this duct can accelerate the wind velocity up

to 30% [67]. As an advantage, it is possible that installing several ducted wind turbines in a

row along the roof and wall edges of a building for more power output. But it should be noted

that a drawback for the duct is the restriction to the angle of incident wind. It was reported

that the ducted wind turbine only operated well in a range of incident wind angle of ±60o

[67]. Hu and Cheng [68] designed a duct with a nozzle outlet as shown in Figure 2.11(b). The

nozzle outlet can increase the passing flow velocity. They found that the maximum flow

velocity was increased by 60% inside the duct, but the performance of the wind turbine was

also limited by the angle of incident wind [68].

(a) (b)

Figure 2.11: Ducted wind turbines: (a) original type [69]; (b) a bucket-shape duct with a
nozzle [68].

Ahmed [70] proposed a duct with a group of guide vanes which could lead air flow

downwards as shown in Figure 2.12. This duct combines some features of diffusers and ducts.

This device can capture wind and lead wind to the vertical direction inside the shroud. The

maximum flow speed inside the shroud was increased by 20-30% [70]. This wind turbine

24
 Chapter 2 Overview of Existing Wind Turbines

type is easy for maintenance since the generator can be placed near ground. But a drawback is

the restriction to the angle of incident wind because of the guide vanes.

Figure 2.12: A shroud present by Ahmed [70].

Chong et al. [9] presented a 5-blade Darrieus wind turbine with surrounded guide vanes. As

shown in Figure 2.13, the guide vanes have simple plate geometry and are installed with a

setting angle. These guide vanes can lead wind from any direction to an optimum direction

for downstream blades. They found that the maximum rotational speed was increased by 182%

with the aid of guide vanes [9]. The power output of the wind turbine with guide vanes was

3.48 times as much as that of a bare one [9]. The passing wind speed was accelerated and the

cut-in wind speed was reduced. So this wind turbine can start at a low wind speed and have

more working hours [9].

25
 Chapter 2 Overview of Existing Wind Turbines

Figure 2.13: An urban VAWT with guide vanes [9].

Employment of shrouds is also available for Savonius wind turbines. Current types of shrouds

include obstacles, curtain arrangements and guide-box tunnels. Mohamed et al. [71]

presented a simple obstacle for shielding the returning blade of a Savonius wind turbine as

shown in Figure 2.14(a), and found that the power coefficient of this Savonius wind turbine

was improved by more than 27% [71]. Altan and Atılgan [72] introduced a curtain

arrangement to improve aerodynamic features of a Savonius wind turbine. As shown in

Figure 2.14(b), the curtain arrangement is placed in front of a rotor to concentrate wind

energy for the rotor. Compared with the obstacle, the curtain arrangement is more complex

but has the better performance. It was reported that the curtain arrangement prevented a

negative torque opposite to the rotor rotation and the maximum power coefficient increased

from about 0.16 to 0.39 [72]. Irabu and Roy [73] designed a rectangular guide-box tunnel as

shown in Figure 2.14(c), and reported that the maximum power coefficient was improved by

23% for a two-blade Savonius rotor with this guide-box tunnel and 50% for a three-blade one

with this tunnel [73]. The above three types of added devices have significant improvements

on power output. However, it should be pointed out that the performances of these added

devices greatly depend on wind directions [71-73].

26
 Chapter 2 Overview of Existing Wind Turbines

(a) (b)

(c)

Figure 2.14: Three types of shrouds for a Savonius wind turbine: (a) obstacle [71]; (b)
curtain arrangement [72]; (c) guide-box tunnel [73].

Different types of shrouds for HAWTs and VAWTs have been reviewed in this subsection.

These shrouds can significantly improve power output and starting capability. However, there

are some drawbacks of these devices, such as additional weight on structures and

performances depending on wind direction.

2.2.3.3 Re-designed Wind Turbines

In addition to the optimisation of conventional wind turbines or utilisation of shrouds on them,

several different wind turbine designs are introduced in this section.

27
 Chapter 2 Overview of Existing Wind Turbines

Figure 2.15: A turbine with a crossflow runner [74].

A wind turbine with a crossflow runner was designed by Dragomirescu [74] for good starting

behaviour and high power coefficient, even under poor wind conditions. As shwon in Figure

2.15, the simple geometry makes it easier to be manufactured. Numerical results showed that

the wind turbine obtained a large value of power coefficient of 0.45, when compared with

other VAWTs [74]. The starting torque coeffcient for this wind turbine was 3.6, which means

the turbine could start fast even under low velocity wind. However, this wind turbine had a

great narrow range of tip speed ratios, which could result in high sensitivity of power

coefficient to wind speed. Condsidering the large torque coeffcient and the ability about

operating under the 360o flow direction, this wind turbine has the potentional to be utilised in

urban areas, but further performance validations are required.

The Sistan wind turbine was employed in the Greater Khorasan and Sistan Basin in the AD

9th century [75]. One of the advantages of Sistan wind turbines is the simplicity of its blades,

compared with Darrieus wind turbines and HAWTs. A Sistan wind turbine with guide vanes

was present to be used on tops of buildings by Chong et al. [76, 77]. Figure 2.16(a) shows the

0.54m-diameter three-blade Sistan wind turbine with a group of guide vanes. It can be seen

28
 Chapter 2 Overview of Existing Wind Turbines

that every guide vane has a different shape that may increase manufacturing cost. Results

showed that the maximum rotational speed was increased by 75.16% and the increment of

power output was 5.8 times under a wind speed of 3 m/s [76]. But it should be noted that its

performances were dependent on wind directions. Mü ller et al. [75] also presented a 0.6m-

diameter Sistan wind turbine with a shroud for building integration as shown in Figure 2.16(b)

and found that the peak efficiency was increased by 48%. The shroud has a simpler profile

than others [76, 77]. However, one drawback is the sensitivity of the wind turbine

performance to the wind direction.

(a) (b)

Figure 2.16: Sistan wind turbine: (a) with guide vanes [76] ; (b) with a shroud [75].

Zephyr wind turbine is a vertical axis wind turbine that employs a Savonius wind turbine

shape and guide vanes with reverse winglets to optimise approaching wind [78]. Figure 2.17

shows its schematic views. Wind properties were affected by guide vanes for a higher

pressure on the pressure surface of the rotor blade [78]. One of advantages of this wind

turbine is that it can capture wind from all directions (360o). It was found that the maximum

power coefficient of the zephyr wind turbine was 0.12 [78]. The structure of this wind turbine

can work with high turbulence wind in urban areas and its performance is independent on

29
 Chapter 2 Overview of Existing Wind Turbines

wind directions, therefore Pope et al. [78] suggested that this wind turbine could be used in

urban areas.

(a) (b)

Figure 2.17: Zephyr wind turbine: (a) 3D view; (b) schematic view [78].

(a) (b)

Figure 2.18: Crossflow wind turbine: (a) concept; (b) architectural integration with
ridge [79].

Sharpe and Proven [79] presented a conceptual crossflex wind turbine. Figure 2.18(a) shows

that the wind turbine has a Darrieus turbine shape and is placed on a frame with a shaft. They

reported that this wind turbine had the low solidity and the low inertial mass design, and

predicted that the produced noise would be low due to absence of blade tips [79]. Loads on

structure mountings and vibration would be lower than those of small-scale HAWTs and

30
 Chapter 2 Overview of Existing Wind Turbines

Darrieus VAWTs [79]. A significant advantage is that the crossflex wind turbine has a good

integration with buildings. As shown in Figure 2.18(b), this wind turbine is suitable to be

integrated with edges or corners of buildings.

Snaker and Tiryakioglu [80] presented a named Aeolum Harvester wind turbine as an

alternative to Savonius wind turbines. Figure 2.19(a) shows the segment of this conceptal

wind turbine. Each segament has an inlet where the speed of air flow can be increased. 3D

view of the wind turbine is shown in Figure 2.19(b). Each segment has an angular offset from

the next segment about the vertical axis. McTavish et al. [81] carried out static and dynamic

analyses about the torque characteristcs and found that this wind turbine generated a

comparable amount of static torque to a Savonius VAWT.

(a) (b)

Figure 2.19: VAWT with segments: (a) wireframe view; (b) 3D model [81].

In this subsection, the literature review shows that several urban wind turbines have the

potential for working in an urban environment, such as starting characteristics, integration

with buildings and simple structures. However, it must be pointed out that there are only

preliminary results about some urban wind turbines, such as the crossflow runner and the

Aeolum Harvester wind turbine, so more investigations about them are needed.

31
 Chapter 2 Overview of Existing Wind Turbines

2.2.3.4 Summary

In this section, various existing designs of wind turbines operating in urban areas have been

discussed, such as conventional wind turbines, wind turbines with shrouds and re-designed

wind turbines. The literature review shows that these wind turbines have drawbacks in the

urban environment, such as the restriction to the wind direction. An omni-flow energy system

has an advantage on accepting the omni-directional wind in the urban environment and its

features are reviewed in the following section.

2.3 Omni-flow Wind Energy System

An omni-flow wind energy system has shown the potential to be used in the urban

environment [14, 15]. The concept of an omni-flow wind energy system was patented by

Sureshan in 2008 [13]. Zhang [15] developed the system by numerical and experimental

studies. The omni-flow wind energy system can be installed on tops of high buildings in

urban areas as shown in Figure 2.20. Rich wind energy can compensate electricity

consumption of high buildings with this system for both economic and environmental

protection purposes. In Zhang’s study, the bottom diameter of the omni-flow wind energy

system is 1 m [15]. Because some facilities like air conditioners on roofs may limit the space

for installations, this omni-flow system has a small size. And the cost of the small size system

can be low. This section reviews key findings about the omni-flow wind energy system from

Zhang’s research [15].

32
 Chapter 2 Overview of Existing Wind Turbines

Figure 2.20: Illustrative view of the omni-flow wind energy system on the top of a
building. The background is a modified view of London by Google Earth.

2.3.1 Geometry of Omni-flow Wind Energy System

Geometry of the omni-flow wind energy system is shown in Figure 2.21. The omni-flow

wind energy system includes a shroud and five chambers. The five chambers are located

along the circumference of a shroud to capture wind from any direction and lead wind to the

vertical outlet through chambers. Thus, the wind energy system is suitable for the wind with

frequent changes in urban areas [15]. A wind turbine can be located near the outlet inside the

system. Key components of a wind turbine, such as turbine blades, can be placed inside the

shroud for protection. The shroud and chambers form a nozzle structure to increase the

velocity of the passing flow.

33
 Chapter 2 Overview of Existing Wind Turbines

(a)

(b)

Figure 2.21: The omni-flow wind energy system: (a) left view; (b) top view.

2.3.2 Flow Characters inside Omni-flow Wind Energy System

Flow characters inside the omni-flow wind energy system have been studied by experiments

and simulations [14, 15]. According to the angle between wind and a reference line, four

wind directions, 0o, 12o, 24o and 36o, are categorised for investigating the omni-flow wind

energy system as shown in Figure 2.22.

34
 Chapter 2 Overview of Existing Wind Turbines

(a) (b)

(c) (d)

Figure 2.22: Four wind directions for the omni-flow wind energy system: (a) 0o; (b) 12o;
(c) 24o; (d) 36o.

Distributions of flow velocity at the outlet of the shroud under four flow directions are shown

in Figure 2.23, when the velocity of free stream is 10 m/s [15]. It can be seen that the

distribution of flow velocity inside the shroud is non-uniform. The flow velocity field can be

divided into two regions: a high velocity region and a low velocity region. The flow in the

high velocity region comes from the chambers that are against to wind. When the wind

direction is at 0o, only one chamber is against to wind, therefore the high velocity region

occupies 1/5 of the outlet area as shown in Figure 2.23(a). Under the wind directions at 12o,

35
 Chapter 2 Overview of Existing Wind Turbines

24o and 36o, the high velocity region is expanded up to 2/5 of the outlet area as shown in

Figures 2.23(b-d).

(a) (b)

(c) (d)

Figure 2.23: Velocity distributions at the outlet of the omni-flow system under four wind
directions: (a) 0o; (b) 12o; (c) 24o; (d) 36o [15].

Due to structure features of the omni-flow system, the passing flow velocity can be increased.

Values of the maximum and average flow velocities at the outlet are shown in Figure 2.24.

With the free stream of 10 m/s, the maximum velocity at the outlet is increased by 46%, 50%,

59% and 69% under wind directions of 0o, 12o, 24o and 36o, respectively [15]. The maximum

velocity of the passing flow has the highest value of 16.9 m/s at 36o and the lowest value of
36
 Chapter 2 Overview of Existing Wind Turbines

14.6 m/s at 0o. The average velocities are increased by 3% – 12% compared with the free

stream velocity [15]. Such an increment on the flow velocity can provide a large increment on

the output power from the wind [65].

17

16
Flow Velocity ( m/s)

15
Maximum Velocity
14
Average Velocity
13

12

11

10
0 12 24 36
Angle of wind direction (degree)

Figure 2.24: The maximum and average velocities at the outlet of the omni-flow system
under four wind directions [15].

The omni-flow wind energy system also affects the flow direction greatly. A vector field on

an axial section plane is shown in Figure 2.25. The vortex behind the system can influence

the passing flow inside the shroud so that the passing flow direction inside the omni-flow

system is not ideal vertical [15]. The vector field at a horizontal section plane inside the

shroud is shown in Figure 2.26. Small vertexes can be seen inside the omni-flow system,

which cause the flow direction non-uniform too.

37
 Chapter 2 Overview of Existing Wind Turbines

Figure 2.25: Flow velocity vector field through the omni-flow system [15].

(a) (b)

(c) (d)

Figure 2.26: Velocity vector field at the outlet under four wind directions: (a) 0o; (b) 12o;
(c) 24o; (d) 36o [15].
38
 Chapter 2 Overview of Existing Wind Turbines

The passing flow inside the omni-flow wind energy system is non-uniform in both the

velocity and direction, which can affect aerodynamic forces on blades and the performances

of a wind turbine. Due to the high velocity region and low velocity region, a blade has to

undergo different aerodynamic forces in each revolution. And the change of aerodynamic

forces on a rotating blade is frequent. Because of such complex wind conditions, the

conventional thin blades of current wind turbines are difficult to work properly. A type of air

turbines, named impulse turbine, could have the potential to work under such complex wind

conditions. In the wave energy field, impulse turbines have shown good performances under

the flow with frequent changes of the wind directions and speeds. Therefore, it is considered

to employ the impulse turbine technology in this project.

2.4 Impulse Turbines in Wave Energy Field

Impulse, also called impulse stage, is defined that there is no pressure drop when flow passing

through the rotor of a turbine [82]. The impulse turbine is a turbine which has the impulse

stage [82]. Typical geometries of impulse turbines can be found in the Pelton wheel and first

stages of steam turbines [82]. In wave energy field, the impulse turbine has been one of most

widely used turbines for converting air flow energy into mechanical energy. A Wells turbine

is another type of turbines used in wave energy field [83]. Compared with Wells turbines, the

impulse turbine has high efficiency, low working noise and good starting characteristics [84,

85]. Therefore, technologies about impulse turbines are reviewed in this section.

Characteristics of air flow in wave energy, geometry features and performances of impulse

turbines in wave energy field are included.

39
 Chapter 2 Overview of Existing Wind Turbines

2.4.1 Flow Characteristics

An impulse turbine works inside an Oscillating Water Column (OWC) plant for wave energy

conversion [86]. Features of the air flow for an impulse turbine are affected by the OWC

plant and waves. The OWC device has a partly submerged structure and opens below the

water surface, inside which air is trapped above the water free surface as shown in Figure

2.27 [87]. Oscillating motions of the water surface caused by incident waves can push air to

flow through an impulse turbine. The OWC converts wave energy into pneumatic energy in a

form of bi-directional air flow [84]. The impulse turbine converts energy from air flow into

mechanical energy and drives a generator to produce electricity.

(a) (b)

Figure 2.27: Two types of OWCs: (a) vertical type; (b) horizontal type [84, 86].

Due to the motion of water wave, there are much more demanding conditions for an impulse

turbine inside an OWC than air turbines in other applications, such as wind turbines [87].

Figure 2.28 shows extremely large variations from wave to wave over a few seconds. Flow

properties are varied frequently by the irregular wave motion. Velocity of the air flow

oscillates with waves inside an OWC. The air flow direction is reversed several times in few

seconds, which is more frequent than changes of nature wind direction.

40
 Chapter 2 Overview of Existing Wind Turbines

Figure 2.28: Height change of the irregular wave in a short time inside an OWC [84].

There are two similar features between the omni-flow wind energy system and OWC. Firstly,

a wind turbine has to work inside the shroud of the omni-flow wind energy system. The

shroud may affect the performance of a wind turbine. An impulse turbine also has to operate

inside the shroud of an OWC [88]. Secondly, air flow produced by the OWC is greatly

unsteady. Blades of an impulse turbine have to suffer the frequently changed flow speed and

direction in one revolution. This feature is similar to that a wind turbine working inside the

omni-flow system. These two features indicate the possibility to use the impulse turbine

technology in a wind turbine for the omni-flow wind energy system.

It should be noted that air flow inside the OWC is a type of internal flow, but wind for a wind

turbine is a type of external flow. With the internal flow, there is a high pressure on front of

an impulse turbine, which can result in high power output. Hence an impulse turbine inside

the OWC can generate more power than a same scale wind turbine. It was reported that an

impulse turbine with a rotor diameter of 1.6 m could produce a maximum power ouput of 400

kW at a flow speed of 20 m/s [89]. But the ideal power output for a wind turbine is only 5.7

kW under the same condition.

41
 Chapter 2 Overview of Existing Wind Turbines

2.4.2 Geometrical Features of Impulse Turbines

The impulse turbine for wave energy was patented by I.A. Babinsten in 1975 [87]. The rotor

of the impulse turbine is basically identical to the rotor of an axial-flow impulse type of

signal-stage steam turbines [87]. 3D view of a classic impulse turbine is shown in Figure 2.29.

In order to suit the bi-directional air flow in an OWC, blades of impulse turbines use

symmetry geometry. In order to optimise the approaching flow, guide vanes are installed in

front of blades. Because air flow is bi-directional in an OWC, guide vanes are placed in both

upstream and downstream of the rotor. Impulse turbines with fixed guide vanes (IFGVs) and

impulse turbines with self-pith-controlled guide vanes (ISGVs) are two major types as shown

in Figure 2.30. Guide vanes of IFGVs are fixed with the chamber walls of OWCs. The ISGVs

use the same rotors as the IFGVs, but ISGVs have self-pith-controlled guide vanes to suit the

change of flow direction [84]. It should be noted that the ISGVs and IFGVs are designed to

suit the bi-directional flow condition.

Figure 2.29: 3D view of an impulse turbine [90].

42
 Chapter 2 Overview of Existing Wind Turbines

(a) (b)

Figure 2.30: Schematics of two types of impulse turbines: (a) IFGV; (b) ISGV [84].

There are two types of guide vane profiles: aerofoil and plate, as shown in Figure 2.31.

Efficiency of an impulse turbine with the aerofoil guide vanes is very close to that with the

plate guide vanes [84]. Considered the manufacturing cost, the plate guide vanes have been

widely used. Due to geometrical features of guide vanes, there is a nozzle effect on the

passing flow to increase the passing flow speed [84].

(a) (b)

Figure 2.31: Two profiles of guide vanes: (a) plate; (b) aerofoil [84].

Figure 2.32 shows two typical blade profiles of impulse turbines. These two types of blade

aerofoils are both symmetric about the centre axis. The simple profile consists of two circular

arcs at mid-chord portion as shown in Figure 2.32(a). In Figure 2.32(b) the elliptic one has a

rounded shape [84]. A comparison about mean efficiency of both types of blade aerofoils

shows that the elliptic blade profile is superior to the simple one [84].
43
 Chapter 2 Overview of Existing Wind Turbines

(a) (b)

Figure 2.32: Two blade profiles: (a) simple; (b) elliptic.

Other key geometrical parameters of an impulse turbine include the hub-to-tip ratio and the

tip clearance. The hub-to-tip ratio, which is defined as a dimensionless ratio between the hub

diameter and rotor diameter, is used to represent the hub size [84]. The hub-to-tip ratio of 0.7

was firstly used in impulse turbines [84, 91]. Then the hub-to-tip ratio of 0.6 was suggested

by Thakker et al. [92]. Tip clearance is the gap between the blade tip and the chamber inner

surface of an OWC. The tip clearance is set for easy installation and maintenance of an

impulse turbine inside an OWC. Thakker et al. [88] investigated influences of the tip

clearance on efficiency of an impulse turbine. Their results showed that the impulse turbine

with a tip clearance that takes 1% of the rotor radius performed the similar efficiency to that

without tip clearance. Additionally, the tip clearance of 1% can reduce the difficulty in

installation of the impulse turbine inside an OWC.

Maeda et al. [93] presented an unidirectional impulse turbine which can work under an

unidirectional flow condition. The unidirectional impulse turbine has the asymmetrical blade

geometry as shown in Figure 2.33(a). It can be seen that the leading edge of the aerofoil is

optimised to suit one fixed flow direction, when compared with the classic impulse turbine

blade. Because the blade aerofoil is asymmetrical, only upstream guide vanes are needed as

shown in Figure 2.33(b). A wave energy system with two unidirectional impulse turbines has
44
 Chapter 2 Overview of Existing Wind Turbines

been presented by Jayashankar et al. [94, 95]. It should be noted that the unidirectional

impulse turbine is designed only for the unidirectional flow, which is the same as a wind

turbine. Therefore, it provides more possibility to employ the impulse turbine technology in

this omni-flow wind energy system.

(a) (b)

Figure 2.33: Unidirectional impulse turbine: (a) asymmetric blade aerofoil; (b)
upstream plate guide vanes and blades [93].

It should be noted that because there are differences about flow conditions between impulse

turbines and wind turbines, the recommended geometrical parameters for an impulse turbine

may not be optimal for a wind turbine and need to be investigated.

2.4.3 Performances of Impulse Turbines

Efficiency is the most important feature of an impulse turbine in the wave energy field.

Efficiency of different impulse turbines is compared in Figure 2.34. It is clear that the

unidirectional impulse turbine has the highest efficiency. Self-pitching guide vanes can

improve flow conditions so the ISGVs have the higher efficiency than the IFGVs. But the

ISGVs require extra machines for self-pith-controlled guide vanes, which results in higher

cost in manufacturing and maintenance. It should be noted that efficiency of an impulse

45
 Chapter 2 Overview of Existing Wind Turbines

turbine cannot be compared directly with power coefficient of a wind turbine since the

efficiency definition of an impulse turbine is different from that of a wind turbine.

Figure 2.34: Efficiency comparison of three impulse turbines [94].

2.5 Summary
In this chapter, a critical review of existing wind turbines has been carried out. Wind

characteristics, requirements from the urban environment and current urban wind turbines

have been discussed. The literature review shows that an omni-flow wind energy system has

the potential to be used in urban areas, but the non-uniform flow condition is difficult for

conventional wind turbines. Based upon the literature review, an impulse turbine has the

potential to work under the complex wind conditions caused by the omni-flow wind energy

system. Therefore, this project proposes a new wind turbine configuration based upon the

impulse turbine technology.

46
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

Chapter 3 Methods for Aerodynamic Investigation of Wind

Turbines

Several methods have been employed to investigate the aerodynamics of a wind turbine, such

as blade element momentum (BEM) theory, computational fluid dynamics (CFD) simulation

and wind tunnel testing. BEM method can estimate aerodynamic performances of a wind

turbine by lift coefficient, drag coefficient and other terms. However, BEM method is not

suitable for the proposed wind turbine. Because the proposed wind turbine has much more

blades than common wind turbines, the interferences among blades are unavoidable and

complex. Another reason is that the upstream guide vanes of the proposed wind turbine have

influences on the flow aerodynamics, which is difficult to be analysed by BEM method.

Therefore, the proposed wind turbine is not studied by the BEM method in this project. In this

chapter, two methods for aerodynamic analysis of wind turbines, CFD simulation and wind

tunnel testing, are reviewed.

3.1 Computational Fluid Dynamics for Investigation of Wind Turbines

With the development of computing power, CFD method has entered into wide applications

such as aircrafts, vehicles and turbomachinery. Generally speaking, CFD simulation can

reduce a period of time and cost, when compared with experiments. Recently, CFD method

has been one of major methods in the aerodynamic investigation of wind turbines. In this

section, the employment of CFD method for investigation of wind turbines is reviewed.

3.1.1 Governing Equations

Governing equations consist of the continuity equation, momentum equation and energy

equation. Continuity equation represents the physical principle that mass is conserved.

Momentum equation is the resulting equation of Newton’s second law. Energy equation
47
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

represents that energy is conserved. Energy equation can be ignored in aerodynamic

investigations of wind turbines since the temperature change is not always considered.

Continuity equation and momentum equation for a compressible Newtonian fluid are shown

as follows [96]:

Continuity equation

𝜕𝜌
⃗)=0
+ 𝑑𝑖𝑣(𝜌𝑈 (3.1)
𝜕𝑡

⃗ is velocity vector.
where 𝜌 is fluid density, t is time and 𝑈

Momentum equation

∂(𝜌𝑢)
⃗ ) = − 𝜕𝑝 + div(𝜇 grad 𝑢) + 𝑆𝑀𝑥
+ div(𝜌𝑢𝑈 (3.2)
∂t 𝜕𝑥

∂(𝜌𝑣)
⃗ ) = − 𝜕𝑝 + div(𝜇 grad 𝑣) + 𝑆𝑀𝑦
+ div(𝜌𝑣𝑈 (3.3)
∂t 𝜕𝑦

∂(𝜌𝑤)
⃗ ) = − 𝜕𝑝 + div(𝜇 grad 𝑤) + 𝑆𝑀𝑧
+ div(𝜌𝑤𝑈 (3.4)
∂t 𝜕𝑧

⃗ in x-axis, y-axis and z-axis, 𝜇 is dynamic

where u, v and w are components of velocity 𝑈

viscosity, 𝑝 is the pressure on a fluid element, 𝑆𝑀𝑥 , 𝑆𝑀𝑦 and 𝑆𝑀𝑧 are momentum source terms

for the viscous stress on a fluid element.

Governing equations need be transferred into discrete counterparts for numerical evaluation

and implementation in calculation. There are three major discretization techniques: finite

difference method, finite volume method and finite element method [97]. Finite difference

method is easy to formulate and suitable for simple geometries and structured meshes [98].

Finite volume method is appropriate for complex geometries and unstructured meshes [98].

Solution with finite volume method needs less time than that with finite element method [97].

Finite volume method has been popularly used in CFD simulations of wind turbines [17]. The

simulations in this project are also based on the finite volume method.

48
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

3.1.2 Mesh Strategies for Wind Turbines

The quality of meshing has significant influences on the rate of convergence, solving time

and solution accuracy. There are two major types of cells in meshing: structured cell and

unstructured cell. The structured cell is simpler than the unstructured cell. A disadvantage of

the structured cell is that it is only suitable for simple geometries and primary for finite

difference method, and the cell distribution is difficult to control [97, 98]. The unstructured

cell is more suitable for complex geometries and is adapted to the finite volume method [97].

The 2D cell shape is always either triangular or quadrilateral. The 3D cell shape is always

tetrahedral, hexahedral or polyhedral. Polyhedral cells used in this project are shown in

Figure 3.1.

Figure 3.1: Polyhedral cells on the proposed wind turbine model in this project.

Unstructured cells are mainly employed on the blade meshing because they are suitable for

the complex profiles of blades. Lanzafame et al. [99] used a large amount of tetrahedral cells

for modelling a two-blade HAWT, and then replaced the tetrahedral cells by polyhedral cells

in order to reduce the cell number and improve accuracy. Thumthae et al. [100] employed a

hybrid mesh strategy including rectangular cells for the region near blade and triangular cells

49
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

for the reset region. They found that the numerical results were quite close to experimental

data.

For a 2D straight-blade Darrieus wind turbine model, Danao et al. [101] used structured O-

type cells as shown in Figure 3.2. They found that the O-type cell was better than the

conventional C-type cell since the O-type cell was good at predicting wakes on the blade. For

a similar Darrieus wind turbine model, Almohammadi et al. [102] employed the structured

quadrilateral cells near the blade region and the unstructured cells in the region far from the

blade in order to balance the computer source and accuracy.

Figure 3.2: O-type cells for a blade aerofoil [101].

For an impulse turbine, Thakker et al. [103] used two mesh strategies: the hybrid unstructured

mesh strategy including hexahedral, tetrahedral and pyramidal cells as shown in Figure 3.3,

and the hexahedral structured mesh strategy. Numerical results produced by two mesh

strategies were both close to experimental results. They found that the hybrid unstructured

cells had the better mesh quality and somewhat were easier to be created than the hexahedral

cells because the hybrid cells were more suitable for complex profiles than the hexahedral

cells [103].

50
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

Figure 3.3: The hybrid mesh strategy for an impulse turbine [103].

3.1.3 Turbulence Models Used for Investigation of Wind Turbines

Turbulent flow is quite difficult to be solved because turbulent flow is highly unsteady and

contains a great deal of vorticity [97]. Recently, turbulent flow can be solved by the solution

with Reynolds-averaged Navier-Stokes (RANS) equations, large eddy simulations (LES) and

direct numerical simulations (DNS) [98, 104]. The solution with RANS equations can

provide results with acceptable levels of accuracy for many engineering purposes with less

computing cost, when compared with LES and DNS [104-107]. The work in this project is

based on the solution with RANS equations.

RANS equations for compressible flow are given as follows [96]:

Continuity equation:
̅
𝜕𝜌
̅) = 0
+ div(𝜌𝑈 (3.5)
𝜕𝑡

̅ is mean velocity vector, 𝜌̅ is mean fluid density and t is time.

where 𝑈

Reynolds equations:

̅̅̅̅̅
′2 ) ̅ ̅̅̅̅̅̅
𝑢′ 𝑣 ′ ) ̅ ̅̅̅̅̅̅̅
𝑢′ 𝑤 ′ )
̅𝑢
∂(𝜌 ̅)
̅) = − 𝜕𝑝 + div(𝜇 grad 𝑢̅) + [− 𝜕(𝜌̅𝑢
+ div(𝜌̅ 𝑢̅𝑈 −
𝜕(𝜌
−
𝜕(𝜌
] + 𝑆𝑀𝑥
∂t 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑧

(3.6)
51
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

̅̅̅̅̅̅
′ ′ ̅̅̅̅̅
′2 ) ̅ ̅̅̅̅̅̅̅
𝑣′𝑤′)
̅ 𝑣̅)
∂(𝜌
̅) = − 𝜕𝑝 + div(𝜇 grad 𝑣̅ ) + [− 𝜕(𝜌̅𝑢 𝑣 ) − 𝜕(𝜌̅𝑣
+ div(𝜌̅ 𝑣̅𝑈 −
𝜕(𝜌
] + 𝑆𝑀𝑦
∂t 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑧

(3.7)

̅𝑤
∂(𝜌 ̅) 𝜕𝑝 ̅ ̅̅̅̅̅̅̅
𝜕(𝜌 𝑢′ 𝑤 ′ ) ̅ ̅̅̅̅̅̅̅
𝜕(𝜌 𝑣′𝑤′) ̅ ̅̅̅̅̅
𝜕(𝜌 𝑤 ′2 )
+ div(𝜌̅ 𝑤
̅𝑈̅) = − + div(𝜇 grad 𝑤
̅) + [− − − ] + 𝑆𝑀𝑧
∂t 𝜕𝑧 𝜕𝑥 𝜕𝑦 𝜕𝑧

(3.8)

̅ are mean velocity components of u, v and w. 𝑢′ , 𝑣 ′ , and 𝑤 ′ are fluctuating

where 𝑢̅, 𝑣̅ and 𝑤

velocity components of u, v and w. 𝑆𝑀𝑥 , 𝑆𝑀𝑦 and 𝑆𝑀𝑧 are momentum source terms for the

viscous stress on a fluid element.

Several turbulence models have been used in simulations of wind turbines, such as Spalart–

Allmaras model, two equation models (Algebraic stress model, 𝑘 - 𝜀 and 𝑘 - 𝜔 turbulence

models) and Reynolds stress model. Spalart–Allmaras model can provide economical

computations, but Benjanirat et al. [108] reported that Spalart–Allmaras model was poor on

predictions of aerodynamic performances of wind turbines, when compared with two

equation models. Reynolds stress model has a potential for being one of the most general

turbulence models, but the model needs more computing cost than two equation models [109].

Algebraic stress model is an approach combining the generality of Reynolds stress model

with the economy of 𝑘-𝜀 models, but it has not been widely validated and used on wind

turbines [96, 110]. Compared with other turbulence models, the families of 𝑘-𝜀 and 𝑘-𝜔

turbulence models are more popular in investigations of wind turbines with acceptable

computer source requirements [17]. The employment of 𝑘-𝜀 and 𝑘-𝜔 turbulence models for

aerodynamic analyses of wind turbines is reviewed in this subsection.

Launder and Spalding [111] have developed the standard 𝑘-𝜀 turbulence model that consists

of two transport equations for the turbulent kinetic energy, 𝑘 and its dissipation rate, 𝜀. The

52
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

standard 𝑘-𝜀 turbulence model has been widely used but this model has some limitations

related to low Reynolds number and flow separation [112]. Wolfe et al. [113] used the

standard 𝑘-𝜀 turbulence model to study a blade aerofoil and found that the numerical results

agreed well with experimental data at low angles of attack. The standard 𝑘-𝜀 turbulence

model can predict performances of a HAWT with acceptable levels of accuracy [108]. Aresti

et al. [114] analysed the flow separation and vortex behind a VAWT by the standard 𝑘-𝜀

turbulence model. Thakker et al. [103, 115, 116] employed the standard 𝑘-𝜀 model to study

impulse turbines and found that the predicted efficiency agreed well with experimental data.

The realisable 𝑘-𝜀 model contains a new transport equation for the turbulent dissipation rate

and a new eddy viscosity formulation based on the realisability constraints, when compared

with the standard 𝑘-𝜀 model [117]. Because the eddy viscosity formulation includes a rotation

rate, the realisable 𝑘-𝜀 model is possible to yield a better performance than the standard 𝑘-𝜀

model when the flow involves separation and rotation [117, 118]. Raciti Castelli et al. [119,

120] reported that the realisable 𝑘-𝜀 model successfully predicted most flow features in a

study of a Darrieus wind turbine. Dragomirescu [74] employed both the realisable 𝑘-𝜀 model

and the SST (shear stress transport) 𝑘-𝜔 model for a wind turbine and found that only the

realisable 𝑘-𝜀 model assured convergence. Mohamed [121] used the realisable 𝑘-𝜀 model to

study the power output performance of a Darrieus wind turbine and found that the numerical

results matched well with experimental data. Abohela et al. [37] suggested that the realisable

𝑘 - 𝜀 model was preferred in simulations of flow around buildings for positioning wind

turbines.

The standard 𝑘-𝜔 model is prominent one of two-equation turbulence models. This model is

𝑘
characterized by the turbulent kinetic energy k and the frequency, 𝜔 = 𝜀 . Wilcox [122]
53
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

pointed that the standard 𝑘-𝜔 model could be integrated into the flow near walls without

wall-damping functions. But Menter [123] pointed out that the standard 𝑘-𝜔 model could fail

for the flow with pressure induced separation. Yang et al. [124] studied a blade aerofoil by

the standard 𝑘-𝜔 model and found that the computational results like pressure coefficients

had good agreement with the experimental ones. In a study of a VAWT, results from the

standard 𝑘-𝜔 model were closer to the experimental data than those from the standard 𝑘-𝜀

model [125]. In studies of impulse turbines, Thakker et al. [103, 115] found that the

numerical results from the standard 𝑘-𝜔 model had a similar trend to the experimental results

although the numerical approach over predicted results at high flow speeds.

Menter [126] found that the near-wall performance of the standard 𝑘 - 𝜔 model was

unsatisfactory for boundary layers with adverse pressure gradients and proposed a SST (shear

stress transport) 𝑘-𝜔 turbulence model. The SST 𝑘-𝜔 model uses a transformation of the 𝑘-𝜀

model into the 𝑘-𝜔 model in the near wall region and keeps the standard 𝑘-𝜀 model in the

fully turbulent region far from the wall. The SST 𝑘-𝜔 model has a good wall treatment and

avoids shortcomings of the 𝑘-𝜀 model and the standard 𝑘-𝜔 model [110, 126]. Le Pape et al.

[127] studied a HAWT by the standard 𝑘-𝜔 turbulence model and SST 𝑘-𝜔 turbulence model.

They found that the SST 𝑘-𝜔 turbulence model had better performances on prediction of the

stall angle than the standard 𝑘-𝜔 turbulence model. Chong et al. [76] investigated a Darrieus

wind turbine with guide vanes by the SST 𝑘-𝜔 model and found that the trend of numerical

results was close to that of experimental data.

3.1.4 Applications of CFD Method

The current employment of CFD method in aerodynamic analyses of wind turbines is

reviewed in this subsection. CFD method has been used to optimise the blade aerofoil design.

54
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

Yang et al. [128] and Wolfe et al. [129] both used CFD method to study aerodynamic

characteristics of NERL aerofoils of wind turbine blades. Fuglsang et al. [130] and

Bertagnolio et al. [131] analysed performances of the Risø airfoil family by CFD simulations.

CFD method has been employed in 3D aerodynamic analyses of wind turbines. Duque et al.

[132] and Sørensen et al. [133] simulated the flow around 3D HAWT models and predicted

power output. Since then, the CFD method has been widely used for various wind turbines.

Howell et al. [27] presented a 3D simulation of a straight-blade Darrieus wind turbine.

Features of other types of wind turbines were also studied by the CFD method before

experiments [70, 74, 76, 81]. Moreover, the tower and the nacelle of a wind turbine have been

analysed by simulations. Li et al. [134] simulated a NREL Phase VI wind turbine model that

included blades, a nacelle and a tower. Lin et al. [135] studied the interference between the

tower and blades. Because the rotor axis of a HAWT is parallel to wind direction, it can be

regarded that the wind features are unchangeable in one revolution. Thus steady solutions are

always used for HAWTs [99]. However, the rotor axis of a VAWT is perpendicular to wind

direction, wind features depend on the position of blades in each revolution. Therefore,

unsteady solutions are always employed for VAWTs [114].

The predicted aerodynamic forces by CFD method have been used in structural analyses of

blade structures [136, 137]. Baxevanou et al. [138] presented a numerical model combining a

Navier-Stokes CFD solver with an elastic model to study the aeroelastic features of wind

turbine blades.

Using the CFD method in studies of wind turbines has shown advantages over experimental

tests in time and costs. Present aerodynamic research of wind turbines has indicated that CFD

approach can provide results with acceptable levels of accuracy. Recently, the CFD method
55
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

has been one of major approaches in aerodynamic investigations of wind turbines. Many

studies of wind turbines have used commercial CFD software packages [139-142]. In this

project the CFD method has been employed to study aerodynamic performances of the

proposed wind turbine.

3.2 Wind Tunnel Testing

A wind tunnel is used to measure the characteristics of the air flow around an object

positioned in a test section, such as flow velocity, flow pressure, aerodynamic force and

moment. A wind tunnel consists of a flow passage, a test section, a fan system and a balance

system. A wind tunnel produces air flow inside the test section with the aid of a powerful fan.

The flow velocity in a wind tunnel is controlled by adjusting rotational speeds of the fan.

Aerodynamic forces and moments on an object are measured by a balance system. Generally

speaking, there are two types of wind tunnels: the open return wind tunnel and the closed

return wind tunnel. There are also two types of test sections: the open test section and the

closed test section.

Wind tunnel testing is one of major methods in aerodynamic research of wind turbines. A

classical measurement system includes a torque transducer and a rotational speed sensor.

Figure 3.4 shows a measure system for testing a HAWT. In tests, both torque and rotational

speeds of a tested wind turbine are obtained from the sensor. As shown in Figure 3.4, a brake

positioned behind the sensor applies loads on the shaft to control the rotational speed of the

turbine.

56
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

Figure 3.4: Measure system in wind tunnel testing [143].

In wind tunnel testing, the torque, rotational speed and power output of a wind turbine over a

range of wind velocities can be measured. Figure 3.5 shows various wind turbines in wind

tunnel tests. Kishore et al. [52] tested a small HAWT inside an open test section.

Distributions of the power coefficient and torque coefficient under various wind velocities

were achieved. Wind tunnel testing has also been used to investigate VAWTs. Danao et al.

[144] studied a straight-blade Darrieus wind turbine inside a closed test section. They

prepared an unsteady wind condition by controlling a shutter drive mechanism inside the

wind tunnel. The power output and acceleration of rotational speeds of the VAWT under the

unsteady condition were tested. Burlando et al. [145] employed a closed return wind tunnel

with a closed test section to study the flow field around a Savonius wind turbine.

Wind tunnel testing has been used to validate the numerical approach by comparing

experimental data with numerical results. Aresti et al. [114] studied performances of a wind

turbine by different turbulent models and then compared the modelling results with data from

wind tunnel testing. Rolland et al. [146] employed smoke as a flow visualisation method in

wind tunnel testing to prove a prediction about a turbulent vortex pattern found in simulations.

57
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

Ahmed et al. [70] presented a shroud to accelerate the flow speed for a wind turbine and

validated its performances in an open return wind tunnel.

Figure 3.5: Wind turbines in wind tunnel tests [47, 143, 147].

Solid blockage is a common phenomenon existed in wind tunnel testing. Solid blockage is

caused by a reduction in the cross area of the test section, when placing an object inside the

wind tunnel. Due to the solid blockage, velocity of the flow inside the test section is increased,

when compared with the free stream. The magnitude of solid blockage can be represented by

a ratio between the front area of the object and the cross area of the test section. Solid

blockage exists both in a closed test section and an open test section. A wind tunnel with a

closed test section is more sensitive to solid blockage than that with an open test section, but

the influence of solid blockage on the flow speed can be corrected [148]. Barlow et al. [148]

suggested a blockage correction method for an unusual shape in a tunnel. The correction

factor, 𝐶𝐵 , can be expressed as:

1 𝑚𝑜𝑑𝑒𝑙 𝑓𝑟𝑜𝑛𝑡𝑎𝑙 𝑎𝑟𝑒𝑎

𝐶𝐵 = 4 (3.9)
𝑡𝑒𝑠𝑡 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑟𝑒𝑎
58
 Chapter 3 Methods for Aerodynamic Investigation of Wind Turbines

The corrected flow speed, 𝑈corrected, can be given by

𝑈𝑐𝑜𝑟𝑟𝑒𝑐𝑡𝑒𝑑 = (1 + 𝐶𝐵 ) ∙ 𝑈𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 (3.10)

A closed return wind tunnel has been used in this project since the quality of flow in a closed

return wind tunnel is easy to be controlled [149]. A measurement system in wind tunnel

testing can record the power output from the proposed wind turbine under various wind

velocities. Meanwhile, the wind tunnel testing can be used to validate a numerical approach

used in this project.

3.3 Summary
The CFD method and the wind tunnel testing for aerodynamic investigations of wind turbines

have been reviewed in this chapter. BEM method is not considered to be used in this project

because the proposed wind turbine has a large number of blades and a group of upstream

guide vanes. Applications of the CFD method and the wind tunnel testing on studies of wind

turbines have been discussed. From the literature review, both the CFD method and the wind

tunnel testing can be used to study the proposed wind turbine. Considering the low cost, the

CFD method is used to analyse flow characteristics near the blades and aerodynamic

performances of the wind turbine in this project. The wind tunnel testing is used to validate the

CFD modelling by comparing computational results with experimental data.

59
 Chapter 4 Proposed Configuration of a Novel Wind Turbine

Chapter 4 Proposed Configuration of a Novel Wind Turbine

As discussed in Section 2.3, both velocity and direction of the flow are non-uniform inside

the omni-flow wind turbine energy system and conventional wind turbines are difficult to suit.

Therefore, a new concept of a wind turbine is needed for the system. This chapter introduces

the configuration of a novel wind turbine based on the impulse turbines technology.

4.1 Introduction
This project proposes a new wind turbine configuration for an omni-flow wind energy system

in urban areas. In Sections 2.2 and 2.3, the requirements of urban wind turbines and the flow

features inside the omni-flow wind energy system have been summarised. The novel wind

turbine configuration is based on the impulse turbine technology. The impulse turbine

technology reviewed in Section 2.4 shows the potential to operate well under the non-uniform

flow condition inside the omni-flow wind energy system. As reviewed in Section 2.4, the

impulse turbine technology includes several characters as follows [84, 93]:

 Blades of impulse turbine utilise specific aerofoils as shown in Figures 2.32 - 2.33.

 Guide vanes placed ahead of blades are needed.

 The rotor has a large number of blades.

 The rotor hub has a significantly large size.

4.2 Geometry of a Novel Wind Turbine

The review from Section 2.4 shows that the unidirectional impulse turbine has good

performances [94] therefore the proposed wind turbine uses a type of asymmetric blade

profiles and only one group of upstream guide vanes. The wind turbine consists of two

components: a stator and a rotor as shown in Figure 4.1(a). The diameter of the proposed

wind turbine is 300 mm. This diameter is chosen due to the dimensions of the omni-flow
60
 Chapter 4 Proposed Configuration of a Novel Wind Turbine

wind energy system presented by Zhang [15]. Considering the internal installation space of

the system, the wind turbine diameter is 300 mm. Both the stator and the rotor have an

identical hub diameter of 135 mm. Figure 4.1(b) shows the proposed wind turbine inside a

cylindrical chamber. Guide vanes are fixed with the chamber, thus the stator is stationary

when the turbine operating. Guide vanes can change the flow direction and speed. When air

flow passing blades, the generated torque drives the rotor to rotate so that wind energy is

converted into mechanical energy. Figure 4.1(b) also shows that there is a clearance of 2 mm

between blades and the chamber wall. This small clearance has a negligible influence on the

performance of an impulse turbine [88].

(a) (b)

Figure 4.1: The proposed wind turbine: (a) 3D view; (b) schematic view of the wind
turbine inside a chamber.

The stator has 20 guide vanes and the profile of one guide vane is shown in Figure 4.2. The

guide vane has a simple plate geometry that includes an arc front and a straight rear as shown

in Figure 4.2(a). Such a plate geometry has been widely used in impulse turbines [84]. The

straight rear of the guide vane has a setting angle of 20o that is an optimal setting angle for

guide vanes of an impulse turbine [93]. The arc part has a radius of 30.02 mm and the straight
61
 Chapter 4 Proposed Configuration of a Novel Wind Turbine

part has a length of 31.52 mm. Setoguchi et al. [84] recommended ratios between the turbine

diameter and dimensions of guide vanes for an impulse turbine. Dimensions of the proposed

guide vanes are selected based upon the recommended ratios. Technical drawings of the

guide vane and the stator hub are given in Appendices A and B.

(a) (b)

Figure 4.2: Guide vane: (a) section view; (b) 3D view.

Maeda et al.[93] suggested that the impulse turbine had good performances when the blade

aerofoil was not asymmetric and the aerofoil trailing edge was thinner than the leading edge.

This suggestion was considered when the configuration of the wind turbine was proposed.

The blade profile is asymmetric as shown in Figure 4.3(a). This aerofoil is utilised on each

blade section and the blade is not twisted as shown in Figure 4.3(b). This simple geometry

can reduce difficulty and cost of manufacturing. The chord length is 49.51 mm which takes

up 33% of the turbine radius. The amount of camber as a percentage of the chord length is

38%. The lengths of chord and camber are chosen due to the dimensional ratios

recommended by Setoguchi et al [84]. There are 20 blades on the rotor. Technical drawing of

62
 Chapter 4 Proposed Configuration of a Novel Wind Turbine

the rotor hub is given in Appendix C. This aerofoil is named as Type 1 aerofoil and its

coordinates are listed in Appendix D.

(a) (b)

Figure 4.3: Schematic views of the blade: (a) aerofoil; (b) 3D view.

4.3 Unique Geometrical Features

Compared with existing wind turbines, this novel wind turbine has several unique geometrical

features as follows:

(1) Influences of the stator. This wind turbine has an extra stationary component – stator –

compared with conventional wind turbines. Air flow passes through the stator firstly and

then blades. Employing such a stator with guide vanes is to optimise the properties of the

passing flow. It can be expected that the stator has influences on both flow direction and

velocity.

(2) Blade number. This wind turbine has a blade number of 20 that is much greater than two

or three blades of general wind turbines. Based upon the definition of solidity in Eq. (2.7),

this novel wind turbine has a solidity of approximately 1 that is 10 times higher than 0.1

of a general wind turbine [18]. It is reported that the blade number has influences on the
63
 Chapter 4 Proposed Configuration of a Novel Wind Turbine

power output and starting capability [57]. The effects of the blade number on

aerodynamics of the proposed wind turbine are studied in following chapters.

(3) Aerofoil. The camber of the blade aerofoil takes 44% of the chord length. The percentage

camber of this aerofoil is much larger than that 0% - 6% of general blade aerofoils. The

ratio between the chord and the blade radius is 0.33 that is also higher than 0.04 - 0.2 of a

general blade [18].

(4) Aspect ratio. The aspect ratio is defined as the blade length to the blade chord. The aspect

ratio of the proposed wind turbine is approximately 3 that is much smaller than 5 - 10 of a

general HAWT [18]. A blade with a smaller aspect ratio has a potential to support more

load if an identical material is used.

(5) Large hub diameter. A hub-to-tip ratio, which is a non-dimensional term to represent the

hub size, is defined as a ratio between the hub diameter and wind turbine diameter. A

conventional wind turbine always has a hub-to-tip ratio of approximately 0.1 that has a

tiny influence on the performance of a conventional wind turbine [18]. But this new wind

turbine has a large hub-to-tip ratio of 0.45. It is suggested that the hub, as a component of

the flow passing passage, can affect the properties of the passing flow.

(6) Working inside a shroud. The conventional wind turbines always work in an open

environment. But the proposed wind turbine operates inside a shroud of the omni-flow

system that can affect the speed and direction of the passing flow.

4.4 Summary
The configuration of a novel wind turbine for the omni-flow wind energy system is proposed

in this chapter. The proposed wind turbine is based on the impulse turbine technology that has

not been utilised in the wind turbine field. This novel wind turbine has different geometrical

features compared with existing wind turbines, such as the blade profile, the number of blades,
64
 Chapter 4 Proposed Configuration of a Novel Wind Turbine

guide vanes and the size of the hub. Aerodynamics of this proposed wind turbine is studied in

following chapters.

65
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

Chapter 3 has reviewed features of the CFD simulation and wind tunnel testing. This chapter

presents aerodynamic analyses of the novel wind turbine by two methods. The test rig in wind

tunnel and the setup of CFD simulation are also presented. The main focus of the chapter is

on aerodynamic features of the proposed wind turbine such as power output, torque and

starting capability, especially on the effect of the stator on flow characteristics.

5.1 Test Rig in Wind Tunnel Testing

5.1.1 Wind Tunnel and Test Rig

Experimental results obtained from wind tunnel testing have been used to validate numerical

approach for the proposed wind turbine. All the wind tunnel tests were carried out in a closed

return wind tunnel as shown in Figure 5.1 in the School of Engineering and Technology at the

University of Hertfordshire. The quality of air flow inside a closed return wind tunnel is easy

to be controlled [149]. The wind tunnel has a closed test section of 1.14 m (width) × 0.84 m

(height), with a maximum operational flow velocity of 25 m/s. The flow uniformity is greater

than 99% and the turbulence intensity is less than 1%.

Figure 5.1: The closed return wind tunnel used in this project.
66
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

In the tests, a prototype of the wind turbine installed on the test rig was positioned at the

centre of the test section. Dimensions of the wind turbine have been given in Chapter 4. As

shown in Figure 5.2, the prototype has 20 guide vanes and 20 blades. The wind turbine

diameter is 300 mm and the hub diameter is 135 mm. Both the stator and the rotor have the

same hub diameter. The stator and the rotor were produced by a Rapid Prototyping machine

(Stratsys Dimension 3D Printer) with acrylonitrile butadiene styrene (ABS) plastic to have

accurate geometries. Blades, guide vanes and hubs were produced separately and then

assembled together. This wind turbine prototype was installed inside a 200 mm long

cylindrical chamber. Figure 5.2(a) shows the stator inside the cylindrical chamber. The guide

vanes were fixed to the chamber wall. Therefore, the stator was stationary when the wind

turbine rotated. There was a clearance of 2 mm between the blade tip and chamber wall. This

clearance is useful in the safe installation of the rotor inside the chamber. The 2 mm clearance

is approximately 1% of the rotor radius, which has a negligible influence on the performance

of an impulse turbine [88].

(a) (b)

Figure 5.2: Wind turbine prototype: (a) stator inside a cylindrical chamber; (b) rotor
with blades.
67
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

Figure 5.3: Schematic for the measurement system of the wind tunnel testing.

Figure 5.4: Test rig inside the wind tunnel test section.

The test rig was used to measure the torque and rotational speed of the proposed wind turbine.

Figure 5.3 shows a schematic view of the measurement system and Figure 5.4 shows the test

rig inside the test section. The review about wind tunnel testing in Chapter 3 has shown that

this type of test rigs is typical and has been used for various wind turbine evaluations. As

68
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

shown in Figure 5.3, this test rig consists of a torque transducer, a DC motor, a power and

supports. The torque transducer can measure the torque generated from the wind turbine. The

torque transducer (Datum Electronics M420) has a test range from 0 to 10 Nm with an

accuracy of 0.1% and maximum rotational speed of 5000 rpm. 100 sample data can be

recorded in one second by this transducer. In the tests, data obtained from the transducer were

collected by an interface (Datum Electronics, Part No. 400150) and a monitor software

package TorqueLog (Datum Electronics, Version 1.8). A DC motor (Crouzet 8286201)

applied a braking load on the wind turbine prototype to control rotational speeds. The load

was adjusted by varying the current in the DC motor. The current for the DC motor came

from a DC Power Supply (ISO-TECH IPS1820D, output current range: 0-20A). Two pairs of

spur polyacetal gears (Davall MA08-30/90) were used in the tests. They have the same

transmission ratio of 3:1 and the module number of 0.8. One pair was used to connect the

wind turbine and the torque transducer to transmit a torque. Another pair was used to connect

the torque transducer and the DC motor so that the load from the DC motor could be applied

to the wind turbine for controlling the rotational speeds. The shaft of the wind turbine was

constrained by house bearings. Both gears and bearings could cause a power loss, which will

be discussed in Section 5.1.2.

Flow velocity in the wind tunnel was measured by a Pitot tube and a pressure meter

(AIRFLOW-PVM100). The size of the Pitot tube used in this study is very small as shown in

Figure 5.5, thus influence on the surrounding air flow can be ignored [15]. This Pitot tube was

handheld to measure the flow velocity during the wind tunnel testing.

69
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

Pitot tube

(a) (b)

Figure 5.5: Flow measurement: (a) tiny Pitot tube; (b) pressure meter

5.1.2 Blockage Correction and Error Analysis

As discussed in Chapter 3, when the wind turbine was tested in the closed test section, a

correction for blockage effect should be introduced. Based on Eq. (3.9), the correction can be

calculated as

1 π × 0.1622 + 0.04
𝐶𝐵 = = 0.032
4 1.14 × 0.84

where π × 0.1622 m2 represents the frontal area of the wind turbine and 0.04 m2 is the frontal

area of supports. Therefore, a flow velocity inside the test section can be corrected as

𝑈𝑐𝑜𝑟𝑟𝑒𝑐𝑡𝑒𝑑 = 1.032𝑈𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 (5.1)

The flow velocity inside the wind tunnel test section was calibrated by the tiny Pitot tube.

Error analysis of the tiny Pitot tube has been carried out by Zhang [15]. The average error in

measuring flow velocity was 0.58%.

In wind tunnel testing, due to the friction between bearings and the shaft, a resistive torque

was generated in the house bearing, which caused power loss. The resistive torque was

measured before wind tunnel testing. As shown in Figure 5.4, a motor (Crouzet 8286201)

with a known power-speed relationship produced an applied power (𝑃𝑎 ) at one end of the

70
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

shaft. The shaft was supported by the house bearing that was used in wind tunnel testing. The

transmitted power (𝑃𝑡 ) was tested on another end of the shaft. The power loss can be obtained

by comparing the applied power and the transmitted power. The resistive torque, 𝑇𝑟 , is

calculated by
𝑃𝑎 −𝑃𝑡
𝑇𝑟 = (5.2)
𝜔

where 𝜔 is the rotational speed of the rotor. The error (e) is calculated as follows [150]:

𝑃𝑚𝑎𝑥 −𝑃𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑃𝑚𝑖𝑛 −𝑃𝑎𝑣𝑒𝑟𝑎𝑔𝑒

𝑒=| | × 100% or 𝑒 = − | | × 100% (5.3)
𝑃𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑃𝑎𝑣𝑒𝑟𝑎𝑔𝑒

where 𝑃𝑚𝑎𝑥 is the power with the maximum resistive torque, 𝑃𝑚𝑖𝑛 is the power with the

minimum resistive torque , and 𝑃𝑎𝑣𝑒𝑟𝑎𝑔𝑒 is the power from the wind turbine with the average

resistive torque.

The minimum, maximum and average values of the resistive torque and errors are listed in

Table 5.1. The power coefficient in Table 5.1 is based on the flow velocity of 8.2 m/s. The

average resistive torque caused by the house bearing is in a range from 0.007 Nm to 0.0163

Nm. It can be seen that the maximum error in CP is ±2% under high rotational speeds and the

average error is ±1% over a range of rotational speeds from 0 rpm to 600 rpm.

Table 5.1: Errors caused by the house bearing on power coefficient.

Rotational speed, rpm Min 𝑇𝑟 Max 𝑇𝑟 Average 𝑇𝑟 Error in 𝐶𝑃

0-150 0.0056 0.0084 0.007 ±0.5%

150-300 0.009 0.0118 0.01 ±0.8%

300-450 0.013 0.0147 0.0138 ±0.9%

450-600 0.0155 0.017 0.0163 ±2%

Average ±1%

71
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

In wind tunnel testing, the test flow velocity is in a range of 3.5 m/s - 8.2 m/s that represents a

typical range of wind velocities in urban areas. With an increase of the flow velocity from 3.5

m/s to 8.2 m/s, the growth rate of the resistive torque is lower than the growth rate of the

produced torque. Hence, the error of resistive torque decreases with the increase of the flow

velocity. As shown in Table 5.2, the average error decreases from ±11.36% to ±1% when the

wind speed increases from 3.5 m/s to 8.2 m/s. The error at the flow velocity of 8.2 m/s is the

minimum one in Table 5.2, therefore the velocity of 8.2 m/s is chosen in subsequent

aerodynamic analyses.

Table 5.2: Errors caused by the house bearing on power coefficient under different flow
velocities.

Flow velocity, m/s 3.5 4.2 5 6.1 7.3 8.2

Average error in 𝐶𝑃 ±11.36% ±5.33% ±5.42% ±3.25% ±1.56% ±1%

Magnitude of the power loss in transmission between gears can be represented by the

transmission efficiency, 𝜂. The transmission efficiency is calculated by comparing the applied

power, 𝑃𝑎 , and the transmitted power through gears, 𝑃𝑡 :

𝑃𝑎
𝜂= (5.4)
𝑃𝑡

Table 5.3: Errors caused by gears on power coefficient.

Rotational speed, rpm Max 𝜂 Min 𝜂 Average 𝜂 Error in 𝐶𝑃

0 – 300 0.92 0.88 0.9 ±1%

300-600 0.92 0.88 0.9 ±2.2%

Average ±1.6%

Table 5.3 lists errors and the maximum, minimum and average values of efficiency of the

gears under different rotational speeds. It can be seen that the maximum error is 2.2% under
72
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

the high rotational speeds. During the rotational speed range of 0 rpm - 600 rpm, the average

error is 1.6%.

The propagation of all errors in measured and derived variables has influences on estimating

power coefficients of the wind turbine prototype in wind tunnel testing. Errors in flow speed,

resistive torque and transmission efficiency are considered together in analysing experimental

results. Considering the averaged error of ±0.58% in testing flow speed, ±1% caused by

bearings and ±1.6% caused by gears, the maximum error in the overall power coefficient is

calculated as

Total error = 0.58% + 1% + 1.6% = 3.18% (5.5)

In this section, the test rig for wind tunnel testing in this project has been introduced. The test

rig is used to measure the flow velocity, rotational speed and torque of the proposed wind

turbine. Errors in experiments are not unavoidable and have been analysed. The calculated

overall error of the test is 3.2% which is not significant and can be accepted.

5.2 CFD Modelling of the Proposed Wind Turbine

This section describes the setup of the CFD modelling including the computational domain,

mesh strategy, boundary conditions and physical models. A three-dimensional (3D) model of

the proposed wind turbine with a computational domain is created by a commercial CFD

software package STAR-CCM+. While a two-dimensional (2D) model can save memory and

CPU time, models are 3D in the study due to the fact that the flow development in the blade

spanwise direction cannot be included in 2D simulations.

73
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

5.2.1 Computational Domain and Boundary Conditions

The computational domain for the proposed wind turbine model is shown in Figure 5.6. The

domain is 16 times as long as the length (L) of the chamber in the axial direction. Both the

domain width and height are 20 times as long as the rotor diameter (D) of the wind turbine.

The domain is large enough to reduce the blockage effect of the domain walls [65]. In

simulations, the inlet boundary condition was set as a uniform velocity inlet and the

downstream boundary was set as a pressure outlet. Surrounding walls of the domain were

specified as slip. Surfaces of the wind turbine and chamber were set as non-slip.

Figure 5.6: Schematic view of the computational domain setup in simulations.

5.2.2 Mesh Setup and Mesh Independence Test

Polyhedral cells were generated for the complex geometry of the proposed wind turbine. The

polyhedral meshing can provide a balanced solution for complex mesh generation problems.

A mesh independence test was conducted in a cell number range from 3,300,000 to more than

7,000,000. Results with different cell numbers are shown in Figure 5.7. When the cell number

is increased from 3,300,000 to 5,800,000, there are significant differences about results and
74
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

computational time. Figure 5.7 also shows that when the cell number is increased from

5,800,000 to 7,180,000, the torque difference is less than 0.1%. But the computational time of

5,800,000 cells is 19% less than that of 6,800,000 cells and 27% less than that of 7,180,000

cells. Considered the balance of mesh independency and computational time, approximately

5,800,000 cells were produced in simulations in this project.

0.134

0.133

0.132
Torque, Nm

0.131

0.13

0.129

0.128

0.127

0.126
3 4 5 6 7 8
Cell number (×106)

Figure 5.7: Torque values under different cell numbers.

The cell size on a surface can be controlled by a minimum size and a target size. The

minimum and target sizes of all the surfaces under the cell number of 5,800,000 are shown in

Table 5.4. Figure 5.8 shows the polyhedral cells employed in volume meshing. Surface

curvature was set as 36 to create 36 cells around a 360o cylindrical surface. The surface

growth rate was set as 1.5 so that an edge of a triangle was 1.5 times its neighbour length

when expanding from a shorter edge length to a longer one [151]. The value of 1.5 can have a

good balance of the mesh quality and the total number of cells.

75
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

Table 5.4: Cell size control on different surfaces.

Surface Min size (mm) Target size (mm)

Domain wall 50 100

Chamber surface 2 5

Stator hub 1 3

Guide vane surfaces 1 1.5

Edges of guide vanes 0.5 1

Rotor hub 3 5

Blade surfaces 0.5 1

Edges of blades 0.4 0.8

Blade tips 0.4 0.6

Interfaces 1.5 2.5

(a) (b)

Figure 5.8: Meshed model: (a) a section view in the flow direction; (b) an axial section
view of the wind turbine model.

A boundary layer close to surfaces has complex aerodynamic characteristics caused by the

viscosity [152]. In the boundary layer the flow velocity has a rapid change. Flow separation is
76
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

also related to the boundary layer, leading to wake formation [152, 153]. To capture the

boundary layer correctly, thin prism layers were generated near surfaces. The mesh strategy

with thin layers has been employed on studies of wind turbines [154].

The distance from the first layer cell to the body surface can be estimated by a dimensionless

wall distance, 𝑦 + , which is defined as [99]

𝑦𝑢𝑡
𝑦+ = (5.6)
𝜐

where 𝑦 is the distance of the first grid layer from the model surface, 𝑢𝑡 is the friction

velocity and 𝜐 is the kinematic viscosity.

Values of 𝑦 + represent three boundary layers close to the model surface: viscous sub-layer,

log-law layer and outer layer [96]. The viscous sub-layer contacts with the model surface and

is extremely thin ( 𝑦 + < 5). Flow properties in this layer are dominated by the viscous effect.

In the sub-layer the shear stress is approximately constant and equals to the wall shear stress,

𝜏𝑤 . Outside the viscous sub-layer (30 < 𝑦 + < 500) there is a log-law layer where both viscous

and turbulent effects are important. When the value of 𝑦 + is greater than 500, the outer layer

exists where turbulent stresses dominate in this region.

It is known that [97]

𝜏
𝑢𝑡 = √ 𝜌𝑤 (5.7)

and [155]

𝜌𝑈 2
𝜏𝑤 = ̅̅̅
𝐶𝑓 ∙ (5.8)
2

where 𝑢𝑡 is the friction velocity, ̅̅̅

𝐶𝑓 is the friction coefficient and 𝑈 is the free stream

velocity.

77
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

The friction coefficient can be estimated from a correlation [156]:

̅̅̅̅
𝐶𝑓 0.036
= 𝑅𝑒 1⁄5 (5.9)
2

In order to obtain 𝑦 + < 5, the distance of the first layer from the blade was set as 0.15 mm.

Eight layers were generated with a growth rate of 1.5. Generated cell layers around blades

and guide vanes are shown in Figure 5.9.

(a) (b)

Figure 5.9: Thin layers near surfaces: (a) cell layers near the blade leading edge; (b) cell
layers near the guide vane edge.

5.2.3 Physical Conditions

Wind has low speeds in urban areas and therefore was assumed as incompressible in

simulations [33, 34]. The rotor axis of the wind turbine was parallel to the wind direction and

the flow properties in simulations were treated as constant in one revolution. Hence, steady

solutions were used in this project. Segregated solver was chosen in the simulation as it

required less memory and CPU time than the coupled solver [157].

In order to simulate the flow rotation, an individual rotating flow region for the rotor was

created. The rotating region was separated from the whole free stream region by the upstream

and downstream interfaces. Figure 5.10 shows the interfaces used in simulations. The

78
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

interfaces belonged to both the rotating region and the free stream region, and maintained the

physics continuity between these two regions. The flow in the rotating region was computed

in a Moving Reference Frame (MRF). The flow in the remaining region was computed in an

inertial frame of reference. Investigations of impulse turbines and wind turbines with the

MRF have produced results with good accuracy [99, 158].

Figure 5.10: Positions of upstream interface and downstream interface in simulations.

A two-layer approach was employed on the k-𝜀 turbulence models in simulations. The two-

layer approach divides the flow into two layers. In the layer closer to surfaces, the turbulent

dissipation rate, k, and the turbulent viscosity, 𝜀 are specified as functions of wall distances.

Far from the wall, the transport equations of turbulence models are calculated. With the aid of

the two-layer approach, the k-𝜀 models have less mesh dependence and more numerical

stability [96].

The review about turbulence models in Section 3.1.3 indicates that four turbulence models,

namely the standard k-𝜀, realisable k-𝜀, standard k-𝜔 and SST k-𝜔 model, all have been used

for investigations of wind turbines. These four turbulent models were employed in modelling

79
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

the performance of the proposed wind turbine. The suitable turbulence model for the

proposed wind turbine can be identified through the experimental validation.

5.3 Aerodynamic Analysis of the Proposed Wind Turbine

This section presents the aerodynamic analyses of the proposed wind turbine model analysed

by both the wind tunnel testing and the CFD simulations. The computational wind turbine

model and the prototype for wind tunnel testing have the same dimensions as presented in

Chapter 4. The power coefficient, torque coefficient and thrust coefficient are represented

quantitatively to illustrate the power output, torque and thrust of the wind turbine.

The power coefficient (𝐶𝑃 ) can evaluate the capability of power output generation for a wind

turbine and is defined as the ratio of the shaft power output from a wind turbine to the power

available from the wind [18]:

𝑃
𝐶𝑃 = 0.5𝜌𝑈 3 𝜋𝑅2 (5.10)

where P is the power output from the wind turbine, 𝜌 is the air density, U is the wind

velocity and R is the blade radius.

Torque coefficient (𝐶𝑇 ) has been used to assess the mechanical torque generated by a wind

turbine. The wind turbine with high 𝐶𝑇 can start and work under low velocity wind [52].

Torque coefficient is defined as [57]:

𝐶𝑃 T
𝐶𝑇 = = 0.5𝜌𝑈 2 𝜋𝑅3 (5.11)
𝜆

where T is torque generated from the wind turbine.

A dimensionless parameter, thrust coefficient (CN), can be used to estimate the thrust of a

wind turbine and is determined as follows [19]:

80
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

𝐹
𝐶𝑁 = 0.5𝜌𝑈𝑁2 𝜋𝑅2 (5.12)

In aerodynamic analyses of wind turbines, wind velocity is normally represented by a

dimensionless term, tip speed ratio (λ), as follows [18]:

𝜔𝑅
𝜆= (5.13)
𝑈

where 𝜔 is the angular speed of a rotor.

5.3.1 Validation of Numerical Results with Experimental Results

Chapter 3 has reviewed features of turbulence models and their applications. To simulate the

rotating flow around this wind turbine, the realisable 𝑘-𝜀 model may be better on prediction

of flow rotation than the standard 𝑘-𝜀 model [117, 118]. When simulating the air flow around

the blade, flow separation may appear near the blade surface that can affect the accuracy of

results [152, 153]. As discussed in Section 3.1.3, the 𝑘-ω model family has an advantage in

prediction of properties of the flow near the surface [122, 123]. Based on the literature review,

it is difficult to draw a conclusion on which turbulence model is more suitable for this

proposed wind turbine. In order to identify the most suitable numerical configuration, the

performances of four turbulence models are compared and validated with experimental results.

Figure 5.11 shows a comparison about power coefficient between numerical results and

experimental results. Based upon error analyses in Section 5.1.2, the free stream velocity is

8.2 m/s in simulations. The error bar on experimental data represents a scattering of ±3.18%.

It can be seen that the numerical results from all four turbulence models have a similar trend

of the CP –𝜆curves to that of the experimental results. Figure 5.11 also shows that the

maximum power coefficient is obtained at 𝜆 = 0.66. It can be seen that except for the standard

𝑘-𝜀 model, numerical results from the other three models agree well with the experimental

results.
81
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

0.2
Experiment
Standard k-ε
Realisable k-ε
0.18
Standard k-ω
SST k-ω
Power coefficient, CP

0.16

0.14

0.12

0.1
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Tip speed ratio, λ

Figure 5.11: Comparison of power coefficient between numerical results and

experimental data at U = 8.2m/s.

As shown in Figure 5.11, values of CP from the standard 𝑘-𝜀 model are lower than those from

experiments, which is caused by the weakness of this model in predication of flow rotation.

The eddy viscosity formulation in the standard 𝑘-𝜀 model cannot describe the rotating flow

well [117, 159]. Compared with the standard 𝑘-𝜀 model, the realisable 𝑘-𝜀 model has better

capacity in predictions of flow rotation [117]. The eddy viscosity formulation of the realisable

𝑘-𝜀 model consists of a mean rotation rate, which enables the model to simulate the rotating

flow [118]. Figure 5.11 also shows that two 𝑘-ω models are also better than the standard 𝑘-𝜀

model for this wind turbine. This can be because the 𝑘-ω model is integrated directly to the

low Reynolds number flow near walls without the aid of wall-damping functions while the

standard 𝑘-𝜀 model has to use functions in solution of the flow near walls [112]. Moreover,

the SST 𝑘-ω model is good at predicting the flow with adverse pressure gradients by avoiding

over estimation of the turbulence length scale in turbulent properties [123]. It is worth nothing

82
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

that results from two 𝑘-ω turbulence models are always higher than the experimental data,

which may be linked to the weakness in capturing effects of system rotation [160]. Based on

the results shown in Figure 5.11, the realisable 𝑘-𝜀 model has been employed in analyses of

the proposed wind turbine.

Simulations with four turbulence models have predicted the pressure distribution on the blade

aerofoil. Figure 5.12 shows the pressure distribution on a blade section selected at r = 140

mm where the blade section has the largest contribution to torque. It can be seen that four

turbulence models have similar pressure distributions, such as the positions of the maximum

pressure and the minimum pressure on the aerofoil. However, the pressure values are

different among the four turbulence models.

Max Min Min

Max

(a) Standard 𝑘-𝜀 (b) Realisable 𝑘-𝜀

Max Min Max Min

(c) Standard 𝑘-ω (d) SST 𝑘-ω

Figure 5.12: Pressure distributions on the aerofoil predicted by four turbulence models
at r = 140 mm and λ = 0.76.
83
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

As shown in Figure 5.12, there is a pressure increment along the aerofoil surface from the

minimum pressure position towards the trailing ledge, which causes an adverse pressure

gradient. The adverse pressure gradient results in the flow separation near the trailing edge of

the aerofoil. The flow separation on the blade was predicted by all turbulence models in

simulations.

Pressure coefficient, a dimensionless term, represents the magnitude of pressure on the blade

aerofoil. The pressure coefficient is calculated with Eq. (2.4) as given in Section 2.1.3. Figure

5.13(a) shows pressure coefficient distributions on the upper and lower surfaces of the blade

aerofoil at r = 140 mm. The upper surface and the lower surface are separated by the leading

edge and the trailing edge. Figure 5.13 shows that at λ = 0.76, the stagnation point just occurs

on the leading edge where pressure coefficient has the maximum value. The enclosed areas of

the pressure coefficient curves represent the magnitude of the normal force acting at a blade

section. As shown in Figure 5.13(a), in the x/c range of 0 - 0.15 corresponding to the leading

edge region of the aerofoil, the pressure coefficient on the upper surface is higher than that on

the lower surface, which leads to a negative enclosed area (marked by −) representing a

negative normal force. The negative normal force can hinder the rotation of a wind turbine. In

the area marked by +, the pressure coefficient on the upper surface is lower than that on the

lower surface, which provides a positive normal force that contributes to power generation.

As shown in Figure 5.13(a), the positive area is much larger than the negative area. Therefore,

the positive force is dominant and this blade section has contributed to the torque and power

generation.

Figure 5.13(b) shows the comparison of pressure coefficients among the four turbulence

models. It can be seen that the standard 𝑘-𝜀 model has the smallest positive area (+) for

84
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

positive normal force and the largest negative area (-) for negative normal force. As a result,

the standard 𝑘-𝜀 model has produced the smallest power coefficient as shown in Figure 5.11.

0.6
Pressure coefficient
Pressure coefficient on
on the
the upper
upper aerofoil
aerofoil surface
surface
0.5
Pressure coefficient on the lower aerofoil surface
Pressure
Pressure coefficient
coefficient on
on the
the lower
lower aerofoil
aerofoil surface
surface
0.4
Pressure coefficient on the lower aerofoil surface
0.3
Pressure coefficient

0.2

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.1

-0.2

-0.3

-0.4
Position, x/c
(a)
0.6
Standard k-ε
0.5 Realisable k-ε
0.4 Standard k-ω
SST k-ω
0.3
Pressure coefficient

0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.1
-0.2
-0.3
-0.4
-0.5
Position, x/c
(b)

Figure 5.13: Pressure coefficients on the blade aerofoil at r = 140 mm and λ = 0.76: (a)
by the standard k-𝜺 model; (b) by four turbulence models.
85
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

Figure 5.14 shows the contributions of different blade sections to the normal forces at λ =

0.66 where the wind turbine has the maximum power coefficient. It can be seen that the

normal force per unit length (dF’) achieves the maximum value at r = 135 mm. The blade

sections in the range of r = 70 mm – 80 mm are close to the hub and have negative values of

dF’ as the pressure on the upper aerofoil surface is higher than that on the lower surface.

0.8
Normal force per unit length, dF'

0.6

0.4

0.2

0
70 80 90 100 110 120 130 140

-0.2
Local radius r, mm

Figure 5.14: Contributions of blade sections to the normal force at 𝝀 = 0.66 by the
realisable k-𝜺 model.

Contributions of blade sections to torque can be obtained by multiplying the normal force per

unit length (dF’) by local radius (r). Figure 5.15 shows values of dT along the blade. It can be

seen that dT rises steadily against radius, r, and has a peak value at r = 140 mm. Hence, the

blade section at r = 140 mm provides the largest contribution to torque and power generation.

That is why the analyses about pressure and pressure coefficient were conducted on the blade

section at r = 140 mm as shown in Figures 5.12 – 5.13.

86
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

0.14

0.12

0.1
Torque per unit length, dT

0.08

0.06

0.04

0.02

0
70 80 90 100 110 120 130 140
-0.02
Radius r, mm

Figure 5.15: Contributions of blade sections to torque at λ = 0.66 by the realisable k-𝜺
model.

5.3.2 Power Output and Power Coefficient

In wind tunnel testing, power output of the wind turbine prototype was obtained under a flow

velocity range of 3.5 m/s - 8.2 m/s. This velocity range represents the typical wind condition

in urban areas. The relationships between the power coefficient (𝐶𝑃 ) and the tip speed ratio (λ)

under different velocities are shown in Figure 5.16. The vertical bars represent the scattering

of the experimental results. It can be seen that the power coefficient increases with the tip

speed ratio to a maximum value at the middle tip speed ratio, then the power coefficient

decreases. A curve fit in Figure 5.16 shows that the power coefficient increases with the wind

speed. The maximum power coefficient is 0.174 at the wind speed of 8.2 m/s. The tip speed

ratio corresponding to the maximum power coefficient is λ = 0.66.

87
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

0.2
y = -0.8562x2 + 1.2239x - 0.2579
0.18 3.5 m/s
4.2 m/s
0.16
5 m/s
0.14 6.1 m/s
Power coefficient, CP

0.12 7.3 m/s

8.2 m/s
0.1

0.08

0.06

0.04

0.02

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Tip speed ratio, λ

Figure 5.16: Distributions of experimental power coefficient of the proposed wind

turbine under different wind velocities.

Figure 5.17 shows the relationships between the power output and rotational speed under

different flow velocities. The curve fit shows that higher wind velocity results in greater

power output. The wind turbine can produce power up to 4.5 W at the flow velocity of 8.2

m/s and the rotational speed corresponding to the maximum power output is approximately

350 rpm.

88
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

5
y = 4E-05x2 + 0.0013x - 0.3574 3.5 m/s
4.5
4.2 m/s
4 5 m/s
6.1 m/s
3.5
7.3 m/s
3 8.2 m/s
Power, P

2.5

1.5

0.5

0
0 100 200 300 400 500 600 700
Rotational speed, rpm

Figure 5.17: Distributions of experimental power output of the proposed wind turbine
under various wind velocities.

Figure 5.18 shows the pressure coefficient distributions obtained from simulations in order to

analyse aerodynamic performances of the proposed wind turbine. With different values of λ,

positions of the stagnation point are also different. Pressure coefficient curves under different

tip speed ratios (λ) are compared. The enclosed area marked by + represents the positive

normal force and has a steady decline when λ is increased. Figure 5.18(d-f) shows that, at λ =

0.66 and beyond, the area (marked by − ) represents the negative normal force and is

increased significantly with an increment of λ. When λ exceeds 0.66, the increased negative

normal force is the major reason for the low power coefficients shown in Figure 5.16. Figure

5.18 also shows that areas (marked by −) occur in the x/c range of 0 – 0.4 that represents the

leading edge region of the aerofoil. It means that the negative normal force is always

produced in the leading edge region of the aerofoil.

89
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

1.8 0.8
1.4 0.6
Upper
Up surface
Surface
Pressure coefficient

1
0.4

Pressure coefficient
Up Surface
Upper surface
0.6 Lower
Low surface
Surface
0.2 Low Surface
Lower surface
0.2
0
-0.2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.2
-0.6
-1 -0.4

-1.4 -0.6
Position, x/c Position, x/c
(a) 𝜆 = 0 (CP = 0) (b) 𝜆 = 0.19 (CP = 0.095)
0.6 Upper
Up surface
Surface 0.6
Up Surface
Upper surface
Lower
Low surface
Surface 0.4
0.4 Lower
Low surface
Surface
Pressure coefficient

Pressure coefficient

0.2 0.2

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.2 -0.2

-0.4 -0.4

-0.6 -0.6
Position, x/c Position, x/c
(c) 𝜆 = 0.47 (CP = 0.164) (d) 𝜆 = 0.66 (CP = 0.176)
0.8 0.8
Up Surface
Upper surface Upper surface
Up Surface
0.6 0.6
Lower
Low surface
Surface Lower surface
Low Surface
Pressure coefficient
Pressure coefficient

0.4 0.4

0.2 0.2

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.2 -0.2

-0.4 -0.4
Position, x/c Position, x/c
(e) 𝜆 = 0.94 (CP = 0.122) (f) 𝜆 = 1.13 (CP = 0.053)

Figure 5.18: Pressure coefficient distributions on the blade section at r = 140 mm under
the flow speed of 8.2 m/s.

90
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

5.3.3 Torque and Starting Wind Velocity

Figure 5.19 shows the torque coefficient distributions under different flow velocities. In wind

tunnel testing, when the wind turbine was stationary (λ = 0), the wind turbine had the

maximum torque coefficient. As shown in Figure 5.19, torque coefficient has the maximum

value of around 0.7 at λ = 0. The maximum torque coefficient of 0.7 is greater than that of

some HAWTs [18]. Results also show that the values of torque coefficient under different

flow velocities are close when the tip speed ratio is below 0.2. In the range of high tip speed

ratios (λ > 0.2), higher wind velocity can result in larger torque coefficient.

0.8
3.5 m/s
0.7 4.2 m/s
5 m/s
0.6 6.1 m/s
Torque coefficient, CT

7.3 m/s
0.5
8.2 m/s
0.4

0.3

0.2

0.1

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Tip speed ratio, 𝜆

Figure 5.19: Distributions of experimental torque coefficient of the proposed wind

turbine under various wind velocities.

A large torque coefficient can allow the wind turbine to start up under low-velocity wind [52].

The starting capability of the wind turbine was tested in the wind tunnel. Figure 5.20 shows

the relationship between the starting wind velocity and the maximum rotational speed of this
91
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

wind turbine. There is a three-order regression analysis in Figure 5.20. The regression

analysis shows that the maximum rotational speed rises quickly with wind speeds. The high

wind speed causes high Reynolds number that can improve aerodynamic performances of the

blades of this wind turbine [49, 50]. It can also be seen that this wind turbine can operate at

the lowest wind velocity of 1.6 m/s. The wind velocity of 1.6 m/s is in the low end of the

range of wind velocities in an urban environment. Hence, the proposed wind turbine has a

good starting capability and can capture wind energy in an urban environment.

800
Max Rotor rotational speed, rpm

700
y = -0.0939x3 + 8.7798x2 + 7.9342x - 24.106
600

500

400

300

200

100

0
0 1 2 3 4 5 6 7 8 9 10
Starting wind velocity, m/s

Figure 5.20: Starting wind velocities of the proposed wind turbine in wind tunnel tests.

5.4 Effects of the Stator on Passing Flow

As discussed in Section 4.3, the stator with guide vanes is a part of the presented wind turbine,

based upon the impulse turbine technology. The air flow passes the stator firstly and then the

rotor. Properties of the passing flow are affected by the stator. Hence, influences of the stator

on flow properties are analysed in this section.

92
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

5.4.1 Effects of the Stator on Flow Velocity

In order to study the effects of the stator on the flow characteristics, the rotor was not

considered in simulations and wind tunnel testing in this section. Figure 5.21(a) shows the

stator model inside the cylindrical chamber. Figure 5.21(b) shows selected positions for

testing flow velocity near the rear of the guide vane. The test position is defined by its local

radius from the centre of the stator. In the wind tunnel testing, flow velocities were measured

by a Pitot tube at these positions.

(a) (b)

Figure 5.21: Stator model in simulations: (a) stator inside the cylindrical chamber; (b)
testing positions both in experiments and simulations.

Figure 5.22 presents the wind speed ratios at the selected positions when free stream velocity

is 5 m/s. The velocity of 5 m/s is a medium value in the general range of wind speeds in urban

areas and has been used in investigations of wind turbines [70]. Studying with the speed of 5

m/s can save both the cost and time of wind tunnel tests. As shown in Figure 5.22, the results

from both the wind tunnel tests and computational simulations correlate well with each other.

The small discrepancies could be caused by the roughness of the prototype surface. It can be

seen that most wind speed ratios are greater than 1.0 except for the positions near the chamber

wall. The maximum flow velocity ratio exceeds 1.2 from r = 82.5 mm to r = 100 mm. The
93
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

wind velocity ratio exceeds 1.0 on the positions from r = 70 mm to r = 132.5 mm that

occupies 75.8% of the guide vane length. It shows that the stator can increase the velocity of

the passing flow significantly in a wide range of the local radius.

1.4

1.2

1
Wind speed ratio

0.8
CFD
0.6
Experimental
0.4

0.2

0
70 90 110 130 150
Local radius of the selected position, r, mm

Figure 5.22: Comparison of wind velocity ratios between the experimental and
numerical results at the selected positions.

It has been observed that changes of the flow characteristics are affected by the guide vane

geometry. Figure 5.23(a) shows a schematic view of the flow passage between two guide

vanes. From the entrance to the exit of the flow passage, the cross sectional area of the flow

passage decreases continuously (the entrance area of flow passage, A1 > the exit cross

sectional area of flow passage, A2) and this feature acts as a nozzle cascade. The nozzle

cascade can increase the speed of the passing flow.

Figure 5.23(b) shows the velocity and pressure of the passing flow at r = 100 mm where the

wind speed has the maximum value. These data were obtained at the test points on the central

line between two guide vanes as shown in Figure 5.23(a). Figure 5.23(b) shows that the air

94
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

flow has the highest pressure and lowest velocity at the entrance of the flow passage due to

the influence of the nozzle cascade. When the distance is less than 15 mm from the entrance

of the flow passage, the velocity and pressure change slowly. But both the velocity and

pressure have changes with large slopes from a distance of x = 15 mm to x = 30 mm. This is

caused by a sharp reduction of the cross sectional area of the flow passage after x = 15 mm.

Distributions of the flow velocity and pressure at other local radii are shown in Appendix E

and similar phenomenon can be found.

7 16
14
6
12
5 10

Pressure change, Pa
Velocity, m/s

Velocity 8
4
Pressure change 6
3 4
2 2
0
1
-2
0 -4
0 10 20 30 40
Distance from the entrance of guide vanes in the
axial direction of the turbine, x, mm

(a) (b)

Figure 5.23: Properties of the flow passing guide vanes: (a) flow passage between two
guide vanes; (b) velocity and pressure along the centre line at r = 100 mm.

5.4.2 Effects of the Stator on Flow Direction

The stator with guide vanes can not only increase the passing flow velocity, but also affect the

flow direction. The flow direction inside the omni-flow wind system with the stator has been

compared with that without the stator. Simulations in this section employed the computational

setup presented by Zhang [15]. Figure 5.24(a) shows the computational domain and Figure
95
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

5.24(b) shows the free stream direction for the omni-flow wind energy system model in

simulations. Inside the omni-flow system, the flow direction was recorded by approximately

300 probes at a height of 260 mm as shown in Figure 5.25(a). At the height of 260 mm, the

flow velocity achieved the highest speed [15]. In the study of flow direction inside the omni-

flow system with the stator, the flow direction was recorded by approximately 300 probes at a

height of 420 mm as shown in Figure 5.25(b). At the height of 420mm, the flow just leaves

the stator.

(a) (b)
Figure 5.24: Study of flow direction: (a) computational domain; (b) the direction of free
stream in simulations [15].

420 mm
260 mm

(a) (b)
Figure 5.25: Positions for recording flow directions: (a) the system model without a
stator; (b) the system model with a stator.
96
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

The flow direction is represented by an angle, φ, referred to the horizontal axis. Figure 5.26

shows the angle φ between the flow direction and the horizontal axis. Inside the omni-flow

system without the stator, the optimal flow direction is vertical so φ optimal = 90o. Figure 5.27

shows values of φ inside the omni-flow wind energy system without the stator. Results show

that values of φ are in a range from -60o to 80o, which means the optimal flow direction φ

optimal = 90o is out of the range (-60o < φ < 80o). For best performances, a conventional wind

turbine has to operate under the optimal flow direction. Therefore, conventional wind turbines

are difficult to work well inside the omni-flow wind energy system.

Figure 5.26: Angle of the flow direction, φ.

When the stator is installed inside the omni-flow system, the optimal flow direction is 20o (φ

optimal = 20o) as the guide vane has a setting angle of 20o. Figure 5.28 shows the values of φ

obtained inside the omni-flow system with the stator, which are in a range from -10o to 30o. It

can be seen that the optimal flow direction φ optimal = 20o is within the range (-10o < φ < 30o).

Therefore, the proposed wind turbine has the potential to operate well inside the omni-flow

system. Compared with the range (-60o < φ < 80o) shown in Figure 5.27, this range (-10o < φ

< 30o) affected by the stator is reduced by 71%.

This section shows the analysis of the flow direction when the free stream is at 0o to the

omni-flow system as shown in Figure 5.24(b). Results indicate that the stator can optimise the

flow direction and improve the flow quality inside the omni-flow system. When the free

97
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

stream is at 12o, 24o and 36o to the omni-flow system, values of φ are given in Appendix F. A

similar conclusion can be drawn.

100
φ optimal = 90o
80
60
40
Angle φ, degree

20
0
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
-20
-40
-60
-80
-100
Position, m

Figure 5.27: Angles of flow directions inside the omni-flow system without the stator.

100
80
60
40
Angle φ, degree

20 φ optimal = 20o

0
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
-20
-40
-60
-80
-100
Position, m

Figure 5.28: Angles of flow directions inside the omni-flow system with the stator.

98
 Chapter 5 Aerodynamic Analysis of the Novel Wind Turbine

5.5 Summary
In this chapter, aerodynamics of this wind turbine has been analysed by both the wind tunnel

testing and CFD simulations. The CFD model was validated by comparing numerical results

with experimental data. It is found that the realisable 𝑘-𝜀 turbulence model can be used to

study aerodynamics of the proposed wind turbine. Results show that the proposed wind

turbine can achieve a maximum power coefficient of approximately 0.17. It is also found that

this wind turbine has the good starting capability. In addition, the effects of the stator on the

velocity and the direction of the flow have been studied. Results indicate that the stator can

improve the flow quality inside the omni-flow system.

99
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

Chapter 6 Aerodynamic Features of the Wind Turbine under

Non-uniform Flow

As shown in Section 2.3, the flow velocity field inside the omni-flow wind energy system is

non-uniform. It is necessary to study aerodynamic features of the wind turbine under non-

uniform flow. This chapter presents the aerodynamic analyses of the proposed wind turbine

under the non-uniform flow condition by CFD simulations. The setup of simulations has been

presented in Chapter 5.

6.1 Wind Turbine Model under Non-uniform Flow

Features of the flow inside the omni-flow wind energy system have been reviewed in Section

2.3.2. A major feature is the non-uniform velocity distribution. Figure 6.1(a) shows the flow

velocity distribution at the outlet of the omni-flow system when the free-stream wind is at 0o

to the omni-flow system. As shown in Figure 6.1(b), the flow velocity distribution at the

outlet can be divided into two regions: high velocity region and low velocity region. The high

velocity region occupies 20% of the outlet area. The low velocity region occupies 80% of the

outlet area. The average velocity in the low velocity region is close to zero and much smaller

than that in the high velocity region. Hence, the flow in the low velocity region can be treated

as zero to reduce the complexities of analyses. In order to simulate these two different

velocity regions, a cover was added in front of the wind turbine model in simulations as

shown in Figure 6.2(a). The cover has a 20% entrance area corresponding to the high velocity

region. Thus, with the aid of this cover, the 20% non-uniform flow condition can be created.

When the wind is at 36o to the omni-flow system model, the high velocity region occupies 40%

of the outlet area as shown in Figures 6.3. In simulations, the cover has a 40% entrance area

that is the same as the high velocity region as shown in Figure 6.2(b).

100
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

High velocity

Low velocity

(a) (b)

Figure 6.1: 20% non-uniform flow condition: (a) wind direction; (b) velocity
distribution at the outlet of the system.

(a) (b)

Figure 6.2: Wind turbine models: (a) for 20% non-uniform flow condition; (b) for 40%
non-uniform flow condition.

High velocity

Low velocity

(a) (b)

Figure 6.3: 40% non-uniform flow condition: (a) wind direction; (b) velocity
distribution at the outlet of the system.
101
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

There are two features about the wind turbine model with an added cover. Firstly, the cover

divides the flow into the flow passing region and the non-flow passing region as shown in

Figure 6.4. Due to the pressure difference between two regions, air flow may move from the

flow passing region to the non-flow passing region. The second feature is that whereby the

partial flow is blocked in the axial direction by the cover, a high pressure can be generated in

front of the cover, which will accelerate the flow velocity at the entrance. Therefore, a

correction about the flow velocity is needed to account for this effect due to the introduction

of the cover.

Figure 6.4: Two flow regions caused by the cover ahead the wind turbine.

The correction factor γ is defined as the ratio of the averaged flow velocities between covered

and uncovered conditions:

𝑈𝑐𝑜𝑣𝑒𝑟𝑒𝑑
𝛾=𝑈 (6.1)
𝑢𝑛𝑐𝑜𝑣𝑒𝑟𝑒𝑑

where Ucovered represents the average flow velocity at the entrance with the cover and

Uuncovered is the average velocity without the cover. This ratio represents the change of flow

velocity due to the cover effect. Table 6.1 lists values of Ucovered and Uuncovered obtained at the

free stream velocity of 8.2 m/s. The correction factor is 1.56 for the 20% non-uniform flow

102
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

condition and 1.2 for the 40% non-uniform flow condition. Therefore, the corrected flow

velocity is 12.8 m/s under the 20% non-uniform flow condition and 9.8 m/s under the 40%

non-uniform flow condition. The corrected flow velocities of 12.8 m/s and 9.8 m/s have been

used in calculations of power, torque and thrust coefficients under the non-uniform flow

condition.

Table 6.1: Values of Ucovered and Uuncovered under different rotational speeds.

Rotational speed, rpm 100 rpm 200 rpm 300 rpm 400 rpm

Uuncovered 3.4 m/s 3.39 m/s 3.34 m/s 3.31 m/s

Ucovered under 20% entrance area 5.3 m/s 5.27 m/s 5.22 m/s 5.18 m/s

Average γ under 20% entrance area 1.56

Ucovered under 40% entrance area 4.12 m/s 4.1 m/s 4.07 m/s 4.03 m/s

Average γ under 40% entrance area 1.2

6.2 Power Coefficient and Torque Coefficient

Power coefficient (CP) and torque coefficient (CT) of this wind turbine under the non-uniform

flow condition are analysed in this section to evaluate its performance. As shown in Figure

6.5, the wind turbine under the non-uniform flow condition has the value of the maximum CP

in a range of 0.06 – 0.126 which is lower than the maximum CP of 0.17 under the uniform

flow condition. Figure 6.6 shows the distributions of torque coefficients under the non-

uniform flow condition. The 40% entrance area results in the higher power coefficient and

torque coefficient compared with the 20% entrance area. It can be seen that the value of the

maximum CT is in a range of 0.4 – 0.6 that is lower than 0.7 of the proposed wind turbine

under the uniform flow condition, but higher than 0.1 of conventional wind turbines [52].

103
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

0.14
Non-uniform 20%
0.12
Non-uniform 40%

0.1
y = -0.6136x2 + 0.5479x + 0.0033
Power coefficient, CP

0.08

0.06

0.04

y = -0.6453x2 + 0.3889x + 0.0002

0.02

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Tip speed ratio, λ

Figure 6.5: Distributions of power coefficient under two non-uniform flow conditions.

0.6
Non-uniform 20%
0.5
Non-uniform 40%
Torque coefficient, CT

0.4

0.3

0.2
y = -0.6396x + 0.5664

0.1

y = -0.6578x + 0.3928
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Tip speed ratio, λ

Figure 6.6: Distributions of torque coefficient under two non-uniform flow conditions.

Due to the non-uniform flow, the torque distribution on blades is also non-uniform. Figure 6.7

shows the torque value on each blade under the 20% non-uniform flow condition. It can be
104
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

seen that blades (No.3 – No.8) in the flow passing region have much greater torque values

than other blades. A blade in the flow passing region can produce larger torque than that in

the non-flow passing region. Results indicate that blades in the flow passing region have the

greater contribution to the overall torque of the wind turbine. Figure 6.7 also shows some

blades (No.10 – No.20) in the non-flow passing region generate negative torque or negligible

torque which lowers the overall performances of the wind turbine.

Figure 6.8 shows the torque value on each blade under the 40% non-uniform flow condition.

Compared with the wind turbine under the 20% non-uniform flow condition, the wind turbine

under the 40% non-uniform flow condition has more blades (No.3 – No.12) to produce large

torque as more blades are rotating in the flow passing region. Therefore, it results in the

higher power and torque coefficients as shown in Figures 6.5 – 6.6. Figure 6.8 also shows that

the blades (No.14 – No.20) produce negative torque or negligible torque as these blades are

rotating in the non-flow passing region.

It is worth noting that six blades (No.3 – No.8) can produce high positive values of torque as

shown in Figure 6.7 while only four blades of the wind turbine are directly exposed to the

flow under the 20% condition as shown in Figure 6.2(a). The reason is that the flow moves

from the flow passing region to the non-flow passing region due to the pressure difference

between two regions. A similar phenomenon can be observed in Figure 6.2(b) and Figure 6.8.

105
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

0.012
𝜆=0.06
0.01 𝜆=0.18
𝜆=0.3
0.008
𝜆=0.43
Torque, Nm

0.006 𝜆=0.55

0.004

0.002

-0.002
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Blade number

Figure 6.7: Torque generated on every blade under the 20% non-uniform flow condition.

0.012
𝜆=0.16
0.01 𝜆=0.24
𝜆=0.32
0.008 𝜆=0.48
𝜆=0.56
Torque, Nm

0.006 𝜆=0.8
𝜆=0.87
0.004

0.002

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

-0.002
Blade number

Figure 6.8: Torque generated on every blade under the 40% non-uniform flow condition.
106
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

The distribution of pressure coefficients on the aerofoil in the flow passing region is

compared with that in the non-flow passing region in Figure 6.9. No.6 blade is in the flow

passing region and No.16 is in the non-flow passing region. It can be seen that No.6 blade has

a much larger positive area about pressure coefficient than No.16 blade. Therefore, the

normal force produced by No.6 blade is larger than that by No.16 blade. Figure 6.9 indicates

that the different distributions of pressure cause different torque values on blades as shown in

Figures 6.7 – 6.8.

0.4 0.1

0.2
Pressure coefficient

0
Pressure coefficient

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 -0.1 Lower Surface
-0.2 Upper Surface
Lower
Up surface
Surface
Upper
Low surface
Surface -0.2
-0.4

-0.6 -0.3
Position, x/c Position, x/c
(a) (b)

Figure 6.9: Pressure coefficient distributions at r = 140 mm and λ = 0.3 under the 20%
non-uniform flow condition: (a) No. 6 blade; (b) No. 16 blade.

6.3 Thrust
The thrust distribution of the proposed wind turbine has been studied. Thrust is a part of the

aerodynamic force and acting in the axial direction. This study shows influences of the non-

uniform flow on the thrust distribution, which is helpful in understanding aerodynamics of the

proposed wind turbine.

Thrust on each blade under the non-uniform flow condition is shown in Figures 6.10 – 6.11.

Figure 6.10 indicates that blades (No.3 – No.8) in the flow passing region have larger thrust

107
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

values than other blades in the non-flow passing region. Thrust can generate a bending

moment on a blade. Considering small values of thrust on each blade, the generated bending

moments will however be fairly small to affect the integrity of the blade structure.

0.035
𝜆=0.06

0.03 𝜆=0.18
𝜆=0.3
0.025 𝜆=0.43
𝜆=0.55
Thrust, N

0.02

0.015

0.01

0.005

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Blade number

Figure 6.10: Thrust on every blade under the 20% non-uniform flow condition.

108
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

0.04
𝜆=0.16
0.035 𝜆=0.24
𝜆=0.32
0.03
𝜆=0.48
0.025 𝜆=0.56
Thrust, N

𝜆=0.8
0.02
𝜆=0.87

0.015

0.01

0.005

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Blade number

Figure 6.11: Thrust on every blade under the 40% non-uniform flow condition.

Results in Figures 6.10 – 6.11 indicate that thrust has a change during one operation cycle,

that is that thrust increases up to the peak value and then decreases. The frequency of the

change of thrust, f, can be calculated as

𝑁
𝑓 = 2 × 60 (6.2)

where N is the rotational speed.

Based upon results in Figure 6.5, the rotational speed at the maximum CP is 250 rpm under

the 20% non-uniform flow condition. Therefore, the frequency of the thrust change is

approximately 8 Hz. When the wind turbine operates under the 40% non-uniform flow

condition, the wind turbine works with higher rotational speeds. Hence the frequency of the

thrust change is about 10 Hz.

109
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

Considering the definition of tip speed ratio,

𝑁
×2𝜋×𝑅
𝜆 = 60 (6.3)
𝑈

the relationship between the thrust change frequency and the rotor radius can be given as

𝜆×𝑈
𝑓 = 𝜋×𝑅 (6.4)

Equation 6.4 states that the frequency of thrust change decreases with increasing the rotor

radius under a constant wind speed. When the wind turbine has a rotor radius of 1 m, the

frequency of thrust change will be decreased to 1.2 Hz under the 20% condition and 1.5 Hz

under the 40% condition as listed in Table 6.2. It can be seen that such changes of frequencies

of thrust are marginal. The maximum value of thrust on a blade for the wind turbine is also

given in Table 6.2.

Table 6.2: Predicted maximum thrust and frequency with different radii.

f at R = 0.15 f at R = Thrust at R = 0.15 Thrust at R = 1

Case at 8.2 m/s 𝜆
m 1m m m

20% entrance area 0.3 8 Hz 1.2 Hz 0.022 N 1.3 N

40% entrance area 0.47 10 Hz 1.5 Hz 0.025 N 1.1 N

6.4 Summary
Aerodynamic performances of the proposed wind turbine under the non-uniform flow

condition have been analysed in this chapter. The non-uniform flow results in poor

performances of the wind turbine. It is found that the maximum CP is in a range from 0.06 to

0.12 and that the maximum CT is in a range of 0.4 – 0.6. The larger the entrance area of the

non-uniform flow is, the higher the power coefficient and torque coefficient will be. It is

found that the torque distribution is non-uniform on blades. The blade in the passing flow
110
 Chapter 6 Aerodynamic Features of the Wind Turbine under Non-uniform Flow

region has the larger torque than that in the non-flow passing region. It is also found that the

non-uniform flow can result in a periodical change of thrust on a blade in one operation cycle.

The frequency of the thrust change is in a range from 8 Hz to 10 Hz for the proposed wind

turbine. It is however worth nothing that the frequency of the thrust change depends on the

size of the turbine. If the wind turbine employs a larger blade radius of 1 m, the frequency can

reduce to approximately 1 Hz.

111
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

Chapter 7 Influences of Geometrical Parameters on Aerodynamics

of the Wind Turbine

7.1 Introduction
Compared with conventional wind turbines, the proposed wind turbine has several exclusive

geometrical features which include the large hub-to-tip ratio, more blades, and the special

blade aerofoil and guide vanes. This chapter analyses influences of these geometrical

parameters on the power out, torque and thrust of the wind turbine by the CFD approach. The

computational setup has been presented in Chapter 5. All simulations in this chapter have an

identical free stream velocity of 8.2 m/s.

7.2 Effects of Hub-to-tip Ratios

The proposed wind turbine has a large hub based upon the impulse turbine technology. The

dimensionless hub-to-tip ratio is used to evaluate the effects of the hub size on the

aerodynamic performance of the turbine. The hub-to-tip ratio is defined as the ratio between

the hub diameter and the wind turbine diameter. Figure 7.1 shows the wind turbine models

with three hub-to-tip ratios of 0.35, 0.45 and 0.6, representing the hub diameter of 105 mm,

135 mm and 180 mm, respectively. The blade length is 95.5 mm for the hub-to-tip ratio of

0.35, 80.5 mm for 0.45, and 58 mm for 0.6. As shown in Figure 7.1, the largest hub-to-tip

ratio of 0.6 has the shortest blade and the smallest ratio of 0.35 has the longest blade. Because

the blade is the component to produce power, performances of the wind turbine will be

affected by the hub-to-tip ratio.

112
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

Hub diameter = 105 mm

(a)

Hub diameter = 135 mm

(b)

Hub diameter = 180 mm

(c)

Figure 7.1: Employed three hub-to-tip ratios and hub diameters in simulations: (a) 0.35;
(b) 0.45; (c) 0.6.

Figure 7.2 shows the distributions of the torque per unit length (dT) with three hub-to-tip

ratios at λ = 0.66 where the wind turbine has the maximum power coefficient. It is found that

for a given blade radius in the range of 90 mm – 140 mm the hub-to-tip ratio of 0.6 has the
113
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

smallest value of dT, which can result in the lowest power output and torque. Figure 7.2 also

shows that the hub-to-tip ratio of 0.35 has a similar distribution of dT to the ratio of 0.45.

However, the ratio of 0.35 has the negative dT in a larger range of r = 55 mm – 80 mm, while

the ratio of 0.45 has the negative dT in a range of r = 70 mm – 80 mm. Therefore, more

negative torque was produced when the hub-to-tip ratio was 0.35.

0.12

0.1
Torque per unit length, dT

0.08

0.06

0.6 Hub-to-tip ratio

0.04
0.45 Hub-to-tip ratio
0.35 Hub-to-tip ratio
0.02

0
50 60 70 80 90 100 110 120 130 140 150

-0.02
Blade radius r, mm

Figure 7.2: Distributions of torque per unit length on the blade with three hub-to-tip
ratios at λ = 0.66.

Figure 7.3 shows the distributions of power coefficient at three hub-to-tip ratios. It can be

seen that three CP–𝜆 curves have a similar trend. The hub-to-tip ratio of 0.6 has the lowest

power coefficient, which is caused by the lowest dT as shown in Figure 7.2. Figure 7.3

indicates that the ratio of 0.45 has the highest power coefficient. And the maximum power

coefficient of the ratio of 0.45 is 3.5% higher than that of 0.35. The reason is that the ratio of

114
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

0.45 has the smaller range of negative dT than the ratio of 0.35 as shown in Figure 7.2.

Negative dT can reduce the power out of the wind turbine.

0.2
0.6 Hub-to-tip ratio
0.18
0.45 Hub-to-tip ratio
0.16
0.35 Hub-to-tip ratio
Power coefficient, CP

0.14

0.12

0.1

0.08

0.06
0.35: y = -0.4214x2 + 0.5134x + 0.0107
0.04
0.45: y = -0.4549x2 + 0.5589x + 0.0033
0.6: y = -0.3834x2 + 0.4701x + 0.0113
0.02

0
0 0.2 0.4 Tip0.6
speed ratio,
0.8𝜆 1 1.2 1.4

Figure 7.3: Distributions of power coefficient of the proposed wind turbine with three
hub-to-tip ratios.

Distributions of torque coefficient with three hub-to-tip ratios are shown in Figure 7.4. For

the hub-to-tip ratio varying from 0.35 to 0.6, the maximum torque coefficient is in the range

of 0.5 - 0.6 which is higher than 0.1 of conventional wind turbines [52]. As the ratio of 0.6

has the lowest value of the maximum CT, the starting capability of the wind turbine with the

ratio of 0.6 would be the worst among the three ratios concerned.

115
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

0.7
0.6 Hub-to-tip ratio
0.6
0.45 Hub-to-tip ratio
0.5
Torque coefficient, CT

0.35 Hub-to-tip ratio

0.4

0.3

0.2
0.35: y = -0.4784x + 0.5683
0.45: y = -0.4659x + 0.5667
0.1
0.6: y = -0.4358x + 0.5233
0
0 0.2 0.4 0.6 0.8 1 1.2
Tip speed ratio, λ

Figure 7.4: Distributions of torque coefficient of the proposed wind turbine with three
hub-to-tip ratios.

Figure 7.5 shows that the distribution of thrust coefficient varies with the hub-to-tip ratio.

Thrust coefficient consists of a positive component from the blades and a negative component

from the rotor hub. As shown in Figure 7.5, at low tip speed ratios (λ) thrust coefficient is

negative because the negative thrust component from the rotor hub is dominant in thrust

generation. At high λ thrust coefficient is positive as the blades are dominant in thrust

generation. The hub-to-tip ratio of 0.6 has the largest negative thrust component from the hub

due to the largest hub size, and the smallest positive thrust component from blades due to the

shortest blade length. Hence, the hub-to-tip ratio of 0.6 has the lowest thrust coefficient and

the hub-to-tip ratio of 0.35 has the highest thrust coefficient as shown in Figure 7.5. The hub-

to-tip ratio of 0.45 has the middle thrust coefficient because of the medium hub size and blade

length.

When the hub-to-tip ratio is 0.35, the blades have more contribution to thrust than the rotor

hub at high λ. However, the negative thrust from hub increases faster than the positive thrust
116
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

from blades, which is shown by the concave 𝐶𝑁 -λ curve in Figure 7.5. On the other hand, the

𝐶𝑁 -λ curve with the hub-to-tip ratio of 0.6 is convex, which indicates that the negative thrust

from the hub increases slower than the positive thrust from blades. A regression analysis for

three hub-to-tip ratios indicates that 𝐶𝑁 rises with the increment of λ, which is similar to the

distribution of 𝐶𝑁 -λ of a HAWT [18]. The values of maximum 𝐶𝑁 of the proposed wind

turbine are however lower than that of a HAWT [18].

0.25
y = 0.0639x3 - 0.4545x2 + 0.7119x - 0.104
0.2

0.15
Thrust coefficient, CN

0.1
y = -0.163x2 + 0.4171x - 0.085
0.05

0
0 0.2 0.4 0.6 0.8 1 1.2
-0.05
0.6 Hub-to-tip ratio
-0.1
y = 0.0772x2 + 0.1747x - 0.166 0.45 Hub-to-tip ratio
-0.15
0.35 Hub-to-tip ratio
-0.2
Tip speed ratio, λ

Figure 7.5: Distributions of thrust coefficient of the proposed wind turbine with three
hub-to-tip ratios.

7.3 Influences of Blade Numbers

The proposed wind turbine has more blades than a conventional wind turbine. The literature

review shows that the blade number is one of most important geometrical parameters for a

wind turbine. In order to study influences of the blade number on aerodynamics of the turbine,

the wind turbine models with 10, 15, 17, 20, 25 and 30 blades have been simulated. Figure

7.6 shows distributions of power coefficient of turbines with different blade numbers. It can
117
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

be seen that the maximum power coefficient has two clear incremental steps of approximately

10% when the blade number has been increased from 15 to 17 and 17 to 20. There is however

no significant improvement of the maximum power coefficient when the blade number is 20

or over.

0.2
15 blades
0.18
17 blades
0.16 20 blades
0.14 25 blades
Power coefficient. CP

0.12 30 blades

0.1

0.08

0.06

0.04

0.02

0
0 0.2 0.4 0.6 0.8 1 1.2
Tip speed ratio, 𝜆

Figure 7.6: Distributions of torque coefficient of the proposed wind turbine with
different blade numbers.

Distributions of pressure coefficient on the aerofoil are shown in Figure 7.7 when the

numbers of blades are 15 and 30. The enclosed area (marked by +) in Figure 7.7 represents

the positive normal force generated on the blade aerofoil. A comparison of the positive areas

in Figure 7.7(a) and Figure 7.7(b) indicates that the positive normal force with the blade

number of 15 is larger than that with the blade number of 30. Meanwhile, with the blade

number of 30, a large negative area (marked by -) that represents the negative normal force

can be found in Figure 7.7(b). This is caused by the narrow space between blades which
118
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

limits the flow moving when the wind turbine has more blades. Thus the normal force with

the blade number of 15 is larger than that with 30. The power output from each individual

blade decreases with the increase of the number of blades. The turbine with 15 blades

however has the lowest power coefficient as shown in Figure 7.6 as the turbine has the least

number of blades to produce power although power generated from each individual blade is

higher.

0.5
On aerofoil upper surface
0.4
On aerofoil lower surface
0.3

0.2
Pressure coefficient

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.1

-0.2

-0.3

-0.4
Position, x/c

(a)

119
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

0.5
On aerofoil upper surface
0.4
On aerofoil lower surface
0.3

0.2
Pressure coefficient

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.1

-0.2

-0.3

-0.4
Position, x/c

(b)

Figure 7.7: Distributions of pressure coefficient on the blade aerofoil at r = 140 mm: (a)
the wind turbine with 15 blades; (b) the wind turbine with 30 blades.

Figure 7.8 shows the relationships between the torque coefficient and the tip speed ratio for

the wind turbines with different blade numbers. Values of the maximum torque coefficients

are in the range of 0.5 to 0.6 which is higher than that of a conventional wind turbine [52].

Results in Figure 7.8 show that the maximum torque coefficient are increased with more

blades. Figure 7.8 indicates that the good starting capability of the wind turbine can be

maintained with a change in the blade number.

120
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

0.7
15 blades
0.6 17 blades

20 blades
0.5
Torque coefficient, CT

25 blades
0.4
30 blades

0.3
15: y = -0.4096x + 0.4901
0.2 17: y = -0.4442x + 0.5358
20: y = -0.4659x + 0.5667
0.1 25: y = -0.4679x + 0.577
30: y = -0.4516x + 0.5684
0
0 0.2 0.4 0.6 0.8 1 1.2
Tip speed ratio, λ

Figure 7.8: Distributions of torque coefficient of the proposed wind turbine with five
blade numbers.

Relationships between the thrust coefficient and the tip speed ratio with different blade

numbers are shown in Figure 7.9. It can be seen that the thrust coefficient increases with the

number of blades. It is known that a wind turbine with a larger solidity can produce a higher

thrust coefficient [18]. The solidity of a wind turbine is defined as

𝐵 𝑐
𝜎 = 2𝜋 𝑅 (2.7)

where B is the blade number, c is the blade chord and R is the blade radius. Compared with

the solidity of 0.1 of a conventional wind turbine, this wind turbine has the large solidity of

1.08 for 15 blades, 1.2 for 17 blades, 1.38 for 20 blades, 1.67 for 25 blades and 1.96 for 30

blades. This wind turbine has the large solidity, but the obtained maximum thrust coefficient

in Figure 7.9 is smaller than that of a conventional wind turbine [18]. This is related to the

low tip speed ratios of the turbine. Thrust increases with larger tip speed ratios but the

121
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

maximum tip speed ratio of this wind turbine is one fifth of that of a conventional wind

turbine [18]. Figure 7.9 also shows how different blade numbers influence the 𝐶𝑁 -λ

relationship. It can be seen that 𝐶𝑁 rises much quicker for the wind turbines with 25 blades

and 30 blades compared with turbines of 20 blades or less.

0.35
15 blades
y = 0.1311e0.8016x
0.3 17 blades
20 blades
0.25 y = 0.0898e0.8794x
25 blades
Thrust coefficient, CN

30 blades
0.2 y = -0.137x2 + 0.3778x - 0.072

0.15
y = -0.1646x2 + 0.3634x - 0.0819
0.1

0.05
y = -0.1229x2 + 0.2699x - 0.0575

0
0 0.2 0.4 0.6 0.8 1 1.2
Tip speed ratio, λ

Figure 7.9: Distributions of thrust coefficient of the proposed wind turbine with five
blade numbers.

7.4 Effects of Aerofoil Profiles

Analyses in Chapter 5 have shown that there is a negative pressure difference on the leading

edge of the aerofoil. The negative pressure difference can result in a negative normal force

and lower power output. Hence, there is the potential to improve the power performance by

changing the leading edge profile of the aerofoil. In this section, aerodynamic performances
122
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

of the wind turbines with three aerofoils are compared. A schematic view of three aerofoils

with different leading edges is shown in Figure 7.10. Coordinates of three aerofoils can be

found in Appendix D.

Type 3

Type 2

Type 1

Figure 7.10: Schematic view of three blade aerofoils employed.

Figure 7.11 shows the power coefficient distributions of three aerofoils. It can be seen that the

aerofoil change does not affect the power coefficient distributions when the tip speed ratio is

less than 0.6. And there are slight improvements on the maximum power coefficient for all

three aerofoil profiles. However, when the tip speed ratio exceeds 0.6, the power coefficient

is increased by 12% and 25% when Type 1 is replaced by Type 2 and Type 3 aerofoil,

respectively. It means that changing the leading edge can improve the power output

performance at high tip speed ratios (λ > 0.6). Figure 7.11 also shows that Type 1 aerofoil has

a λ range of 0.4 - 0.76 where the power coefficient exceeds 0.16. When using Type 3 aerofoil,

the λ range is 0.4 - 0.9 which is 38.9 % wider. The wider λ range means that the wind turbine

can produce power in a wider range of wind speeds.

123
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

0.2
Type 1
0.18
Type 2
0.16 Type 3
0.14
Power coefficient, CP

0.12

0.1

0.08

0.06
Type 1: y = -0.4549x2 + 0.5589x + 0.0033
0.04
Type 2: y = -0.4139x2 + 0.5432x + 0.0006
0.02
Type 3: y = -0.4048x2 + 0.5453x - 0.0012
0
0 0.2 0.4 0.6 0.8 1 1.2
Tip speed ratio, λ

Figure 7.11: Distributions of power coefficient of the proposed wind turbine with three
aerofoils.

Figure 7.12 shows distributions of pressure coefficient with different aerofoils at λ = 0.76

where the difference about CP is significant. It can be seen that the enclosed area (marked by

−) representing the negative normal force decreases when Type 1 is replaced by Type 2.

There is no negative area (−) when Type 3 is used, meaning that no negative force is

generated on Type 3 aerofoil. Hence, the wind turbine with Type 3 aerofoil has the best

power performance at high λ (shown in Figure 7.11).

124
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

0.5
On Type 1 aerofoil upper surface
0.4
On Type 1 aerofoil lower surface
0.3

0.2
Pressure coefficient

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.1

-0.2

-0.3

-0.4

-0.5
Position, x/c

(a)

0.5
On Type 2 aerofoil upper surface
0.4
On Type 2 aerofoil lower surface
0.3

0.2
Pressure coefficient

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.1

-0.2

-0.3

-0.4

-0.5
Position, x/c

(b)

125
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

0.5
On Type 3 aerofoil upper surface
0.4
On Type 3 aerofoil lower surface
0.3

0.2
Pressure coefficient

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.1

-0.2

-0.3

-0.4

-0.5
Position, x/c
(c)

Figure 7.12: Distributions of pressure coefficient on the blade aerofoil at r = 140 mm: (a)
Type 1; (b) Type 2; (c) Type 3.

Figure 7.13 shows distributions of torque coefficient for the three aerofoils. It can be seen that

the difference of the torque coefficient among the three aerofoils is not significant. It indicates

that the starting capability of three turbines is similar. Regression analysis result in Figure

7.14 shows however that there are two trends about thrust coefficient distributions. For Type

1 aerofoil, thrust coefficient increases steadily. For Type 2 and 3 aerofoils, the thrust

coefficient varies over a much narrower band centred around 0.1. A comparison of the two

trends in Figure 7.14 demonstrates that the leading edge is an important part in thrust

generation.

126
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

0.7
Type 1
0.6 Type 2
Torque coefficient, CT

0.5 Type 3

0.4

0.3

0.2
Type 1: y = -0.4659x + 0.5667
0.1
Type 2: y = -0.4009x + 0.5276
Type 3: y = -0.3769x + 0.5146
0
0 0.2 0.4 0.6 0.8 1 1.2
Tip speed ratio, λ

Figure 7.13: Distributions of torque coefficient of the proposed wind turbine with three
aerofoils.

0.2

0.15
Thrust coefficient, CN

0.1

0.05

0
0 0.2 0.4 0.6 0.8 1 1.2
-0.05
Type 1 y = -0.163x2 + 0.4171x - 0.085

-0.1 Type 2 y = -0.2385x3 + 0.3885x2 - 0.1087x + 0.067

Type 3 y = -0.1943x3 + 0.4297x2 - 0.298x + 0.1481
-0.15
Tip speed ratio, λ

Figure 7.14: Distributions of thrust coefficient of the proposed wind turbine with three
aerofoils.

127
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

7.5 Influences of Guide Vane Numbers

Analyses in Chapter 5 have shown that the stator affects characteristics of the passing flow. In

this section, influences of guide vane numbers on aerodynamic performances of the wind

turbine are studied. Figure 7.15 shows the stator models with 5, 10, 15 and 20 guide vanes,

respectively.

(a) 5 (b) 10

(c) 15 (d) 20

Figure 7.15: The stator model with four guide vane numbers: (a) 5; (b) 10; (c) 15; (d) 20.

As discussed in Section 5.5.1, guide vanes increase the speed of the passing flow, which can

improve the aerodynamic performance of the turbine. Changing the guide vane number

affects this potentially beneficial effect. Table 7.1 lists values of the maximum flow speed
128
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

against four guide vane numbers. Results show that the maximum speed increases with the

number of guide vanes. It indicates that the beneficial effect of guide vanes is enhanced with

the increase of the number of the guide vanes.

Table 7.1: Maximum speeds of the approaching flow with different guide vane numbers.

Number of guide vanes 5 10 15 20

Maximum flow speed (m/s) 8.24 9.56 10.08 10.45

On the other hand, changing guide vane number also affects the mass flow rate of the wind

turbine. The mass flow rate represents mass of the flow per unit time passing through the

wind turbine. Table 7.2 shows the mass flow rate against four guide vane numbers. It can be

seen that the mass flow rate decreases with the number of guide vanes due to the restriction of

the guide vane to flow moving.

Table 7.2: Mass flow rates for the wind turbine with different guide vane numbers.

Number of guide vanes 5 10 15 20

Mass flow rate (kg/s) 0.226 0.21 0.18 0.16

Tables 7.1-7.2 show that wind turbines with 10 guide vanes and 15 guide vanes can achieve

moderate values in both the maximum flow speed and the mass flow rate. It means that guide

vane number between 10 and 15 can achieve a balance between the flow speed and the mass

flow rate.

Distributions of power coefficient with four guide vane numbers are shown in Figure 7.16.

The wind turbines with 10 guide vanes and 15 guide vanes have good power coefficient

performances due to the good balance between the flow speed and the mass flow rate. Figure

7.16 also shows that the wind turbine with 5 guide vanes has the lowest power coefficient,

129
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

which is caused by the smallest speed of the passing flow as listed in Table 7.1. It is

suggested that the guide vane number of 5 has the weakest effect on increasing the flow speed.

It can be seen that performances of power coefficient with 20 guide vanes are also poor. This

is caused by the lowest mass flow rate as listed in Table 7.2. 20 guide vanes greatly restrain

wind from passing through the wind turbine.

0.25
5 guide vanes

10 guide vanes
0.2 15 guide vane

20 guide vanes
Power coefficient, CP

0.15

0.1

5 guide vanes: y = -0.4261x2 + 0.4979x - 0.0033

0.05
10 guide vanes: y = -0.5145x2 + 0.6434x + 0.0029
15 guide vanes: y = -0.5022x2 + 0.631x + 0.0032
20 guide vanes: y = -0.4549x2 + 0.5589x + 0.0033
0
0 0.2 0.4 0.6 0.8 1 1.2
Tip speed ratio, λ

Figure 7.16: Distributions of power coefficient of the proposed wind turbine with four
guide vane numbers.

Torque coefficient distributions with four guide vane numbers are shown in Figure 7.17. The

values of maximum torque coefficients are in a range of 0.5 - 0.67. It is worth noting that,

while the maximum torque coefficient reaches the lowest value for the turbine with 5 guide

vanes due to the smallest flow speed, the minimum value of 0.5 is high compared with those

130
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

of conventional wind turbines [52], demonstrating the potential of the proposed wind turbine

in terms of good starting capability.

0.8
5 guide vanes
0.7
10 guide vanes
0.6 15 guide vanes
Torque coefficient, CT

0.5 20 guide vanes

0.4

0.3

0.2
5 guide vanes: y = -0.4175x + 0.4859
10 guide vanes: y = -0.5365x + 0.664
0.1
15 guide vanes: y = -0.5281x + 0.6568
20 guide vanes: y = -0.4659x + 0.5667
0
0 0.2 0.4 0.6 0.8 1 1.2
Tip speed ratio, λ

Figure 7.17: Distributions of torque coefficient of the proposed wind turbine with four
guide vane numbers.

Figure 7.18 shows distributions of thrust coefficient of wind turbines with four guide vane

numbers. It can be seen that thrust coefficient with 5 guide vanes is the highest. Influences of

guide vanes on the properties of passing flow have been studied in Section 5.4. When the

wind turbine has less guide vanes, guide vanes will have less influence on the flow direction.

Therefore, the flow cannot pass blades with an optimal direction, which generates large thrust

on blades. More guide vanes have stronger effects on the passing flow and result in lower

thrust on blades.

131
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

The guide vane number also influences the trend of 𝐶𝑁 -λ relationship. A regression analysis

shows that the value of 𝐶𝑁 rises slowly when the proposed wind turbine has 15 guide vanes or

more. This indicates that guide vanes can restrict the increment of 𝐶𝑁 and lead to the small 𝐶𝑁

value.

0.8
y = 0.3506x2 - 0.2318x + 0.5426
0.7

0.6
Thrust coefficient , CN

5 guide vanes
0.5
y = 0.0487x2 + 0.1625x + 0.152 10 guide vanes
0.4
15 guide vanes
0.3
y = -0.1488x2 + 0.3289x - 0.0175 20 guide vanes
0.2

0.1
y = -0.163x2 + 0.4171x - 0.085
0
0 0.2 0.4 0.6 0.8 1 1.2
-0.1

-0.2
Tip speed ratio, λ

Figure 7.18: Distributions of thrust coefficient of the proposed wind turbine with four
guide vane numbers.

7.6 Summary
In this chapter, influences of geometry parameters, such as hub-to-tip ratios, numbers of

blades, aerofoil profiles and numbers of guide vanes, on aerodynamic performances of the

wind turbine have been analysed. Results show that varying these geometrical parameters can

affect the maximum power coefficient significantly. When the wind turbine has 10 guide

vanes, the power coefficient can exceed 0.2 which is the highest value obtained by the wind

132
 Chapter 7 Influences of Geometrical Parameters on Aerodynamics of the Wind Turbine

turbines investigated in the study. It is also found that the proposed wind turbine has higher

values of the maximum torque coefficient compared with conventional wind turbines.

133
 Chapter 8 Discussion

Chapter 8 Discussion

A novel wind turbine based on the impulse turbine technology has been proposed. The

aerodynamics of the proposed wind turbine has been analysed by the wind tunnel testing and

CFD simulations. Aerodynamic performances of the proposed wind turbine are discussed in

this chapter.

8.1 Comparison about Aerodynamics between the Proposed Wind Turbine

and Conventional Wind Turbines

Aerodynamic performances of the proposed wind turbine are compared with those of

conventional wind turbines in this section. The power coefficient, torque coefficient and

thrust coefficient represent power output, starting capability and axial loads on the rotor. The

proposed wind turbine has a diameter of 300 mm. The diameter depends on the size of the

omni-flow wind energy system [15]. Therefore, this proposed wind turbine is compared with

the similar scale conventional wind turbines, along with aerodynamics under low wind speeds,

due to the urban environment where it is situated.

8.1.1 Power Coefficient

The maximum power coefficient (CP) of the proposed wind turbine is similar to that of a

conventional wind turbine. In the wind speed range of 5 m/s – 6 m/s, the proposed wind

turbine has the maximum CP of approximately 0.16 as shown in Figure 5.16. From the

literature review, the maximum CP of a 391mm-diameter HAWT is 0.14 [47]. The maximum

CP of a 330mm-diameter Savonius VAWT is 0.13 [161]. In the wind speed range of 7 m/s -

8.2 m/s, the proposed wind turbine can achieve the maximum CP of 0.2, as shown in Figure

7.16. From the literature review, a 234mm-diameter HAWT has a peak CP of 0.18 [162]. The

134
 Chapter 8 Discussion

maximum CP of a 240mm-diameter VAWT is 0.14 [163]. Therefore, the proposed wind

turbine has a competitive power generation capability, in the urban areas.

8.1.2 Torque Coefficient

The maximum torque coefficient (CT) of the proposed wind turbine is approximately 0.7 as

shown in Figure 5.19. From research performed, a small-scale HAWT has the maximum CT

of 0.11 and a VAWT has the maximum CT of 0.4 [52, 163]. The proposed wind turbine has

the greater value of the maximum CT, when compared with conventional wind turbines. This

indicates that the proposed wind turbine has a good starting capability. Experimental results

in Figure 5.20 show that the wind turbine can start to work at wind speeds as low as 1.6 m/s.

Hence, the wind turbine can capture wind energy with low velocity wind. This feature is

important for an urban wind turbine because wind in an urban environment occurs at low

speeds.

8.1.3 Thrust Coefficient

Results in Figure 7.5, Figure 7.9 and Figure 7.14 show that the thrust coefficient (CN)

corresponding to the maximum CP is always less than 0.2. The CN of a conventional wind

turbine is high up to 1 [18]. The proposed wind turbine has 20 blades so thrust on one blade is

much smaller than that of a conventional wind turbine. The small thrust can allow a reduction

in strength and weight of the materials used in the turbine.

8.2 Comparison about Aerodynamics of the Proposed Wind Turbine under

Uniform Flow and Non-uniform Flow

Recently, researchers have mainly focused on studying a wind turbine under the uniform flow

condition [17]. In this project, the proposed wind turbine has been investigated under both

uniform and non-uniform flow conditions. Compared with the maximum CP under the
135
 Chapter 8 Discussion

uniform flow condition, the maximum CP decreases by 65% under the 20% non-uniform flow

condition and by 29% under the 40% non-uniform flow condition. Hence, the power output

capability of the wind turbine is weakened under the non-uniform flow conditions.

The maximum CT decreases from 0.7 to 0.4 under the 20% non-uniform flow condition and to

0.58 under the 40% non-uniform flow condition. Therefore, the starting capability of the wind

turbine is lowered under the non-uniform flow conditions, when compared to the uniform

flow condition. However, values of maximum CT under the non-uniform flow conditions are

higher than those of conventional wind turbines [52, 163]. Hence, the proposed wind turbine

has the potential of good starting capability even operating under the non-uniform flow

conditions.

8.3 Effects of Geometrical Parameters on Aerodynamics

Influences of four geometrical parameters on aerodynamics of the proposed wind turbine

have been studied under the uniform flow condition. Values of the maximum CP, maximum

CT and maximum CN with different geometrical parameters are listed in Tables 8.1 – 8.4. The

ratio of 0.45 is the best of three ratios, due to the highest value of the maximum CP shown in

Table 8.1.

Table 8.1: Key aerodynamic features of the proposed wind turbine with three hub-to-tip
ratios.

Hub-to-tip ratio 0.35 0.45 0.6

Maximum CP 0.17 0.176 0.157

Maximum CT 0.59 0.59 0.54

Maximum CN 0.22 0.18 0.13

136
 Chapter 8 Discussion

Table 8.2 shows that the wind turbine with 20 blades has a high value of the maximum CP.

When the blade number is greater than 20, there is no improvement on the maximum CP and

CT. However, the thrust performance would be compromised. Table 8.2 indicates that the

aerodynamics of the wind turbine cannot be improved significantly with more than 20 blades.

Table 8.2: Key aerodynamic features of the proposed wind turbine with five blades
numbers.

Number of blades 15 17 20 25 30

Maximum CP 0.148 0.162 0.176 0.177 0.177

Maximum CT 0.5 0.54 0.59 0.59 0.59

Maximum CN 0.09 0.12 0.18 0.24 0.32

When replacing Type 1 aerofoil with Type 2 or Type 3 aerofoils, there are slight

improvements on the maximum CP as shown in Table 8.3. On the other hand, Type 3 aerofoil

has the highest CP at a high tip speed ratio (λ = 0.76). Type 3 aerofoil also has a high value of

maximum CT for good startup capability. Table 8.3 also shows that Type 3 aerofoil has the

good performance on thrust due to the improved leading edge. The comparison in Table 8.3

indicates that changing the leading edge of the blade aerofoil can improve the aerodynamic

properties of the proposed wind turbine.

Table 8.3: Key aerodynamic features of the proposed wind turbine with three aerofoils.

Aerofoil type Type 1 Type 2 Type 3

Maximum CP 0.176 0.178 0.181

CP at λ = 0.76 0.161 0.175 0.18

Maximum CT 0.59 0.52 0.5

Maximum CN 0.18 0.1 0.08

137
 Chapter 8 Discussion

Table 8.4 lists key aerodynamic features of the wind turbines with four guide vane numbers

(5, 10, 15 and 20) that have been studied. The wind turbines with 10 and 15 guide vanes have

high values for maximum CP and CT. The 15 guide vane turbine is more efficient than the 10

guide vane turbine as it has a smaller CN.

Table 8.4: Key aerodynamic features of the proposed wind turbine with four numbers
of guide vanes.

Number of guide vanes 5 10 15 20

Maximum CP 0.142 0.2 0.2 0.176

Maximum CT 0.5 0.67 0.67 0.59

Maximum CN 0.73 0.41 0.18 0.18

Although geometrical parameters affect values of CT, the maximum CT is in a high range

from 0.5 to 0.67. This ensures the good starting capability of the proposed wind turbine with

different geometrical parameters, when compared with conventional wind turbines.

8.4 Approaches Adopted in this Project

In this study, the CFD approach and wind tunnel testing have been utilised to investigate

aerodynamics of the proposed wind turbine. The computational wind turbine model has the

same dimensions to the experimental prototype, therefore, there is no scale effect when

comparing numerical results with experimental data. The comparison has shown that the CFD

approach can provide results with acceptable levels of accuracy, and that the realisable 𝑘-𝜀

model is suitable to investigate the proposed wind turbine.

When compared with the wind tunnel testing, the CFD approach can save the cost of the test

rig. Furthermore, through the use of CFD simulations, the cost of prototypes also can be

saved, especially when studying influences of geometrical parameters. Visual indications

138
 Chapter 8 Discussion

were found by CFD simulations and then employed in analyses such as, flow pressure

distributions on the blade aerofoil, shown in Chapter 5. Because the blades were rotating in

the wind tunnel testing, pressure on the blade aerofoil was very difficult to be detected.

139
 Chapter 9 Conclusions and Recommendations for Future Work

Chapter 9 Conclusions and Recommendations for Future Work

9.1 Conclusions
A new wind turbine that employs the impulse turbine technology has been proposed. The

aerodynamics of the proposed wind turbine subjected to the uniform and non-uniform flow

conditions has been investigated. Influences of several key geometrical parameters on

aerodynamic characteristics have also been studied. The project aim stated in Chapter 1 has

been fulfilled.

Findings about aerodynamics of this wind turbine under the uniform flow condition are listed

as follows:

 The wind turbine has the maximum power coefficient of 0.17, which means that the

wind turbine has the potential to be used. The wind turbine has the maximum torque

coefficient of 0.7 that means good self-starting capability to operate with low velocity

wind in an urban environment.

 Guide vanes can increase the velocity of the passing flow. Guide vanes also can

optimise directions of the passing flow inside the omni-flow wind energy system.

Employment of such guide vanes can improve the potential of the wind turbine in

power generation and starting capability.

 The numerical results have agreed well with experimental data. The realisable k-𝜀

turbulence model is suitable to model the proposed wind turbine, compared with

standard k-𝜀, standard k-𝜔 and SST k-𝜔 turbulence models.

140
 Chapter 9 Conclusions and Recommendations for Future Work

The investigation under the non-uniform flow condition has revealed the following:

 Aerodynamic performances of this wind turbine under the non-uniform flow condition

are poor. Due to the generation of the negative normal force on blades, the wind

turbine has low values (0.06 – 0.126) of the maximum power coefficient and low

values (0.4 – 0.58) of the maximum torque coefficient. It is found that the larger

entrance area of the non-uniform flow, the higher the power coefficient and torque

coefficient.

 The non-uniform flow leads to non-uniform thrust on the blades. It is found that thrust

is subject to periodical changes of loading with the frequency of 8 Hz – 10 Hz.

Findings about investigations of geometrical parameters can be addressed as follows:

 Geometrical parameters affect the performances of power coefficient. The hub-to-tip

ratio of 0.45 has the better distribution of power coefficient than the other two ratios.

An increase of blade number improves the maximum power coefficient significantly,

but there is only slight improvement when the blade number exceeds 20. The guide

vane number of 15 can produce good performances on both power and thrust

coefficients.

 A change of the leading edge of blade aerofoil helps to reduce the negative pressure

difference that results in the negative normal force and lowering power output. With

the improved aerofoil profile, power coefficient of the wind turbine is significantly

improved at high tip speed ratios.

 The proposed wind turbine can maintain high values of the maximum torque

coefficients with different geometrical parameters.

141
 Chapter 9 Conclusions and Recommendations for Future Work

In summary, a new type of wind turbine has been presented in this thesis. On the basis of the

aerodynamic analyses performed, there are indications that the wind turbine can be used in

urban areas.

9.2 Recommendations for Future Work

The thesis provides aerodynamic analyses about the new wind turbine. The following

recommendations are presented for future studies:

9.2.1 Study of the Whole Energy System

A study of the whole system, combining the novel wind turbine and the omni-flow system,

was not carried out as part of this project. Due to the size of the omni-flow system, it is not

feasible to test a prototype of the whole system in the wind tunnel. It is recommended for

future researchers to implement an aerodynamic study of the whole system.

9.2.2 Optimisation of the Geometry

Because of the limited amount of time, the geometry of the proposed wind turbine has not

been exhaustively investigated. From the literature review, optimisation of geometrical

parameters, such as the twist angle of blades, the guide vane profile and the setting angle of

guide vanes, can improve aerodynamics of a wind turbine. Therefore, it is recommended to

optimise the geometry in further research.

9.2.3 Study of the Wind Turbine with a Diffuser

This project has carried out aerodynamic analyses for the proposed wind turbine working

inside a cylindrical shroud that is based upon the impulse turbine technology. But the

literature review shows that using a diffuser can improve aerodynamic performances of a

142
 Chapter 9 Conclusions and Recommendations for Future Work

wind turbine. Therefore, it is suggested to study the aerodynamics when a diffuser is

introduced into the proposed wind turbine.

143
 References

References

[1] A. Arapogianni and J. Moccia, "Avoiding fossil fuel costs with wind energy,"
European Wind Energy Association, 2014.

[2] D. W. I. Association, "Denmark - Wind Energy Hub," 2014.

[3] "Wind energys cenarios for 2030," the European Wind Energy Association, 2015.

[4] "Wind Energy in the UK," RenewableUK, 2014.

[5] "Small and Medium Wind UK Market Report," RenewableUK, 2015.

[6] "Improving the energy efficiency of our homes and buildings - energy certificates and
air-conditioning inspections for buildings," UK Department for Communities and
Local Government, 2008.

[7] "Climate Chang Act 2008," UK Parliament, 2008.

[8] A. D. Peacock, D. Jenkins, M. Ahadzi, A. Berry, et al., "Micro wind turbines in the
UK domestic sector," Energy and Buildings, vol. 40, pp. 1324-1333, 2008.

[9] W. T. Chong, A. Fazlizan, S. C. Poh, K. C. Pan, et al., "The design, simulation and
testing of an urban vertical axis wind turbine with the omni-direction-guide-vane,"
Applied Energy, vol. 112, pp. 601-609, 2013.

[10] F. Balduzzi, A. Bianchini, E. A. Carnevale, L. Ferrari, et al., "Feasibility analysis of a

Darrieus vertical-axis wind turbine installation in the rooftop of a building," Applied
Energy, vol. 97, pp. 921-929, 2012.

[11] S. Mertens, Wind energy in the built environment, Brentwood, UK: Multi-Science,
2006.

[12] "Global wind report," Global Wind Energy Council, 2014.

[13] V. Sureshan, "Omni-directional wind power station," Patent no. WO2008017106A1,

2008.

[14] X. Zhang, Y. K. Chen, and R. Calay, "Modelling and analysis of a novel wind turbine
structure," International Journal of Modelling, Identification and Control, vol. 19, pp.
142-149, 2013.

[15] X. Zhang, "Analysis and optimisation of a novel wind turbine turbine," Ph.D. thesis,
University of Hertfordshire, UK, 2013.

144
 References

[16] P. Ying, Y. K. Chen, and Y. G. Xu, "An aerodynamic analysis of a novel small wind
turbine based on impulse turbine principles," Renewable Energy, vol. 75, pp. 37-43,
2015.

[17] A. Miller, B. Chang, R. Issa, and G. Chen, "Review of computer-aided numerical

simulation in wind energy," Renewable and Sustainable Energy Reviews, vol. 25, pp.
122-134, 2013.

[18] T. Burton, D. Sharpe, N. Jenkins, and E. Bossanyi, Wind energy handbook, Chichester,
UK: John Wiley & Sons, Ltd., 2001.

[19] J. F. Manwell, J. G. McGowan, and A. L. Rogers, Wind Energy Explained - Theory,

Design and Application, Chichester, UK: John Wiley & Sons, Ltd., 2002.

[20] "Global Wind Report Annual Market Update 2013," Global Wind Energy Concil,
2013.

[21] P. Molly, "Zukunft der Windenergie weltweit DEWI GmbH,Erneuerbare

Großprojekte Wind und Wasser," in Stiftung Energie & Klimaschutz Baden-
Württemberg, Stuttgart, Germany 2009.

[22] Y. Kyozuka, "An experimental study on the Darrieus-Savonius turbine for the tidal
current power generation," Journal of Fluid Science and Technology, vol. 3, pp. 439-
449, 2008.

[23] M. M. Aslam Bhutta, N. Hayat, A. U. Farooq, Z. Ali, et al., "Vertical axis wind
turbine – a review of various configurations and design techniques," Renewable and
Sustainable Energy Reviews, vol. 16, pp. 1926-1939, 2012.

[24] S. Eriksson, H. Bernhoff, and M. Leijon, "Evaluation of different turbine concepts for
wind power," Renewable and Sustainable Energy Reviews, vol. 12, pp. 1419-1434,
2008.

[25] G. J. M. Darrieus, "Turbine having its rotating shaft trans-verse to the flow of the
current," US Patent no.1835018, 1931.

[26] M. R. Patel, Wind and Solar Power Systems: design, analysis and operation, Boca
Raton, US: CRC Press, 2005.

[27] R. Howell, N. Qin, J. Edwards, and N. Durrani, "Wind tunnel and numerical study of
a small vertical axis wind turbine," Renewable Energy, vol. 35, pp. 412-422, 2010.

[28] J. L. Tangler and D. M. Somers, "NREL Airfoil Families for HAWTs," NREL, 1995.

[29] W. A. Timmer and R. P. J. O. M. van Rooij, "Summary of the Delft University wind
turbine dedicated airfoils," Journal of Solar Energy Engineering, vol. 125, pp. 488-
496, 2003.

145
 References

[30] P. Fuglsang and C. Bak, "Development of the Riso wind turbine airfoils," Wind
Energy, vol. 7, pp. 145-162, 2004.

[31] E. L. Houghton and P. W. Carpenter, Aerodynamics for Engineering Students, Fourth

ed, London, UK: Edward Arnold, 1993.

[32] A. Makkawi, A. Celik, and T. Muneer, "Evaluation of micro-wind turbine

aerodynamics, wind speed sampling interval and its spatial variation," Building
Services Engineering Research and Technology, vol. 30, pp. 7-14, 2009.

[33] D. R. Drew, J. F. Barlow, and T. T. Cockerill, "Estimating the potential yield of small
wind turbines in urban areas: a case study for Greater London, UK," Journal of Wind
Engineering and Industrial Aerodynamics, vol. 115, pp. 104-111, 2013.

[34] J. T. Millward-Hopkins, A. S. Tomlin, L. Ma, D. B. Ingham, et al., "Assessing the

potential of urban wind energy in a major UK city using an analytical model,"
Renewable Energy, vol. 60, pp. 701-710, 2013.

[35] F. Balduzzi, A. Bianchini, and L. Ferrari, "Microeolic turbines in the built

environment: Influence of the installation site on the potential energy yield,"
Renewable Energy, vol. 45, pp. 163-174, 2012.

[36] P. Blackmore, "Siting micro-wind turbines on house roofs," BRE Press, 2008.

[37] I. Abohela, N. Hamza, and S. Dudek, "Effect of roof shape, wind direction, building
height and urban configuration on the energy yield and positioning of roof mounted
wind turbines," Renewable Energy, vol. 50, pp. 1106-1118, 2013.

[38] L. Ledo, P. B. Kosasih, and P. Cooper, "Roof mounting site analysis for micro-wind
turbines," Renewable Energy, vol. 36, pp. 1379-1391, 2011.

[39] M. A. Kotb and H. A. Soliman, "Performance analysis of a horizontal axis wind

turbine under non uniform flow with swirl," Journal of Wind Engineering and
Industrial Aerodynamics, vol. 37, pp. 103-111, 1991.

[40] S. Murakami and A. Mochida, "3-D numerical simulation of airflow around a cubic
model by means of the k-ϵ model," Journal of Wind Engineering and Industrial
Aerodynamics, vol. 31, pp. 283-303, 1988.

[41] S. Wagner, R. Bareib, and G. Guidati, Wind Turbine Noise, Berlin, Germany:
Springer-Verlag, 1996.

[42] L. D. Knopper and C. A. Ollson, "Health effects and wind turbines: a review of the
literature," Environmental Health, vol. 10, pp. 78, 2011.

[43] E. Pedersen and K. Persson Waye, "Perception and annoyance due to wind turbine
noise—a dose–response relationship," The Journal of the Acoustical Society of
America, vol. 116, pp. 3460, 2004.
146
 References

[44] A. L. Rogers, J. F. Manwell, and S. Wright, "Wind Tubrine Acoustic Noise,"

Renewable Energy Research Laboratory, University of Massachusetts at Amherst,
2002.

[45] B. Vick and S. Broneske, "Effect of blade flutter and electrical loading on small wind
turbine noise," Renewable Energy, vol. 50, pp. 1044-1052, 2013.

[46] G. J. W. Van Bussel, "Small wind turbines for the built environment," in 4th Eurpean
and Asian Wind Engineering Conference, Prague, Czech, 2005.

[47] R. A. Kishore, T. Coudron, and S. Priya, "Small-scale wind energy portable turbine
(SWEPT)," Journal of Wind Engineering and Industrial Aerodynamics, vol. 116, pp.
21-31, 2013.

[48] R. K. Singh, M. R. Ahmed, M. A. Zullah, and Y.-H. Lee, "Design of a low Reynolds
number airfoil for small horizontal axis wind turbines," Renewable Energy, vol. 42, pp.
66-76, 2012.

[49] R. K. Singh and M. R. Ahmed, "Blade design and performance testing of a small wind
turbine rotor for low wind speed applications," Renewable Energy, vol. 50, pp. 812-
819, 2013.

[50] O. Ozgener and L. Ozgener, "Exergy and reliability analysis of wind turbine systems:
A case study," Renewable and Sustainable Energy Reviews, vol. 11, pp. 1811-1826,
2007.

[51] "Small Wind Systems UK Market Report," RenewableUK, 2010.

[52] R. A. Kishore and S. Priya, "Design and experimental verification of a high efficiency
small wind energy portable turbine (SWEPT)," Journal of Wind Engineering and
Industrial Aerodynamics, vol. 118, pp. 12-19, 2013.

[53] M. S. Selig and B. D. McGranahan, "Wind tunnel aerodynamic tests of six airfoils for
use on small wind turbines," Journal of Solar Energy Engineering, vol. 126, pp. 986,
2004.

[54] P. Giguere and M. S. Selig, "Low Reynoldes number airfoils for small horizontal axis
wind turbines," Wind Engineering, vol. 21, pp. 367-380, 1997.

[55] J. C. C. Henriques, F. Marques da Silva, A. I. Estanqueiro, and L. M. C. Gato,

"Design of a new urban wind turbine airfoil using a pressure-load inverse method,"
Renewable Energy, vol. 34, pp. 2728-2734, 2009.

[56] P. R. Ebert and D. H. Wood, "Observations of the starting behavior of a small

horizontal axis wind turbine," Renewable Energy, vol. 12, pp. 245-257, 1997.

147
 References

[57] M. M. Duquette and K. D. Visser, "Numerical implications of solidity and blade

number on rotor performance of horizontal-axis wind turbines," Journal of Solar
Energy Engineering, vol. 125, pp. 425-432, 2003.

[58] E. Dayan, "Wind energy in buildings: Power generation from wind in the urban
environment-where it is needed most," Refocus, vol. 7, pp. 33-38, 2006.

[59] G. M. Joselin Herbert, S. Iniyan, and D. Amutha, "A review of technical issues on the
development of wind farms," Renewable and Sustainable Energy Reviews, vol. 32, pp.
619-641, 2014.

[60] K. Pope, I. Dincer, and G. F. Naterer, "Energy and exergy efficiency comparison of
horizontal and vertical axis wind turbines," Renewable Energy, vol. 35, pp. 2102-2113,
2010.

[61] J. V. Akwa, H. A. Vielmo, and A. P. Petry, "A review on the performance of Savonius
wind turbines," Renewable and Sustainable Energy Reviews, vol. 16, pp. 3054-3064,
2012.

[62] B. L. Gilbert, R. A. Oman, and K. M. Foreman, "Fluid dynamics of diffuser-

augmented wind turbines," Journal of Energy, vol. 2, pp. 368-374, 1978.

[63] O. Igra, "Research and development for shrouded wind turbines," Energy Conversion
and Management, vol. 21, pp. 13-48, 1981.

[64] B. L. Gilbert and K. Foreman, "Experiments with a diffuser-augmented model wind

turbine," Journal of Energy Resources Technology, vol. 105, pp. 46-53, 1983.

[65] K. Abe, M. Nishida, A. Sakurai, Y. Ohya, et al., "Experimental and numerical

investigations of flow fields behind a small wind turbine with a flanged diffuser,"
Journal of Wind Engineering and Industrial Aerodynamics, vol. 93, pp. 951-970,
2005.

[66] Y. Ohya, T. Karasudani, A. Sakurai, K. Abe, et al., "Development of a shrouded wind

turbine with a flanged diffuser," Journal of Wind Engineering and Industrial
Aerodynamics, vol. 96, pp. 524-539, 2008.

[67] R. K. W. Dannecker and A. D. Grant, "Investigations of a building-integrated ducted

wind turbine module," Wind Energy, vol. 5, pp. 53-71, 2002.

[68] S. Y. Hu and J. H. Cheng, "Innovatory designs for ducted wind turbines," Renewable
Energy, vol. 33, pp. 1491-1498, 2008.

[69] G. W. Webster, "Devices for utilizing the power of the wind," Patent No. US4154556,
1979.

[70] N. A. Ahmed, "A novel small scale efficient wind turbine for power generation,"
Renewable Energy, vol. 57, pp. 79-85, 2013.
148
 References

[71] M. H. Mohamed, G. Janiga, E. Pap, and D. Thévenin, "Optimization of Savonius

turbines using an obstacle shielding the returning blade," Renewable Energy, vol. 35,
pp. 2618-2626, 2010.

[72] B. D. Altan and M. Atılgan, "The use of a curtain design to increase the performance
level of a Savonius wind rotors," Renewable Energy, vol. 35, pp. 821-829, 2010.

[73] K. Irabu and J. N. Roy, "Characteristics of wind power on Savonius rotor using a
guide-box tunnel," Experimental Thermal and Fluid Science, vol. 32, pp. 580-586,
2007.

[74] A. Dragomirescu, "Performance assessment of a small wind turbine with crossflow

runner by numerical simulations," Renewable Energy, vol. 36, pp. 957-965, 2011.

[75] G. Müller, M. F. Jentsch, and E. Stoddart, "Vertical axis resistance type wind turbines
for use in buildings," Renewable Energy, vol. 34, pp. 1407-1412, 2009.

[76] W. T. Chong, K. C. Pan, S. C. Poh, A. Fazlizan, et al., "Performance investigation of a

power augmented vertical axis wind turbine for urban high-rise application,"
Renewable Energy, vol. 51, pp. 388-397, 2013.

[77] W. T. Chong, A. Fazlizan, S. C. Poh, K. C. Pan, et al., "Early development of an

innovative building integrated wind, solar and rain water harvester for urban high rise
application," Energy and Buildings, vol. 47, pp. 201-207, 2012.

[78] K. Pope, V. Rodrigues, R. Doyle, A. Tsopelas, et al., "Effects of stator vanes on

power coefficients of a zephyr vertical axis wind turbine," Renewable Energy, vol. 35,
pp. 1043-1051, 2010.

[79] T. Sharpe and G. Proven, "Crossflex: Concept and early development of a true
building integrated wind turbine," Energy and Buildings, vol. 42, pp. 2365-2375, 2010.

[80] T. Snaker and M. Tiryakioglu, "Design and power characterisation of a novel vertical
axis wind energy conversion system (VAWECS)," Wind Engineering, vol. 32, pp.
559-73, 2008.

[81] S. McTavish, D. Feszty, and T. Sankar, "Steady and rotating computational fluid
dynamics simulations of a novel vertical axis wind turbine for small-scale power
generation," Renewable Energy, vol. 41, pp. 171-179, 2012.

[82] D. Japikse and N. C. Baines, Introduction to Turbomachinery, Concepts ETI, 1997.

[83] L. M. C. Gato and A. F. O. Falcão, "Aerodynamics of the wells turbine," International

Journal of Mechanical Sciences, vol. 30, pp. 383-395, 1988.

[84] T. Setoguchi, S. Santhakumar, H. Maeda, M. Takao, et al., "A review of impulse

turbines for wave energy conversion," Renewable Energy, vol. 23, pp. 261-292, 2001.

149
 References

[85] T. Setoguchi and M. Takao, "Current status of self rectifying air turbines for wave
energy conversion," Energy Conversion and Management, vol. 47, pp. 2382-2396,
2006.

[86] D. L. O'Sullivan and A. W. Lewis, "Generator selection and comparative performance

in offshore oscillating water column ocean wave energy converters," IEEE
Transactions on Energy Convertion, vol. 26, pp. 603-613, 2011.

[87] A. F. O. Falcão, "Wave energy utilization: A review of the technologies," Renewable

and Sustainable Energy Reviews, vol. 14, pp. 899-918, 2010.

[88] A. Thakker and T. S. Dhanasekaran, "Computed effects of tip clearance on

performance of impulse turbine for wave energy conversion," Renewable Energy, vol.
29, pp. 529-547, 2003.

[89] A. Thakker, J. Jarvis, and A. Sahed, "Design charts for impulse turbine wave energy
extraction using experimental data," Renewable Energy, vol. 34, pp. 2264-2270, 2009.

[90] I. López, J. Andreu, S. Ceballos, I. Martínez de Alegría, et al., "Review of wave

energy technologies and the necessary power-equipment," Renewable and Sustainable
Energy Reviews, vol. 27, pp. 413-434, 2013.

[91] H. Maeda, S. Santhakumar, T. Setoguchi, M. Takao, et al., "Performance of impulse

turbine with fixed guide vanes for wave power conversion," Renewable Energy, vol.
17, pp. 533-547, 1999.

[92] A. Thakker and F. Hourigan, "Modeling and scaling of the impulse turbine for wave
power applications," Renewable Energy, vol. 29, pp. 305-317, 2004.

[93] H. Maeda, M. Takao, T. Setoguchi, K. Kaneko, et al., "Impulse turbine for wave
power conversion with air flow rectification system," Proc. of the Eleventh
International Offshore and Polar Engineering Conference, Stavanger (Norway), 2001.

[94] V. Jayashankar, S. Anand, T. Geetha, S. Santhakumar, et al., "A twin unidirectional

impulse turbine topology for OWC based wave energy plants," Renewable Energy,
vol. 34, pp. 692-698, 2009.

[95] K. Mala, J. Jayaraj, V. Jayashankar, T. M. Muruganandam, et al., "A twin

unidirectional impulse turbine topology for OWC based wave energy plants –
Experimental validation and scaling," Renewable Energy, vol. 36, pp. 307-314, 2011.

[96] H. K. Versteeg and W. Malalasekera, An introduction to computational fluid dynamics

- the finite volume method, Second ed, Harlow, UK: Pearson Education Ltd., 2007.

[97] J. H. Ferziger and M. Peric, Computational Methods for Fluid Dynamics, Third ed,
Springer, 2001.

150
 References

[98] T. J. Chung, Computational Fluid Dynamics, Second ed, Cambridge University Press,
2002.

[99] R. Lanzafame, S. Mauro, and M. Messina, "Wind turbine CFD modeling using a
correlation-based transitional model," Renewable Energy, vol. 52, pp. 31-39, 2013.

[100] C. Thumthae and T. Chitsomboon, "Optimal angle of attack for untwisted blade wind
turbine," Renewable Energy, vol. 34, pp. 1279-1284, 2009.

[101] L. A. Danao, J. Edwards, O. Eboibi, and R. Howell, "A numerical investigation into
the influence of unsteady wind on the performance and aerodynamics of a vertical
axis wind turbine," Applied Energy, vol. 116, pp. 111-124, 2014.

[102] K. M. Almohammadi, D. B. Ingham, L. Ma, and M. Pourkashan, "Computational

fluid dynamics (CFD) mesh independency techniques for a straight blade vertical axis
wind turbine," Energy, vol. 58, pp. 483-493, 2013.

[103] A. Thakker and F. Hourigan, "A comparison of two meshing schemes for CFD
analysis of the impulse turbine for wave energy applications," Renewable Energy, vol.
30, pp. 1401-1410, 2005.

[104] B. Geurts, Elements of Direct and Large-eddy Simulation, R.T.Ewards, 2004.

[105] W. Rodi, "Comparison of LES and RANS calculations of the flow around bluff
bodies," Journal of Wind Engineering and Industrial Aerodynamics, vol. 69, pp. 55-
75, 1997.

[106] C. J. Simão Ferreira, A. van Zuijlen, H. Bijl, G. van Bussel, et al., "Simulating
dynamic stall in a two-dimensional vertical-axis wind turbine: verification and
validation with particle image velocimetry data," Wind Energy, vol. 13, pp. 1-17, 2010.

[107] M. Lesieur, O. Métais, and P. Comte, Large-Eddy Simulations of Turbulence,

Cambridge University Press, 2005.

[108] S. Benjanirat, L. N. Sankar, and G. Xu, "Evaluation of turbulence models for the
prediction of wind turbine aerodynamics," Proc. of ASME 2003 Wind Energy
Symposium, Reno, Nevada, USA, 2003.

[109] A. Makridis and J. Chick, "Validation of a CFD model of wind turbine wakes with
terrain effects," Journal of Wind Engineering and Industrial Aerodynamics, vol. 123,
pp. 12-29, 2013.

[110] F. Balduzzi, A. Bianchini, R. Maleci, G. Ferrara, et al., "Critical issues in the CFD
simulation of Darrieus wind turbines," Renewable Energy, vol. 85, pp. 419-435, 2016.

[111] B. E. Launder and D. B. Spalding, "The numerical compution of turbulent flows,"

Computer Methods in Applied Mechanics and Engineering, vol. 3, pp. 269-289, 1974.

151
 References

[112] D. C. Wilcox, "Comparison of two-equation turbulence models for boundary layers

with pressure gradients," AIAA Journal, vol. 31, pp. 1414-1421, 1993.

[113] W. Wolfe and S. S. Ochs, "CFD calculations of S809 aerodynamic characteristics," in

35th AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada, USA, 1997.

[114] L. Aresti, M. Tutar, Y. Chen, and R. K. Calay, "Computational study of a small scale
vertical axis wind turbine (VAWT): comparative performance of various turbulence
models," Wind and Structures, vol. 17, pp. 647-670, 2013.

[115] A. Thakker and T. S. Dhanasekaran, "Experimental and computational analysis on

guide vane losses of impulse turbine for wave energy conversion," Renewable Energy,
vol. 30, pp. 1359-1372, 2005.

[116] A. Thakker, T. S. Dhanasekaran, and J. Ryan, "Experimental studies on effect of

guide vane shape on performance of impulse turbine for wave energy conversion,"
Renewable Energy, vol. 30, pp. 2203-2219, 2005.

[117] T. Shih, W. W. Liou, A. Shabbir, Z. Yang, et al., "A new k- eddy viscosity model for
high reynolds number turbulent flows," Computers & Fluids, vol. 24, pp. 227-238,
1995.

[118] T. Shih, "Some developments in computational modeling of turbulent flows," Fluid

Dynamics Research, vol. 20, pp. 67-96, 1997.

[119] M. Raciti Castelli, A. Englaro, and E. Benini, "The Darrieus wind turbine: Proposal
for a new performance prediction model based on CFD," Energy, vol. 36, pp. 4919-
4934, 2011.

[120] M. Raciti Castelli, G. Pavesi, L. Battisti, E. Benini, et al., "Modeling strategy and
numerical validation for a Darrieus vertical axis micro-wind turbine," Proc. of ASME
ASME 2010 International Mechanical Engineering Congress & Exposition,
Vancouver, British Columbia, Canada, 2010.

[121] M. H. Mohamed, "Impacts of solidity and hybrid system in small wind turbines
performance," Energy, vol. 57, pp. 495-504, 2013.

[122] D. C. Wilcox, "Reassessment of the scale-determining equation for advanmced

turbulence models," AIAA Journal, vol. 26, pp. 1299-1310, 1988.

[123] F. R. Menter, "Zonal two-equation k-w turbulence models for aerodynamic flows,"
AIAA Paper 2906, vol., 1993.

[124] S. L. Yang, Y. L. Chang, and O. Arici, "Navier-Stokes computations of the NREL

airfoil ssing a κ — ω turbulent model at high angles of attack," Journal of Solar
Energy Engineering, vol. 117, pp. 304-310, 1995.

152
 References

[125] B. Yang and C. Lawn, "Fluid dynamic performance of a vertical axis turbine for tidal
currents," Renewable Energy, vol. 36, pp. 3355-3366, 2011.

[126] F. Menter, "Two-equation eddy-viscosity turbulence model for engineering

applications," AIAA Journal, vol. 32, pp. 1598-1605, 1994.

[127] A. L. Pape and J. Lecanu, "3D Navier-Stokes computations of a stall-regulated wind

turbine," Wind Energy, vol. 7, pp. 309-324, 2004.

[128] S. Yang, Y. Chang, and O. Arici, "Navier-Stokes computations of the NREL airfoil
using a k-w turbulent model at high ang;es of attack," Journal of Solar Energy
Engineering, vol. 117, pp. 304-310, 1995.

[129] W. Wolfe and S. Ochs, "Predicting aerodynamic characteristics of typical wind

turbine airfoils using CFD," Sandia National Laboratories, 1997.

[130] P. Fuglsang, C. Bak, M. Gaunaa, and I. Antoniu, "Designa nd verification of the Riso-
B1 airfoil family for wind turbines," Journal of Solar Energy Engineering, vol. 126,
pp. 1002-1010, 2006.

[131] F. Bertagnolio, N. Sorensen, and J. Johansen, "Profile catalogue for airfoil sections
based on 3D computations," Riso, 2006.

[132] E. Duque, W. Johnson, C. van Dam, R. Cortes, et al., "Numerical predictions of wind
turbine power and aerodynamic loads for the NREL phase II combined experiment
rotor," in ASME Wind Energy Symposium, Reno, Nevada, US, 2000.

[133] N. Sorensen and M. O. L. Hansen, "Rotor performance predictions using a Navier-

Stokes method," in ASME Wind Energy Symposium, Reno, Nevada, US, 1998.

[134] Y. Li, K. J. Paik, T. Xing, and P. M. Carrica, "Dynamic overset CFD simulations of
wind turbine aerodynamics," Renewable Energy, vol. 37, pp. 285-298, 2012.

[135] S. Lin and T. Shieh, "Study of aerodynamical interference for a wind turbine,"
International Communications in Heat and Mass Transfer, vol. 37, pp. 1044-1047,
2010.

[136] Y. Bazilevs, M. C. Hsu, I. Akkerman, S. Wright, et al., "3D simulation of wind

turbine rotors at full scale. Part I: Geometry modeling and aerodynamics,"
International Journal for Numerical Methods in Fluids, vol. 65, pp. 207-235, 2011.

[137] Y. Bazilevs, M. C. Hsu, J. Kiendl, R. Wüchner, et al., "3D simulation of wind turbine
rotors at full scale. Part II: Fluid-structure interaction modeling with composite
blades," International Journal for Numerical Methods in Fluids, vol. 65, pp. 236-253,
2011.

[138] C. A. Baxevanou, P. K. Chaviaropoulos, S. G. Voutsinas, and N. S. Vlachos,

"Evaluation study of a Navier–Stokes CFD aeroelastic model of wind turbine airfoils
153
 References

in classical flutter," Journal of Wind Engineering and Industrial Aerodynamics, vol.

96, pp. 1425-1443, 2008.

[139] N. Tachos, C. Zioutis, A. Filios, and D. Margaris, "Aerodynamic computational

analysis for a HAWT rotor," in European Wind Energy Conference, Athens, Greece,
2006.

[140] N. Mandas, F. Cambuli, and C. Carcangiu, "Numerical prediction of horizontal axis

wind turbine flow," in European Wind Energy Conference, Athens, Greece, 2006.

[141] D. Hartwanger and A. Horvat, "3D modelling of a wind turbine using CFD," in
NAFEMS Conference, Cheltenham, UK, 2008.

[142] F. Bet and H. Grassmann, "Upgrading conventional wind turbines," Renewable

Energy, vol. 28, pp. 71-78, 2003.

[143] T. Y. Chen, Y. T. Liao, and C. C. Cheng, "Development of small wind turbines for
moving vehicles: Effects of flanged diffusers on rotor performance," Experimental
Thermal and Fluid Science, vol. 42, pp. 136-142, 2012.

[144] L. A. Danao, O. Eboibi, and R. Howell, "An experimental investigation into the
influence of unsteady wind on the performance of a vertical axis wind turbine,"
Applied Energy, vol. 107, pp. 403-411, 2013.

[145] M. Burlando, A. Ricci, A. Freda, and M. P. Repetto, "Numerical and experimental

methods to investigate the behaviour of vertical-axis wind turbines with stators,"
Journal of Wind Engineering and Industrial Aerodynamics, vol. 144, pp. 125-133,
2015.

[146] S. Rolland, W. Newton, A. J. Williams, T. N. Croft, et al., "Simulations technique for

the design of a vertical axis wind turbine device with experimental validation,"
Applied Energy, vol. 111, pp. 1195-1203, 2013.

[147] R. Bravo, S. Tullis, and S. Ziada, "Performance testing of a small vertical-axis wind
turbine," in 21st Canadian Congress of Applied Mechanics, Toronto, Canada, 2007.

[148] J. B. Barlow, W. H. Jr. Rae, and A. Pope, Low-speed Wind Tunnel Testing, Third ed,
New York, US: John Wiley & Sons Ltd., 1999.

[149] W. H. Rae and A. Pope, Low-speed Wind Tunnel Testing, Second ed, John Wiley &
Sons Ltd., 1984.

[150] L. A. M. Danao, "The influence of unsteadywind on the performance and

aerodynamics of veritical axis wind turbines," Ph.D. thesis, University of Sheffield,
2012.

[151] CD-adapco, Star-CCM+ User Guide 6.04, 2011.

154
 References

[152] P. S. Bernard, Fluid Dynamics, Cambridge Unversity Press, 2015.

[153] V. L. Streeter, E. B. Wylie, and K. B. Bedford, Fluid Mechanics, Ninth ed, McGrew
Hill, 1998.

[154] J. Chen, H. Yang, M. Yang, and H. Xu, "The effect of the opening ratio and location
on the performance of a novel vertical axis Darrieus turbine," Energy, vol. 89, pp.
819-834, 2015.

[155] Y. Peet and P. Sagaut, "Theoretical prediction of turbulent skin friction on

geometrically complex surfaces," Physics of Fluids, vol. 21, pp. 105105, 2009.

[156] I. H. Abbott and A. E. Von Doenhoff, Theory of Wind Sections, Including a Summary
of Airfoil Data, Dover, 1959.

[157] CD-adapco, Star-CCM+ Training, US, 2011.

[158] A. Thakker and F. Hourigan, "Computational fluid dynamics analysis of a 0.6m, 0.6
hub-to-tip ratio impulse turbine with fixed guide vanes," Renewable Energy, vol. 30,
pp. 1387-1399, 2005.

[159] T. Shih, J. Zhu, and J. L. Lumley, "A realizable Reynolds stress algebraic equation
model," National Aeronaustic and Space Administration, 1993.

[160] P. E. Smirnov and F. R. Menter, "Sensitization of the SST turbulence model to

rotation and curvature by applying the Spalart-Shur correction term," Journal of
Turbomachinery, vol. 131, pp. 041010-1-7, 2009.

[161] T. Hayashi, Y. Li, and Y. Hara, "Wind tunnel tests on a different phase three-stage
Savonius rotor," JSME International Journal Series B: Fluids and Thermal
Engineering, vol. 48, pp. 9-16, 2005.

[162] D. Y. C. Leung, Y. Deng, and M. K. H. Leung, "Design optimization of a cost-

effective micro wind turbine," Proc. of the World Congress on Engineering, London,
UK, 2010.

[163] U. K. Saha and M. J. Rajkumar, "On the performance analysis of Savonius rotor with
twisted blades," Renewable Energy, vol. 31, pp. 1776-1788, 2006.

155
 Appendix

Appendix A – Drawing of Guide Vane

156
 Appendix

Appendix B – Drawing of Stator Hub

157
 Appendix

Appendix C – Drawing of Rotor Hub

158
 Appendix

Appendix D – Coordinates of Three Aerofoils

Type 1

x (mm) y (mm) x (mm) y (mm) x (mm) y (mm) x (mm) y (mm)

0 0.77 25 21.306 50.476 0.056 22 13.777
0.04 1.343 26 21.05 50.24 0 21 13.777
0.06 1.605 27 20.788 50.124 0.016 20 13.441
0.08 1.843 28 20.429 50 0.056 19 13.217
0.1 2.064 29 20.062 49.875 0.133 18 12.956
0.2 2.992 30 19.655 49.79 0.22 17 12.655
0.3 3.745 31 19.207 49.5 0.626 16 12.31
0.4 4.395 32 18.715 49 1.332 15 11.929
0.6 5.502 33 187.178 48 2.734 14 11.502
1 7.276 345 17.591 47 4.01 13 11.028
1.5 9.047 35 16.953 46 5.147 12 10.505
2 10.531 36 16.258 45 6.169 11 9.931
3 12.988 37 15.507 44 7.093 10 9.3
4 14.833 38 14.727 43 7.933 9 8.61
5 16.133 39 13.926 42 8.696 8 7.853
6 17.213 40 13.299 41 9.391 7 7.024
7 18.158 41 12.251 40 10.024 6 6.113
8 18.96 42 11.349 39 10.599 5 5.11
9 19.645 43 10.389 38 11.121 4 3.998
10 20.229 44 9.368 37 11.593 3 2.758
11 20.723 45 8.279 36 12.017 2 1.359
12 21.138 46 7.115 35 12.397 1.5 0.586
13 21.478 47 5.864 34 12.733 1.2 0.171
14 21.748 48 4.538 33 13.027 1 0.051
15 21.953 49 3.204 32 13.282 0.9 0.02
16 22.094 49.5 2.537 31 13.497 0.7 0
17 22.144 50 1.87 30 13.675 0.5 0.051
18 22.153 50.4 1.337 29 13.815 0.4 0.079
19 22.129 50.7 0.936 28 13.918 0.3 0.141
20 22.279 50.77 0.819 27 13.985 0.2 0.229
21 21.986 50.818 0.609 26 14.18 0.1 0.364
22 21.866 50.8 0.438 25 14.01 0.06 0.444
23 21.713 50.732 0.281 24 13.973 0.04 0.497
24 21.526 50.611 0.141 23 13.891 0 0.77

159
 Appendix

Type 2

x (mm) y (mm) x (mm) y (mm) x (mm) y (mm) x (mm) y (mm)

0 0.786 28.5 17.914 52.15 0.185 23 10.762
0.069 1.108 29.5 17.654 52.073 0.231 22 10.522
0.25 1.507 30.5 17.359 51.95 0.35 20 9.981
0.5 2.048 31.5 17.026 51.5 0.854 18 9.355
1 3.092 32.5 16.655 51 1.418 17 9.009
1.5 4.091 33.5 16.244 50 2.539 16 8.641
2.5 5.971 34.5 15.791 49 3.574 15 8.25
3.5 7.733 35.5 15.294 48 4.509 14 7.836
5 10.122 36.5 14.752 47 5.358 13 7.398
5.5 10.875 37.5 14.15 46 6.13 11 6.447
6 11.607 38.5 13.547 45 6.832 9 5.39
7.5 13.475 39.5 12.916 44 7.472 7 4.222
8.5 14.462 40.5 12.268 43 8.054 6 3.594
9.5 15.3 41.5 11.601 42 8.582 5 2.933
10.5 16.013 42.5 10.9 41 9.059 4 2.24
11.5 16.621 43.5 10.158 40 9.489 3 1.513
12.5 17.139 44.5 9.371 39 9.874 2 0.749
13.5 17.573 45.5 8.537 38 10.215 1.5 0.353
14.5 17.932 46.5 7.653 37 10.514 1.25 0.152
15.5 18.221 47.5 6.713 36 10.774 0.9 0.009
16.5 18.429 48.5 5.713 35 10.994 0.8 0
17.5 18.565 49.5 4.656 34 11.175 0.7 0.005
18.5 18.668 50.5 3.588 33 11.32 0.6 0.022
19.5 18.738 51.5 2.521 32 11.427 0.5 0.054
20.5 18.776 52.5 1.454 31 11.498 0.4 0.101
21.5 18.782 52.83 1.103 30 11.533 0.3 0.168
22.5 18.756 52.97 0.856 29 11.532 0.2 0.262
23.5 18.698 52.97 0.557 28 11.495 0.15 0.324
24.5 18.608 52.83 0.31 27 11.421 0.1 0.402
25.5 18.485 52.63 0.173 26 11.311 0.1 0.03
26.5 18.328 52.41 0 25 11.165 0.05 0.51
27.5 18.138 52.25 0.147 24 10.98 0 0.786

160
 Appendix

Type 3

x (mm) y (mm) x (mm) y (mm) x (mm) y (mm) x (mm) y (mm)

0 0.723 29 15.83 54.21 0 24 8.341
0.02 0.875 30 15.655 54.09 0.04 23 8.17
0.05 0.978 31 15.446 53.97 0.077 22 7.984
0.08 1.047 32 15.203 53.84 0.161 21 7.781
0.15 1.171 33 14.924 53.79 0.213 20 7.563
0.3 1.428 34 14.608 53.5 0.487 19 7.328
0.4 1.596 35 14.254 53 0.967 18 7.077
0.6 1.925 36 13.861 52.5 1.446 17 6.809
0.8 2.246 37 13.427 52 1.922 16 6.525
2 4.015 38 12.951 51 2.838 15 6.223
3 5.32 39 12.45 50 3.667 14 5.904
4 6.502 40 11.929 49 4.418 13 5.568
5 7.578 41 11.395 48 5.098 12.29 5.32
6 8.564 42 10.847 47 5.712 11 4.842
7 9.47 43 10.277 46 6.265 10 4.451
8 10.305 44 9.672 45 6.762 9 4.042
9 11.076 45 9.032 44 7.206 8 3.614
10 11.788 46 8.353 43 7.601 7 3.165
11 12.444 47 7.634 42 7.95 6 2.697
12 13.038 48 9.872 41 8.251 5 2.209
13 13.57 49 6.064 40 8.509 4 1.7
14 14.046 50 5.208 39 8.725 3 1.168
15 14.468 51 4.307 38 8.9 2 0.615
16 14.84 52 3.399 37 9.035 1 0.048
17 15.164 53 2.491 36 9.13 0.9 0.016
18 15.441 54 1.583 35 9.185 0.8 0
19 15.675 54.5 1.129 34 9.205 0.5 0.034
20 15.865 54.6 1.0335 33 9.2 0.4 0.076
22 16.119 54.687 0.941 32 9.2 0.25 0.178
23 16.184 54.765 0.784 31 9.104 0.15 0.286
24 16.209 54.793 0.57 30 9.04 0.1 0.361
25 16.198 54.748 0.379 29 8.96 0.08 0.4
26 16.155 54.67 0.243 28 8.869 0.05 0.469
27 16.079 54.56 0.141 27 8.76 0.02 0.571
28 15.971 54.42 0.063 26 8.636 0 0.723

161
 Appendix

Appendix E – Flow Velocity and Pressure Change When Passing

Guide Vanes

r = 70mm

7 20

6 15

5 10
Velocity, m/s

Pressure, Pa
4 5
Velocity
3 0
Pressure

2 -5

1 -10

0 -15
0 10 20 30 40
Distance from the entrance of guide vanes in the axial direction of
the turbine, mm

r = 147mm
6 20
18
5
16
14
4
Velocity, m/s

Pressure, Pa

12
Velocity
3 10
Pressure
8
2
6
4
1
2
0 0
0 10 20 30 40
Distance from the entrance of guide vanes in the axial direction
of the turbine, mm

162
 Appendix

Appendix F – Flow Directions inside the Omni-flow System with

and without a Stator
(a) When free stream wind is at 12o to the omni-flow system without the stator

100
φoptimal = 90o
80

60

40

20
Angle φ, degree

0
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
-20

-40

-60

-80

-100
Position, m

(b) When free stream wind is at 12o to the omni-flow system with the stator

100

80

60

40
φoptimal = 20o
Angle φ, degree

20

0
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
-20

-40

-60

-80

-100
Position, m
163
 Appendix

(c) When free stream wind is at 24o to the omni-flow system without the stator

100
φoptimal = 90o
80

60

40
Angle φ, degree

20

0
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
-20

-40

-60

-80

-100
Position, m

(d) When free-stream wind is at 24o to the omni-flow system with the stator

100

80

60

40
Angle φ, degree

20 φoptimal = 20o

0
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
-20

-40

-60

-80

-100
Position, m

164
 Appendix

(e) When free stream wind is at 36o to the omni-flow system without the stator

200

160

120
Angle φ, degree

φoptimal = 90o
80

40

0
-0.3 -0.2 -0.1 0 0.1 0.2 0.3

-40
Position, m

(f) When free stream wind is at 36o to the omni-flow system with the stator

200

160
Angle φ, degree

120

80

40
φoptimal = 20o

0
-0.3 -0.2 -0.1 0 0.1 0.2 0.3

-40
Position, m

165