Wind Turbines

CFD Open Series
Patch 2.20
Wind Turbines
Edited by:
Ideen Sadrehaghighi, Ph.D.
Off Shore Wind

Farm
Horizontal
Axis Wind
Turbine
Vertical Axis
Wind Turbine
(VAWT)
ANNAPOLIS, MD
2
Contents
List of Figures: ..................................................................................................................................................................... 6
1 Introduction ................................................................................................................................ 12
1.1 Some Preliminaries & Definitions ................................................................................................................ 12
1.2 History ..................................................................................................................................................................... 12
1.2.1 Types of Wind Turbines ...................................................................................................................... 14
1.2.1.1 Horizontal Axis ............................................................................................................................ 14
1.2.1.2 Vertical Axis .................................................................................................................................. 15
1.2.1.2.1 Darrieus Wind Turbine ...................................................................................................... 15
1.2.1.2.2 Giromill ..................................................................................................................................... 16
1.2.1.2.3 Savonius Wind Turbine...................................................................................................... 16
1.2.1.2.4 Parallel ...................................................................................................................................... 16
1.2.2 References.................................................................................................................................................. 16
2 Wind Turbine Design and Working Principle................................................................. 19

2.1 Airfoils and Blade Design ................................................................................................................................. 19
2.1.1 Blade Twist ................................................................................................................................................ 20
2.1.2 Tip Speed Ratio (TSR) ........................................................................................................................... 21
2.1.3 Wind Turbine Operation...................................................................................................................... 22
2.1.4 Wind Turbine Aerodynamics ............................................................................................................. 23
2.1.5 Actuator Disk Concept .......................................................................................................................... 23
2.2 Case Study 1 - Effect of Blade Angles on the Performance of Axial Six Blades Wind
Turbine ................................................................................................................................................................................ 25
2.2.1 Introduction & Literature Survey .................................................................................................... 25
2.2.2 Experimental Work ................................................................................................................................ 26
2.2.3 Theoretical Analysis .............................................................................................................................. 28
2.2.4 Results and Discussion ......................................................................................................................... 30
2.2.5 Conclusions ............................................................................................................................................... 33
2.2.6 References.................................................................................................................................................. 33
2.3 Case Study 2 - Blade Thickness Effect on The Aerodynamic Performance of An
Asymmetric NACA Six Series Blade Vertical Axis Wind Turbine in Low Wind Speed ........................ 34
2.3.1 Abstract ....................................................................................................................................................... 34
2.3.2 Introduction .............................................................................................................................................. 34
2.3.3 Computational Settings ........................................................................................................................ 35
2.3.3.1 Geometrical and Operational Details of Reference Turbine Case .......................... 35
2.3.3.2 Rationale of Choosing 2D CFD Model ................................................................................. 36
2.3.3.3 Computational Domain and Grid ......................................................................................... 37
2.3.3.4 Numerical Setting ....................................................................................................................... 37
2.3.3.5 Sensitivity and Validation ....................................................................................................... 38
2.3.5 Conclusions & Recommendations.................................................................................................... 45
2.3.6 Acknowledgments .................................................................................................................................. 45
2.3.7 References.................................................................................................................................................. 45
2.4 Wind Profile........................................................................................................................................................... 49
2.5 Wind Turbine Cost Figures ............................................................................................................................. 49
3 Case Studies Involving Horizontal Axis Wind Turbine (HAWTs)............................. 51

3
3.1 Case Study 1 - Flow Around a Wind Turbine........................................................................................... 51

3.1.1 Near-Wake ................................................................................................................................................. 53
3.1.1.1 Tip and Root Vortices ............................................................................................................... 53
3.1.1.2 Hub Vortex .................................................................................................................................... 53
3.1.1.3 Mean Flow Distribution ........................................................................................................... 54
3.1.2 Far-Wake .................................................................................................................................................... 54
3.1.2.1 Mean Flow Velocity Distribution ......................................................................................... 54
3.1.2.2 Turbulence Distribution .......................................................................................................... 56
3.1.2.2.1 Streamwise Turbulence Intensity ................................................................................. 56
3.1.2.2.2 Turbulent Momentum Flux .............................................................................................. 57
3.1.2.2.3 Turbulence Kinetic Energy (TKE) ................................................................................. 57
3.1.2.3 Wake Twisting ............................................................................................................................. 57
3.1.2.4 Analytical Wake Modelling ..................................................................................................... 58
3.1.3 Summary & Conclusions ...................................................................................................................... 62
3.1.4 References.................................................................................................................................................. 62
3.2 Case Study 2 - Design of the AOC 15/50 Rotor Blade for Wind Turbine ..................................... 65
3.2.1 Blade Design in ANSYS® DesignModeler....................................................................................... 65
3.2.2 Design Variables and Design Space ................................................................................................. 66
3.2.3 Constructing the Response Surface Method (RSM) .................................................................. 66
3.2.4 CFD Analysis on the AOC 15/50 HAWT Blade ............................................................................ 68
3.2.5 CFD Domain Mesh and Numerical Model for Rotating Bodies............................................. 68
3.2.6 Moving Reference Frame Model ....................................................................................................... 68
3.2.7 Computational Domain (Grid) for the Turbine Blade Model ................................................ 70
3.2.7.1 Boundary Condition .................................................................................................................. 70
3.2.7.2 Grid Independence Study ........................................................................................................ 70
3.2.8 Flow Simulation over the Rotor Blade ........................................................................................... 71
3.2.9 Results for Flow Simulation over Rotor Blade............................................................................ 72
3.2.10 Optimization Method ............................................................................................................................ 72
3.2.10.1 Routine 1 ........................................................................................................................................ 72
3.2.10.2 Routine 2 ........................................................................................................................................ 72
3.2.11 Optimization Results ............................................................................................................................. 72
3.2.11.1 Routine 1 ........................................................................................................................................ 72
3.2.11.1.1 Routine 2 .................................................................................................................................. 73
3.2.12 Conclusion.................................................................................................................................................. 74
3.3 Case Study 3 - Multi-Element Ducts for Ducted Wind Turbines ...................................................... 75
3.3.1 Introduction .............................................................................................................................................. 75
3.3.2 Multi-Element Duct–AD Flow Model .............................................................................................. 76
3.3.3 Numerical Validation............................................................................................................................. 78
3.3.4 Numerical Approach.............................................................................................................................. 80
3.3.4.1 Panel Method................................................................................................................................ 80
3.3.4.2 RANS Method ............................................................................................................................... 81
3.3.5.1 Multi-Element Duct Geometry .............................................................................................. 82
3.3.5.2 Duct Force Coefficient............................................................................................................... 83
3.3.5.3 Power Augmentation ................................................................................................................ 85
3.3.6 Conclusions ............................................................................................................................................... 86
4 Case Studies Involving Vertical Axis Wind Turbines (VAWTs) ................................. 87

4
4.1 Case Study 1 - Vertical Axis Wind Turbine (VAWTs) - An Efficient Way To Harness
Power : A Review ............................................................................................................................................................. 87
4.1.1 Design Optimizations ............................................................................................................................ 87
4.1.2 Operation Analysis and Analysis Techniques and Optimizations ...................................... 87
4.1.3 Novel Design Attachments and Configuration ............................................................................ 87
4.1.4 Conclusions ............................................................................................................................................... 91
4.1.5 References.................................................................................................................................................. 91
4.2 Case Study 2 - Airfoil Optimization to Improve Power Performance of a High-Solidity
Vertical Axis Wind Turbine (VAWT) at a Moderate Tip Speed Ratio (TSR)............................................ 93
4.2.1 Background and Introduction ........................................................................................................... 93
4.2.2 Types of VAWTs and Their Characteristics.................................................................................. 93
4.2.3 Literature Survey and Procedure ..................................................................................................... 94
4.2.4 Numerical Model and Selection of CFD Model ............................................................................ 95
4.2.5 Parametric Modelling ............................................................................................................................ 95
4.2.6 Computational Domain & Meshing .................................................................................................. 98
4.2.7 Solver Setting ............................................................................................................................................ 99
4.2.8 Verification and Validation of CFD Model..................................................................................... 99
4.2.8.1 Grid Independence Study ........................................................................................................ 99
4.2.8.2 Sensitivity W.R.T. Time Step .................................................................................................. 99
4.2.9 Comparison with Experimental Results ..................................................................................... 100
4.2.10 Optimization System .......................................................................................................................... 101
4.2.11 Optimization System Procedure .................................................................................................... 102
4.2.12 Results of Geometric Profile Evolution ....................................................................................... 103
4.2.13 Power Performance ............................................................................................................................ 104
4.2.14 Comparison of Pressure Distribution .......................................................................................... 106
4.2.15 Discussion ............................................................................................................................................... 106
4.2.16 Limitations of Current Study........................................................................................................... 107
4.2.17 Conclusion............................................................................................................................................... 108
4.3 Case Study 3 - Winglet Design for Vertical Axis Wind Turbines Based on a Design of
Experiment (DoE) and CFD ...................................................................................................................................... 109
4.3.1 Introduction ........................................................................................................................................... 109
4.3.2 Current Study on VAWTs .................................................................................................................. 110
4.3.3 Influence of 3D Effects on the Performance of VAWTs ........................................................ 111
4.3.4 The Novelty in the Design of the VAWT’s Winglet ................................................................. 111
4.3.5 Physical and Numerical Model ....................................................................................................... 112
4.3.5.1 Physical Model Introduction ............................................................................................... 112
4.3.5.2 Modeling Strategy ................................................................................................................... 114
4.3.6 Code Validation and Grid Independency Analysis ................................................................. 115
4.3.7 3D VAWT CFD Validation Using Experimental Data ............................................................. 117
4.3.8 Orthogonal Experimental Design (OED) .................................................................................... 118
4.3.8.1 A Brief Introduction of the OED Approach ................................................................... 118
4.3.8.2 Design Variables, Levels and Objective of the Winglet Design ............................. 119
4.3.8.3 L25(56) Orthogonal Array...................................................................................................... 119
4.3.9 Results and Discussion ...................................................................................................................... 121
4.3.9.1 Analysis of the OED Results................................................................................................. 121
4.3.10 Analysis of the Variance (ANOVA) of the OED Results ......................................................... 121
4.3.11 Working Mechanism Analysis of the Winglet .......................................................................... 122
4.3.12 Comparison of the Best Result and the Reference Blade .................................................... 122
4.3.12.1 Full Rotor Model Simulation of the Winglet in the OED Result ............................ 125
4.3.13 Discussion of the Results .................................................................................................................. 127
5
4.3.14 Conclusion............................................................................................................................................... 127

4.3.15 References............................................................................................................................................... 128
4.4 Case Study 4 - Flow Characteristics And Dynamic Responses of a Parked Straight
bladed Vertical Axis Wind Turbine (VAWT) ..................................................................................................... 131
4.4.1 Introduction ........................................................................................................................................... 131
4.4.2 Numerical Simulation Method........................................................................................................ 133
4.4.2.1 CFD ................................................................................................................................................ 133
4.4.2.2 Dynamic Responses By Finite Element Analysis ........................................................ 133
4.4.2.3 Two‐Way Coupling Procedure ........................................................................................... 133
4.4.3 Validation of the Two Way ............................................................................................................... 134
4.4.3.1 Coupling Method ...................................................................................................................... 134
4.4.4 Numerical Model .................................................................................................................................. 135
4.4.4.1 Wind Turbine Model .............................................................................................................. 136
4.4.4.2 Computational Fluid Dynamics Model............................................................................ 137
4.4.4.3 Finite Element Model ............................................................................................................. 138
4.4.4.4 Mesh Independence Test...................................................................................................... 139
4.4.4.5 Time Step Test .......................................................................................................................... 139
4.4.4.6 Two‐Way Coupling Simulation Setup ............................................................................. 140
4.4.4.7 Solver Settings .......................................................................................................................... 140
4.4.5.1 Surface Pressure Analysis .................................................................................................... 141
4.4.5.2 Flow Characteristics Analysis............................................................................................. 142
4.4.5.3 Dynamic Responses Analysis.............................................................................................. 144
4.4.6 Conclusions ............................................................................................................................................ 147
4.4.7 References............................................................................................................................................... 147
4.5 Case Study 5 - Mesh Sensitivity of Vertical Axis Turbine Wakes For Farm Simulations .... 149
4.5.1 Introduction ........................................................................................................................................... 149
4.5.2 Methods ................................................................................................................................................... 149
4.5.2.1 Turbine And Operating Conditions .................................................................................. 149
4.5.2.2 Numerical Model...................................................................................................................... 150
4.5.2.2.1 Mesh ........................................................................................................................................ 150
4.5.2.2.2 Turbulence Model and Numerical Procedure ....................................................... 152
4.5.3.1 Turbine Performance............................................................................................................. 152
4.5.3.2 Mesh Sensitivity of The Turbine Wake ........................................................................... 153
4.5.3.3 Comparison to Experimental Data ................................................................................... 154
4.5.4 Conclusions ............................................................................................................................................ 156
4.5.5 References............................................................................................................................................... 156
4.6 Case Study 6 - Numerical Investigations of the Savonius Turbine with Deformable
Blades ................................................................................................................................................................................ 157
4.6.1 Introduction ........................................................................................................................................... 157
4.6.2 Savonius Rotor with Deformable Blades ................................................................................... 158
4.6.3 Materials and Methods ...................................................................................................................... 160
4.6.3.1 Definition of Structural Simulations ................................................................................ 162
4.6.3.2 Definition of Flow Simulations........................................................................................... 163
4.6.3.3 Validation of the Numerical Procedure .......................................................................... 164
4.6.4.1 Blade Loading and Torque Generation ........................................................................... 167
4.6.5 Summary and Conclusions ............................................................................................................... 172
4.6.6 References............................................................................................................................................... 173
6
List of Tables:
Table 2.1 Geometrical and operational details of VAWT of the reference case ...................................... 35
Table 2.2 Details of Grids Used In VAWT ................................................................................................................ 39
Table 3.1 Design Space.................................................................................................................................................... 67
Table 3.2 Table of CFD Solver Settings ..................................................................................................................... 71
Table 3.3 Table Showing Optimized Candidate Point for Routine 1 ............................................................ 73
Table 3.4 Table Showing Optimized Candidate Point for Routine 2 ............................................................ 73
Table 3.5 Grid statistics for the grid independence study of the reference case ................................... 82
Table 4.1 Design Optimization Summary ................................................................................................................ 88
Table 4.2 Summary of Operation Analysis and Analysis Techniques and Optimizations ................... 89
Table 4.3 Summary of Novel Design Attachments and Configurations ...................................................... 90
Table 4.4 Summary of Novel Design Attachments and Configurations (Continued) ............................ 91
Table 4.5 Geometrical and Operational Parameters of the VAWT – [Courtesy of Ma et al. )............ 96
Table 4.6 Ranges of variables and specific values of variables for NACA0018 airfoil - [Courtesy of
Ma et al.) .................................................................................................................................................................................... 97
Table 4.7 Sensitivity test of grid size at TSR = 1.0 - [Courtesy of Ma et al.) .............................................. 99
Table 4.8 Settings of multi-island genetic algorithm [Courtesy of Ma et al.)......................................... 102
Table 4.9 The Operational Conditions of the VAWT Model .......................................................................... 112
Table 4.10 The Geometrical Features of the VAWT Model .......................................................................... 113
Table 4.11 Three Sets of Meshes ............................................................................................................................. 116
Table 4.13 The L25(56) Standard Orthogonal Array for the Parametric Study of the Winglet and the
Power Coefficients ............................................................................................................................................................. 120
Table 4.12 Design Parameters and Design Levels of the Winglet .............................................................. 120
Table 4.14 Mean Cp Distribution as a Function of the Different Design Parameters and Different
Parameter Levels ................................................................................................................................................................ 121
Table 4.15 The Data Employed in the Analysis Steps in the Study of the Parameter Significance
.................................................................................................................................................................................................... 122
Table 4.16 Oscillating plate model parameters ................................................................................................. 134
Table 4.17 Mesh Independence Test Based on The Relative Errors of Dynamic Responses of The
Top Section............................................................................................................................................................................ 139
Variable Section................................................................................................................................................................... 139
Table 4.19 Time step test based on the relative errors of dynamic responses of the variable section
.................................................................................................................................................................................................... 140
Table 4.20 Time step test based on the relative errors of dynamic responses of the top section 140
Table 4.21 Representative dynamic responses data of the support tower in the first parked
condition ................................................................................................................................................................................ 145
Table 4.22 Comparison of the representative dynamic responses data of the support tower in three
parked conditions under V2 = 10 m/s ........................................................................................................................ 147
Table 4.23 Details of the four mesh resolutions used for the outer domain ......................................... 151
Table 4.24 Main numerical schemes used for every simulation in this study ...................................... 152
Table 4.25 Coefficients of Power (CP) and Thrust (CT ) Obtained with Different Grid Resolutions in
the Turbine Wake. Values are Averaged Over the 80th Revolution. ............................................................ 153
Table 4.26 Mesh Size Dependence Study ............................................................................................................. 165
List of Figures:
Figure 1.1 Early Wind Turbines (left) Nashtifan wind turbines in Sistan, Iran. And (Right)-James
Blyth's electricity-generating wind turbine, photographed in 1891 ............................................................... 12
7
Figure 1.2 The first automatically operated wind turbine, built in Cleveland in 1887 by Charles F.
Brush. It was 60 feet (18 m) tall, weighed 4 tons (3.6 metric tons) and powered a 12 kW generator.[8]
....................................................................................................................................................................................................... 13
Figure 1.3 3 Primary Types of Wind Turbines ...................................................................................................... 14
Figure 1.4 The rotor of a gearless wind turbine being set. This particular turbine was prefabricated
in Germany, before being shipped to the U.S. for assembly. ................................................................................ 14
Figure 1.5 An Example of Vertical Axis Wind Turbines (VAWTs) ................................................................ 15
Figure 1.6 A vertical Axis Twisted Savonius type Turbine .............................................................................. 16
Figure 2.1 Schematic of a Wind Turbine Generation System- (Courtesy of Gaurav Kapoor)............ 19
Figure 2.2 Profiles of Flat-back and Sharp Trailing Edge Airfoils ................................................................. 19
Figure 2.3 Schematic of Internal Components of a Modern HAWT- (Courtesy of Gaurav Kapoor) 20
Figure 2.4 Blade Twist at Span-wise Sections (Airfoils) and Apparent Wind Angles ........................... 20
Figure 2.5 Swirling Flow in the Wind Turbine Wake ......................................................................................... 21
Figure 2.6 Typical Wind Turbine Blade Planform View.................................................................................... 21
Figure 2.7 Typical Wind Turbine Power Output Curve ..................................................................................... 22
Figure 2.8 Actuator Disk Concept for Wind Turbine Rotor ............................................................................. 23
Figure 2.9 Actuator Disk Concept, Pressure and Velocity Profiles ............................................................... 24
Figure 2.10 Test Rig of Wind-Turbine: (a) Schematic Diagram (b) Photo ................................................ 27
Figure 2.11 Scheme of Axial Type Wind-Turbine ................................................................................................ 28
Figure 2.12 Wind-Power Versus the Wind Velocity ........................................................................................... 30
Figure 2.13 Turbine Speed Versus the Wind Velocity, at Different Blade Angles .................................. 30
Figure 2.14 Total Generated Power Versus the Air Velocity, at Different Angles of Blade ................ 31
Figure 2.15 Modified Power-Coefficient Versus the Air Velocity, at Different Angles of Blade........ 32
Figure 2.16 Modified Power-Coefficients Versus Generated Power at Different Wind Velocities .. 32
Figure 2.17 Computational domain for VAWT (not drawn in scale). .......................................................... 36
Figure 2.18 Computational domain and grid: (a) grid near the rotating cell zone, (b) grid near the
airfoil, (c) grid near airfoil trailing edge ...................................................................................................................... 37
Figure 2.19 (a) Variation of moment coefficient with number of time steps, (b) Variation power
coefficient with number of revolutions of turbine .................................................................................................. 38
Figure 2.20 Variation of power coefficient with azimuthal increment ....................................................... 39
Figure 2.21 Validation of CFD Results with Established Experimental Results ...................................... 40
Figure 2.22 Variation of (a) average and (b) instantaneous moment coefficient with azimuthal angle
at U = 6.0 m/s. ......................................................................................................................................................................... 40
Figure 2.23 Variation of (a) power coefficient for different t/c turbine with TSR (b) power
coefficient at blade t/c = 0.3 for different wind speed ........................................................................................... 41
Figure 2.24 Variation of drag coefficient with azimuthal angle for different wind speeds (U = 6.0
and 7.0 m/s). ........................................................................................................................................................................... 42
Figure 2.25 Comparison of vortex structures in the flow field around a single blade for θ = 78° at
different t/c at U = 6.0 m/s. ............................................................................................................................................... 42
Figure 2.26 Pressure Coefficient Distribution for Different t/c Turbine at U = 6.0 m/s. .................... 43
Figure 2.27 Variation of moment coefficient for turbine having optimal t/c at different wind speeds.
....................................................................................................................................................................................................... 43
Figure 2.28 Comparison of vortex structures in the flow field around a single blade at different wind
speeds at 135° azimuthal position ................................................................................................................................. 44
speeds at 254° azimuthal position ................................................................................................................................. 44
Figure 2.30 Variations of wind speeds with heights, surfaces & times of a day ..................................... 49
Figure 2.31 Capital cost breakdown of an offshore wind farm [16] ............................................................ 50
8
Figure 3.1 Schematic figure showing the flow regions resulting from the interaction of a wind
turbine and incoming turbulent boundary layer. Depicted are the most characteristic instantaneous
(top) and time-averaged (bottom) flow features ..................................................................................................... 51
Figure 3.2 Flow visualizations of the 3D helical vortex structures behind a turbine rotor for different
values of tip-speed ratio λ (figure reprinted from Okulov et al. (2014) with permission of Cambridge
University Press) ................................................................................................................................................................... 52
Figure 3.3 Contours of the time-averaged streamwise velocity component (in m s−1) in the vertical
plane normal to the rotor plane, at zero span, for different roughness lengths. Figure taken from Wu
and Porté-Agel (2012), in accordance with the Creative Commons Attribution (CC BY) license) ...... 55
Figure 3.4 Distribution of, a streamwise turbulence intensity I = σu/ ūh, and b normalized kinematic
vertical turbulent momentum flux Bar(u’ w’)/ ū2h , in a vertical plane at zero span. Figure reprinted
from Barlas et al. (2016) with the permission of Springer Nature ................................................................... 56
Figure 3.5 Wake temporal oscillations at three rotor diameters downwind of a turbine. Velocity
contours obtained from lidar measurements in the field are shown in greyscale, and the red line
indicates the temporal variation of the wake center predicted by the DWM model. Figure reprinted
from Bingöl et al. (2010) with the permission of John Wiley and Sons, Inc.................................................. 57
Figure 3.6 Lateral (top) and vertical (bottom) profiles of the normalized velocity deficit through
the hub level at different downwind locations. The data obtained from wind-tunnel measurements
(Bastankhah and Porté-Agel 2017b) are shown by black solid lines. The predictions of the analytical
models developed by Jensen (1983), Frandsen et al. (2006) and Bastankhah and Porté-Agel (2014b)
are shown by red dashed lines, green dash-dot lines and blue dashed lines, respectively. ................... 61
Figure 3.7 AOC 15/50 Blade Geometry .................................................................................................................... 65
Figure 3.8 Turbine Blade Showing Various Radial Stations in ANSYS® DesignModeler.................... 66
Figure 3.9 Response Surface Showing Variation of P3, P9 with respect to P12 (Torque) .................. 68
Figure 3.10 Rotating Body in the Inertial Reference Frame............................................................................ 69
Figure 3.11 Grid Independence Study ...................................................................................................................... 71
Figure 3.12 Torque (P12) vs Twist_Station3 (P9) for Optimization Routine 2 ....................................... 73
Figure 3.13 Schematic Cross-Sectional Layout of the Three Dimensional Experimental Model used
for the Numerical Validation study. ............................................................................................................................... 78
Figure 3.14 Velocity contours colored with normalized free-stream velocity obtained using the (a)
panel method, (b) RANS method and (c) URANS method for Model C (ii) flap ........................................... 79
Figure 3.15 Effect of yawed inflow on the power augmentation factor: comparison between the
experiments (Igra, 1977), panel method and RANS method for different duct configurations. .......... 80
Figure 3.16 Panel distribution along the duct surface and the wake region used for the inviscid
panel method calculations ................................................................................................................................................. 81
Figure 3.17 Computational domain showing the boundary conditions employed. The length is
indicated in terms of duct chord length c (representative, not to scale). The computational grid along
the duct and flap leading edge is zoomed in to show the C-grid and boundary layer refinement. ..... 82
Figure 3.18 A schematic cross section of the multi-element duct– .............................................................. 83
panel method with no flap, (b) RANS method with no flap, (c) panel method with ζ = 5% and θ=10∘,
(d) RANS method with ζ = 5% and θ=10∘, (e) panel method with ζ = 5% and θ = 70∘, and (f) RANS
method with ζ= 5% and θ = 70∘ ....................................................................................................................................... 84
Figure 3.20 Comparing the Effect of the Variable Radial Gap and Deflection Angle of ........................ 85
Figure 3.21 Comparing the Effect of the Variable Radial Gap and Deflection Angle of the flap on the
power augmentation factor using Panel and RANS Methods................................................................................ 85
Figure 4.1 Schematic diagram of an H-rotor type VAWT: (a) 3D view; (b) top view - [Courtesy of Ma
et al) ............................................................................................................................................................................................ 96
Figure 4.2 Plan Views of the Computational Domain and Boundary Conditions - [Courtesy of Ma et
al.) ................................................................................................................................................................................................ 97
9
Figure 4.3 Mesh Setup (a) Topology; & BCs. (b) Grids at the Mid-Height Plane - [Courtesy of Ma et
al.) ................................................................................................................................................................................................ 98
Figure 4.4 Time-varying power coefficients in a rotational period when adopting different time
steps - [Courtesy of Ma et al.) ........................................................................................................................................ 100
Figure 4.5 Comparison of average power coefficients between the experiment and simulation
(NACA0018, freestream velocity V∞ = 8 m/s) as well as relative errors of them - [Courtesy of Ma et
al.) ............................................................................................................................................................................................. 100
Figure 4.6 The scheme of the optimization system - [Courtesy of Ma et al.)......................................... 102
Figure 4.7 Profiles of the initial airfoil and optimized airfoil ....................................................................... 103
Figure 4.8 Relative thickness distribution of the initial and optimized airfoil profiles - [Courtesy of
Ma et al.) ................................................................................................................................................................................. 103
Figure 4.9 Comparison of average power coefficients and the growth rate of CP between the initial
and optimized airfoils at different TSRs ................................................................................................................... 104
Figure 4.10 Instantaneous torque values during one rotational period - [Courtesy of Ma et al.) 105
Figure 4.11 Instantaneous Contours of Vorticity Magnitude at TSR = 0.9 for the VAWT with
Optimized Airfoils - [Courtesy of Ma et al.).............................................................................................................. 107
Figure 4.12 Schematic of the VAWT Geometry .................................................................................................. 112
Figure 4.13 Relationships Among Different Coordinate Systems on the Blade of VAWT ................ 113
Figure 4.14 Schematic of the computational domain for the one blade model (dimensions are not
to scale)................................................................................................................................................................................... 115
Figure 4.15 An illustration of the baseline computational mesh, showing the mesh topologies
around a blade and an adjacent arm on different vertical and horizontal sectional planes ............... 116
Figure 4.16 The prediction of both the power coefficient contribution of the turbine mid-span
section and the turbine overall power coefficients for the different ........................................................... 117
Figure 4.17 A comparison between the experimental and the numerical data of the torque
coefficient contribution at the turbine mid-span at a TSR of 2.29 ................................................................. 117
Figure 4.18 Schematic Diagram of the Parameters that Define the Winglet Geometry ..................... 119
Figure 4.20 Comparison of the tangential and normal force coefficients as a function of the
azimuthal angle between the reference blade and the blade with the optimized winglet. ................. 122
Figure 4.19 The designed winglet using the optimal variable level arrangement .............................. 122
Figure 4.21 16. The flow field characteristics when the azimuthal angle is 180°. (Left side: the
reference blade, right side: the blade with the optimal winglet). .................................................................. 123
Figure 4.22 The flow field characteristics when the azimuthal angle is 90°. (Left side: the reference
blade, right side: the blade with the optimal winglet). ....................................................................................... 123
Figure 4.23 The contours of the turbulence kinetic energy in normal and lateral planes when θ=90°
.................................................................................................................................................................................................... 124
Figure 4.24 An illustration of the swept area of the turbine with the selected winglet ................... 125
Figure 4.25 The influence of the winglet on the turbine power coefficient at different TSRs. ...... 126
Figure 4.26 Comparison Between the Torque Coefficient for Different Spanwise Locations at a TSR
of 2.29 with and without the Proposed Winglet.................................................................................................... 126
Figure 4.27 Flowchart illustrating the Two‐Way Coupling Between STAR‐CCM+ and ABAQUS .. 134
Figure 4.28 Computational model: A, computational domain; B, over view of the grids; C, grids
around the plate; D, 3D view of the finite element model ................................................................................. 135
Figure 4.29 Validation of the two‐way coupling method between the simulation and literature25
.................................................................................................................................................................................................... 135
Figure 4.30 Straight‐bladed VAWT model: A, 3D view; B, top view (the first parked condition, θ =
0°); C, top view (the second parked condition, θ = 90°); D, top view (the third parked condition, θ =
135°) ........................................................................................................................................................................................ 136
Figure 4.31 Computational domain and boundary conditions: A, 3D view; B, top view; C, front view
.................................................................................................................................................................................................... 137
10
Figure 4.32 3D view of the finite element model of the VAWT ................................................................... 138
Figure 4.33 Mesh topology of the computational domain: A, over view of the grids; B, mesh topology
around the VAWT; C, boundary layer grids of the support tower; D, boundary layer grids of the blade
.................................................................................................................................................................................................... 138
Figure 4.34 Contours of the instantaneous pressure distribution on the windward surface of the
VAWT in the first parked condition: A, V1 = 5 m/s; B, V2 = 10 m/s; C, V3 = 20 m/s ................................. 141
Figure 4.35 Contours of the instantaneous pressure distribution on the windward surface of the
VAWT in three parked conditions at V2 = 10 m/s: A, the first parked condition, θ = 0°; B, the second
parked condition, θ = 90°; C, the third parked condition, θ = 135° .............................................................. 142
Figure 4.36 Instantaneous contours of vorticity magnitude of the VAWT at different wind velocities:
A, B, V1 = 5 m/s; C, D, E, V2 = 10 m/s; F, G, V3 = 20 m/s....................................................................................... 143
Figure 4.37 Instantaneous contours of vorticity magnitude of the VAWT in three parked conditions
at V ............................................................................................................................................................................................ 144
Figure 4.38 Displacement responses in the x‐direction at critical sections of the support tower in
the first parked condition: A, D, V1 = 5 m/s; B, E, V2 = 10 m/s; C, F, V3 = 20 m/s..................................... 145
Figure 4.39 Velocity responses in the x‐direction at critical sections of the support tower in the first
parked condition: A, D, V1 = 5 m/s; B, E, V2 = 10 m/s; C, F, V3 = 20 m/s ...................................................... 146
Figure 4.40 Picture of the experimental turbine (a) and schematic of a horizontal section of the
rotor ......................................................................................................................................................................................... 150
Figure 4.41 Close view of the mesh in the rotor region ................................................................................. 150
Figure 4.42 D10 mesh around the turbine and in the near wake. Water flows from left to right.151
Figure 4.43 Comparison of measured and simulated coefficients of power (left) and coefficients of
thrust (right)......................................................................................................................................................................... 152
Figure 4.44 Instantaneous streamwise velocity (left) and turbulence kinetic energy (right) fields
after 40 revolutions for grids D10 to D80. The line plotted on the right side of the figures is located at
x = 6D. ...................................................................................................................................................................................... 154
Figure 4.45 Comparison of non-dimensioned streamwise velocity profiles (a) and turbulence
kinetic energy profiles (b) for grids D10, D20, D40 and D80 at five locations downstream of the
turbine..................................................................................................................................................................................... 155
Figure 4.46 Comparison of measured and simulated (with gridD40) non-dimensioned streamwise
velocity profiles (a) and turbulence kinetic energy profiles (b) at five locations downstream the
turbine..................................................................................................................................................................................... 155
Figure 4.47 Principle Of The Turbine Operation............................................................................................... 159
Figure 4.48 Computational domain scheme (external domain dimensions do not correspond to the
scale) ........................................................................................................................................................................................ 161
Figure 4.49 Turbine Motion Constrains In The Structural Simulation .................................................... 162
Figure 4.50 Computational mesh in the external and internal domains with details of its refinement
in the blade vicinity presented in the successive magnifications ................................................................... 163
Figure 4.51 Comparison of the power coefficient Cp for one revolution of the non-deformable rotor
from simulations with the rotating inner domain and the deformed/re-meshed one. Subsequent
positions of the turbine illustrated in the schemes above the graph correspond to the turbine
revolution angle. ................................................................................................................................................................. 165
Figure 4.52 Comparison of the averaged values of the power coefficient Cp for one revolution of
the ............................................................................................................................................................................................. 167
Figure 4.53 Comparison of the averaged values of the power coefficient Cp for each rotor blade for
one ............................................................................................................................................................................................ 168
Figure 4.54 Torque derivative (dT/dL) variation along the blade for selected instants ( 55° and
105°) ........................................................................................................................................................................................ 169
Figure 4.55 Pressure fields for selected instants (15∘, 55∘ and 105∘) of one revolution of the rotors
with non-deformable and deformable blade .......................................................................................................... 171
11
12
1 Introduction
1.1 Some Preliminaries & Definitions
According to Department of Energy, Wind turbines work on a simple principle: instead of using
electricity to make wind, like a fan, wind turbines use wind to make electricity. Wind turns the
propeller-like blades of a turbine around a rotor, which spins a generator, which creates
electricity. A wind turbine turns wind energy into electricity using the aerodynamic force from the
rotor blades, which work like an airplane wing or helicopter rotor blade. When wind flows across the
blade, the air pressure on one side of the blade decreases. The difference in air pressure across the
two sides of the blade creates both lift and drag. The force of the lift is stronger than the drag and this
causes the rotor to spin. The rotor connects to the generator, either directly (if it’s a direct drive
turbine) or through a shaft and a series of gears (a gearbox) that speed up the rotation and allow for
a physically smaller generator. This translation of aerodynamic force to rotation of a generator
creates electricity 1.
1.2 History
Main Source: Wikipedia
Main article: History of wind power
The wind wheel of Hero of Alexandria (10 AD–70 AD) marks one of the first recorded instances of
wind powering a machine in history [2][3]. However, the first known practical wind power plants were
built in Sistan, an Eastern province of Persia (now Iran), from the 7th century (see frame left of
Figure 1.1). These "Panemone" were vertical axle windmills, which had long vertical drive
shafts with rectangular blades [4]. Made of six to twelve sails covered in reed matting or cloth
material, these windmills were used to grind grain or draw up water, and were used in the grist
milling and sugarcane industries [5].
Wind power first appeared in Europe during the Middle Ages. The first historical records of their use
in England date to the 11th or 12th centuries and there are reports of German crusaders taking their
windmill-making skills to Syria around 1190 [6]. By the 14th century, Dutch windmills were in use to
drain areas of the Rhine delta. Advanced wind turbines were described by Croatian inventor [Fausto
Veranzio]. In his book Machinae Novae (1595) he described vertical axis wind turbines with curved
or V-shaped blades.
Figure 1.1 Early Wind Turbines (left) Nashtifan wind turbines in Sistan, Iran. And (Right)-James
Blyth's electricity-generating wind turbine, photographed in 1891
1 Enrgy.Gov , Office Of Energy Efficiency & Renewable Energy

13
The first electricity-generating wind turbine

was a battery charging machine installed in
July 1887 by Scottish academic James Blyth to
light his holiday home in Marykirk, Scotland
(see frame right of Figure 1.1). Some months
later American inventor [Charles F.
Brush] was able to build the first automatically
operated wind turbine after consulting local
University professors and colleagues Jacob S.
Gibbs and Brinsley Coleberd and successfully
getting the blueprints peer-reviewed for
electricity production in Cleveland,
Ohio.[3] Although Blyth's turbine was
considered uneconomical in the United
Kingdom,[7] electricity generation by wind
turbines was more cost effective in countries
with widely scattered populations.[6]
The first automatically operated wind turbine,
built in Cleveland in 1887 by Charles F. Brush.
It was 60 feet (18 m) tall, weighed 4 tons (3.6 Figure 1.2 The first automatically operated wind
metric tons) and powered a 12 kW generator. turbine, built in Cleveland in 1887 by Charles F.
(See Figure 1.2) . [8] Brush. It was 60 feet (18 m) tall, weighed 4 tons (3.6
In Denmark by 1900, there were about 2500 metric tons) and powered a 12 kW generator.[8]
windmills for mechanical loads such as pumps
and mills, producing an estimated combined peak power of about 30 MW. The largest machines were
on 24-meter (79 ft) towers with four-bladed 23-meter (75 ft) diameter rotors. By 1908, there were
72 wind-driven electric generators operating in the United States from 5 kW to 25 kW. Around the
time of World War I, American windmill makers were producing 100,000 farm windmills each year,
mostly for water-pumping.[9]
By the 1930s, wind generators for electricity were common on farms, mostly in the United States
where distribution systems had not yet been installed. In this period, high-tensile steel was cheap,
and the generators were placed atop prefabricated open steel lattice towers.
A forerunner of modern horizontal-axis wind generators was in service at Yalta, USSR in 1931. This
was a 100 kW generator on a 30-meter (98 ft) tower, connected to the local 6.3 kV distribution
system. It was reported to have an annual capacity factor of 32 percent, not much different from
current wind machines.[10][11]
In the autumn of 1941, the first megawatt-class wind turbine was synchronized to a utility grid
in Vermont. The [Smith–Putnam] wind turbine only ran for 1,100 hours before suffering a critical
failure. The unit was not repaired, because of a shortage of materials during the war.
The first utility grid-connected wind turbine to operate in the UK was built by John Brown &
Company in 1951 in the Orkney Islands.[7][12]
Despite these diverse developments, developments in fossil fuel systems almost entirely eliminated
any wind turbine systems larger than supermicro size. In the early 1970s, however, anti-nuclear
protests in Denmark spurred artisan mechanics to develop microturbines of 22 kW. Organizing
owners into associations and co-operatives lead to the lobbying of the government and utilities and
provided incentives for larger turbines throughout the 1980s and later. Local activists in Germany,
nascent turbine manufacturers in Spain, and large investors in the United States in the early 1990s
then lobbied for policies that stimulated the industry in those countries.
It has been argued that expanding use of wind power will lead to increasing geopolitical competition
over critical materials for wind turbines such as rare earth elements neodymium, praseodymium,
14
and dysprosium. But this perspective has been criticized for failing to recognize that most wind
turbines do not use permanent magnets and for underestimating the power of economic incentives
for expanded production of these minerals.[13]
1.2.1 Types of Wind Turbines
The three primary types are: VAWT Savonius, HAWT towered; VAWT Darrieus as they appear in
operation Wind turbines (Figure 1.3) can rotate about either a horizontal or a vertical axis, the
former being both older and
more commonn[25]. They can
also include blades, or be
bladeless [26]. Vertical designs
produce less power and are less
common.[27]
1.2.1.1 Horizontal Axis
Large three-bladed horizontal-
axis wind turbines (HAWT) with
the blades upwind of the tower
produce the overwhelming
majority of wind power in the
world today (see Figure 1.4).
These turbines have the Figure 1.3 3 Primary Types of Wind Turbines
main rotor shaft and electrical
generator at the top of a tower,
and must be pointed into the wind. Small turbines are pointed by a simple wind vane, while large
turbines generally use a wind sensor coupled with a yaw system. Most have a gearbox, which turns
the slow rotation of the blades into a
quicker rotation that is more suitable
to drive an electrical
generator.[28] Some turbines use a
different type of generator suited to
slower rotational speed input. These
don't need a gearbox and are called
direct-drive, meaning they couple the
rotor directly to the generator with no
gearbox in between. While permanent
magnet direct-drive generators can be
more costly due to the rare earth
materials required,
these gearless turbines are sometimes
preferred over gearbox generators
because they "eliminate the gear-speed
increaser, which is susceptible to Figure 1.4 The rotor of a gearless wind turbine being set.
significant accumulated fatigue torque This particular turbine was prefabricated in Germany,
loading, related reliability issues, and before being shipped to the U.S. for assembly.
maintenance costs." There is also
[29]
the pseudo direct drive mechanism, which has some advantages over the permanent magnet direct
drive mechanism.[30][31]. The rotor of a gearless wind turbine being set. This particular turbine was
prefabricated in Germany, before being shipped to the U.S. for assembly.
Most horizontal axis turbines have their rotors upwind of the supporting tower. Downwind machines
have been built, because they don't need an additional mechanism for keeping them in line with the
15
wind. In high winds, the blades can also be allowed to bend, which reduces their swept area and thus
their wind resistance. Despite these advantages, upwind designs are preferred, because the change
in loading from the wind as each blade passes behind the supporting tower can cause damage to the
turbine.
Turbines used in wind farms for commercial production of electric power are usually three-bladed.
These have low torque ripple, which contributes to good reliability. The blades are usually colored
white for daytime visibility by aircraft and range in length from 20 to 80 meters (66 to 262 ft). The
size and height of turbines increase year by year. Offshore wind turbines are built up to 8 MW today
and have a blade length up to 80 meters (260 ft). Designs with 10 to 12 MW are in
preparation.[32] Usual multi megawatt turbines have tubular steel towers with a height of 70 m to
120 m and in extremes up to 160 m.
1.2.1.2 Vertical Axis
Vertical-axis wind turbines (or VAWTs) have the
main rotor shaft arranged vertically (Figure 1.5).
One advantage of this arrangement is that the
turbine does not need to be pointed into the wind
to be effective, which is an advantage on a site
where the wind direction is highly variable. It is
also an advantage when the turbine is integrated
into a building because it is inherently less
steerable. Also, the generator and gearbox can be
placed near the ground, using a direct drive from
the rotor assembly to the ground-based gearbox,
improving accessibility for maintenance. However,
these designs produce much less energy averaged
over time, which is a major drawback.[27][33]
The key disadvantages include the relatively low
rotational speed with the consequential
higher torque and hence higher cost of the drive
train, the inherently lower power coefficient, the Figure 1.5 An Example of Vertical Axis Wind
360-degree rotation of the aero foil within the Turbines (VAWTs)
wind flow during each cycle and hence the highly
dynamic loading on the blade, the pulsating torque
generated by some rotor designs on the drive train, and the difficulty of modelling the wind flow
accurately and hence the challenges of analyzing and designing the rotor prior to fabricating a
prototype.[34]
When a turbine is mounted on a rooftop the building generally redirects wind over the roof and this
can double the wind speed at the turbine. If the height of a rooftop mounted turbine tower is
approximately 50% of the building height it is near the optimum for maximum wind energy and
minimum wind turbulence. While wind speeds within the built environment are generally much
lower than at exposed rural sites,[35][36] noise may be a concern and an existing structure may not
adequately resist the additional stress. Sub-types of the vertical axis design are include below.
1.2.1.2.1 Darrieus Wind Turbine
"Eggbeater" turbines, or Darrieus turbines, were named after the French inventor, Georges Darrieus
[37]. They have good efficiency, but produce large torque ripple and cyclical stress on the tower, which
contributes to poor reliability. They also generally require some external power source, or an
additional Savonius rotor to start turning, because the starting torque is very low. The torque ripple
is reduced by using three or more blades, which results in greater solidity of the rotor. Solidity is
16
measured by blade area divided by the rotor area. Newer Darrieus type turbines are not held up
by guy-wires but have an external superstructure connected to the top bearing.[38]
1.2.1.2.2 Giromill
A subtype of Darrieus turbine with straight, as opposed to curved, blades. The cycloturbine variety
has variable pitch to reduce the torque pulsation and is self-starting.[39] The advantages of variable
pitch are: high starting torque; a wide, relatively flat torque curve; a higher coefficient of
performance; more efficient operation in turbulent
winds; and a lower blade speed ratio which lowers blade
bending stresses. Straight, V, or curved blades may be
used.[40]
1.2.1.2.3 Savonius Wind Turbine
These are drag-type devices with two (or more) scoops
that are used in anemometers, Flettner vents (commonly
seen on bus and van roofs), and in some high-reliability
low-efficiency power turbines. They are always self-
starting if there are at least three scoops. Twisted
Savonius is a modified savonius, with long helical scoops
to provide smooth torque. This is often used as a rooftop
wind turbine and has even been adapted for ships.[41].
(see Figure 1.6). For a numerical study of Savonius
turbine, reader are encourage to consult the work by
[Hassan Saeed et al.]2.
1.2.1.2.4 Parallel
The parallel turbine is similar to the crossflow fan or
centrifugal fan. It uses the ground effect. Vertical axis
Figure 1.6 A vertical Axis Twisted
turbines of this type have been tried for many years: a Savonius type Turbine
unit producing 10 kW was built by Israeli wind pioneer
Bruce Brill in the 1980s.[42]
1.2.2 References
[1] Evans, Annette; Strezov, Vladimir; Evans, Tim (June 2009). "Assessment of sustainability indicators
for renewable energy technologies". Renewable and Sustainable Energy Reviews..
[2] Drachmann, A.G. (1961). "Heron's Windmill". Centaurus. 7: 145–151.
[3] Dietrich Lohrmann, "Von der östlichen zur westlichen Windmühle", Archiv für Kulturgeschichte,
(1995).
[4] Ahmad Y Hassan, Donald Routledge Hill (1986). Islamic Technology: An illustrated history, p.
4. Cambridge University Press. ISBN 0-521-42239-6.
[5] Donald Routledge Hill, "Mechanical Engineering in the Medieval Near East", Scientific American,
May 1991, pp. 64–69. (cf. Donald Routledge Hill, Mechanical Engineering) en.wind-turbine-
models.com.
[6] Morthorst, Poul Erik; Redlinger, Robert Y.; Andersen, Per (2002). Wind energy in the 21st century:
economics, policy, technology and the changing electricity industry. Houndmills, Basingstoke,
Hampshire: Palgrave/UNEP. ISBN 978-0-333-79248-3.
[7] Price, Trevor J. (2004). "Blyth, James (1839–1906)". Oxford Dictionary of National
Biography (online ed.). Oxford University Press.
2 Hassan A.Hassan Saeed, Ahmed M. Nagib Elmekawy, Sadek Z. Kassab, “Numerical study of improving Savonius
turbine power coefficient by various blade shapes”, Alexandria Engineering Journal (2019) 58, 429–441.
17
[8] A Wind Energy Pioneer: Charles F. Brush. Danish Wind Industry Association. Archived from the
original on 8 September 2008. Retrieved 28 December 2008.
[9] "Quirky old-style contraptions make water from wind on the mesas of West Texas". Archived
from the original on 3 February 2008.
[10] Alan Wyatt: Electric Power: Challenges and Choices. Book Press Ltd., Toronto 1986, ISBN 0-
920650-00-7
[11] "Bauer, Lucas. "Krasnovsky WIME D-30 – 100,00 kW – Wind turbine".
[12] Anon. "Costa Head Experimental Wind Turbine". Orkney Sustainable Energy Website. Orkney
Sustainable Energy Ltd. Retrieved 19 December 2010.
[13] Overland, Indra (1 March 2019). "The geopolitics of renewable energy: Debunking four emerging
myths". Energy Research & Social Science.
[14] "NREL: Dynamic Maps, GIS Data, and Analysis Tools – Wind Maps". Nrel.gov. 3 September 2013.
Retrieved 6 November 2013.
[15] Appendix II IEC Classification of Wind Turbines. Wind Resource Assessment and Micro-siting,
Science and Engineering. 2015. pp. 269–
270. doi:10.1002/9781118900116.app2. ISBN 9781118900116.
[16] "The Physics of Wind Turbines Kira Grogg Carleton College, 2005, p. 8"(PDF). Retrieved 6
November 2013.
[17] "Wind Energy Basics". Bureau of Land Management. Retrieved 23 April2016.
[18] "Enercon E-family, 330 Kw to 7.5 MW, Wind Turbine Specification"(PDF). Archived from the
original (PDF) on 16 May 2011.
[19] Tony Burton et al., (ed), Wind Energy Handbook, John Wiley and Sons
2001 ISBN 0471489972 page 65.
[20] Sanne Wittrup. "11 years of wind data shows surprising production decrease" (in
Danish) Ingeniøren, 1 November 2013. Retrieved 2 November 2013.
[21] Barber, S.; Wang, Y.; Jafari, S.; Chokani, N.; Abhari, R. S. (28 January 2011). "The Impact of Ice
Formation on Wind Turbine Performance and Aerodynamics". Journal of Solar Energy
Engineering. 133 (1): 011007–011007–9. doi:10.1115/1.4003187. ISSN 0199-6231.
[22] Hau., Wind Turbines: Fundamentals, Technologies, Application, Economics. Springer. Germany.
2006
[23] "Atmospheric stability and topography effects on wind turbine performance and wake properties
in complex terrain". doi:10.1016/j.renene.2018.03.048.
[24] Ozdamar, G. (2018). "Numerical Comparison of the Effect of Blade Material on Wind Turbine
Efficiency". Acta Physica Polonica A. 134: 156–158. doi:10.12693/APhysPolA.134.156.
[25] "Wind Energy Basics". American Wind Energy Association. Archived from the original on 23
September 2010. Retrieved 24 September 2009.
[26] Elizabeth Stinson (15 May 2015). "The Future of Wind Turbines? No Blades". Wired.
[27] Jump up to:a Paul Gipe (7 May 2014). "News & Articles on Household-Size (Small) Wind
Turbines". Wind-works.org.
[28] "Wind Turbine Components". Danish Wind Industry Association. 10 May 2003. Archived from the
original on 2008.
[29] G. Bywaters; P. Mattila; D. Costin; J. Stowell; V. John; S. Hoskins; J. Lynch; T. Cole; A. Cate; C.
Badger; B. Freeman (October 2007). "Northern Power NW 1500 Direct-Drive Generator" (PDF).
National Renewable Energy Laboratory.
[30] Magnetic Pseudo direct drive generator
[31] Innwind: Overview of the project and research
[32] "MHI Vestas Launches World's First* 10 Megawatt Wind Turbine". 26 September 2018.
[33] Michael Barnard (7 April 2014). "Vertical Axis Wind Turbines: Great In 1890, Also-rans In
2014". CleanTechnica.
18
[34] Michael C Brower; Nicholas M Robinson; Erik Hale (May 2010). "Wind Flow Modeling
Uncertainty" (PDF). AWS Truepower. Archived from the original on 2 May 2013.
[35] Hugh Piggott (6 January 2007). "Windspeed in the city – reality versus the DTI database".
Scoraigwind.com. Retrieved 6 November 2013.
[36] "Urban Wind Turbines" (PDF).
[37] "Vertical-Axis Wind Turbines". Symscape. 7 July 2008. Retrieved 6 November 2013.
[38] Exploit Nature-Renewable Energy Technologies by Gurmit Singh, Aditya Books, pp 378
[39] Eric Eggleston & AWEA Staff. "What Are Vertical-Axis Wind Turbines (VAWTs)?". American Wind
Energy Association. Archived from the original on 3 April 2005.
[40] Marloff, R.H. (January 1978). "Stresses in turbine-blade tenons subjected to
bending". Experimental Mechanics. 18 (1): 19–24. doi:10.1007/BF02326553.
[41] Rob Varnon (2 December 2010). "Derecktor converting boat into hybrid passenger
ferry". Connecticut Post. Retrieved 25 April 2012.
[42] "Modular wind energy device – Brill, Bruce I". Freepatentsonline.com. 19 November 2002.
Retrieved 6 November 2013.
[43] Hewitt, Sam; Margetts, Lee & Revell, Alistair (18 April 2017). "Building a digital wind
farm". Archives of Computational Methods in Engineering. 25(4): 879–899.
[44] Navid Goudarzi (June 2013). "A Review on the Development of the Wind Turbine Generators
across the World". International Journal of Dynamics and Control. 1 (2): 192–
202. doi:10.1007/s40435-013-0016-y.
[45] Navid Goudarzi; Weidong Zhu (November 2012). "A Review of the Development of Wind Turbine
Generators Across the World". ASME 2012 International Mechanical Engineering Congress and
Exposition. 4 – Paper No: IMECE2012-88615: 1257–1265.
19
2 Wind Turbine Design and Working Principle

The structure itself is rather simple and fairly common nowadays (see Figure 2.1). The rotor is
made of generally three blades fixed to a hub. The hub is responsible for the blade control and for
connecting the rotor mechanism to the rotor shaft (and consequently to the electrical generator). The
nacelle (see Figure 2.3) is the enclosure that holds all mechanical organs of the machine (gearbox,
rotor brake, bearings, etc.) as well as the generator and control systems. The bedplate, which
connects the nacelle to the tower, is responsible for a very important movement of the WT, the yaw
system, allowing the HAWT to face the direction of the wind flow. Finally, the tower is the structure
that holds the machine in place and that connects the HAWT to the electrical grid.
Figure 2.1 Schematic of a Wind Turbine Generation System- (Courtesy of Gaurav Kapoor)
2.1 Airfoils and Blade Design

The most important factor in designing a wind turbine is the choice of airfoils from which the blade
gets it aerodynamic shape, as the entire blade is shape lofted from these airfoils sections. The lift
generated from these airfoils at every section causes the rotation of the blade, also the performance
of the blade is highly dependent on airfoil performance. The airfoil near the blade root are usually
thicker and are flat-back (or rounded trailing edge) to make the blade thicker at the root section. The
airfoil section at the tip of the blade has a sharp trailing edge for achieving higher tip speed ratio. The
airfoil sections closer to the
tip of the blade generate
higher lift force due to the
speed variation in the
relative wind, the purpose
of airfoils at the root of
blade is mainly structural,
having a minimal
contribution to the Figure 2.2 Profiles of Flat-back and Sharp Trailing Edge Airfoils
aerodynamic performance
of the blade. Thus the root section of the wind turbine blade is thicker and stronger than its tip section
(Figure 2.2). Wind turbine blades are shaped to extract maximum power from the wind at the
20
minimum cost involved. Primarily the blade design is driven by the aerodynamic and performance
requirements. But in true sense, the economics mean that the blade shape is a compromise to keep
the cost of construction, operation and maintenance to a minimum. The blade design procedure
starts with obtaining a solution set for both aerodynamic and structural efficiency. The best blade
design is a tradeoff between both aerodynamic performance and structural stiffness.
Figure 2.3 Schematic of Internal Components of a Modern HAWT- (Courtesy of Gaurav Kapoor)
2.1.1 Blade Twist

Analogous to an airplane wing, wind turbine blades work by generating lift force due to their airfoil
shape. The more curved side generates low air pressures while high pressure air pushes on the
pressure side of the airfoil. The net result of this pressure difference on either side of the blade
surface is a lift force perpendicular to the direction of flow of the air. Since the turbine blade is in
motion, the true wind is
incident on it from a different
angle. This is called apparent
wind as shown in Figure 2.4.
The apparent wind is stronger
than the true wind but its
angle is less favorable to
generate a driving force on the
blade. This also means that the
lift force contributes to the
thrust the rotor. To maintain
an effective angle of attack to
generate lift, the blade must be Figure 2.4 Blade Twist at Span-wise Sections (Airfoils) and Apparent
turned further from the true Wind Angles
21
wind angle which gives twist to the blade from root to tip. It can be seen that the blade tip is moving
faster through the air compared to the blade region closer to the root, hence the tip is operating at a
greater apparent wind angle. Thus, the blade needs to be turned further at the tips than at the root,
which essentially means it must be built with an inherent twist along is length. The requirement to
twist the blades has implications on the manufacturing processes.
2.1.2 Tip Speed Ratio (TSR)
The rotational speed at
which the turbine
operates is a
fundamental choice in
the blade design. It is
defined in terms of the
speed of the blade tips
relative to the free wind
speed. This is called the
tip speed ratio (λ) and
its definition as shown:
ωR
λ=
v0
Eq. 2.1
Where, 𝜔 is the angular
velocity of the wind
turbine rotor, 𝑅 is
radius of the rotor and Figure 2.5 Swirling Flow in the Wind Turbine Wake
𝑣0 is the free wind
speed. A higher tip speed ratio (TSR) induces the net aerodynamic force on the blade (component of
lift and drag) to be approximately parallel to the rotor axis . The lift to drag ratio can be affected
severely by presence of dirt or roughness on the blade surfaces3. Low tip speed ratio unfortunately
results is lower
aerodynamic efficiency
due to two effects. Since
the lift force on the
blade generates torque,
according to the laws of
motion, it has an equal
but opposite effect on
the incident wind,
tending to push it
around tangentially in
the other direction. As a Figure 2.6 Typical Wind Turbine Blade Planform View
result, the air
downwind of the turbine has a swirl, i.e. it spins in the opposite direction to the blade rotation, as
depicted in Figure 2.5. This swirl represents lost power which reduces the available power that can
be extracted from the incident wind. Lower rotational speed requires higher torque to maintain the
same power output, so lower tip speed ratio results in greater wake swirl losses.
3 Nianxin Ren , Jinping Ou, “Dust Effect on the Performance of Wind Turbine Airfoils”,
J.Electromagnetic Analysis
& Applications, 2009, 1: 102-107, doi:10.4236/jemaa.2009.12016, Published Online June 2009.
22
The other reason for the reduction in aerodynamic efficiency at low tip speed ratio is due to the tip
losses, where high-pressure air from the upwind side of the blade escapes around the blade tip to the
low-pressure side, thereby wasting energy. Since power is a product of blade torque and rotational
speed, at slower rotational speed the blades need to generate more lift force to maintain the same
power output. In order to generate greater lift for a given length, the blade has to be wider,
geometrically speaking, a greater proportion of the blade’s width is designed to be close to the tip
(Figure 2.6). The higher lift force on a wider blade translates to greater structural loads on the outer
components such as the hub and bearings. There are practical limits on the absolute tip speed ratio
as well. At these speeds, bird impacts and rain erosion starts to decrease the longevity of the blades
and noise increases dramatically with the tip speed.
2.1.3 Wind Turbine Operation
Wind turbine operating condition depends on the speed of free stream wind speed; generally, it can
be divided into three operation modes (Figure 2.7),
• Cut-in speed -
the minimum
wind speed at
which the
turbine
blades
overcome
frictional
force and
begin to
rotate.
• Operation
mode - the
range of wind
Figure 2.7 Typical Wind Turbine Power Output Curve
speeds
within which
the wind turbine actively
generates power.
• Cut-out mode - the speed at which the turbine is brought to rest to avoid structural
damage due to high wind speeds.
23
2.1.4 Wind Turbine Aerodynamics

A wind turbine extracts mechanical energy from the kinetic energy of the wind by slowing down the
wind. It can either be a Horizontal-Axis Wind Turbine (HAWT) or a Vertical-Axis Wind Turbine
(VAWT), depending on either it rotates around its horizontal axis or vertical axis, respectively As
discussed earlier, many methods for computing the aerodynamic performance of wind turbines exist.
In 1935, [Betz and Glauert]4 derived the classical analysis method, the Blade Element Momentum
Theory (BEMT), which combines the Blade Element and Momentum theories. But at the present,
only flow equations from the Actuator Disc concept are used and the same will be discussed below.
Figure 2.8 Actuator Disk Concept for Wind Turbine Rotor
2.1.5 Actuator Disk Concept

The actuator disk concept is widely used to define the basic aerodynamic flow around the wind
turbine. According to this concept, the wind turbine is considered as an ideal actuator disk:
frictionless, with an infinite number of blades and with no rotational velocity component in the wake
downstream of the turbine. The flow around the turbine is assumed to be homogeneous and steady,
while the air is considered incompressible. If the mass of air passing through the turbine is assumed
to be separated from the mass that does not pass, the separated part of the flow field remains a long
stream tube lying up and downstream of the turbine. As the flow approaches the wind turbine, it
suffers a velocity drop, and in order to compensate for this drop, the stream tube expands (Figure
2.8). From Figure 2.9, the non-dimensionalized difference between the free stream velocity 𝑣0 and
axial induced velocity 𝑢, the axial induction factor is defined as:
V0 − U
a=
V0
Eq. 2.2
The shaft power 𝑃𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 is calculated by using the energy equation on a control volume defined
by the stream tube and assuming no change in the internal energy of the flow (since it is assumed to
be frictionless). The power available is;
4 Albert Betz, “Betz Law for Wind Turbines”, http://en.wikipedia.org/wiki/Betz%27s_law.

24
P = 2ρv0 a(1 − a)AR

Eq. 2.3
where 𝐴𝑅 is the area of the rotor and which
is often non-dimensionalized with respect
to 𝑃𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 as a power coefficient 𝐶𝑃,
Pavialable
CP =
(1/2)AR ρv03
Eq. 2.4
The power coefficient for the ideal wind
turbine may also be written as:
CP = 4a (1 − a)2
dCP
= a(1 − a)(1 − 3a)
da
Eq. 2.5
The maximum value of 𝐶𝑃 = 16⁄27 = ~
0.593 is obtained for 𝑎 = 1⁄3. This
theoretical maximum value is known as the
Betz Limit and it is not possible to design a
wind turbine that goes beyond this
theoretical limit. In other words, according Figure 2.9 Actuator Disk Concept, Pressure and
to the Betz's law, no turbine can capture Velocity Profiles
more than 16/27 (~ 59.3%) of the kinetic
energy in wind.
25
2.2 Case Study 1 - Effect of Blade Angles on the Performance of Axial Six Blades Wind
Turbine
Authors : Mishaal A. AbdulKareem, Ammar A. Hussain, and Raid S. Fahad.
Appeared in : International Journal of Engineering & Technology, 8 (1.5) (2019) 580-586.
Source : Website: www.sciencepubco.com/index.php/IJET
In this paper, the performance of a six blades axial type wind turbine has been studied experimentally
to estimate the wind power, the electrical generated power and-the modified power-coefficient of
the wind-turbine. This study was conducted under different operating conditions assuming steady-
state, incompressible and isothermal air flow through the wind-turbine. The range of operating
condition was (2 to 5.6 m/s wind speed), (10% to 100% of electrical load that is applied on the
terminals of the electrical generator) and (10° to 80° blades angle of the wind-turbine). A good
agreement was obtained when comparing the results of the present work with those of a previously
published article. The predicted results showed that increasing the wind speed and-the blades
angle of the wind-turbine will increase the generated power from the wind-turbine. The
maximum-value of the modified power-coefficient was (0.57) at a wind velocity value of (5.6
m/s) and at a blades angle value of (80°). It is found that it’s not recommended to operate the
wind-turbine at (80°) blades angle associated with a wind speed range that is above (3.8 m/s)
due to a high level of wind-turbine vibration.
2.2.1 Introduction & Literature Survey
The new technology of power-generation from wind energy is growing rapidly and it has technical
and financial challenges including the issues of low performance and high capital costs. These
technical challenges are mainly to overcome the operation of the power unit at frequent change of
wind direction and at extremely low wind velocity. In general, a wind-turbine is defined as a
mechanical device that converts the kinetic-energy of wind into mechanical-energy and then to
electrical-power to be utilized to drive water pumps, wind mills or other applications.
In 2009, few countries have achieved high levels of wind power generation, such as the 20% of power
production in Denmark, 14% in Ireland and Portugal, the 11% in Spain, and 8% in Germany. In May
2009, all countries around the world was using wind energy on a commercial basis. At the end of
2009, the overall capacity of wind-power generators was 159.2 GW (Gigawatts). In 2014, the global
capacity of wind energy increased up to 369,553 GW. The production of wind power is growing
rapidly and has reached approximately 4% of the overall electrical power usage in the last years. Few
regions in Iraq have a wind velocity range of (3.5 - 5 m/s) which is useful for the wind-energy
application. These local areas are more suited for this application, due to being remote and generally
small and scattered population.
(Medici) Investigated experimentally the effect of a wake behind a wind-turbine model with one, two
and three blades to calculate the output power from the turbines. The yaw-angle of the blade was
varied from 0 to 30 degrees. Three velocity-components were measured for the flow field. It was
found that a yawed-turbine deflects the wake to the side. The output power of the turbine depends
on the rotational-speed of the wind-turbine rotor (Khalil). Investigated the performance analysis of
a wind-turbine on the basis of a different modelling approach of wind-turbine. The dynamics of wind-
turbine has been investigated to avoid unpredictable outputs and to ensure that continuous and
efficient power is supplied according to the load requirements. The work was the initial step of the
control of wind-turbine. The performance and transfer function of wind-turbine were derived by
using the commercial software MATLAB. The research showed different approach of wind-turbine
modelling.
The design of a position control system for the blade pitch angle of a variable speed wind-turbine
generators were studied by (Ahmed). The quality of power has been enhanced by using suitable
control technique in the system. The fluctuating output of wind power generator has been controlled
26
to study the dynamic characteristics of the wind-turbine. Blade pitch control has been discussed by
using rotary potentiometers for position control and some experimental tests as a simulation of the
position control.
(Jasim) investigated theoretically the dynamics of a small wind-turbine. The aerodynamic forces
were predicted at any variable speed wind-turbine blade. The results were presented by using a code
of FORTRAN 90. In this code, the torque and the output power were both estimated at any wind-
speed. The results were compared with those predicted using Betz theorem. Additionally, the
numbers of revolutions of the small turbine were evaluated for any wind-speed. The suiting tip speed
ratio and wind-speed have been determined from this model.
(Toshimitsu et al.) Predicted the performance of a compact type flanged diffuser shroud wind-
turbine in a sinusoidal oscillating and fluctuating velocity for both stable and un-stable wind flow.
The results show that the efficiency of compact type wind-turbine is high when compared with that
having only rotor. The performance of both wind-turbines for unsteady flow were studied
numerically and experimentally. Predicted results were visualized. The accuracy of the predict
results of the power-coefficient in the oscillating flows was (94 – 102 %).
Experimental studies of a shrouded, horizontal axis micro-wind-turbine with diffuser augmented
were investigated by (Kosasih) to enhance the wind-turbine performance. The enhancement of
performance depends on many parameters including diffuser geometry, blade airfoil, and-the wind-
speed. Three different shapes of diffusers were selected. It was found that the performance
enhancement ratio with diffuser was only 60% when compared with that of the bare turbine while it
was 63% for nozzle-diffuser enhancement.
(Rachman) Simulated the yearly-energy-production of a wind-turbine. This simulation was
developed using a mathematical model of blade momentum. (Abdin et al.) Manufactured a three
blade PVC wind turbine and tested its efficiency at three values of wind speed (4.0, 5.1 and 6.1 m/s)
and at five values of turbine blade pitch angles (10o, 15o, 20o, 25o and 30o). It was found that there
is no homogeneous trend of voltage and power increase with the increase in turbine blade pitch
angle. In addition, the best turbine performance showed at 15o turbine pitch angle.
(Aldair) Performed a study of the pitch-angle control design of wind-turbines using Fuzzy-ART
network. The power extracted from the wind-turbine can be enhanced by adjusting the blades pitch-
angle of the wind-turbine. In the work, the fuzzy- ART (Adaptive Resonance Theory) system has been
utilized to control the angle between the approaching wind direction and-the chord-line of the blade.
The results show that the system controller is very effective to adjust the blade angles.
(Dakeev) Performed a study of analyses of wind-power-generation with application of wind tunnel
attachment on wind-turbine unit. The wind tunnel apparatus is used for enhancement the power-
generation efficiency of a wind-turbine. The power generated by the wind-turbine at various wind-
speeds was utilized to develop a characteristic performance curve. The experimental work included
operating the wind-turbines at variable wind velocity values with and without the wind tunnel
attachment. Experimental results indicate that the power generated when using the wind tunnel
attachment was 60% higher when compared to that without the wind tunnel attachment.
In this paper, the performance of a six blades axial type wind-turbine will be studied experimentally
to estimate the wind-power, the electrical generated power and-the modified power-coefficient of
the wind-turbine. In addition, it will investigate the effect of altering the wind-speed, the electrical
load that is applied on the terminals of the electrical generator and altering the blades angle of the
wind turbine to estimate the optimum operating condition of the wind-turbine. To validate the
results of the present work, it will be compared with results of a previously published article.
2.2.2 Experimental Work
Figure 2.10-a illustrates the schematic diagram of experimental apparatus. It consists basically of
an axial fan, tunnel, rotor of six blades, speed sensor, wattmeter and loads module. The axial flux fan
introduces to supply air in the tunnel of (2 to 5.6 m/s), to simulate the same wind-speed range in
27
many regions in Iraq that have a wind-speed range of (3.5 to 5 m/s) which is useful for the wind-
energy applications (Darwish and Sayigh) and (Darwish and Sayigh). The tunnel is manufactured
from a stainless steel of (2 x 0.55 x 0.55m) approximately as shown in Figure 2.10-b.
(a)
(b)
Figure 2.10 Test Rig of Wind-Turbine: (a) Schematic Diagram (b) Photo
The turbine is a simple injection model that joins the tips of the six blades were shown in Figure
2.11-b. There is no twist in blades due to very small blade size. The aspect ratio of the rotor diameter
of the turbine to the diameter of the wind tunnel test section is (d/D =16 cm/52 cm = 0.3) which is
less than (0.5) to avoid the boundary layer effect. The blade is placed along a perpendicular direction
to that of the wind, and can be changing the angular position of blades (0° to 90°), as shown in Figure
2.11-a. The speed of air is measured by using a sensor placed in the tunnel and also measure the
rotational-speed of the wind-turbine rotor. There is one temperature probe before the rotor, in order
28
to measure the temperature of the air. The wattmeter is to measure the value of voltage and current
given produced by the electrical generator.
(a) Turbine
Blade (b) Scheme of
turbine
(c) The control (d) The swept-

volume area
Figure 2.11 Scheme of Axial Type Wind-Turbine
2.2.3 Theoretical Analysis

The higher the amount of kinetic-energy extracted from wind is, the slower the outlet air speed will
be. If kinetic-energy is completely extracted, the air speed in the outlet will be zero, the air will not
be able to get out of the turbine and no energy will be obtained. If the air passes through the turbine
with no speed change, then no energy would be obtained. In an ideal aero generator, the outlet air
speed is about 2/3 of the inlet air speed. In order to demonstrate this, fundamental aero dynamics
law is used (Betz-Law). Betz-Law says that less than 59% of the kinetic-energy of the wind can be
converted into mechanical-energy when using an aero generator (Darwish and Sayigh) and (Darwish
and Sayigh). To analyses the wind-turbine performance. The following assumptions apply:
➢ Steady-state flow.
➢ Isothermal flow.
➢ Incompressible flow.
In control volume section AA' as shown in Figure 2.11-c, a tube, with this section fixed into space.
The fluid in A section moves with speed V of constant module and direction perpendicular. The length
(L) travelled by A' section will be obtained in Eq. 2.6:
L = V∆t
Eq. 2.6
The fluid contained in the volume over the tube defined by this length (L) has a mass (M):
29
M = ρAV∆t
Eq. 2.7
Where ρ = air density, (kg/m3), A= rotor or swept-area, (m2), see Figure 2.11-d. Δt = Period of time,
(sec), and V = wind-speed, (m/s). The kinetic-energy of the wind that corresponds to this mass
definition is:
1
Ek = . (ρAV 3 ∆t)
2
Eq. 2.8
Dividing the kinetic-energy (Ek) of the wind by the time spent (Δt) in travelling along the tube will
give us the total wind-power Pw , as follows:
1
Pw = . (ρAV 3 )
2
Eq. 2.9
The power-coefficient (Cp) is the ratio of the total hydraulic power (PT ) extracted by the wind-turbine
to the total wind-power (Pw) that is stored in the wind flow:
PT
Cp =
Pw
Eq. 2.10
Therefore, the total hydraulic power (PT) extracted by the wind-turbine is:
1
PT = C𝑝 ρAV 3
2
Eq. 2.11
However, mechanical and electrical energy losses are present. Therefore, mechanical efficiency of
the rotor axis transmission and electrical efficiency of the electrical generator is introduced to define
the total actual generator power (PG) extracted from the wind. Hence, the total actual generator
power extracted from the wind is given as:
1
PG = ηmec ηelec Cp ρAV 3
2
Eq. 2.12
Also, the total actual generator power (PG) can be obtained from the produced electrical-power. The
output power of the wind-turbine can simply be measured from the current and voltage across the
generator:
PG = Iv
Eq. 2.13
Where I = The current, (Ampere) and v = The voltage, (volt). Merging the mechanical efficiency
(ηmech), the electrical efficiency (ηelec) and the power-coefficient (Cp) into one coefficient will introduce
the modified power-coefficient C*p as follows:
PG Iv
Cp∗ = ηmec ηelec Cp or Cp∗ = =
Pw 1 ρAV 3
2
Eq. 2.14
30
2.2.4 Results and Discussion

An experimental performance study of
a six blades axial type wind-turbine
was conducted to estimate the wind-
power, the electrical generated power
and-the modified power-coefficient of
the wind-turbine. This study was
focused on the effect of altering the
wind-speed and electrical load that is
applied at the terminals of the
electrical generator. In addition, the
effect of altering the blades angle of
the wind-turbine to estimate the
optimum operating condition of the Figure 2.12 Wind-Power Versus the Wind Velocity
wind-turbine. Figure 2.12 shows the
relation between the wind velocity
and its power. The predicted results of the experiments indicated that the wind power increases
when the wind-speed increase. Also, it shows that the maximum wind-power is (21 watt).
Figure 2.13 shows the relation between the measured values of rotational-speed of the wind-
turbine model with different wind velocity. In general, the rotational-speed of the wind-turbine
increases when the increase of wind velocity. A high level of wind-turbine vibration will start to grow
as the blade angles is increased above (10°- 20°) associated with a high value of wind velocity. It was
found that when operating the wind-turbine at (80°) blades angle associated with a wind-speed
range that is above (3.8 m/s), a high level of wind-turbine vibration was observed to reach a critical
point. It’s clear that this high level of wind-turbine vibration was caused by the increase of boundary
layer separation due to the increase of angle of attack of turbine blades. As a result, a vortex shedding
will be generated downstream the rotating blades of the wind-turbine causing a pulsating pressure
Figure 2.13 Turbine Speed Versus the Wind Velocity, at Different Blade Angles
31
difference between either side of the wind-turbine rotating blades. Hence, a pulsating drag force is
generated and causing a high level of wind-turbine vibration.
Figure 2.14-a shows the relation between the total generated power from the wind-turbine model
for different wind velocity. It’s clear that, at any value of blade angle, no power is generated when
the wind velocity is less than (2 m/s). The power generated is increased with both; increasing the
value of blade angles between (60° to 80°) and increasing the wind velocity. The maximum-value of
the generated power is (11.34 watt). The same behavior of the generated power from the wind-
turbine for different wind velocity is illustrated in Figure 2.14-b, (Abdin et al., 2012).
Figure 2.14 Total Generated Power Versus the Air Velocity, at Different Angles of Blade
Figure 2.15 shows the variations in modified power-coefficient with the air velocity. It shows that
the values of will increase with the increase of both; the wind velocity and-the value of blade angle.
32
Figure 2.15 Modified Power-Coefficient Versus the Air Velocity, at Different Angles of Blade
In addition, when the value of blade angles is less than (50°), the generated power is neglected. The
results show that the maximum-value of at a wind velocity value of about (5.6 m/s) and a blade angle
value of (80°). This increase in is cause by both; increasing the value of blade angles between (60° to
80°) and increasing the wind velocity. In Figure 2.16, the variation in the modified power-
Figure 2.16 Modified Power-Coefficients Versus Generated Power at Different Wind Velocities
33
coefficient and-the generated power is shown for the wind-turbine model at blade angle equal 80°.
It is observed that the modified power-coefficient increases with the increment in the generated
power.
2.2.5 Conclusions
In the present work, the performance of a six blades axial type wind-turbine was investigated
experimentally to estimate the wind-power, the electrical generated power and-the modified power-
coefficient of the wind-turbine. This experimental investigation focused on the effect of altering the
wind-speed, the electrical load that is applied on the terminals of the electrical generator and altering
the blades angle of the wind-turbine to estimate the optimum operating condition of the wind-
turbine. This study has the following conclusions:
1 Increasing the wind-speed and-the blades angle of the wind-turbine will increase the
generated power from the wind-turbine.
2 It’s not recommended to operate the wind-turbine at (80°) blades angle associated with a
wind-speed range that is above (3.8 m/s) due to a high level of wind-turbine vibration.
3 No power is generated from the wind-turbine at wind velocity less than (2 m/s) for different
blade angle.
4 The maximum generated power from the wind-turbine is obtained when the blade angle
value is equal to 80° and wind velocity value at 5.6 m/s.
5 The maximum modified power-coefficient was observed at a wind velocity value of about (5.6
6 m/s) and at a blade angle equal to (80°).
2.2.6 References
[1] A.S.K. Darwish, A.A.M. Sayigh (1988 a) ‘Wind energy potential in Iraq’, Journal of Wind
Engineering and Industrial Aerodynamics, 27(1–3), pp. 179–189.
[2] A.S.K. Darwish, A.A.M. Sayigh (1988 b) ‘Wind energy potential in Iraq’, Solar & Wind Technology,
5(3), pp. 215–222.
[3] Z. Abdin, M. A. Alim, M. A. Khairul and M. Mostaqur Rahman (2012) ‘Effect of Blade Pitch Angle
on the Performance of a Wind Turbine’, Engineering e-Transaction (ISSN 1823-6379), 7(2).
[4] Aditya Rachman (2014) ‘Simulating the Influences of the Blade Number and the Rotation on the
Annual Energy Performance of the Wind Machine’, 5(6), pp. 35–42.
[5] Aldair, A. A. abdulhameed (2014) ‘Pitch Angle Control Design of Wind Turbine Using Fuzzy-Art
Network’, Journal of Engineering and Development, 18(4).
[6] Buyung Kosasih, A. T. (2012) ‘Experimental study of shrouded micro-wind turbine’, Procedia
Engineering, 49, pp. 92–98. doi: 10.1016/j.proeng.2012.10.116.
[7] Kazuhiko Toshimitsu, Hironori Kikugawa, Kohei Sato, T. S. (2012) ‘Experimental Investigation of
Performance of the Wind Turbine with the Flanged-Diffuser Shroud in Sinusoidally Oscillating and
Fluctuating Velocity Flows’, Open Journal of Fluid Dynamics, 02(04), pp. 215–221.
[8] Medici, D. (2005) Experimental Studies of Wind Turbine Wakes - Power Optimization and
Meandering. Royal Institute of Technology.
[9] Noor M. Jasim (2010) ‘Investigating the Productive Energy and the Number of Revs of a Small Wind
Turbine At a Variable Wind Speeds .’, Al-Qadisiya Journal For Engineering Sciences, 3(1), pp. 64–78.
[10] Roshen T. Ahmed (2009) ‘Design a position control of the blade pitch angle for variable speed
wind turbine generators’, Eng. & Tech. Journal, 27(13).
[11] Ulan Dakeev (2013) Analysis of wind power generation with application of wind tunnel
attachment. University of Northern Iowa.
[12] Wissam Hashim Khalil (2007) ‘Modeling and Performance of a Wind Turbine’, Anbar Journal for
Engineering Sciences, pp. 116–130.
34
2.3 Case Study 2 - Blade Thickness Effect on The Aerodynamic Performance of An

Asymmetric NACA Six Series Blade Vertical Axis Wind Turbine in Low Wind
Speed
Authors : Hussain Mahamed Sahed Mostafa Mazarbhuiya, Agnimitra Biswas & Kaushal Kumar Sharma
Affiliation : Department of Mechanical Engineering, National Institute of Technology, Silchar, India
Title : Blade thickness effect on the aerodynamic performance of an asymmetric NACA six series blade
vertical axis wind turbine in low wind speed
Original Appearance : International Journal of Green Energy · January 2020.
Citation : Hussain Mahamed Sahed Mostafa Mazarbhuiya, Agnimitra Biswas & Kaushal Kumar
Sharma (2020) Blade thickness effect on the aerodynamic performance of an asymmetric NACA six
series blade vertical axis wind turbine in low wind speed, International Journal of Green
Energy, 17:2, 171-179, DOI: 10.1080/15435075.2020.1712214
2.3.1 Abstract
Vertical axis wind turbine (VAWT) is an economic and widely used energy converter for converting
wind energy into useful form of energy, like mechanical and electrical energy [Mazarbhuiya et al.]5.
For efficient energy conversion in low wind speed and to have improved power coefficient of
asymmetric blade VAWT, selection of optimum blade thickness is needed thus entailing its detailed
investigation with respect to different operating wind speed conditions. Present study methodically
explores the impact of thickness to chord (t/c) ratio on aerodynamic performance of a three bladed
asymmetrical blade H-Darrieus VAWT at different low wind speed conditions by using 2D unsteady
CFD simulations. The optimal t/c is obtained on the basis of maximum power coefficient and average
moment coefficient of the turbine. The aerodynamic performance curves are obtained at different
operating and t/c conditions and the performance insights are corroborated with the findings from
the flow physics study to come to some concrete conclusions on the effects of the thickness to chord
ratio. The present study identifies large blade curvature to create a large diverging passage on the
blade suction surface as the prominent reason for aerodynamic performance drop at a high t/c ratio.
2.3.2 Introduction
Depletion of nonrenewable energy sources and increase of global warming compel the mankind
across the world to think of alternative or renewable energy sources. Researches have been
progressed for developing efficient wind turbine for offshore installation where wind speed is
generally high and also for the built environment of urban locations where wind speed is
comparatively less. Recently, small scale wind turbines basically H-Darrieus and Savonius turbines
have become topic of investigation for installation in the built environment (Kumbernuss et al. 2012;
Mazarbhuiya, Biswas, and Sharma 2018; Roy and Saha 2014; Sengupta, Biswas, and Gupta 2016).
The problem facing the researchers and developers for VAWT’s installation in urban locations is the
wind speed, which is lower than offshore locations.
The performance of VAWT in low wind speed can be influenced by blade thickness effect (Danao, Qin,
and Howell 2012). The blade thickness effects for different sets of airfoils from different series have
been investigated by (Danao, Qin, and Howell 2012). Several blade configurations (t/c = 15–25%)
have been recommended based on different design and operational parameters (Dereng 1981; Liang
et al. 2014; Seki 2005).
Due to better stall characteristics, thicker airfoil has capability to perform better at high solidity
(Ghasemian, Ashrafi, and Sedaghat 2017). But, maximizing airfoil thickness becomes detrimental to
5 Hussain Mahamed Sahed Mostafa Mazarbhuiya, Agnimitra Biswas & Kaushal Kumar Sharma (2020) “Blade
thickness effect on the aerodynamic performance of an asymmetric NACA six series blade vertical axis wind turbine
in low wind speed”, International Journal of Green Energy, 17:2, 171-179, DOI:
10.1080/15435075.2020.1712214
35
the turbine performance due to increase in drag force on turbine blades (Meana-Fernández et al.
2018), although it enhances the structural strength (Liang and Li 2018). Asr et al. (2016) found an
optimum start-up characteristic for medium thickness airfoil. Increasing airfoil thickness may
improve the aerodynamic performance at lower TSR range (Kirke 1998). Chen and Lian (2015)
investigated the effect of blade thickness considering two symmetrical blade profiles namely NACA
0022 and NACA 0015.
The peak torque was noticed for NACA 0022 blade at low TSR due to thickness effect. Subramanian
et al. (2017) reported that thicker airfoils perform better at low TSR range due to longer attached
flows. Both moment coefficient (Cm) and tangential force coefficient (Ct) can be increased with
increase in airfoil thickness (López et al. 2016).
From the above, it is seen that there is a limited literature on the investigation of blade thickness
effect. It is very important to keep an optimal thickness depending upon the design and operational
parameters as too much increase of thickness may lower the aerodynamic performance, although the
same improves the latter at a lower TSR range. Further, the effect of t/c ratio for a single specified
blade by changing its thickness for getting an optimal blade design has yet not been investigated,
especially in low wind speed conditions. Therefore, in the present paper an asymmetric NACA 63–
415 (a six-digit airfoil) blade H-Darrieus turbine is selected and it has been attempted to change its
t/c ratio to attain an optimal ratio for improved aerodynamic performance of the turbine at a
characteristic low wind speed of 6.0 m/s common in the built environment. Five different t/c ratios
namely, 0.22, 0.26, 0.30, 0.33, 0.37 are considered in this investigation. After getting the optimal t/c
ratio, effects of two more wind speeds (5.0, 7.0 m/s) on the aerodynamic performance of the turbine
are also studied. Series of simulations are performed using 2D unsteady Reynolds-averaged Navier-
Stokes (URANS) viscous CFD model and the computational results are validated with the
experimental results.
The aerodynamic performance insights are corroborated with the findings from the flow physics
study to come to some concrete conclusions on the effects of the thickness to chord ratio. The present
paper exhibits how blade curvature change with increasing t/c controls the aerodynamic
performance.
Table 2.1 Geometrical and operational details of VAWT of the reference case
2.3.3 Computational Settings

2.3.3.1 Geometrical and Operational Details of Reference Turbine Case
A three bladed H-Darrieus VAWT with asymmetric NACA 63–415 airfoil blades and having turbine
diameter 0.5 m is considered. This blade profile is selected as it is reported to be a very potential
asymmetric blade profile in low wind speed condition (Mohamed, Ali, and Hafiz 2015). A three bladed
turbine gives higher uniformity in the turbine output power compared to a two bladed turbines due
to less torque ripple effect (Tjiu et al. 2015). The turbine rotates in the anticlockwise direction by
36
virtue of upstream velocity (U = 6.0 m/s). The turbine for reference case has solidity (σ) 0.3, blade
chord length (c) 0.05 m and the same is to be operated with wind speed 6.0 m/s. (Wang et al. 2016)
reported that a solidity of 0.32 resulted in maximum efficiency of H-Darrieus VAWT. A high solidity
turbine (i.e. more than 0.2) is also recommended in low TSR range for power enhancement (Roh and
Kang 2013). The turbine solidity can be calculated using:
Nc
σ=
D
Eq. 2.15
where, N denotes the number of turbine blades and D denotes the turbine diameter. The geometrical
and operational details for the reference turbine are described in Table 2.1.
2.3.3.2 Rationale of Choosing 2D CFD Model
➢ A 2D CFD model can yield acceptable results while investigating the effect of blade thickness
at a considerably less computational time and cost (Danao, Qin, and Howell 2012). Since,
using 2D simulation the focus of the study is on the mid plane of the turbine and since the
blade aspect ratio is also high (H/c = 10) in the present case, the complex blade tip effects can
be avoided (Rezaeiha, Montazeri, and Blocken 2018b) and hence 3D simulations are not
considered in the present study. Moreover, 2D CFD simulation has been found trustworthy
in revealing the underlying flow physics responsible for VAWT’s performance in its plane of
rotation. There are many papers in the available literature, which have successfully
investigated the complex aerodynamic performance of VAWTs using 2D CFD simulation for
e.g. the recent works of Bausas and Danao (2015);, Rezaeiha, Montazeri, and Blocken
(2018b);, Wang et al. (2016). Other prominent factors for adopting 2D CFD model are also
discussed below.
➢ An extensive comparative studies of Bianchini et al. (2017); Rezaeiha, Montazeri, and Blocken
(2018a) reveal that a 2D numerical simulation can provide trustworthy representation of the
flow in the turbine mid-plane. 2D CFD simulation is also capable of reducing computational
time and physical requirements without sacrificing accuracy of results (Kalluvila and Sreejith
2018).
➢ A negligible difference in turbine performance is found between 2D and 2.5D simulation
results (Rezaeiha,
Kalkman, and Blocken
2017a).
➢ Due to high
computational cost of
3D and 2.5D
simulations, it makes
infeasible to use
2.5D/3D models for an
extensive analysis of
turbine performance as
reported by Rezaeiha,
Montazeri, and Blocken
(2019).
Hence, authors firmly believe
that choosing a validated 2D
CFD model can delineate the Figure 2.17 Computational domain for VAWT (not drawn in scale).
37
flow physics and the turbine performance in a trustworthy manner.

2.3.3.3 Computational Domain and Grid
The computational domain consists of one rotating domain where the turbine is located and a
stationary subdomain surrounding the rotating domain as shown in Figure 2.17. The aerodynamic
moment center is at 25% of the blade chord length. The rotating domain is twice the radius of the
turbine, and the fixed domain is 30 times of turbine radius in length and 20 times of turbine radius
in width. The distance between turbine center and inlet is 10 times the turbine radius. The width of
the computational domain is large enough to ensure fully developed flow and avoid any blockage
effect. The term blockage ratio is the ratio between turbine diameter to the width of the domain. Asr
et al. (2016) considered a blockage ratio of 27% for CFD simulation of a three bladed VAWT whereas
in the present case the blockage ratio is only 10%, which is less than Asr et al. (2016). This blockage
ratio is reduced in present case by enlarging the domain size.
Figure 2.18 Computational domain and grid: (a) grid near the rotating cell zone, (b) grid near the airfoil,
(c) grid near airfoil trailing edge
The computational grid is shown in Figure 2.18. A non-conformal interface is created using sliding
mesh technique between the rotating and stationary domain to empower the turbine rotation. A
triangular mesh is used for both rotating and stationary domain. A structured mesh is used near the
blade periphery and refined to meet the wall boundary conditions. To resolve the viscous sub layer,
25 layers of quadrilateral element are created over the turbine blades. The height of the first cell is
calculated as 0.02 mm based on the highest y+ less than 1.5.
2.3.3.4 Numerical Setting
The flow is simulated around the VAWT using finite volume based CFD software package ANSYS
Fluent. The incompressible URANS (Unsteady Reynolds Averaged Navier Stokes) equations are
solved using the SIMPLE (Semi implicit method for pressure linked equations) pressure velocity
coupling algorithm.
A second order spatial discretization is used with standard pressure scheme. Symmetry boundary
condition is used on top and bottom side of rectangular domain (Figure 2.17). A constant velocity
inlet and average gauge pressure outlet is used on left and right side of the rectangular domain,
respectively, a no slip wall condition is selected for the airfoils (Figure 2.17). A well-known 4
equation turbulence model also known as γ-θ SST turbulence model (Menter et al. 2006) is used for
turbulence modeling. This turbulence model is an improvement over 2 equation k-ω SST turbulence
model (Menter 1994; Menter, Kuntz, and Langtry 2003). It is very critical to accurately model the
boundary layer development and calculation of loads on the blades. Hence two more equations such
38
as intermittency γ and momentum equation Reynolds number Reθ are solved to provide more
accurate prediction toward laminar-turbulent transition for wall bounded flows (Menter et al. 2005).
This intermittency based model is widely used for modeling laminar-turbulent transition flow on
airfoil (Cutrone et al. 2008; Leylek and Walters 2004; Suzen et al. 2003) and give promising
prediction.
In the numerical setting, each time step consists of 20 iterations to scale the residual below 10−5 for
continuity, velocity and turbulent properties. The time step size (Δt) can be calculated using:
π
∆t =
ω × 180
Eq. 2.16
where ω denotes angular velocity of H-Darrieus VAWT. The number of time steps is 1800, which
completes five complete rotation of the turbine. The power coefficient (Cp) of the turbine is
calculated using the instantaneous moment coefficient (Cm) of fifth revolution of the turbine.
Governing equations namely continuity equation Eq. 2.17, x-momentum Eq. 2.18 and y-momentum
Eq. 2.19 are solved to calculate the performance of the present turbine.
⃗)=0
∇. (ρV
Eq. 2.17
∂(ρu) ∂p ∂τ ∂τ
⃗ ) = − + xx + yx
+ ∇. (ρuV
∂t ∂x ∂x ∂y
Eq. 2.18
∂(ρv) ∂p ∂τ ∂τ
⃗ ) = − + yy + xy
+ ∇. (ρvV
∂t ∂y ∂y ∂x
Eq. 2.19
2.3.3.5 Sensitivity and Validation
Determination of number of revolutions of the turbine for performing a simulation is important
because unnecessarily excess turbine revolutions increase the number of time steps that in turn
increase computational cost. Initially turbine is rotated for 10 revolutions to investigate the variation
Figure 2.19 (a) Variation of moment coefficient with number of time steps, (b) Variation power
coefficient with number of revolutions of turbine
39
of Cm, which is shown in Figure 2.19(a) and the corresponding plot of Cp with number of revolutions
of the turbine is shown in Figure 2.19(b). With starting of the turbine, the abrupt change of Cp is
visible up to the fourth revolution.
A significant over estimation of turbine

performance would be the reason during data
sampling. A very little deflection (0.06%) in Cp
is obtained between the fourth and fifth
turbine revolution. Similar trend of variation
in Cp is also reported by Rezaeiha, Kalkman,
and Blocken (2017b). The average Cp of fifth
revolution of the turbine is considered for Table 2.2 Details of Grids Used In VAWT
further calculations. Grid sizes of three
different resolutions are tested, which are
refined uniformly and the details are given in Table 2.2. Grids are finer near the edge of all three
blades.
A very less difference in Cm is obtained between medium and fine size grid. Hence a medium grid
size is selected for further simulation, where element sizing near the wall of blades is 0.1 mm. An
average orthogonal quality close to 1 reflects good meshing; in medium grid size the orthogonal
quality is 0.97, which ensures good meshing quality. Sengupta, Biswas, and Gupta (2019) considered
an orthogonal quality of 0.96 generating a mesh of good quality with an optimum number of 106,127
mesh elements. In order to ensure
good prediction of flow
characteristics, turbine
performances at different azimuthal
increment (dθ) are analyzed. Four
different azimuthal increments
within 2°–0.5° are considered for the
sensitivity analysis. With decrease in
azimuthal increment, a precise
solutionis obtained. The variation of
Cp is reduced with decrease in
azimuthal increment as shown in
Figure 2.20. However, if the
azimuthal increment is reduced,
then the simulation time is increased
which increases the computational
cost.
A very less difference (0.3%) in Cp
value is obtained between 1° and Figure 2.20 Variation of power coefficient with azimuthal
0.5°, but the time consumed for increment
simulation is doubled with 0.5°
azimuthal increment. Hence, a
reasonable dθ is required for predictable results and for the present work dθ = 1° is selected for
further simulations. A validation study is performed to ensure the accuracy of numerical results. The
experimental result of Mazarbhuiya, Biswas, and Sharma (2018) is used to compare the Cp values
calculated from numerical simulation as shown in Figure 2.21. In general, the trend of power
coefficient of a 2D CFD model over predicts in contrast to experimental results and computationally
expensive 3D CFD model (Raciti Castelli, Englaro, and Benini 2011; Hand and Cashman 2018). The
overestimation of Cp values is due to neglecting the blade tip losses, strut losses, for the 2D CFD
40
simulation done in the present work,

which are also reported in Castelli,
Marco, and Benini (2011); Rezaeiha,
Montazeri, and Blocken (2019).
However, since the blade aspect ratio is
high (H/c = 10) in the present case,
blades tip effects are insignificant for
which computationally expensive 3D
CFD simulations are avoided (Rezaeiha,
Montazeri, and Blocken 2018b), and 2D
simulation is successfully used to
capture the flow physics of the turbine.
The over-estimation of Cp in the present
validation is less compared to that
obtained by Castelli, Marco, and Benini
(2011); Gosselin, Dumas, and Boudreau
(2016); Hashem and Mohamed (2018); Figure 2.21 Validation of CFD Results with Established
Wang et al. (2018). The root mean Experimental Results
square error (RMSE) of their validations
were 25%, 6.8%, 7% and 8.5%,
respectively, whereas the RMSE of the present CFD simulation is 4.22%, which is less and which
shows the accuracy of the present 2D CFD results. The RMSE is calculated using :
∑N
i=1(Peridictedi − Actuali )
2
RMSE = √
N
Eq. 2.20
Figure 2.22 Variation of (a) average and (b) instantaneous moment coefficient with azimuthal angle at
U = 6.0 m/s.
41

In order to calculate the average and instantaneous moment coefficient (Cm), the fifth revolution of
the turbine is considered. The variation of average and instantaneous Cm with different azimuthal
position is shown in Figure 2.22. A cyclic variation is obtained where wavelengths are at an
azimuthal gap of 120° (Figure 2.22(a)), which is the case of a three bladed turbine. Negative average
Cm is found in certain azimuthal range which is not good for turbine performance. Turbine with t/c
= 0.225, 0.262 and 0.375 has negative average Cm in some of the azimuthal position. Turbine with t/c
= 0.3 does not have negative Cm, which is an indication of better turbine performance at this t/c.
Turbine with t/c = 0.337 has negative average Cm at a small azimuthal range. The instantaneous Cm
of the turbine for different t/c = 0.225, 0.262, 0.3, 0.337 and 0.375 is compared as shown in Figure
2.22(b).
The value of instantaneous Cm increases toward upwind position with increase in blade thickness
and becomes higher for t/c = 0.3; however, with further increase in blade thickness leads to decrease
in Cm, which can be attributed to increase in drag force. Although drag force may help the turbine to
start initially especially in low wind speed at urban region (built environment), the amount of power
extracted from wind is reduced at nominal speed due to drag force. Hence, in spite of increase in
blade t/c ratio from 0.3 to a higher value i.e. 0.375, higher Cm is not obtained. The same insight can
also be corroborated from the results of flow physics study discussed subsequently. Power
coefficient (Cp) of the turbine can be calculated by multiplying the TSR with average Cm during one
revolution of turbine by using
CP = Cm × λ
Eq. 2.21
A higher value of Cp is found at TSR 2.4 for t/c = 0.3 turbine. The maximum value of Cp is gradually
increased with increase in t/c ratio up to 0.3 (optimum value) as shown in Figure 2.23(a) at TSR
2.4. However, with further increase of blade thickness the drag force is increased and thereby reduces
Cp values. The Cp values are also compared with established results of (Li et al. 2016), which follow
a similar trend with TSR like (Li et al. 2016) and the maximum Cp obtained is 0.271 for t/c = 0.3
which is higher than the established results of Li et al. (2016). Further, turbine performance is
evaluated for optimum t/c = 0.3 at different wind speeds. Cp values are calculated at different wind
speeds (5.0 and 7.0 m/s) and compared with the results obtained with wind speed 6.0 m/s as shown
in Figure 2.23(b).
Figure 2.23 Variation of (a) power coefficient for different t/c turbine with TSR (b) power coefficient at
blade t/c = 0.3 for different wind speed
42
The performance of the turbine is drastically reduced for the present turbine configuration with
increased wind speed of 7.0 m/s. This may have caused due to smaller diameter of turbine
configuration for which losses due to drag coefficient are increased at higher wind speed. The
variations of drag coefficient with azimuthal angle for two wind speeds (U = 6.0 and 7.0 m/s) are
compared as shown in Figure 2.24.
It shows that the drag coefficients are
higher for majority of the azimuthal
positions at higher wind speed (7.0 m/s).
In the upwind azimuthal position (67° to
110°), the drag coefficient is higher in case
of higher wind speed (7.0 m/s) thereby
reducing the turbine’s performance in the
power stroke. In the returning stroke,
turbine tries to recover the wake losses for
which drag coefficient decreases in the
leeward side (around 180° azimuthal
position); however, at higher wind speed of
7.0 m/s, the drag coefficient is also higher
for the azimuthal range 247–326° for
which the turbine cannot overcome the
increased losses at higher wind speed
compared to lower wind speed and Figure 2.24 Variation of drag coefficient with
performance also decreases. azimuthal angle for different wind speeds (U = 6.0 and
A turbine with same solidity but different 7.0 m/s).
chord lengths and diameters may increase
Cp at higher wind speeds. In spite of being
the same solidity a turbine Cp differs with wind speed due to different flow physics. Reasons are also
illustrated in later sections. The vorticity patterns near the advancing blade of the turbines with t/c
= 0.225, 0.3 and 0.375 at azimuthal position 78° are compared as shown in Figure 2.25. As seen
from the earlier discussion of Cm that performance of the turbine changes with blade thickness, the
same can now be corroborated from the generated vorticity patterns in the flow physics study. At the
blade suction surface, a large eddy is formed (Figure 2.25(a)) due to less blade thickness and early
Figure 2.25 Comparison of vortex structures in the flow field around a single blade for θ = 78° at
different t/c at U = 6.0 m/s.
43
flow separation, which transforms into

a strong recirculation zone by
consuming energy from the mean flow
thereby reducing turbine performance.
With increase in blade thickness, a

delay in separation occurs as the flow
streamlines are attached with the blade
suction surface over a considerable
chord length distance as shown in
Figure 2.25 (b) thereby increasing
the turbine performance at a larger t/c
of 0.3. This can also be attributed to a
larger curvature of the blade suction
surface at increased t/c, which can also
be corroborated with the findings of
Sengupta, Biswas, and Gupta (2019). Figure 2.26 Pressure Coefficient Distribution for Different
However, with further increase in t/c Turbine at U = 6.0 m/s.
blade thickness from 0.3 to 0.375, flow
streamlines are scattered from blade
suction surface (Figure 2.25(c)) as the flow attachment length on the suction surface decreases due
to a more divergent cone created on the blade suction surface starting at around 1/4th chord length
from leading edge and ending at the tip of the trailing edge thus reducing the turbine performance at
t/c = 0.375. The divergent cone angle on the blade suction surface is more for t/c = 0.375 as the blade
curvature of the suction surface is also more at this higher t/c. The distributions of pressure
coefficient for t/c = 0.225, 0.3 and 0.375 at 78° azimuthal angle are shown in Figure 2.26 for wind
speed 6.0 m/s. The pressure gradient is less for t/c = 0.225 due to the strong eddy generation toward
suction side, which reduces
the turbine performance.
A larger pressure gradient is
visible for t/c = 0.3, which
increases the turbine
performance and hence can
be considered as an optimal
t/c for the present turbine. A
remarkable difference in the
variation of Cm with
azimuthal angle can be seen
in Figure 2.27 for the
turbine with optimal t/c at
different wind speeds. At
135° azimuthal position, the
value of Cm is drastically
reduced when the turbine is
operated at 7.0 m/s wind
speed; due to this Cp values
are also reduced (as shown in
Figure 2.23(b) earlier). Figure 2.27 Variation of moment coefficient for turbine having
The Cm values of the turbine optimal t/c at different wind speeds.
at the other two wind speeds
44
are almost same with marginally higher peak obtained at 90° azimuthal position, which is also
reflected in the result of Figure 2.23(b). The Cm values in the downwind positions are much lower
due to slowed down flow as the energy is extracted by the turbine in the upwind position. The
reduction of Cm at higher wind speed of 7.0 m/s
speeds at 135° azimuthal position
is basically due to more vortex shading that occurs at this wind speed compared to wind speed 6.0
m/s, as can also be corroborated from Figure 2.28. A strong recirculation zone (stall vortices) is
visible near the suction side trailing edge of the blade at 7.0 m/s as shown in Figure 2.28 (b), which
speeds at 254° azimuthal position
45
consumes energy from the incoming flow thereby reducing the turbine performance. In addition, the
shear layer at the blade leading edge is extended to a very long distance.
Finally, the effect of wind speeds on the flow physics of the turbine having optimal t/c in the
downwind position is compared as shown in Figure 2.29. At downwind position (θ = 254°) due to
increase in free-stream velocity from 6.0 to 7.0 m/s, blade momentum increases, which increases the
Cm. However, a small stall vortex is created as shown in Figure 2.29 (b), which when progressed
causes fluctuations in Cm for U = 7.0 m/s, which can also be corroborated from the Cm variation plot
of Figure 2.27 as discussed earlier.
2.3.5 Conclusions & Recommendations
In the present study 2D unsteady RANS simulations were carried out to find out the effect of blade
thickness-to-chord ratio on the aerodynamic performance of an asymmetric NACA six series blade
vertical axis H-Darrieus wind turbine at different low wind speed conditions. The aerodynamic
performance curves are obtained at different operating and thickness-to-chord conditions and the
performance insights are corroborated with the findings from the flow physics study to come to some
concrete conclusions on the effects of the thickness to chord ratio. From the study, the following
conclusions can be drawn:
A. The average moment coefficient is higher for turbine having blade t/c ratio of 0.3. The power
coefficient of the turbine increases with increase in blade t/c ratio and become maximum at
t/c = 0.3 at 2.4 tip speed ratio. The optimum design for efficient turbine has t/c = 0.3 which
has maximum power coefficient of 0.271.
B. The present study identifies large blade curvature to create a large diverging passage on the
blade suction surface as the prominent reason for aerodynamic performance drop at a high
t/c ratio, which in the present design is 0.375.
C. If the free stream velocity is increased the maximum power coefficient is decreased for the
optimal t/c = 0.3, due to large vortex shading from the suction side trailing edge of the blade.
Future study could entail investigating the effect of asymmetric blade twist angle at the trailing edge
on the aerodynamic performance of the H-Darrieus turbine with consideration of different t/c and
different operating low wind speed conditions. All such attempts including the
Figure 10. Pressure coefficient distribution for different t/c turbine at U = 6.0 m/s.
Figure 11. Variation of moment coefficient for turbine having optimal t/c at
different wind speeds.
INTERNATIONAL JOURNAL OF GREEN ENERGY 177
present work should lead to an improved H-Darrieus turbine design applicable to work in the low
wind speed condition of the built environment.
2.3.6 Acknowledgments
Author would like to thank and acknowledge the computational Lab facility, Department of Mechanical
Engineering, NIT Silchar, India.
2.3.7 References
Asr, M. T., E. Z. Nezhad, F. Mustapha, and S. Wiriadidjaja. 2016. Study on Start-up Characteristics of
H-Darrieus Vertical Axis Wind Turbines Comprising NACA 4-Digit Series Blade Airfoils. Energy
112:528–37. doi:10.1016/j.energy.2016.06.059.
Bausas, M. D., and L. A. M. Danao. 2015. The Aerodynamics of a Camber-Bladed Vertical Axis Wind
Turbine in Unsteady Wind. Energy 93:1155–64. doi:10.1016/j.energy.2015.09.120.
46
Bianchini, A., F. Balduzzi, P. Bachant, G. Ferrara, and L. Ferrari. 2017. Effectiveness of Two-
Dimensional CFD Simulations for Darrieus VAWTs: A Combined Numerical and Experimental
Assessment. Energy Conversion and Management 136:318–28.
doi:10.1016/j. enconman.2017.01.026.
Castelli, R., A. E. Marco, and E. Benini. 2011. The Darrieus Wind Turbine: Proposal for a New
Performance Prediction Model Based on CFD. Energy 36 (8):4919–34
doi:10.1016/j.energy.2011.05.036.
Chen, Y., and Y. Lian. 2015. Numerical Investigation of Vortex Dynamics in an H-Rotor Vertical Axis
Wind Turbine. Engineering Applications of Computational Fluid Mechanics 9 (1):21–32.
doi:10.1080/19942060.2015.1004790.
Cutrone, L., P. De Palma, G. Pascazio, and M. Napolitano. 2008. Predicting Transition in Two- and
Three-Dimensional Separated Flows. International Journal of Heat and Fluid Flow 29 (2):504–26.
doi:10.1016/j.ijheatfluidflow.2007.11.005.
Danao, L. A., N. Qin, and R. Howell. 2012. A Numerical Study of Blade Thickness and Camber Effects
on Vertical Axis Wind Turbines. Proceedings of the Institution of Mechanical Engineers, Part A:
Journal of Power and Energy 226 (7):867–81.
doi:10.1177/0957650912454403.
Dereng, V. G. 1981. “Fixed Geometry Self-Starting Transverse Axis Wind Turbine.” United States
Patent No. 4,264,279, April 28.
Ghasemian, M., Z. N. Ashrafi, and A. Sedaghat. 2017. A Review on Computational Fluid Dynamic
Simulation Techniques for Darrieus Vertical Axis Wind Turbines. Energy Conversion and
Management 149:87–100. doi:10.1016/j.enconman.2017.07.016.
Gosselin, R., G. Dumas, and M. Boudreau. 2016. Parametric Study of H-Darrieus Vertical-Axis
Turbines Using CFD Simulations. Journal of Renewable and Sustainable Energy 8:5.
doi:10.1063/1.4963240.
Hand, B., and A. Cashman. 2018. Aerodynamic Modeling Methods for a Large-Scale Vertical Axis Wind
Turbine: A Comparative Study. Renewable Energy 129:12–31. doi:10.1016/j.renene.2018.05.078.
Hashem, I., and M. H. Mohamed. 2018. Aerodynamic Performance Enhancements of H-Rotor Darrieus
Wind Turbine. Energy 142:531–45. doi:10.1016/j.energy.2017.10.036.
Kalluvila, J. B. S., and B. Sreejith. 2018. Numerical and Experimental Study on a Modified Savonius
Rotor with Guide Blades. International Journal of Green Energy 15 (12):744–57. doi:10.1080/
15435075.2018.1529574.
Kirke, B. K. 1998. “Evaluation of Self-Starting Vertical Axis Wind Turbines for Stand-Alone
Applications.” Australia: School of Engineering, Griffith University Kumbernuss, J., J. Chen, H. X. Yang,
and L. Lu. 2012. Investigation into the Relationship of the Overlap Ratio and Shift Angle of Double
Stage Three Bladed Vertical Axis Wind Turbine (VAWT). Journal of Wind Engineering and Industrial
Aerodynamics 107–108:57–75. doi:10.1016/j.jweia.2012.03.021.
Leylek, J., and D. Walters. 2004. A New Model for Boundary Layer Transition Using a Single-Point
RANS Approach (2002-HT-32740). Journal of Turbomachinery 126 (1):193–202.
doi:10.1115/1.1622709.
Li, Q., T. Maeda, Y. Kamada, J. Murata, M. Yamamoto, T. Ogasawara, K. Shimizu, and T. Kogaki. 2016.
Study on Power Performance for Straight-Bladed Vertical Axis Wind Turbine by Field and Wind
Tunnel Test. Renewable Energy. doi:10.1016/j.renene.2016.01.002.
Liang, C., and L. Huaxing. 2018. Effects of Optimized Airfoil on Vertical Axis Wind Turbine
Aerodynamic Performance. Journal of the Brazilian Society of Mechanical Sciences and Engineering
40:2. doi:10.1007/s40430-017-0926-2.
Liang, Y. B., L. X. Zhang, E. X. Li, X. H. Liu, and Y. Yang. 2014. Design Considerations of Rotor
Configuration for Straight-Bladed Vertical Axis Wind Turbines. Advances in Mechanical Engineering
2014. doi:10.1155/2014/534906.
47
López, O., D. Meneses, B. Quintero, and S. Laín. 2016. Computational Study of Transient Flow around
Darrieus Type Cross Flow Water Turbines. Journal of Renewable and Sustainable Energy 8:1.
doi:10.1063/1.4940023.
Mazarbhuiya, H. M. S. M., A. Biswas, and K. K. Sharma. 2018. Performance Investigations of Modified
Asymmetric Blade H-Darrieus VAWT Rotors. Journal of Renewable and Sustainable Energy 033302.
doi:10.1063/1.5026857.
Meana-Fernández, A., I. Solís-Gallego, J. M. Fernández Oro, K. M. Argüelles Díaz, and S. Velarde-Suárez.
2018. Parametrical Evaluation of the Aerodynamic Performance of Vertical Axis Wind Turbines for
the Proposal of Optimized Designs. Energy 147:504–17. doi:10.1016/j.energy.2018.01.062.
Menter, F. R. 1994. Two-Equation Eddy-Viscosity Turbulence Models for Engineering Applications.
AIAA Journal 32 (8):1598–605. doi:10.2514/3.12149.
Menter, F. R., M. Kuntz, and R. Langtry. 2003. Ten Years of Industrial Experience with the SST
Turbulence Model. Turbulence Heat and Mass Transfer 4 (4):625–32.
doi:10.4028/www.scientific.net/AMR.576.60.
Menter, F. R., R. Langtry, S. Völker, and P. G. Huang. 2005. Transition Modelling for General Purpose
CFD Codes. Engineering Turbulence Modelling and Experiments 6 (August):31–48.
doi:10.1016/B978-008044544-1/50003-0.
Menter, F. R., R. B. Langtry, S. R. Likki, Y. B. Suzen, P. G. Huang, and S. Völker. 2006. A Correlation-
Based Transition Model Using Local Variables - Part I: Model Formulation. Journal of
Turbomachinery 128 (3):413–22. doi:10.1115/1.2184352.
Mohamed, M. H., A. M. Ali, and A. A. Hafiz. 2015. CFD Analysis for H-Rotor Darrieus Turbine as a Low
Speed Wind Energy Converter. Engineering Science and Technology, an International Journal 18
(1):1–13. doi:10.1016/j.jestch.2014.08.002.
Rezaeiha, A., I. Kalkman, and B. Blocken. 2017a. CFD Simulation of a Vertical Axis Wind Turbine
Operating at a Moderate Tip Speed Ratio: Guidelines for Minimum Domain Size and Azimuthal
Increment. Renewable Energy 107:373–85. doi:10.1016/j.renene.2017.02.006.
Rezaeiha, A., I. Kalkman, and B. Blocken. 2017b. Effect of Pitch Angle on Power Performance and
Aerodynamics of a Vertical Axis Wind Turbine. Applied Energy 197:132–50.
doi:10.1016/j.apenergy.2017.03.128.
Rezaeiha, A., H. Montazeri, and B. Blocken. 2018a. Characterization of Aerodynamic Performance of
Vertical Axis Wind Turbines: Impact of Operational Parameters. Energy Conversion and Management
169 (February):45–77. doi:10.1016/j.enconman.2018.05.042.
Rezaeiha, A., H. Montazeri, and B. Blocken. 2018b. Towards Optimal Aerodynamic Design of Vertical
Axis Wind Turbines : Impact of Solidity and Number of Blades. Energy 165:1129–48. doi:10.1016/j.
energy.2018.09.192.
Rezaeiha, A., H. Montazeri, and B. Blocken. 2019. On the Accuracy of Turbulence Models for CFD
Simulations of Vertical Axis Wind Turbines. Energy 180:838–57. doi:10.1016/j.energy.2019.05.053.
Roh, S. C., and S. H. Kang. 2013. Effects of a Blade Profile, the Reynolds Number, and the Solidity on
the Performance of a Straight Bladed Vertical Axis Wind Turbine. Journal of Mechanical Science and
Technology 27 (11):3299–307. doi:10.1007/s12206-013-0852-x.
Roy, S., and U. K. Saha. 2014. An Adapted Blockage Factor Correlation Approach in Wind Tunnel
Experiments of a Savonius-Style Wind Turbine. Energy Conversion and Management 86:418–27.
doi:10.1016/j. enconman.2014.05.039.
Seki, K. 2005. “Straight Wing Type Wind and Water Turbine.” United States Patent, No. 6,974,309 B2,
December 13.
Sengupta, A. R., A. Biswas, and R. Gupta. 2016. Studies of Some High Solidity Symmetrical and
Unsymmetrical Blade H-Darrieus Rotors with Respect to Starting Characteristics, Dynamic
Performances and Flow Physics in Low Wind Streams. Renewable Energy 93:536–47.
doi:10.1016/j.renene.2016.03.029.
48
Sengupta, A. R., A. Biswas, and R. Gupta. 2019. Comparison of Low Wind Speed Aerodynamics of
Unsymmetrical Blade H-Darrieus Rotors-Blade Camber and Curvature Signatures for Performance
Improvement. Renewable Energy 139:1412–27. doi:10.1016/j.renene.2019.03.054.
Subramanian, A. S., A. Yogesh, H. Sivanandan, A. Giri, M. Vasudevan, V. Mugundhan, and R. K. Velamati.
2017. Effect of Airfoil and Solidity on Performance of Small Scale Vertical Axis Wind Turbine Using
Three Dimensional CFD Model. Energy 133:179–90. doi:10.1016/j.energy.2017.05.118.
Suzen, Y. B., P. G. Huang, L. S. Hultgren, and D. E. Ashpis. 2003. Predictions of Separated and
Transitional Boundary Layers under Low-Pressure Turbine Airfoil Conditions Using an
Intermittency Transport Equation. Journal of Turbomachinery 125 (3):455–64.
doi:10.1115/1.1580159.
Tjiu, W., T. Marnoto, S. Mat, M. H. Ruslan, and K. Sopian. 2015. Darrieus Vertical Axis Wind Turbine
for Power Generation I: Assessment of Darrieus VAWT Configurations. Renewable Energy 75:50–67.
doi:10.1016/j.renene.2014.09.038.
Wang, Y., S. Shen, L. Gaohui, D. Huang, and Z. Zheng. 2018. Investigation on Aerodynamic Performance
of Vertical Axis Wind Turbine with Different Series Airfoil Shapes. Renewable Energy 126:801–18.
doi:10.1016/j.renene.2018.02.095.
Wang, Y., X. Sun, X. Dong, B. Zhu, D. Huang, and Z. Zheng. 2016. Numerical Investigation on
Aerodynamic Performance of a Novel Vertical Axis Wind Turbine with Adaptive Blades. Energy
Conversion and Management 108:275–86. doi:10.1016/j.enconman.2015.11.003.
49
2.4 Wind Profile6

Wind is unpredictable in nature and wind speed has a dependency on time and location. Wind speed
varies during the day as well as with the surface roughness of different sites. Wind profile is a key
factor in deciding for the capacities of the wind turbines to be installed. The terrains that don’t have
a lot of hindrances like trees, houses or buildings will have less roughness length and can have higher
wind speeds, whereas, urban areas are usually the regions with low wind speeds. On seas or oceans,
much higher wind speeds are faced compared to wind speeds on land. On sunny days, wind speed is
more as compared to the overcast or cloudy days. At low altitudes, wind speed variation is more
significant on the oceans than on the land. During day time, wind speeds are generally lower than
during the nights. At lower altitudes, wind speeds are higher in day time than in nights and at higher
altitudes, wind speeds are higher in nights than in the day time. This different behavior can be
counted for the more temperature change near the surface than the higher altitudes where
temperature exchange between different air layers is not significant (see Figure 2.30).
Figure 2.30 Variations of wind speeds with heights, surfaces & times of a day
2.5 Wind Turbine Cost Figures7

The cost of a wind turbine varies from each design and specification but the biggest cost is the turbine
itself. This is a capital investment that buyers have to pay upfront which is typically around 75% of
the total cost of the project. As wind is a free source of powering wind turbines, only operational and
maintenance expenditures remain accountable after the wind turbine is set up. These O&M costs are
minimal as compared to the overall cost of the project. For the wind turbine, the largest cost
components are the rotor blades, the tower and the gearbox which altogether contributes to around
50 to 60% of the cost of a wind turbine. The electrical components like generator, transformer and
power converters accounts for about 13% of the turbine costs. The cost of onshore wind turbines is
6 Usman Zafar, Literature Review of Wind Turbines, Chair of Geotechnical Engineering, Bauhaus Universität, Weimar, 2018.
7Usman Zafar, Literature Review of Wind Turbines, Chair of Geotechnical Engineering, Bauhaus Universität,
Weimar, 2018.
50
Figure 2.31 Capital cost breakdown of an offshore wind farm [16]
less than the offshore wind turbines. According to International Renewable Energy Agency, the
typical installation range of an onshore wind turbine in 2010 was between USD 1800/kW and
2200/kW, while offshore wind turbine stood between 4000/kW and 4500/kW. The prices of wind
turbines from 2004 to 2010 continued to rise however, since 2010 a reduction in cost has been
observed. The reasons for this reduction is the improved design and overall performance of turbine
components and reduction in steel and carbon prices in the global markets (see Figure 2.31)8.
8International Renewable Energy Agency. Renewable Energy Cost Analysis - Wind Power. 1: Power Sector.
Available from: http://www.irena.org/publications/2012/Jun/Renewable-Energy-Cost-Analysis---Wind-
Power.
51
3 Case Studies Involving Horizontal Axis Wind Turbine (HAWTs)

3.1 Case Study 1 - Flow Around a Wind Turbine
Authors : Fernando Porté-Agel · Majid Bastankhah · Sina Shamsoddin
Published in : Boundary-Layer Meteorology, Springer Nature, Sept 2019.
Citation : Porté-Agel, F., Bastankhah, M. & Shamsoddin, S. Wind-Turbine and Wind-Farm Flows: A
Review. Boundary-Layer Meteorol 174, 1–59 (2020). https://doi.org/10.1007/s10546-019-00473-0
One of the main challenges in optimizing the design, of wind farms is the prediction of their
performance, owing to the complex multiscale two-way interactions between wind farms and the
turbulent Atmospheric Boundary Layer (ABL), as investigated by [Porté-Agel et al.]9. The article
reviews recent theoretical, experimental, and computational research on wind-turbine and wind-
farm flows, with emphasis on turbine wakes and their interaction with the ABL. The presence of a
wind turbine affects the airflow both upwind and downwind of the turbine [Wilson et al. ; Spera;
Burton et al.]. The upwind region affected by the turbine is called the induction region. Prior studies
(e.g., Medici et al. ; Simley et al.) have shown that the main impact of the turbine on that region is a
reduction in wind speed, which can be estimated acceptably with the following simple relationship
based on the vortex sheet theory [Medici et al.],
Figure 3.1 Schematic figure showing the flow regions resulting from the interaction of a wind turbine
and incoming turbulent boundary layer. Depicted are the most characteristic instantaneous (top) and
time-averaged (bottom) flow features
9Fernando Porté-Agel, Majid Bastankhah, Sina Shamsoddin, “Wind-Turbine and Wind-Farm Flows: A Review”,
Boundary-Layer Meteorology, 2019.
52
−0.5
u̅ 2x 2x 2
= 1 − a {1 + [1 + ( ) ] }
u̅∞ d d
Eq. 3.1
where u is the streamwise velocity component along the rotor axis (the overbar denotes time
averaging), x is the streamwise position (being zero at the turbine and negative upwind), u∞ is the
streamwise velocity component far upwind, d is the rotor diameter, and a is the rotor induction
factor. The region downwind of the turbine is called the wake. The wind-turbine wake itself is
generally divided into two regions (Vermeer et al.):
• Region immediately downwind of the turbine with a length of 2-4 rotor diameters, called the
near-wake,
• Region further downstream, called the far-wake.
Figure 3.1 shows a schematic of the different regions affected by the presence of the wind turbine.
The near-wake is directly influenced by the presence of the wind turbine, so characteristics of the
turbine, such as the blade profile, hub and nacelle geometry, can affect the flow field in this region.
As a result, the near-wake is characterized by highly complex, three-dimensional (3D), and
heterogeneous flow distribution. In contrast, the far-wake region is less influenced by detailed
features of the wind turbine. Instead, global wind-turbine parameters, such as the thrust and power
coefficient, and incoming flow conditions, are likely enough to predict the mean flow distribution in
Figure 3.2 Flow visualizations of the 3D helical vortex structures behind a turbine rotor for different
values of tip-speed ratio λ (figure reprinted from Okulov et al. (2014) with permission of Cambridge
University Press)
53
this region. In the following, we provide an overview of the aerodynamic research on wakes (both
near- and far-wake regions) of single turbines in horizontally-homogeneous boundary layers.
3.1.1 Near-Wake
3.1.1.1 Tip and Root Vortices
The most striking features of turbine near-wakes are perhaps the periodic helicoidal vortex
structures shedding from the tip and the root of the rotor blades (Figure 3.1). The presence of tip
and root vortices in the near-wake of wind turbines has been widely demonstrated in the literature
(see Figure 3.2). These vortex structures are caused by the difference in pressure between the
pressure and suction sides of the rotor blade (Andersen et al.). Consequently, their shedding
frequency is three times of the rotor rotational frequency for a three-bladed HAWT. While the helix
pitch (i.e., the streamwise distance between two consecutive vortices) of tip vortices is evidently
greater than the pitch of the root vortices, both decrease with the increase of tip-speed ratio (i.e., the
ratio between the velocity of the blade tip to that of the unperturbed incoming flow at hub height).
The evolution and stability of tip and root vortices have received extensive attention in the literature
both numerically and experimentally.
The main focus has been given to the study of tip vortices as they are more persistent (Sherry et al.).
Moreover, tip vortices can reduce flow entrainment in the near wake by separating this region from
the outer flow. Therefore, it is of great interest to understand the underlying mechanisms that lead
to the evolution and breakdown of tip vortices. To this end, several wind-tunnel studies have been
performed based on high-resolution particle-image velocimetry measurements (both phase-locked
and free-run) to visualize the tip vortices at different locations and time instants. These studies
reported that tip vortices have some random fluctuations around their statistically-averaged
positions. These random motions are referred to as vortex wandering or vortex jittering (Heyes et
al.), and their amplitude increases with the vortex age and the incoming turbulence intensity.
Different mechanisms have been proposed to be responsible for the breakdown of helical vortex
filaments. The mutual inductance instability is, however, considered as the dominant mode of
instability for helical vortex filaments when the helix pitch decreases beyond a certain limit. The
mutual inductance instability results in the pairing of tip vortices and ultimately their breakdown.
The decrease of helix pitch intensifies the mutual inductance instability, so the breakdown of tip
vortices occurs faster at higher tip-speed ratios (Sørensen et al.). It is also important to note that,
under turbulent boundary-layer inflow conditions, the lifetime of tip vortices is significantly reduced
due to the relatively high turbulence intensity and wind shear (Lu and Porté-Agel).
3.1.1.2 Hub Vortex
The presence of the so-called hub vortex, a vortical structure located at the central part of the near-
wake and elongated in the streamwise direction, has recently received some attention. Several wind-
tunnel and numerical studies, by various researchers, have shown that the hub vortex is
characterized by a single helix counter-winding instability, which interacts with the tip-vortex layer.
This helical vortex structure induces periodic motions in the central part of the near-wake. Similar
periodic motions in the central part of the near-wake have been also associated to vortex shedding
(e.g., Medici and Alfredsson), commonly seen behind bluff bodies (e.g., cylinders). It is a common
practice to describe the frequency of periodic oscillations by the dimensionless Strouhal number
St, which is given by
fd
St =
u̅h
Eq. 3.2
where f is the oscillation frequency, d is the rotor diameter, and ūh is the mean incoming wind speed
at hub height. A relatively large discrepancy exists between the values of St reported by different
numerical and wind-tunnel studies, ranging between 0.12 and 0.85. This emphasizes the need for
54
further study to elucidate the underlying mechanisms leading to the development of the hub vortex.
It should also be mentioned that all the above studies were performed with laboratory-scale wind
turbines; therefore, it is of interest to investigate if the same periodic motions can be observed in the
wake of utility-scale turbines, for which the ratio of the nacelle to the rotor is smaller than that of
laboratory-scale turbines.
Finally, it is also important to point out that these periodic motions observed in the central part of
the near-wake are different from the random oscillations of the turbine far-wakes, often referred to
as wake meandering. Meandering of turbine far-wakes is mainly caused by very large turbulent
structures in the incoming boundary layer, and is discussed in detail in Sect. 3.1.2.2.
3.1.1.3 Mean Flow Distribution
Based on the conservation of angular momentum, the near-wake rotates in the opposite direction
from that of the turbine blades [Manwell et al.], and the amount of the rotation decreases with
increasing downstream distance [Zhang et al.]. A speed-up region is also observed in the central part
of the near-wake, particularly at higher tip-speed ratios (Bastankhah and Porté-Agel). In spite of this
complex nature, for the sake of simplicity, the near-wake has been modelled in some studies with a
uniform velocity-deficit distribution in the central part, and a varying velocity deficit in the side shear
layers, as shown in Figure 3.1 (dashed lines). Based on this simplified description, the side shear
layers expand downstream until the central region with the uniform velocity deficit ultimately
vanishes. Further downstream, the far-wake region, characterized by a self-similar Gaussian velocity-
deficit distribution, is found. The length of the near-wake is influenced by a range of parameters such
as the turbulence intensity of the incoming flow [Wu and Porté-Agel], the mechanical shear generated
by the turbine (Vermeulen 1980), and the turbine tip-speed ratio. Different models have been
proposed in the literature to predict the length of the turbine near-wake. Based on the model
proposed by [Sørensen et al.], the normalized near-wake length xn/d is given by
xn 1 16u̅3c
= − [( ) ln(0.3𝐼) + 5.5 ln(𝐼)]
d 2 Nb λCT
Eq. 3.3
where ūc is the mean convective velocity of the tip vortices normalized by the incoming flow speed
(typically within the range of 0.73–0.78), Nb is the number of blades, λ is the tip-speed ratio, CT is the
thrust coefficient, and I is the incoming streamwise turbulence intensity.
3.1.2 Far-Wake
3.1.2.1 Mean Flow Velocity Distribution
In contrast to the near-wake region, the far-wake region has more universal characteristics as it is
less influenced by the detailed features of the rotor. Given the fact that turbine spacing in wind farms
usually falls within the range of 3 to 10 rotor diameters, wind turbines commonly operate in the far-
wake of upwind turbines. As a result, understanding turbine far-wakes is essential for improving the
prediction and optimization of wind-turbine power output in wind farms. In recent years, a great
deal of attention has been paid to studying mean flow distribution in turbine far-wakes by means of
field measurements , laboratory experiments [Medici and Alfredsson; Chamorro and Porté-Agel],
among many others, and numerical simulations. Due to the entrainment of the outer flow, the wake
is found to grow in both lateral and vertical directions as it moves downstream, and the value of the
streamwise velocity component increases until the wake completely recovers far downstream.
Early studies of wind-turbine wakes in uniform inflows showed that the streamwise velocity profiles
have an axisymmetric Gaussian distribution in this region. In the case of boundary-layer flows,
although later studies showed that wake velocity profiles lose the Gaussian shape due to the incoming
shear and the presence of the ground (see the schematic in Figure 3.1), profiles of the velocity deficit
Δū (i.e., difference between the incoming flow speed and that of the wake) still retain the Gaussian
55
distribution, except at the edge of the wake. The slight disagreement between the velocity-deficit
profiles and the Gaussian distribution seen at the wake edges has also been reported for other types
of wake flows [Pope; Johansson et al.; Okulov et al.].
One of the inherent characteristics of Gaussian profiles is self-similarity, implying that the profile of
velocity deficit (normalized by its maximum value) as a function of the distance from the wake center
(normalized by the wake width σ) is constant with streamwise position. Far-wake self-similarity
facilitates the development of simple analytical models for the prediction of the mean flow
distribution in this region, see Sect. 3.1.2.4.
Classical theoretical studies on three-dimensional wakes of bluff bodies (e.g., disks) have shown that
the wake velocity deficit Δū decays with x−2/3 along the rotor axis while the increase of the wake width
σ with the streamwise distance is proportional to x1/3.
These theoretical analyses are based on the assumption that shear-generated turbulence due to the
wake is mainly responsible for the wake recovery, and the effect of the incoming flow turbulence is
negligible. This theoretical result is confirmed by experimental studies of turbine far-wakes under
laminar inflow conditions. In more realistic situations when the ambient turbulence is present,
Figure 3.3 Contours of the time-averaged streamwise velocity component (in m s−1) in the vertical
plane normal to the rotor plane, at zero span, for different roughness lengths. Figure taken from Wu and
Porté-Agel (2012), in accordance with the Creative Commons Attribution (CC BY) license)
however, wake recovery deviates considerably from the aforementioned theory. Several LES, wind-
tunnel, and field studies of turbine wakes have shown that the wake width increases approximately
56
linearly with x, and its recovery rate, denoted by k, is larger for boundary layers with higher
turbulence intensity [e.g. Bastankhah & Porté-Agel ; Fuertes et al.].
This is the main reason why turbine wakes in a rough boundary layer recover more rapidly than
those in a smooth boundary layer [Chamorro and Porté-Agel ; Wu and Porté-Agel ]. This is illustrated
in Figure 3.3, showing contours of the time-averaged streamwise velocity component for the wake
of a wind turbine installed over flat terrain with different roughness lengths. This effect explains why,
in general, the capacity density of offshore wind farms is smaller than that of their onshore
counterparts.
3.1.2.2 Turbulence Distribution
In addition to the far-wake mean velocity distribution, turbulence characteristics of far-wakes have
been extensively studied in the literature. Specifically, the following turbulence quantities are mostly
considered:
3.1.2.2.1 Streamwise Turbulence Intensity
Wind turbine far-wakes are known to have a high turbulence intensity with respect to the incoming
flow, in particular the upper part of the wake. The increased turbulence intensity in far-wakes has
received considerable attention in the literature as it can induce harmful unsteady loads on
downwind turbines. The turbulence intensity (I), and turbulence intensity added by the turbine ΔI,
is given by [Frandsen],
σ𝑢 2 − I2
I= , ∆I = √Iw
u̅ℎ
Eq 3.4
where Iw is the streamwise turbulence intensity in the wake. Under uniform inflow conditions, Iw has
a double Gaussian profile with the maximum values occurring at the edge of the wakes. In boundary-
layer flows, while the maximum value of the turbulence intensity usually occurs close to the upper
edge of the wake as shown in Figure 3.4-a, the turbulence is suppressed by the turbine in regions
close to the ground. The value of ΔI reaches its maximum in the range of two to four rotor diameters
downstream at the top-tip level, coinciding with the transition between the near-wake and the far-
wake. The peak of I therefore occurs earlier for incoming boundary layers with higher turbulence
Figure 3.4 Distribution of, a streamwise turbulence intensity I = σu/ ūh, and b normalized kinematic
vertical turbulent momentum flux Bar(u’ w’)/ ū2h , in a vertical plane at zero span. Figure reprinted from
Barlas et al. (2016) with the permission of Springer Nature
57
intensity since the near-wake is shorter in this case [Wu and Porté-Agel]. Further downstream, the
value of turbulence intensity monotonically decreases with x in the far-wake. Different empirical and
semi-empirical models have been proposed in the literature to predict the variation of ΔI with x in
turbine far-wakes.
3.1.2.2.2 Turbulent Momentum Flux
The spatial distribution of the turbulent momentum flux in turbine wakes reflects the entrainment
of air from the outer flow towards the wake center, and given by
Turbulent Momentum Flux = ⏟̅̅̅̅̅

ρu ′ v′ , ⏟̅̅̅̅̅̅
ρu ′w′
Lateral Direction Vertical Direction

Eq. 3.5
where primes indicate turbulent fluctuations. Similar to the streamwise turbulence intensity, the
magnitude of the momentum flux is greater at the edges of the far-wake, especially close to the upper
edge of the wake where the wind shear is greater, as seen in Figure 3.4-b.
3.1.2.2.3 Turbulence Kinetic Energy (TKE)
The analysis of the TKE provides insights into the production and transportation of turbulence
structures in wind-turbine wakes, as given
1 ̅̅̅̅ ̅̅̅̅
′2 + v ̅̅̅̅̅
′2 + w ′2)
TKE = (u
2
Eq. 3.6
Prior studies showed that the TKE production has high values in the near-wake, particularly in the
upper edge of the wake, where mean shear and turbulent fluxes are significant. The generated TKE
at the edge of the turbine wake is then advected by the mean wind downstream.
3.1.2.3 Wake Twisting
Wake meandering relates to the random unsteady oscillations of the entire wake with respect to the
time-averaged wake centerline. These random oscillations lead to strong turbulence generation and
consequently harmful unsteady loads on downwind turbines. There is almost unanimous agreement
in the wind energy community that wake meandering is caused by very large turbulent eddies in the
incoming boundary layer. [Ainslie] is perhaps the first study to incorporate the effect of wake
meandering into the wake-flow prediction. Later, [Larsen et al.] postulated that, while the wake
recovery is governed by small turbulent eddies, the whole wake is advected passively by turbulent
Figure 3.5 Wake temporal oscillations at three rotor diameters downwind of a turbine. Velocity
contours obtained from lidar measurements in the field are shown in greyscale, and the red line
indicates the temporal variation of the wake center predicted by the DWM model. Figure reprinted from
Bingöl et al. (2010) with the permission of John Wiley and Sons, Inc
58
eddies larger than twice the rotor diameter. Therefore, if the low frequency variation of the incoming
flow is known, one can model random oscillations of the turbine wake as a passive scalar. This study
became the basis of the dynamic wake meandering (DWM) model that was later validated and used
to predict instantaneous wake center position and unsteady loads on downwind turbines in field.
Instead of the incoming flow speed, estimated the wake transportation based on the wake model.
Although this assumption is not consistent with the passive scalar hypothesis, they reported a better
agreement between DMW predictions and field measurements. The DMW predictions in comparison
with field measurements are shown in Figure 3.5.
The connection between the incoming flow characteristics and wake meandering has been further
studied in a series of recent wind-tunnel studies. [España et al.] experimentally confirmed that wake
meandering does not occur unless turbulent eddies much larger than the turbine rotor diameter exist
in the incoming flow. One of the commonly reported characteristics of wake meandering is that
lateral displacements are much more pronounced than vertical ones. The same author argued that
this difference is due to the higher value of σv than σw in turbulent boundary-layer flows.
[Bastankhah and Porté-Agel] hypothesized that this difference is due to the lateral meandering
tendency of very-large-scale motions (VLSMs) present in the incoming boundary layer. VSLMs or
superstructures are very long low- and high-momentum structures observed both in the atmospheric
surface layer and the logarithmic region of a laboratory scale boundary layer. The length scale of
VLSMs can exceed 20δ, where δ is the boundary-layer thickness [Fang and Porté-Agel], and they are
very energetic structures since they account for a considerable share of the TKE and shear stress. The
interaction of VLSMs with wind-turbine wakes might explain another feature of turbine wake
meandering: namely, the fact that the mean wake cross-section is not stretched laterally in spite of
large meandering motions in the lateral direction.
3.1.2.4 Analytical Wake Modelling
Some applications such as wind-farm-layout optimization require the prediction of wake flows for
many (on the order of thousands or more, depending on the optimization technique) scenarios
including, but not limited to, multiple layouts and variations in wind direction, wind speed, and
thermal stratification. Such optimization can only be achieved using simple and computationally
inexpensive wake models. These models can be divided into two main categories:
➢ Empirical Models,
➢ Analytical Models.
Empirical models have been used mainly to estimate the variation of the wake-center velocity deficit
with the streamwise distance from the turbine rotor. Based on these models, the velocity deficit is
generally assumed to have a power-law relationship with x, which is written as
∆u̅ x n
= A( )
u̅∞ d
Eq. 3.7
where A and n are coefficients obtained from experimental and numerical data. Unlike empirical
models, whose model equation is obtained solely by fitting experimental or numerical data, analytical
wake models are derived based on flow governing equations and, therefore, have a superior ability
to capture the physics. The wind-energy literature is enriched with many studies aimed at developing
analytical models for wind-turbine wakes. For the sake of brevity, here, we review those that
attracted the most attention: [Jensen], [Frandsen et al.] and [Bastankhah and Porté-Agel].
Jensen developed a pioneering turbine-wake model, which has been extensively used in the literature
and commercial software (e.g., WAsP, Wind PRO, Wind Sim, Wind-Farmer, and Open Wind). The
Jensen model is obtained by applying the conservation of mass to a control volume downwind of the
59
wind turbine, and then using the so-called Betz theory to relate the wind speed just behind the rotor
to the turbine thrust coefficient CT. It also assumes a top-hat distribution for the velocity deficit in the
wake for the sake of simplicity. The normalized velocity deficit based on this model is given by
∆u̅ 1 − √ 1 − CT
=
u̅∞ (1 + 2k t x/d)2
Eq. 3.8
where CT is the thrust coefficient of the turbine, ū∞ is the mean incoming flow speed, and Δū = ū∞ − ū.
The wake width is assumed to grow linearly with downwind distance and, therefore, the wake
growth rate, kt , is constant. Jensen suggested that kt = 0.1, whereas values of 0.04 or 0.05 for kt in off-
shore cases and 0.075 for onshore cases are suggested in the later literature. Alternatively, kt can be
estimated by the ratio of the friction velocity to the streamwise velocity component at the hub height
for the incoming boundary layer [Frandsen]. For a logarithmic wind profile, this approximately gives
0.5
kt ≈ z
ln (zh )
0
Eq. 3.9
where zh and z0 are the turbine hub height and the roughness length, respectively. [Peña and
Rathmann] extended the above relationship to account for the effect of thermal stratifications on the
wake growth rate. [Frandsen] used the conservation of mass and momentum for a control volume
around the turbine, with the same top-hat shape assumed for velocity-deficit profiles in the wake.
Based on this work, the normalized velocity deficit is given by
∆u 1 2CT
= (1 − √1 − x)
u̅∞ 2 β+α
d
Eq. 3.10
where α is of order of 10kt and
1 + √ 1 − CT
𝛽=
2 √1 − C T
Eq. 3.11
Note that β is meaningful only for values of CT smaller than one. As a result of the assumption of a
top-hat distribution for wake velocity-deficit profiles, these models tend to underestimate the
velocity deficit at the wake center and overestimate it at the edges of the wake. Moreover,
[Bastankhah and Porté-Agel] showed that top-hat models make the power predictions of downwind
turbines unrealistically sensitive to the lateral position of turbines with respect to each other.
Different numerical and experimental data were used by same authors, to show that self-similar
Gaussian distribution can acceptably represent velocity-deficit profiles in turbine far-wakes. The
normalized velocity deficit is therefore given by
𝑟 2
∆u (− )
= C(x)e 2𝜎2
u̅∞
Eq. 3.12
60
where σ is the wake width. A linear wake growth rate is assumed for the wake, since this is in
agreement with wind-tunnel measurements. Hence, σ is given by
σ x
=k +ϵ
d d
Eq. 3.13
where k is the wake growth rate, and is the initial wake width, equal to 0.2√β. The conservation of
mass and momentum in an integral form is expressed by
πd2
T= ρCT u̅2∞ = ρ ∫ u̅ (u̅∞ − u̅) dA
8
Eq. 3.14
where T is the turbine thrust force. Inserting Eq. 3.12 into Eq. 3.14 yields
∆u̅ CT y 2 + (z − zh )2
= 1 − √1 − + exp (− )
u∞ σ 2 2σ2
8( )
d
( )
Eq. 3.15
where σ is given by Eq. 3.13. In order to use this model to predict the wake velocity distribution, the
value of the wake growth rate k has to be estimated for each case. Note that the original version of
the model expressed by Eq. 3.15 assumes that the wake growth rate k is the same in both lateral and
vertical directions. [Abkar and Porté-Agel] and [Xie & Archer] showed, however, that the wake width
in the vertical direction can be different from that in the lateral direction due to the effect of the
ground or thermal stratification. Hence, for the sake of generality, the model can be written as
∆u̅ CT y 2 + (z − zh )2
= (1 − √1 − σ 𝜎 ) + exp (− )
u∞ 𝑦 𝑧
8( 2 ) 2σ σ
𝑦 𝑧
d
Eq. 3.16
The data obtained from wind-tunnel measurements [Bastankhah and Porté-Agel] are shown by black
solid lines. The predictions of the analytical models developed by [ Jensen], [Frandsen et al.] and
[Bastankhah and Porté-Agel ] are shown by red dashed lines, green dash-dot lines and blue dashed
lines, respectively where σy and σz are given by
σ𝑦 x σ𝑧 x
= k𝑦 + ϵ , = k𝑧 + ϵ
d d d d
Eq. 3.17
Here, ky and kz are wake growth rates in the y and z directions, respectively, and as mentioned earlier,
∈ = 0.2√β. Figure 3.6 shows the predictions of the analytical models reviewed above in comparison
with the wind-tunnel data recent. The model inputs are determined based on the incoming boundary-
layer flow conditions as well as turbine operating conditions reported in the mentioned study. The
growth rate of the top hat wake kt is calculated according to Eq. 3.9, while the wake growth rate k
for the last model with a Gaussian velocity deficit profile is estimated to be 0.022 based on the wind-
tunnel data.
61
A key parameter of this empirical model is the wake growth rate k, which depends on the turbulence
intensity in the incoming flow. [Niayifar and Porté-Agel] used LES data to propose the following
empirical linear relation to estimate k as a function of the streamwise turbulence intensity I (for 0.06
< I < 0.15),
k ≈ α1 I + α2
Eq. 3.18
with α1 = 0.38 and α2 = 0.004. A recent field study of wind-turbine wakes using two nacelle mounted
lidars has reported a reasonable fit of the measurements using Eq. 3.18 for the growth rate (with α1
= 0.35 and α2 = 0). It should be mentioned that, even though the streamwise turbulence intensity is
extensively used in analytical modelling of wind-turbine wakes (as discussed before), some studies
have suggested that the spanwise and vertical velocity component fluctuations play a dominant role
on the structure and dynamics of wind-turbine wakes. Considering this, [Cheng and Porté-Agel]
proposed a physics-based analytical model for the wake expansion based on Taylor’s diffusion theory
(Taylor).
Figure 3.6 Lateral (top) and vertical (bottom) profiles of the normalized velocity deficit through the
hub level at different downwind locations. The data obtained from wind-tunnel measurements
(Bastankhah and Porté-Agel 2017b) are shown by black solid lines. The predictions of the analytical
models developed by Jensen (1983), Frandsen et al. (2006) and Bastankhah and Porté-Agel (2014b) are
shown by red dashed lines, green dash-dot lines and blue dashed lines, respectively.
62
3.1.3 Summary & Conclusions

We, [Porté-Agel et al.]10, have reviewed the relevant literature on experimental, computational, and
theoretical studies of the interactions of ABL flow with wind turbines and wind farms. Emphasis has
been placed on the current state of our understanding and ability to model wind-turbine wake, flows
and their impact on ABL structure and wind-farm performance. This knowledge is essential for
optimizing the design and control of wind farms. The focused was on the simplest case of the
interaction between a stand-alone horizontal-axis wind turbine and the ABL over homogeneous flat
terrain. The structure and dynamics of the main flow regions (induction, near-wake, and far-wake
regions) are discussed, with emphasis on the role of atmospheric turbulence. The main conclusions
can be summarized as follows:
• The near-wake region, whose structure and dynamics (e.g., tip and hub vortices) are affected
by the geometry and operation of the wind turbine, has a length of about two to four rotor
diameters, depending on the turbulence intensity in the ABL.
• The mean flow velocity in the far-wake region, which depends only on global turbine
performance parameters (mainly CT ) and atmospheric turbulence, can be analytically
modelled using conservation of mass and momentum, together with the assumptions of a
Gaussian distribution of the velocity deficit and a nearly-linear wake expansion.
• Recent attempts have been made to estimate the role of atmospheric turbulence on the
growth rate of the far-wake by using empirical relations as well as theoretical developments
based on the analogy with passive scalar plumes.
• The above-mentioned analytical framework for the far-wake flow has been extended to the
case of turbines working under yawed conditions by using conservation of momentum in
both the streamwise and spanwise direction. Experimental and analytical evidence suggests
that yawing can be used as an effective wake mitigation strategy.
• Meandering of the far-wake has been associated with the dynamics of relatively large (larger
than twice the rotor diameter) turbulent eddy motions in the ABL. This connection has been
used to develop models for the position of the instantaneous wake center and the unsteady
loads on downwind turbines.
Notice : In the original paper, authors reviewed an extensive literature survey of subject, which some
been ignored here. For a complete information, readers are encouraged to consult the original source
[Porté-Agel et al.]11.
3.1.4 References
1. Wilson RE, Lissaman PB, Walker SN (1976) Aerodynamic performance of wind turbines. Final
report. Technical report, Oregon State University, Corvallis (USA). Department of Mechanical
Engineering.
2. Spera DA (1994) Wind turbine technology: fundamental concepts of wind turbine engineering.
ASME Press, New York
3. Burton T, Sharpe D, Jenkins N, Bossanyi E (1995) Wind energy handbook, 1st ed. Wiley, New
York.
4. Medici D, Ivanell S, Dahlberg JÅ, Alfredsson PH (2011) The upstream flow of a wind turbine:
blockage effect. Wind Energy 14(5):691–697
10 Fernando Porté-Agel, Majid Bastankhah, Sina Shamsoddin, “Wind-Turbine and Wind-Farm Flows: A Review”,
Boundary-Layer Meteorology, 2019.
11 Same as previous.
63
5. Simley E, Angelou N, Mikkelsen T, Sjöholm M, Mann J, Pao LY (2016) Characterization of wind

velocities in the upstream induction zone of a wind turbine using scanning continuous-wave
lidars. J Renew Sus. Energy 8(1):013,301
6. Medici D, Ivanell S, Dahlberg JÅ, Alfredsson PH (2011) The upstream flow of a wind turbine:
blockage effect. Wind Energy 14(5):691–697
7. Vermeer L, Sørensen J, Crespo A (2003) Wind turbine wake aerodynamics. Prog Aerospace Sci
39:467–510.
8. Andersen SJ, Sørensen JN, Mikkelsen R (2013) Simulation of the inherent turbulence and wake
interaction inside an infinitely long row of wind turbines. J Turbulence 14(4):1–24.
9. Sherry M, Sheridan J, Jacono DL (2013b) Characterization of a horizontal axis wind turbine’s
tip and root vortices. Exp Fluids 54(3):1–19.
10. Heyes A, Jones R, Smith D (2004) Wandering of wing-tip vortices. Proceedings of the 12th
international symposium on applications of laser techniques to fluid mechanics, pp 1–20.
11. Sørensen JN, Mikkelsen RF, Henningson DS, Ivanell S, Sarmast S, Andersen SJ (2015)
Simulation of wind turbine wakes using the actuator line technique. Philos Trans R Soc
373(2035):20140,071.
12. Lu H, Porté-Agel F (2011) Large-eddy simulation of a very large wind farm in a stable
atmospheric boundary layer. Phys Fluids 23(065):101.
13. Medici D, Alfredsson P (2006) Measurement on a wind turbine wake: 3D effects and bluff body
vortex shedding. Wind Energy 9:219–236.
14. Manwell JF, McGowan JG, Rogers AL (2010) Wind energy explained: theory, design and
application. Wiley, New York.
15. Bastankhah M, Porté-Agel F (2017c) Wind tunnel study of the wind turbine interaction with a
boundary-layer flow: upwind region, turbine performance, and wake region. Phys Fluids
29(065):105.
16. Wu YT, Porté-Agel F (2012) Atmospheric turbulence effects on wind-turbine wakes: an LES
study. Energies 5(12):5340–5362.
17. Sørensen JN, Mikkelsen RF, Henningson DS, Ivanell S, Sarmast S, Andersen SJ (2015)
Simulation of wind turbine wakes using the actuator line technique. Philos Trans R Soc
373(2035):20140,07.
18. Medici D, Alfredsson P (2006) Measurement on a wind turbine wake: 3D effects and bluff body
vortex shedding. Wind Energy 9:219–236.
19. Chamorro LP, Porté-Agel F (2009) A wind-tunnel investigation of wind-turbine wakes:
boundary-layer turbulence effects. Boundary-Layer Meteorol 132:129–149.
20. Pope SB (2000) Turbulent flows. Cambridge University Press, Cambridge.
21. Johansson PBV, GeorgeWK, Gourlay MJ (2003) Equilibrium similarity, effects of initial
conditions and local Reynolds number on the axisymmetric wake. Phys Fluids 15(3):603–617.
22. Okulov VL,Naumov IV, MikkelsenRF, Sørensen JN (2015) Wake effect on a uniform flow behind
wind-turbine model. Journal of Physics: Conference Series 625(1):012,011.
23. BastankhahM, Porté-Agel F (2014b) A new analytical model for wind-turbine wakes. Renew
Energy 70:116–123.
24. Fuertes FC, Markfort CD, Porté-Agel F (2018) Wind turbine wake characterization with
nacelle-mounted wind lidars for analytical wake model validation. Remote Sens 10(5):668.
25. Chamorro LP, Porté-Agel F (2009) A wind-tunnel investigation of wind-turbine wakes:
boundary-layer turbulence effects. Boundary-Layer Meteorol 132:129–149.
27. Frandsen ST (2007) Turbulence and turbulence-generated structural loading in wind turbine
clusters. Ph.D. thesis, Risø National Laboratory.
64
29. Ainslie J (1988) Calculating the flow field in the wake of wind turbines. J Wind Eng. Ind Aero
27(1):213–224.
30. Larsen GC,Madsen HA, ThomsenK, Larsen TJ (2008) Wake meandering: a pragmatic
approach. Wind Energy 11(4):377–395.
31. España G, Aubrun S, Loyer S, Devinant P (2011) Spatial study of the wake meandering using
modelled windnturbines in a wind tunnel. Wind Energy 14(7):923–937.
32. Zhang W, Markfort CD, Porté-Agel F (2012) Near-wake flow structure downwind of a wind
turbine in a turbulent boundary layer. Exp Fluids 52(5):1219–1235.
33. BastankhahM, Porté-Agel F (2017b) A new miniature wind turbine for wind tunnel
experiments. Part II: wake structure and flow dynamics. Energies 10(7):923.
34. Fang J, Porté-Agel F (2015) Large-eddy simulation of very-large-scale motions in the neutrally
stratified atmospheric boundary layer. Boundary-Layer Meteorol 155(3):397–416.
35. Jensen NO (1983) A note on wind turbine interaction. Technical report. Risø-M-2411, Risoe
National Laboratory, Roskilde, Denmark.
36. Frandsen ST, Barthelmie R, Pryor S, Rathmann O, Larsen S, Højstrup J, Thøgersen M (2006)
Analytical modelling of wind speed deficit in large offshore wind farms. Wind Energy 9:39–53.
37. BastankhahM, Porté-Agel F (2014b) A new analytical model for wind-turbine wakes. Renew
Energy 70:116–123.
38. Frandsen S (1992) On the wind speed reduction in the center of large clusters of wind turbines.
J Wind Eng. Ind Aero 39(1):251–265.
39. Peña A, Rathmann O (2014) Atmospheric stability-dependent infinite wind-farm models and
the wake-decay coefficient. Wind Energy 17(8):1269–1285.
40. Abkar M, Porté-Agel F (2015a) Influence of atmospheric stability on wind-turbine wakes: a
large-eddy simulation study. Phys Fluids 27(3):035,104.
41. Xie S, Archer C (2015) Self-similarity and turbulence characteristics of wind turbine wakes via
large-eddy simulation. Wind Energy 18:1815–1838.
42. Niayifar A, Porté-Agel F (2016) Analytical modeling of wind farms: a new approach for power
prediction. Energies 9(9):741.
43. ChengWC, Porté-Agel F (2018) A simple physically-based model for wind-turbine wake growth
in a turbulent boundary layer. Boundary-Layer Meteorol.
44. Taylor GI (1922) Diffusion by continuous movements. Proc London Math Soc s2–20(1):196–
212.
65
3.2 Case Study 2 - Design of the AOC 15/50 Rotor Blade for Wind Turbine
The main purpose of this study is to conduct a parametric sensitivity study on the blade design of
AOC 15/50 wind turbine based on a CFD approach and optimize the blade design for maximizing the
power output. [Kapoor] . The Fluent® flow solver using the k-ω SST turbulence model was validated
by simulating the flow over two dimensional airfoils comprising the AOC 15/50 wind turbine blade.
The CFD results have shown a considerable agreement with the experimental data for the airfoils.
Parametric correlation study and sensitivity analysis were conducted by performing actual flow
simulations over the turbine blade using ANSYS® Fluent®. This illustrates the dependence of power
output on the blade design parameters. Parametric correlation study reveals that the blade design
variables on the outer 40% of the blade span have a predominant effect on the power output of the
blade, while the obtained scatter plots and determination matrix indicate the blade optimization
problem setup as non-linear and quadratic fit.
The most sensitive design parameters are used to formulate the flow optimization problem. A
response surface optimization (RSO) methodology is employed for carrying out the blade shape
optimization process. Design of Experiments (DoE) using the Latin Hypercube Sampling (LHS)
algorithm is used to construct a robust response surface model, which is then searched for the
optimized design using the Nonlinear Programming by Quadratic Lagrangian (NLPQL) technique.
Two optimization routines are carried out by varying the geometric constraints on the blade. First
optimization routine constrained the blade length and maximum chord occurring at a 40% span
location from the hub to be fixed, yielding a design that performs marginally well up to the wind
speed of 9.2 m/s with a maximum power increment of 7.55 % occurring at the 8.03 m/s wind speed.
The search for the second optimization routine was initialized in the design space with the best
candidate point obtained from the first optimization routine. Second optimization routine generated
a design configuration that resulted in an increased blade length and surface area, thus leading to an
overall lift force augmentation producing a 25.26% increase in the power output. Both the optimized
candidates obtained were validated using the flow solver to verify the optimized design for
maximized power output. The coefficient of pressure plots at various span locations of the blade
bolster the claim that most of the mechanical power is produced in the outer 30-40% of the blade.
3.2.1 Blade Design in ANSYS® DesignModeler
At each station along the length of the blade, the airfoil shapes are the same as that for the AOC 15/50
wood-epoxy blade (Figure 3.7)
used in the test configuration, which
has a length of 7.5 m (≈ 295 in). The
root of the AOC 15/50 blade starts at
the hub-blade connection, at a
radius 11 inches from the center of
the hub. At the root end of the blade,
the cross-sectional shape is
relatively oval and is only semi-
aerodynamic. From the root region,
the blade transitions from an oval
shape to an aerodynamic shape at
40% of the tip radius as defined by
the SERI 821 airfoil shape. Outboard
from the root region, the shape
transition continues span-wise to a
shape is based on a SERI 819 airfoil
at 75% of the tip radius and a shape
Figure 3.7 AOC 15/50 Blade Geometry
66
that is based on a SERI 820 airfoil at 95% of the tip radius.

The blade was designed in the ANSYS® DesignModeler by using the curve generation function to
import the three different airfoil profiles (S819, S820 and S821) and then the 3D blade was modeled
by using the skin/loft feature. Since the hub does not hold any importance in this case study, it was
modeled to be a simple circular extrusion to which another two blades were duplicated at 120°
angular symmetry using the pattern feature. The blade root section was twisted towards the feather
at 1.54° and the blade tip was given a feather angle of -1.54°(away from the feather) to represent the
same blade geometric features as used in the Power Performance Test Report for AOC 15/50. Also
the blade was imparted a 6° of positive (downwind wind turbine) cone angle. Figure 3.8 shows
Various Radial Stations.
Figure 3.8 Turbine Blade Showing Various Radial Stations in ANSYS® DesignModeler
3.2.2 Design Variables and Design Space

From an aerodynamic shape optimization point of view, the system is basically the blade geometry
that has to be optimized for a specific operating condition (wind speed). The design points are the
design variables that completely define the blade geometry. In this problem formulation, there are
11 design variables, namely: radial sectional fraction (r/R) of three airfoil sections, chord length (c)
of the three airfoil sections, twist angle (θ) of the three airfoil sections from which the entire blade is
lofted span-wise. The blade cone angle (ϕ) is the tenth and the attach angle (α) also called as the pitch
angle is the eleventh design variable. The design space is the region bounded by the upper and lower
limits of the design variables. This implies that the design variables are allowed to vary only within
the limits defined by the design space. It is defined such that, overly unusual or unrealistic shapes are
not attained. (see Table 3.1).
3.2.3 Constructing the Response Surface Method (RSM)
RSM builds a response model by calculating data points with experimental design theory to prescribe
a response of a system with independent variables. The relationship can be written in a general form
as follows:
y = F(𝐗) + ε
Eq. 3.19
where 𝜖 represents the total error, which is often assumed to have a normal distribution with a zero
67
Table 3.1 Design Space
mean. Consider a sampling plan 𝐗 and a set of 𝑁𝑠 observed values comprising the responses obtained
from the computer simulations:
𝐗11 𝐗12 ⋯ 𝐗1Nvar → y1

𝐗 21 𝐗 22 … 𝐗 2Nvar → y2
𝐗 DoE = [ ⋮ ]
⋮ ⋮ ⋱ ⋮ ⋮
[𝐗 NS1 𝐗 NS 2 ⋯ 𝐗 NSNvar ] → yN S
Eq. 3.20
The polynomial approximation of order m (degree 𝑚−1) of a function f is, essentially, a Taylor series
expansion of f truncated after 𝑚−1 terms. This suggests that a higher order expansion will usually
yield a more accurate approximation. However, the greater the number of terms, the more flexible
the model becomes and there is a danger of over-fitting the noise that may be corrupting the
underlying response values, thereby introducing truncation errors in the predicted output function
value. A full quadratic polynomial (degree 2, order 3) approximation of F can be written as:
Nvar Nvar Nvar −1 Nvar
𝐲̂ = ̂f(𝐱, 𝛃) = β1 + ∑ βi xi + ∑ βjj xj + ∑ ∑ βij xi xj

i=1 j=1 i=1 j=i+1
Eq. 3.21
Here 𝛽1, 𝛽𝑖, 𝛽𝑖𝑗 etc. are the regression coefficients of the polynomial. The total number of these
coefficients is 𝑛𝑡 = (Nvar+1)(N var+2)/2. These values can be determined using the standard least square
fitting regression of an over determined problem:
68
𝐲 = 𝚽𝛃
Eq. 3.22
Here y is the initial response matrix [𝑦1, 𝑦2,…,]𝑇 and
𝚽 is the [Vandermonde] matrix of size (N𝑠 × Nvar)
given in [Kapoor]12. Figure 3.9 displays response
surface variation of chord and twist angle at station 4
and 3 respectively, w.r.t Torque.
3.2.4 CFD Analysis on the AOC 15/50 HAWT Blade
Physical flow analysis of turbine rotor blades using
wind tunnel would be possible for small scale rotors,
but the increase in diameters has called for the use of
computational fluid dynamics for fluid flow over
blades and predication of loads. In this research
work, a compressible Navier-Stokes (N-S) solver
ANSYS® Fluent was used to predict the
aerodynamics of the blade. The main aim of this Figure 3.9 Response Surface Showing
research is to develop and validate a numerical Variation of P3, P9 with respect to P12
methodology for predicting the torque on the AOC (Torque)
15/50 HAWT blade. Simulations were performed
with the commercial software ANSYS® Fluent, using a k-ω SST turbulence model.
3.2.5 CFD Domain Mesh and Numerical Model for Rotating Bodies
This subtopic gives an insight into the CFD numerical models for turbo-machinery applications. The
im of this paragraph is providing the numerical basis to perform CFD simulation of rotating bodies.
The main challenge in turbo-machinery applications is the introduction of a rotating body to apply
forces on the fluid (e.g. compression or expansion). From an analytical point of view the rotation
should be introduced into constitutive equations of motion, and there are mainly two approaches:
the Moving Reference Frame (MRF) and the Sliding Mesh (SLM). The first one consists of rewriting
N-S equations in a rotating frame, while the second one introduces rotation assigning a rotational
component of velocity to all nodes of the domain (physical grid rotation). It is immediately
understandable that SLM approach is more realistic that MRF, but also more CPU demanding as the
computational model needs re-meshing at every time advancement during the simulation procedure.
Since the rotation of grid intrinsically depends on time-evolution of simulation, this approach is not
recommended for steady state simulations as the solution obtained is not time-dependent. In other
words, a time steady calculation performed with MRF approach according to the evidence in most of
the turbo-machinery problems, does not compute a time-accurate solution.
3.2.6 Moving Reference Frame Model
Moving Reference Frame (MRF) model solves the equations of motion of a steady formulation in a
moving frame. For a rotating frame with constant rotational speed, it is possible to transform the
equations of motion to the rotating frame such that steady-state solutions are possible. This approach
is based on the assumption that in most of cases of practical interest, steady solutions are required
for rotating bodies, without taking into account the unsteady details of the flow field (e.g. vortex
shedding from a bluff body). On the other hand, an unsteady solution using the MRF model can also
be computed to simulate the unsteady details. Consider a coordinate system which is rotating with
an angular velocity ω relative to a stationary (inertial) reference frame, as illustrated in Figure 3.10.
12 Gaurav Kapoor, “Exploration of a Computational Fluid Dynamics Integrated Design Methodology for Potential
Application to a Wind Turbine Blade”, Thesis Submitted to the Department of Aerospace Engineering, College
of Engineering, Embry-Riddle Aeronautical University, Daytona Beach, December 2014.
69
Figure 3.10 Rotating Body in the Inertial Reference Frame
The origin of the rotating system is given by a position vector ro. In accordance to the MRF method,
the computational domain for the CFD problem can then be defined with respect to the rotating
reference frame, such that an arbitrary point in the CFD domain is located by a position vector 𝑟 from
the origin of the rotating frame. The fluid velocities can be transformed from the stationary frame to
the rotating frame using the relation,
vr = v − (ω × r)
Eq. 3.23
In the above equation, vr is the relative velocity (the velocity as viewed from the rotating frame) while
𝑣 is the absolute velocity (the velocity as viewed from the stationary frame). When the equations of
motion are solved in the rotating reference frame, the acceleration of the fluid is increased by the
additional terms that appear in the momentum equation. Moreover, the equations can be formulated
expressing absolute or relative velocity as dependent variable of momentum equation. Constitutive
N-S equations for which the solution is being calculated according to the relative velocity formulation
for continuity, momentum and energy respectively are as follows:
∂ρ
+ ∇. ρ𝐯r = 0
∂t
∂(ρ𝐯𝐫 )
+ ∇. (ρ𝐯𝐫 . 𝐯r ) + ρ(2𝛚
⏟ × 𝐯r + ⏟
𝛚 × 𝛚 × 𝐯r ) = −∇p + ∇. 𝛕 + 𝐅
∂t
Coriolis Centripetal
∂(ρEr )
+ ∇. (ρ𝐯𝐫 . Hr ) + ∇. (k∇T + 𝛕. 𝐯𝐫 ) + 𝐒H
∂t
Eq. 3.24
The momentum equation formulated above contains two additional acceleration terms, the Coriolis
component of acceleration (2 𝜔 x 𝑣𝑟 ) and the centripetal acceleration (𝜔 x 𝜔 x 𝑣𝑟 ). In addition, viscous
stress tensor does not change with respect to the MRF equation, except for the introduction of the
70
relative velocity. Energy equation is written in the form of internal energy Er, introducing the total
enthalpy Hr of the system in consideration. MRF model can be applied to different zones in the
domain (both rotating and nonrotating), solving RANS formulation of equations. Moreover
translational or rotational periodic boundaries can be applied wherever periodic surfaces are
present in the domain. For these reasons MRF model is widely used for industrial applications, being
one of the most versatile and low CPU-demanding approaches for turbo-machinery simulation.
3.2.7 Computational Domain (Grid) for the Turbine Blade Model
The meshing is performed in the ANSYS® Meshing module after importing the respective blade
geometry and creating a flow domain around the airfoil cross section by using the Boolean feature
for the volume extraction. The full rotor three bladed model can be reduced to a symmetric model of
a single blade with a 120 degree rotational symmetry along the global Y-axis. In order to simply our
CFD model and save computational resources, simulations are performed on a wedge shaped
computational domain (120º periodicity) with rotational periodic boundary conditions applied to
the wedged faces of the domain. It implies that the velocities going out from the left symmetry
boundary can enter the right boundary on the other side in an infinite loop. It was further assumed
that the flow conditions on either side of the 120º wedge are fully symmetric. A hybrid mesh topology
is used as the computational domain around the blade which extends to 10 times the blade length in
the upstream direction and 30 blade lengths in the downstream direction, as measured from the
global origin. The rotational periodic boundary conditions applied to the wedged faces of the
computational domain. The grid thus obtained is a combination of structured grid with hexahedral
elements in the far-field region and tetrahedral elements in the near-field region of the blade. An
inflation layer of 25 structured prismatic cells stacked one on another is used to capture the boundary
layer effects. The thickness of the first cell to the wall was kept at 6.3 x 10-5 m so that the y+ value
falls between 1 and 3. Patch dependent geometry controls were set for the meshing algorithm to
make sure that during successive meshing iterations, the mesh grows outward from the blade surface
and fully captures the geometric details of the blade. The overall mesh has a geometric grid growth
rate of 1.20. A sequence of 20 smoothing iterations were carried out post meshing to repair the grid
and bring down the average skewness to 0.88. The grid points are clustered in the proximity and the
wake region of the blade to capture the flow physics accurately.
3.2.7.1 Boundary Condition
The boundary conditions for the computational domain are set as listed below;
➢ Velocity Inlet – The upstream surface of the domain is set to the velocity inlet condition as the
free stream velocity to be simulated in the computational domain is
known beforehand.
➢ Wall – The blade upper and lower surfaces are selected as wall with no-slip condition.
Periodic Boundary – The edges of the computational domain on either side of the wedge are
selected to be the periodic boundaries. The velocities going out from the left symmetry
boundary can enter the right boundary on the other side in an infinite loop.
➢ Pressure Outlet – The surface of the computational domain downstream from the blade is
set to a pressure outlet condition. This gives a better prediction of the exit pressure
distribution and thus results in better accuracy of the overall solution. The pressure at the
outlet was set to be atmospheric pressure.
➢ Symmetry – The curved surface of the computational domain is selected to be the symmetry
type. This just means that these boundaries do not affect the flow in any possible way.
3.2.7.2 Grid Independence Study
An initial grid independence study was performed in order to be sure that the flow solutions obtained
in the later sensitivity analysis were consistent and independent of the grid used for discretizing the
flow domain. Three grid topologies; coarse (3.9 M elements), medium (6.6 M elements) and fine (9
71
M elements) were used for obtaining the

initial solution. The cell count was differed
by clustering more prismatic cell layers
near the blade surface where the
boundary layer effects take place. The
thickness of the first cell to the wall was
kept at 6.3 x 10-5 m so that the y+ value
falls between 1 and 3. Such range of y+ is
suitable for the tested turbulence models.
Since torque acting on the blade is of
primary concern for this study, the torque
on the blade was the deciding factor for
finding the optimum grid for this flow
problem. Medium grid quality was chosen
to be the best candidate as it exhibited Figure 3.11 Grid Independence Study
grid independence to the next iteration
towards a finer grid, as seen from Figure
3.11.
3.2.8 Flow Simulation over the Rotor Blade
Simulations were performed with the commercial software ANSYS® Fluent, using a RANS model. A
pressure based compressible flow solver with k-ω SST turbulence model was used for the flow
Blade AOC 15/50 Atlantic Orient Corporation

Solver Pressure-based
Transient Formulation Second Order Implicit
Velocity Formulation Absolute
Time Steady and Unsteady
Time Step Size 0.01 sec
Time Stepping Method Fixed
Turbulence Model k-ω SST
Fluid Material Air
Moving Reference Frame (Frame Motion) Symmetric about global Y-axis
Rotational Velocity 65 rpm ≈ 6.8067 rad/s (clockwise)
Temperature 288.16 K
Velocity 5.96, 7.0, 8.03, 10.98, 12.02 m/s
Density 1.225 Kg/m3
Pressure 101325 Pa
Dynamic Viscosity (μ) 1.7894e-05 Kg/m-s
Ratio of Specific Heats (γ) 1.4
Pressure-Velocity Coupling Scheme SIMPLE
Spatial Discretization & Interpolation Scheme
Gradient Least Squares Cell Based
Pressure STANDARD
Momentum Second Order Upwind
Turbulent Kinetic Energy Second Order Upwind
Specific Dissipation Rate Second Order Upwind
Table 3.2 Table of CFD Solver Settings

72
simulation. For simplifying the computational model, the atmospheric boundary layer effects in the
inflow, the near and the far wake modelling and their subsequent interactions with the mean flow
were neglected in the simulations. Since the study focusses on the dependence of blade geometry on
the torque produced, a uniform inflow velocity profile was modelled for all CFD simulations for
parametric study and sensitivity analysis. All simulations were computed in steady state until
convergence or till the end of prescribed iterations to allow developed flows in the domain. Then in
order to maintain computational stability, the simulations were switched to transient solver.
Convergence was monitored looking at the thrust force time histories over different revolutions and
reached in a few cycles (about 2 to 3) for all wind conditions tested. Also the residual tolerance of
10-6 was reached for all velocity and energy terms to ascertain the robustness of the obtained flow
parameters. Furthermore, the difference in the mass flow at the inlet and the outlet of the
computational domain showed a negligible error (order 10-6). Additionally, a vertex point was
created on the symmetry axis at one blade length downstream of the blade to track the history of
average velocity at the vertex point over the course of simulations. The simulations were stopped
when the average velocity at this vertex was fairly constant and did not show any appreciable change.
All the above four conditions were satisfied as per the best practices to be followed in ANSYS® Fluent
for obtaining an accurate and converged solution. (See Table 3.2).
3.2.9 Results for Flow Simulation over Rotor Blade
Flow simulations were carried at five different wind speeds: 5.96, 7.0, 8.03, 10.98 and 12.02 m/s. The
power obtained is calculated from the product of torque (τ) and angular velocity (ω).
Power = 𝛕. 𝛚
Eq. 3.25
Coefficient of power (CP), a measure of how efficiently a wind turbine converts the energy available
in the wind to electricity.
3.2.10 Optimization Method
The optimization algorithm used in this study employs Nonlinear Programming by Quadratic
Lagrangian (NLPQL) technique based on Latin Hypercube Sampling and Kriging Response
Surface. This is a gradient based algorithm to provide a global optimization result. Since, we have a
single objective to achieve, this technique is best as it can deal with multiple constraints and aims at
finding the global optimum. Our optimization problem is now reduced to Objective: Maximize
Torque - Two optimization routines are carried out as following:
3.2.10.1 Routine 1
The total length of the blade (7.5 m) and the maximum chord (0.749 m) occurring at Station 2 are
kept a constant (constrained) with an aim to optimize the existing blade within the length
requirements.
3.2.10.2 Routine 2
The starting point of this routine is taken as the best candidate point of Routine 1 to begin the search
on the response surface. The constraints applied are bounded by the design space spanning (+ -) 10%
from the base value of the design variables P3, P4 and P9. Further details are available in13.
3.2.11 Optimization Results
3.2.11.1 Routine 1
There is no change in the values of design variables P3 and P4 after the optimization routine 1. But
the optimized value of P9 turns out to be 2.66 degrees instead of the baseline value of 0 degrees. This
essentially means that the 7.85% increase in power output from the blade is solely the result of
13 See Previous.
73
optimum value of the twist at station 3 (SERI 819 airfoil). As evident from the optimization results,
the baseline design of the blade is highly engineered for maximum power output. The graph
(obtained from what-if scenario study) is in agreement with the above optimization result. The graph
Table 3.3 Table Showing Optimized Candidate Point for Routine 1
clearly shows that if all the other input parameters are held constant, P9 (Twist Station 3) at a value
of approximately 2.66 degrees gives the maximum blade torque output of about 856.14 Nm. The
graph below verifies the optimization routine 1 carried out. (See Table 3.3).
3.2.11.1.1 Routine 2
There is a change in the values of
design variables P3, P4 and P9
after the optimization routine 2.
The optimized value of P3 turns
out to be 0.43578 m (+7.33%)
instead of the baseline value of
0.406 m. Also, the optimized
values of P4 and P9 are 5.214 m
(+10%) and 2.9549º (+10.87%)
respectively. This also indicates
that the total blade length has
been increased by 10%, which
results in the augmented power
of 1069.5 Nm (+25.26%). The Figure 3.12 Torque (P12) vs Twist_Station3 (P9) for
graph (obtained from what-if Optimization Routine 2
scenario study) in Figure 3.12
above is in agreement with the above optimization result. The graph clearly shows that if all the other
Table 3.4 Table Showing Optimized Candidate Point for Routine 2

74
input parameters are held constant, the optimum design values for P3, P4 and P9 (Twist Station 3)
at a value of approximately 2.9 degrees gives the maximum blade torque output of about 1069 Nm.
The graph also verifies the optimization routine 2 carried out. (See Table 3.4).
3.2.12 Conclusion
In this research, the flow around the airfoils comprising the HAWT blade and the three dimensional
rotor blade is established using the commercial solver ANSYS® Fluent. A pressure based
compressible flow solver with k-ω SST turbulence model was used for all the flow simulations. To
study the dependence (sensitivity) of blade geometric/design parameters (what-if scenario) on the
power generated using ANSYS® Fluent, the Design of Experiments (DOE) approach of ANSYS
DesignXplorer was used. Parameter correlation study and sensitivity analysis conducted gave an
insight to how the changes in the blade geometry would affect the power output of the blade. The
blade aerodynamic optimization inclined toward the non-linear or quadratic relationship between
parameters, clearly indicated by the scatter plots and the quadratic determination matrix. This
parametric correlation study reveals that the blade design variables on the outer 40% of the blade
span have a predominant effect on the power output of the blade. Only the most sensitive design
variables are used for the blade optimization problem.
Using the results obtained from CFD simulations, a full quadratic polynomial response surface model
(RSM) is constructed, which is then optimized using the Nonlinear Programming by Quadratic
Lagrangian (NLPQL) technique to obtain the optimum values of the design variables. For
constructing the RSM, the Latin Hypercube Sampling (LHS) design is used to obtain the Design of
Experiments (DoE) plan. The main advantage of using this approach for shape optimization
problems is that values obtained from commercially available flow solvers can directly be used in the
optimization process, without making any changes to the solver’s code. Also the noise and non-
smoothness issues associated with CFD results are smoothened out by using the RSM which is
quadratic polynomial in terms of the design variables. Thus the optimization process can be
performed effectively and smoothly without any sudden divergence issues associated with the CFD
results. As evident from the CFD validations carried out on the optimum candidate point, the
optimization algorithm generated a design configuration that resulted in a localized optimum design
that had increased power output (+7.55%) at wind speed of 8.03 m/s only. The algorithm thus
resulted in a local optimum solution rather than a global optimum. Achieving a global optimum
solution to this problem would require several data points to be generated for obtaining a complete
and well established response surface spanning the entire operating wind spectrum of the turbine,
this is a costly affair in terms of the computational resources available. The Cp plots at various span
locations also bolster the claim that only the outer (from tip) 30-40% of the blade contributes most
towards the power output.
75
3.3 Case Study 3 - Multi-Element Ducts for Ducted Wind Turbines

Authors : Vinit V. Dighe1, Francesco Avallone1, Ozer Igra2, and Gerard van Bussel1
Affiliation : 1Wind Energy Research Group, Faculty of Aerospace Engineering, TU Delft, Delft, the
Netherlands
2Department of Mechanical Engineering, Ben-Gurion University of Negev, Beersheva, Israel
Citation : Dighe, V. V., Avallone, F., Igra, O., and van Bussel, G.: Multi-element ducts for ducted wind
turbines: a numerical study, Wind Energy. Sci., 4, 439–449, https://doi.org/10.5194/wes-4-439-2019,
2019.
Multi-element ducts are used to improve the aerodynamic performance of Ducted Wind Turbines
(DWTs)14. Steady-state, 2D (CFD) simulations are performed for a multi-element duct geometry
consisting of a duct and a flap. The goal is to evaluate the effects on the aerodynamic performance of
the radial gap length and the deflection angle of the flap. Solutions from inviscid and viscous flow
calculations are compared. It is found that increasing the radial gap length results in an augmentation
of the total thrust generated by the DWT, whereas a larger deflection angle has an opposite effect.
Reasonable to good agreement is seen between the inviscid and viscous flow calculations, except for
multi-element duct configurations characterized by large flap deflection angles. The viscous effects
become stronger at large flap deflection angles, and the inviscid calculations are incapable of taking
this phenomenon into account.
3.3.1 Introduction
Ducted wind turbines (DWTs) represent an interesting technological solution for increasing energy
extraction with respect to conventional horizontal-axis wind turbines (HAWTs) for a given rotor
radius and free-stream velocity [de Vries]15. DWTs consist of a rotor and a duct; the role of the latter
is to increase the mass flow rate through the rotor relative to a similar rotor operating in the open
atmosphere, thereby increasing the generated power. There is more than one explanation for how
this occurs. One explanation, as stated by [van Bussel]16, is that the duct forces an expansion of flow
downstream of the turbine beyond what is attainable for a bare wind turbine. This provides reduced
pressure on the downstream of the turbine, thereby increasing the total mass flow through the
turbine.
A second explanation, as argued by [de Vries], is that if the sectional lift force of the duct is directed
towards the turbine plane, then the associated circulation of the duct induces an increased mass flow
through the turbine. This solution is particularly suited for urban areas where the radius of the rotor
is a constraint and the free-stream velocity is low due to the presence of buildings. [van Busse], using
a one-dimensional momentum theory approach, found that the maximum power coefficient obtained
by a DWT can exceed the Betz limit by a factor of 2.5.
The best aerodynamic performance for a DWT can be achieved by increasing the duct expansion ratio
[Liley and Rainbird]17; [Foreman et al.]18; [Loeffler and Vanderbilt]19; [Foreman and Gilbert,]20;
14 Vinit V. Dighe, Francesco Avallone, Ozer Igra, and Gerard van Bussel,, “Multi-element ducts for ducted wind
turbines: a numerical study”, Wind Eng. Sci., 4, 439–449, 2019.
15 de Vries, O.: Fluid dynamic aspects of wind energy conversion, No. AGARD-AG-243, Advisory Group for
Aerospace Research and Development NEUILLY-SUR-SEINE, France, 1979.

16 van Bussel, G. J.W.: The science of making more torque from wind: Diffuser experiments and theory revisited,
Journal of Physics: Conference Series, Vol. 75, p. 012010, IOP Publishing, 2007.
17 Lilley, G. M. and Rainbird, W. J.: Preliminary report on the design and performance of ducted windmills, College
of Aeronautics, Cranfield, 1956.

18 Foreman, K. M., Gilbert, B., and Oma, R. A.: Diffuser augmentation of wind turbines, Solar Energy, 1978.
19 Loeffler, A. L. and Vanderbilt, D.: Inviscid flow through wide-angle diffuser with actuator disk, AIAA J. 1978.
20 Loeffler, A. L. and Vanderbilt, D.: Inviscid flow through wide-angle diffuser with actuator disk, AIAA J, 1978.
76
[Samson and Katebi]21. As a drawback, for a duct with a large ratio of duct outlet to rotor area, flow
separation along the duct inner walls might be present [Aranake et al]22; [Dighe et al.]23. [Tang et al.]24
experimentally investigated the effects of variable duct expansion ratios on the aerodynamic
performance of DWTs. They found that increasing the expansion ratio over a certain limit results in
a power reduction. This was linked to the appearance of flow separation within the inner walls of the
duct.
An alternative approach to improve the aerodynamic performance is to increase the camber of the
airfoil, used as a cross section of the duct, until separation occurs behind the rotor plane. Since, as
expected, flow separation has an undesired effect, solutions to prevent flow separation via active
boundary layer control techniques have been proposed by various researchers. However, Published
by Copernicus Publications on behalf of the European Academy of Wind Energy e.V. 440 V. V. Dighe
et al.: Multi-element ducts for ducted wind turbines: a numerical study the performance benefits
were limited by the cost of the active system and its installation. Another possible solution to improve
the aerodynamic performances of DWTs, which has only been explored to a very limited extent,
consists of using a duct with a flap (i.e., a multi-element duct). The flap is realized as a secondary duct
with a small chord airfoil cross section mimicking high-lift devices for aircraft wing (see Error! R
eference source not found.).
A first theoretical and experimental analysis of DWTs with a flap was carried out by [Foreman et al.]25
and [Igra]26. The latter found that the addition of a flap improves the DWT aerodynamic performance
by 25% with respect to a single duct. The flap inhibits flow separation along the inner duct wall and
increases the camber of the equivalent airfoil, thus being beneficial for the aerodynamic
performances of the DWT [Dighe et al.]27. The literature, however, is missing a detailed parametric
study that investigates the effect of flap installation setting, i.e., the radial location and its angle of
attack, on the total power generated by a DWT.
The goal of this paper is to conduct a parametric study to investigate the effect of the installation
settings of the flap on the aerodynamic performance of a multi-element DWT. This is performed using
computational fluid dynamics (CFD). To this aim, a reference multi-element duct is selected and the
rotor is simulated by a uniformly loaded Actuator Disk (AD) model.
3.3.2 Multi-Element Duct–AD Flow Model
The incompressible flow past a wind turbine is computed by substituting the rotor with an AD of
infinitesimal width. The AD exerts a uniform thrust force TAD per unit area. Then, the
nondimensional thrust force coefficient is
21 Samson, J. and Katebi, R.: Shroud design criteria for a lighter than air Wind Energy System, Journal of Physics:
Conference Series, Vol. 524, p. 012079, IOP Publishing, 2014.
22 Aranake, A. C., Lakshminarayan, V. K., and Duraisamy, K.: Computational analysis of shrouded wind turbine
configurations using a 3-dimensional RANS solver, Renew. Energy, 75, 818–832, 2015.
23 Dighe, V. V., de Oliveira, G., Avallone, F., and van Bussel, G. J.W.: Towards improving the aerodynamic
performance of a ducted wind turbine: A numerical study, in: Journal of Physics: Conference Series, vol. 1037, p.
022016, IOP Publishing, 2018a.
24 Tang, J., Avallone, F., Bontempo, R., van Bussel, G. J., and Manna, M.: Experimental investigation on the effect of
the duct geometrical parameters on the performance of a ducted wind turbine, Journal of Physics: Conference
Series, Vol. 1037, p. 022034, IOP Publishing, 2018.
25 Foreman, K. M. and Gilbert, B.: Diffuser for augmenting a wind turbine, US Patent No. 4,482,290, 1984.
26 Igra, O.: Research and development for shrouded wind turbines, Energy Convers. Manage, 1981.
27 Dighe, V. V., de Oliveira, G., Avallone, F., and van Bussel, G. J. W.: Characterization of aerodynamic performance
of ducted wind turbines: A numerical study, Wind Energy, https://doi.org/10.1002/we.2388, in press, 2019.
77
TAD
CT,AD =
1 2
ρU S
2 ∞ AD
Eq. 3.26
where ρ is the fluid density, U∞ is the free-stream velocity and SAD is the AD area. TAD is obtained by
forcing a uniform pressure drop across the AD, TAD = Δp x SAD. The pressure drop Δp is taken from
experiments and is given as an input parameter to the numerical simulations. The mean velocity
across the AD, UAD, is obtained by integrating the difference of the streamwise velocity component
across the AD surface Ux :
1
UAD = ∮ U ds
SAD SAD x
Eq. 3.27
Then, the power coefficient is
p0 UAD
Cp0 = = C
1 3
ρU S U∞ T,AD
2 ∞ AD
Eq. 3.28
For a multi-element duct–AD configuration, additional thrust forces are exerted by the duct and the
flap. Then, the total thrust force T is the vectoral sum of the AD thrust force TAD and the axial thrust
force exerted by the duct TD and the flap TF. It can be written as
T = TAD + TD + TF = TAD + TM
Eq. 3.29
The total thrust coefficient is then defined as
CT = CT,AD + CT,M
Eq. 3.30
where CT,M is the multi-element duct thrust coefficient. To highlight the relative contribution of the
multi-element duct thrust TM and the AD thrust TAD to the total thrust T, a dimensionless thrust factor
τ is introduced [Bontempo and Manna]28:
TM CT,M
τ= =
TAD CT,AD
Eq. 3.31
so that the total thrust coefficient can be written as
CT = (1 − τ) CT,AD
Eq. 3.32
Following [Bontempo and Manna], the normalized axial velocity at the AD for a ducted configuration
can be also expressed as a function of the thrust coefficient:
28 Bontempo,R. and Manna, M.: Effects of the duct thrust on the performance of ducted wind turbines, Energy, 99,
274–287, 2016.
78
UAD 1 + τ
= (1 + √1 − CT,AD )
U∞ 2
Eq. 3.33
Using Eq. 3.28 and Eq. 3.33, the power coefficient of the multielement duct–AD model considering
SAD as the reference area can be written as
1+τ
Cp = (1 + √1 − CT,AD ) CT,AD
2
Eq. 3.34
In Eq. 3.34, CP indicates the power coefficient of the multielement duct–AD model. The above
relation is also valid for a simple AD model setting τ= 0:
1
Cp0 = (1 + √1 − CT,AD ) CT,AD
2
Eq. 3.35
Eq. 3.34 and Eq. 3.35 can be used to evaluate the contribution of the multi-element duct through a
power augmentation parameter r:
Cp CT,M
r= 1+τ= 1+
C p0 CT,AD
Eq. 3.36
Eq. 3.36 states that r for a multi-element duct AD model is proportional to the ratio between the
multielement duct thrust coefficient CT;M and the AD thrust coefficient CT,AD. Thus, if τ > 0, then a
higher power coefficient can be obtained for a multi-element DWT in comparison to a HAWT with
the same rotor. The performance coefficients described above were evaluated by means of axial
momentum theory (AMT) for DWTs by [van Bussel]. However, the AMT cannot be used to estimate
the performance of the duct–AD model for a
prescribed CT,AD and a given duct geometry.
This problem can be solved using numerical
solutions based on panel and RANS
(Reynolds-averaged Navier–Stokes)
methods (Bontempo and Manna]; [Dighe et
al.].
3.3.3 Numerical Validation
For validating the numerical methods,
experimental data reported by [Igra] were
simulated. Igra’s experiments were
conducted in the subsonic wind tunnel of
Israel Aerospace Industries (formerly Israel
Aircraft Industry); this tunnel has a large
test section that measures 3.6m x 2.6 m.
Eight different geometries were
investigated experimentally, but only two Figure 3.13 Schematic Cross-Sectional Layout of the
geometries are used for the validation Three Dimensional Experimental Model used for the
study. The two geometries are a duct–AD Numerical Validation study.
model with CT;AD = 0.434 (Model B) and a
79
multi-element duct–AD model with CT;AD = 0.550 (Model C (ii) flap).

A schematic of the cross sections of the two geometries is shown in Figure 3.13. The longitudinal
cross section of the duct and of the flap is a NACA 4412 airfoil. The leading edges of both duct
geometries are identical. For Model C (ii) flap, the trailing edge of the duct is radially stretched,
resulting in a duct expansion ratio Ae/AAD= 1.84; this ratio is 1.71 for Model B. The flap chord
measures 35% of the duct chord length, c, and the deflection angle θ =30∘ with respect to the free-
stream direction. The experimental dataset consists of static pressure distribution at different axial
and radial positions, as well as forces generated by the duct and the flap surfaces. During the
experiments, the inflow velocity was set at U∞ =32m/s, corresponding to Re ≈ 4.5x105. Following
[Igra], the wall interference and blockage correction can be ignored.
For numerical validation, three numerical methods are considered: a two-dimensional panel method
based on Euler equations, a two-dimensional steady RANS method based on time-averaged RANS
equations and a two-dimensional unsteady RANS method based on time-filtered URANS equations.
A detailed description of the differences in the governing equations is beyond the scope of the current
discussion. The reader may refer to [Versteeg and Malalasekera]29. For both steady and unsteady
RANS methods, the k-ω SST (shear stress transport) model is used as a turbulence model.
Preliminary investigations showed good agreement with the experiments (Dighe et al.]. A numerical
stability analysis is performed in the context of time-filtered URANS simulations. The physical time
step corresponding to a Courant–Friedrichs–Lewy (CFL) number of 1 in the finest mesh resolution
level is 1.67x105 s. More details on the grid settings and the boundary conditions will be presented
in Sect 3.3.4.
In Figure 3.15, the power augmentation factor r is plotted as a function of the inflow yaw angle α.
CFD results, obtained using panel, steady RANS and unsteady RANS methods, are compared with the
experimental data. A very good agreement between the CFD simulations and the experimental
findings is found for Model B. On the other hand, the deviation between the CFD and the experimental
findings is larger for Model C (ii)+ flap, in particular for α ≠ 0. The discrepancies might be due to
three-dimensional effects not accounted for in the two-dimensional simulations. At non yawed
conditions (α = 0∘), however, the CFD results agree well with the experimental findings, wherein the
maximum deviation of the CFD results from the experimental data is 1.2% and 4.3% for Model B and
Model C (ii)+flap, respectively. The differences in the CFD results can be explained by looking at the
flow field. Figure 3.14 shows the contours of nondimensional axial velocity Ux/U∞ computed with
the three numerical methods for Model C (ii)+flap.
panel method, (b) RANS method and (c) URANS method for Model C (ii) flap
29Versteeg, H. K. and Malalasekera, W.: An introduction to computational fluid dynamics: the finite volume
method. Pearson education, 2007.
80
Figure 3.15 Effect of yawed inflow on the power augmentation factor: comparison between the
experiments (Igra, 1977), panel method and RANS method for different duct configurations.
The velocity contour from the panel method is plotted on the left; steady RANS is in the center, and
unsteady RANS is on the right. A clear difference in the flow field between the panel and the RANS
(steady and unsteady) methods could be identified from the contour plots. Neglecting viscosity, as in
the panel method solution, the flow remains attached over the suction side of the duct. As a result,
the magnitude of velocity on the suction side, and ultimately the value of r (Figure 3.15), is larger
for the panel method solution in comparison to the RANS (steady and unsteady) method solutions.
On the other hand, the flow fields obtained using steady and unsteady RANS methods are almost
identical. Both the steady and unsteady RANS solutions show flow separation along the inner walls
of the flap. Subtle differences appear in the flow separation region, where the velocity contour
patterns differ slightly in their spatial organization. In the URANS solution, the turbulent flow
structures, which evolve in time and space, are explicitly computed.
These flow structures are temporally averaged in the RANS solution. The net result of such different
formulations explains the difference in the value of r in Figure 3.15 calculated using steady and
unsteady RANS methods. Although URANS simulations increase the level of description of the
unsteady flow due to the multi-element duct–AD interaction, the computing cost incurred by going
from RANS to URANS does not justify the scope of the current study, in which the effects of distributed
AD loading, wake rotation, divergence and inflow yaw angle are totally ignored.
3.3.4 Numerical Approach
In this section, the numerical methods employed will be briefly described. For an in-depth
description, the reader can refer to [Dighe et al.].
3.3.4.1 Panel Method
A two-dimensional potential flow panel method has been used to compute the steady iso-entropic
incompressible flow field around the multi-element duct–AD model following the work of [de
81
Oliveira et al.]30. The governing flow equations are a simple form of the Euler equations. The AD is
represented by a pair of symmetric counter-rotating vortices.
The duct and the flap geometries are defined using a distribution of vortices located on the panels to
reproduce the desired cross-sectional shape. A uniform distribution of vorticity on the panels is
assigned by assuming the Kutta condition. The assumption of uniform vorticity distribution over the
panels represents a simplification of real physics and prevents flow separation on the multi-element
duct surface, even for larger pressure gradients. The duct and the flap surface discretization is based
on the constant spacing approach. The streamwise discretization in the near and far wake is
nonuniform, with initial panel length equal to 1.0%of duct chord length c, just behind the AD, and
increasing gradually in length as the wake expansion settles further downstream (see Figure 3.16).
Figure 3.16 Panel distribution along the duct surface and the wake region used for the inviscid panel
method calculations
The panel method is particularly appealing for routine design analysis due to its short execution time.
A typical converged panel method solution is obtained in roughly 0.05 h on a multicore workstation
desktop computer.
3.3.4.2 RANS Method
A commercial CFD solver, ANSYS Workbench®, has been used for a complete viscous solution of
steady incompressible flow around the multi-element duct–AD model. The governing flow equations
are the Reynolds-averaged Navier–Stokes (RANS) equations. The 2D computational domain is shown
in Figure 3.17, where the distances from the AD location to the domain inlet and outlet are 12c and
24c, respectively.
The computational grid consists of quadrilateral cells with a minimum y+ value of 1 on the duct walls.
Boundary conditions are a uniform velocity at the inlet, zero gauge static pressure at the outlet, and
no-slip walls for duct and flap surfaces. A symmetric boundary condition is applied along the center-
line axis, while a fan boundary condition is used for the AD. Experimental data from the wind tunnel
[Dighe et al.] in the form of pressure drop against velocity normal to the AD plane were extrapolated
to determine the input parameter for the fan boundary condition. RANS solutions are considerably
more reliable and accurate than the panel method solution but at the expense of computational cost.
30de Oliveira, G., Pereira, R. B., Ragni, D., Avallone, F., and van Bussel, G. J. W.: How does the presence of a body
affect the performance of an actuator disk?, Journal of Physics: Conference Series, Vol. 753, IOP Publishing, 2016.
82
A typical converged RANS solution

with approximately 0.1 M mesh
elements is obtained in roughly 0.5
h on a multicore workstation
desktop computer. RANS solutions
are sensitive to the discretization of
the computational domain. For the
present computations, a C grid
structured zonal approach is
chosen (Figure 3.17), which
proved advantageous in the case of
a curved boundary (duct and flap
leading edge). The C-shaped loop
terminates in the wake region. Grid
independence analysis has been
carried out using three grid sizes;
the refinement factor in each
direction is approximately 1.5. The
refinement factor is defined as the
Figure 3.17 Computational domain showing the boundary
rate at which the grid size increases
conditions employed. The length is indicated in terms of duct
far from the object. The multi- chord length c (representative, not to scale). The computational
element duct thrust force grid along the duct and flap leading edge is zoomed in to show
coefficient, CT;M, is taken as a the C-grid and boundary layer refinement.
reference for the convergence
analysis. The results of the grid
independence study are shown in Table 3.5. Convergence is reached for the medium grid, which is
then used in the rest of the paper.
3.3.5.1 Multi-Element Duct Geometry
In the following sections, the effects of flap
installation settings on the aerodynamic
performance of the multi-element duct–AD model
are described. The multi-element duct–AD
configuration investigated in the present work is Table 3.5 Grid statistics for the grid
shown in Figure 3.18. The longitudinal cross independence study of the reference case
section of the duct is a DonQi D5 airfoil; the profile is
chosen based on the duct shape parametrization
study conducted by the authors [Dighe et al.]. For the DonQi D5 duct an optimal CT;AD = 0.7 was
obtained. This value is employed throughout the present discussion.
A NACA 4412 longitudinal cross section, measuring 0.35c, is chosen for the flap following [Igra]. The
flap installation settings are the radial gap ζ and the deflection angle θ. The radial gap ζ , indicated as
a percentage of duct chord length c, is defined as the distance from the trailing edge of the duct to the
leading edge of the flap. A positive value of radial gap ( ζ > 0) indicates that the leading edge of the
flap is positioned below the trailing edge of the duct. A positive deflection angle (θ > 0) corresponds
to a downward flap deflection, whereby the angle is defined relative to the free-stream direction.
The axial gap between the trailing edge of the duct to the leading edge of the flap is zero based on the
findings of [Igra]. The numerical study is performed at a fixed Re of 4.5x105 as in the experiments. In
Sect. 3.3.5.2 and 3.3.5.3, the changes in the aerodynamic performance coefficients with respect to
the flap’s geometric orientation are quantified.
83
3.3.5.2 Duct Force Coefficient

Contours of the multi-element
duct force coefficient CT;M,
obtained from panel and RANS
methods, are shown in Figure
3.20 (a-b), where CT;M as a
function of the radial gap ζ and
the deflection angle θ are
reported. The figures show that
CT;M increases for larger ζ .
Conversely, CT;M decreases with
increasing ζ. The maximum CT;M
obtained from both the
numerical methods lies in the
same region, i.e., ζ ≈ 5% and θ ≈
10∘. The differences between
results obtained using the panel
and RANS methods are highly
contrasting for θ ≥ 60∘.
The differences can be explained
by looking at the flow field.
Contours of nondimensional Figure 3.18 A schematic cross section of the multi-element duct–
axial velocity Ux/U∞ from both AD model with the variable flap parameters used for the flap
methods are reported in Figure installation study.
3.19 a–f. Results from the panel
method are plotted on the left, while the ones from RANS are on the right. Contours for the no-flap
configuration are shown in Figure 3.19(a–b). Two flap settings, in order to explain the
aerodynamics behind Figure 3.20 (a-b), are shown: ζ = 5% and θ = 10∘ in Figure 3.19 (c-d) as well
as ζ= 5% and θ = 70 in Figure 3.19 (e-f). Contour plots show a higher velocity at the rotor plane for
the configuration with flap in comparison with the no flap configuration. This is due to the additional
aerodynamic thrust force generated by the flap. The presence of a radial gap between the duct and
the flap accelerates the flow over the flap. This reduces the pressure recovery demands on the multi-
element duct, thereby reducing flow separation. Obviously, flow separation is seen for RANS
contours only. The overall integral contribution of the viscous forces increases the CT;M magnitude in
the RANS solutions relative to the panel solutions, a trend that can be clearly observed by comparing
Figure 3.20 (a - b). For the flap configuration with ζ = 5% and θ = 70∘ (Figure 3.19 f), the flow
over the flap’s inner walls separates completely. The separation along the inner walls of the multi-
element duct reduces the CT;M, which rapidly becomes large and negative at higher flap deflection
angles as seen in Figure 3.20 (b). For panel solutions, however, the drop in the CT;M magnitude for
higher flap deflection angles is gradual (see Figure 3.20 (a)) because viscous effects are neglected
(see Figure 3.19 e).
84
Figure 3.19 Velocity contours colored with normalized free-stream velocity obtained using the (a) panel
method with no flap, (b) RANS method with no flap, (c) panel method with ζ = 5% and θ=10 ∘, (d) RANS
method with ζ = 5% and θ=10∘, (e) panel method with ζ = 5% and θ = 70∘, and (f) RANS method with ζ=
5% and θ = 70∘
85
(a) Panel method (b) RANS method
Figure 3.20 Comparing the Effect of the Variable Radial Gap and Deflection Angle of
the flap on the duct thrust force coefficient using Panel and RANS Methods
3.3.5.3 Power Augmentation

Figure 3.21 (a-b) represent contours of power augmentation factor r using the panel and RANS
solutions, respectively, as a function of radial gap ζ and deflection angle θ. Recall from Figure 3.21
(a) that an r gain for a multi-element duct–AD model can be attained by increasing the CT;M
magnitude for a constant CT;AD. Evidence of this is provided in Figure 3.21 (a-b), which exhibit the
r maximum in the same region of CT;M maximum, as in Figure 3.20 (a-b). Then, the maximum power
augmentation factor rmax = 1.25 and 1.38, obtained for panel and RANS calculations, respectively,
corresponds to ζ ≈ 5% and θ ≈10∘.
(a) Panel method (b) RANS method
Figure 3.21 Comparing the Effect of the Variable Radial Gap and Deflection Angle of the flap on the
power augmentation factor using Panel and RANS Methods
A comparison of rmax obtained from CFD simulations with the one-dimensional axial momentum
theory[van Bussel is carried out. The analysis of AMT is not reported herein for the sake of brevity; a
86
description of the AMT theory applied is reported in Appendix of [Dighe et al.]31. The velocity and
pressure values within a DWT, using the AMT approach, are determined by applying the relations of
duct expansion ratio τ and back-pressure velocity ratio. With the AMT approach, the maximum
power augmentation factor of rmax = 1.50 for ζ ≈ 5% and θ ≈ 10∘ is obtained. In comparison to the CFD
methods, the AMT approach does not take into account the effects of multi-element duct geometry
and the nonlinear mutual interaction between the multi-element duct and the AD. As a result, the
values of CP and the power augmentation factor r, determined using the AMT approach, are
overestimated in comparison to the CFD methods shown in the above discussion.
3.3.6 Conclusions
In this work, the aerodynamic performance of a multielement DWT is studied using a numerical
approach. To this aim, 2D numerical calculations using the panel method and the RANS method are
employed. A simplified AD model is used for simulating the rotor. Based on the existing studies
conducted by the authors, the multi-element duct geometry consists of a DonQi D5 airfoil and a NACA
4415 airfoil for the duct and the flap cross sections, respectively. To validate the numerical methods,
the present simulations are compared with similar experimental data. In order to deepen the design
principles of multi-element ducts, the effects of radial gap ζ and the flap deflection angle θ on the
global performance of DWTs are investigated. Clear trends of the multi-element duct thrust force
coefficients CT;M and the power augmentation factor r, are observed across a range of multi-element
duct configurations. An increase in the flap deflection angle θ results in a decrease in CT;M, whereas
an increase in the radial gap ζ shows an increase in CT;M. The analysis of flow field shows that flow
separation in the multi-element duct inner walls increases for higher values of θ.
This phenomenon determines the reduction in CT;M and ultimately the augmentation factor r. As
expected, the RANS method is more suitable for representing solutions for highly deflected flap
configurations. The viscous effects become stronger at higher flap deflection angles, and the panel
method is inherently incapable of taking this into account. Regarding prediction of the near-optimal
multi-element duct configuration, both the numerical methods show good agreement. Moreover, in
comparison to the AMT approach, the CFD method fully takes into account the mutual interactions
between the multi-element duct and the AD.
31Vinit V. Dighe, Francesco Avallone, Ozer Igra, and Gerard van Bussel,, “Multi-element ducts for ducted wind
turbines: a numerical study”, Wind Eng. Sci., 4, 439–449, 2019.
87
4 Case Studies Involving Vertical Axis Wind Turbines (VAWTs)

4.1 Case Study 1 - Vertical Axis Wind Turbine (VAWTs) - An Efficient Way To
Harness Power : A Review
Authors : Sanshodhan Shende, Amol Andhare, Falgun Bhelkar, Naval Rajas, and Vaibhav Thakur
Appeared in : International Journal of Mechanical and Production, Engineering Research and
Development (IJMPERD), ISSN (P): 2249-6890; ISSN (E): 2249-8001, Vol. 9, Special Issue, Jun 2019.
Source : www.tjprc.org
Vertical axis wind turbine (VAWTs) has proved itself to be more efficient and environmentally
friendly for small as well as large power requirements. VAWTs are more popularly used irrespective
of wind flow conditions and are suitable for its use in wind farms. VAWTs are more economical as
compared to horizontal axis wind turbine as the system components used are less for establishing
control over pitch, yaw and vibrations generated. VAWT has proven solutions and had overcome
drawbacks and limitations over a period of time, as presented by researchers. The researchers had
investigated upon areas like power output, power coefficient, tip speed ratio, solidity, thrust and
optimization of blade design. This paper analyses various developments and highlights the potential
prospects that VAWTs have to offer and implementing it pragmatically through the performed
experiments, simulations run and the theoretical analysis which governs the power performance of
the VAWTs. [Shende et al.]32.
4.1.1 Design Optimizations
Optimization of the design comprises mainly the development of the blade on which end plates;
quarter plates are mounted to increase the performance. Serrations on the leading edge are also
commonly used to improve the performance. This is accompanied by the momentous research in the
structural parameters such as the supporting arms, mast, etc. All the pertinent research and their
results are summarized in Table 4.1.
4.1.2 Operation Analysis and Analysis Techniques and Optimizations
Structural, Mathematical, Computational analysis of the VAWTs are done and effects of physical
quantities such as the solidity, pitch angle; other miscellaneous factors such as rain on the
performance are studied. Optimization of the simulation techniques and factors on which they can
give actual and useful results are studied and presented in
Table 4.2.
4.1.3 Novel Design Attachments and Configuration
Novel attachments such as the deflectors, passive flow control shields for decreasing the stall and all
the different configurations which increase the efficiency of the turbine are studied and results are
summarized in Table 4.3 and Table 4.4.
32 Sanshodhan Shende, Amol Andhare, Falgun Bhelkar, Naval Rajas, and Vaibhav Thakur, “Vertical Axis Wind
Turbine An Efficient Way To Harness Power: A Review”, International Journal of Mechanical and Production,
Engineering Research and Development (IJ MPERD), ISSN (P): 2249-6890; ISSN (E): 2249-8001, Vol. 9, Special
Issue, Jun 2019.
88
Table 4.1 Design Optimization Summary

89
Table 4.2 Summary of Operation Analysis and Analysis Techniques and Optimizations
90
Table 4.3 Summary of Novel Design Attachments and Configurations

91
Table 4.4 Summary of Novel Design Attachments and Configurations (Continued)
4.1.4 Conclusions
A comprehensive review of the latest research in the field of vertical axis wind turbine was
conducted. Only the most consequential research is compiled and categorized into three fields:
optimizations made to the design of the complete system, analysis of the various configurations
under disparate operating conditions (with novel simulation methods) and newfound accretions to
design to improve efficiency. All the significant findings, results are presented in tabular form which
makes it easier to refer. Also, it was found that Darrieus VAWT provides more scope of research based
on power output and reliability. This study as a whole act as a guideline for further research and
concludes on the basis of the studies that VAWTs are a great energy efficient way of harnessing power
from wind.
4.1.5 References
1. Subramanian, Abhishek, S. Arun Yogesh, Hrishikesh Sivanandan, Abhijit Giri, Madhavan
Vasudevan, Vivek Mugundhan, and Ratna Kishore Velamati."Effect of airfoil and solidity on
performance of small scale vertical axis wind turbine using three dimensional CFD model." Energy 133
(2017): 179-190.
2. Didane, Djamal Hissein, Nurhayati Rosly, Mohd Fadhli Zulkafli, and Syariful Syafiq Shamsudin.
"Numerical investigation of a novel contra-rotating vertical axis wind turbine." Sustainable Energy
Technologies and Assessments31 (2019): 43-53.
3. Verkinderen, E., and B. Imam. "A simplified dynamic model for mast design of H-Darrieus vertical
axis wind turbines (VAWTs)." Engineering Structures 100 (2015): 564-576.
4. Posa, Antonio. "Wake characterization of coupled configurations of vertical axis wind turbines using
Large Eddy Simulation." International Journal of Heat and Fluid Flow 75 (2019): 27-43.
5. Ma, Ning, Hang Lei, Zhaolong Han, Dai Zhou, Yan Bao, Kai Zhang, Lei Zhou, and Caiyong Chen.
"Airfoil optimization to improve power performance of a high-solidity vertical axis wind turbine at a
moderate tip speed ratio." Energy 150 (2018): 236-252.
6. Wang, Zhenyu, Yuchen Wang, and Mei Zhuang. "Improvement of the aerodynamic performance of
vertical axis wind turbines with leading-edge serrations and helical blades using CFD and Taguchi
method." Energy Conversion and Management 177 (2018): 107-121.
92
7. Lin, Jinghua, Leo KK Leung, You-Lin Xu, Sheng Zhan, and Songye Zhu. "Field measurement, model
updating, and response prediction of a large-scale straight-bladed vertical axis wind turbine structure."
Measurement 130 (2018): 57-70.
8. Premkumar, T. Micha, Seralathan Sivamani, E. Kirthees, V. Hariram, and T. Mohan. "Data set on the
experimental investigations of a helical Savonius style VAWT with and without end plates." Data in brief
19 (2018): 1925-1932.
9. Chaisiriroj, Pongchalat, Nuttapat Tinnachote, Saowapak Usajantragul, and Thananchai
Leephakpreeda. "Experimental Performance Investigation of Optimal Vertical Axis Wind Turbines
under Actual Wind Conditions in Thailand." Energy Procedia 138 (2017): 651-656.
10. Ostos, Iván, Iván Ruiz, Maja Gajic, William Gómez, Adriana Bonilla, and Carlos Collazos. "A
modified novel blade configuration proposal for a more efficient VAWT using CFD tools." Energy
Conversion and Management 180 (2019): 733-746.
11. Govind, Bala. "Increasing the operational capability of a horizontal axis wind turbine by its
integration with a vertical axis wind turbine." Applied Energy 199 (2017): 479-494.
12. Kim, Daegyoum, and Morteza Gharib. "Efficiency improvement of straight-bladed vertical-axis
wind turbines with an upstream deflector." Journal of Wind Engineering and Industrial Aerodynamics
115 (2013): 48-52.
13. Müller, Gerald, Mert Chavushoglu, Mark Kerri, and Toru Tsuzaki. "A resistance type vertical axis
wind turbine for building integration." Renewable Energy 111 (2017): 803-814.
14. Li, Jinyi, Yang Cao, Guoqing Wu, Zifan Miao, and Jiawei Qi. "Aerodynamic stability of airfoils in lift-
type vertical axis wind turbine in steady solver." Renewable energy 111 (2017): 676-687.
15. Wu, Zhenlong, Yihua Cao, Shuai Nie, and Yue Yang. "Effects of rain on vertical axis wind turbine
performance." Journal of Wind Engineering and Industrial Aerodynamics 170 (2017): 128-140.
16. Wang, Wei-Cheng, Wen Tong Chong, and Tien-Hsin Chao. "Performance analysis of a cross-axis
wind turbine from wind tunnel experiments." Journal of Wind Engineering and Industrial
Aerodynamics 174 (2018): 312-329.
17. Wang, Zhenyu, and Mei Zhuang. "Leading-edge serrations for performance improvement on a
vertical-axis wind turbine at low tip-speed-ratios." Applied Energy 208 (2017): 1184-1197.
18. Nobile, Rosario, Maria Vahdati, Janet F. Barlow, and Anthony Mewburn-Crook. "Unsteady flow
simulation of a vertical axis augmented wind turbine: A two-dimensional study." Journal of Wind
Engineering and Industrial Aerodynamics 125 (2014):168-179.
19. Hand, Brian, and Andrew Cashman. "Conceptual design of a large-scale floating offshore vertical
axis wind turbine." Energy Procedia 142 (2017): 83-88.
20. Sunny, Kalakanda Alfred, and Nallapaneni Manoj Kumar. "Vertical axis wind turbine: Aerodynamic
modelling and its testing in wind tunnel." Procedia Computer Science 93 (2016): 1017-1023.
21. Tian, Wenlong, Zhaoyong Mao, and Hao Ding. "Numerical study of a passive-pitch shield for the
efficiency improvement of vertical axis wind turbines." Energy Conversion and Management (2019).
22. Rezaeiha, Abdolrahim, Ivo Kalkman, and Bert Blocken. "Effect of pitch angle on power performance
and aerodynamics of a vertical axis wind turbine." Applied energy 197 (2017): 132-150.
93
4.2 Case Study 2 - Airfoil Optimization to Improve Power Performance of a High-

Solidity Vertical Axis Wind Turbine (VAWT) at a Moderate Tip Speed Ratio
(TSR)
Authors : [Ning Ma, Hang Le, Zhaolong Han, Dai Zhou, Yan Bao , Kai Zhang, Lei Zhou, Caiyong Chen] 33
Citation : Ning Ma, Hang Lei, Zhaolong Han, Dai Zhou, Yan Bao, Kai Zhang, Lei Zhou, Caiyong Chen,
Airfoil optimization to improve power performance of a high-solidity vertical axis wind turbine at a
moderate tip speed ratio ,Energy, Volume 150, 2018, Pages 236-252, ISSN 0360-5442,
https://doi.org/10.1016/j.energy.2018.02.115.
The main purpose of the present study is to develop an automatic airfoil profile optimization system
to improve the power performance of a VAWT as studied by [Ma et al]34. A three-bladed high-solidity
VAWT is adopted as the research object with its chord length, blade span and rotor diameter being
0.2 m, 0.8m and 0.8 m, respectively. The optimization is conducted at a moderate tip speed ratio
(TSR) with a value of 1.0 and the method of coupled CFD simulations with genetic algorithms is
employed. The following points make this paper different from previous studies:
➢ Introducing Multi-Island Genetic Algorithm to optimize airfoils for VAWTs;
➢ Investigating the airfoil as part of the VAWT rather than as a single isolated body with 3D
simulations.
The results show that the power coefficient of the VAWT, equipped with the optimized blades,
improves at all TSRs from 0.4 to 1.5. Moreover, it was shown that and the maximum growth rate
occurs at TSR = 0.9, with a value of 26.82%.
4.2.1 Background and Introduction
According to the axis of rotation, wind turbines can be classified into Horizontal Axis Wind Turbine
(HAWT) and Vertical Axis Wind Turbine (VAWT). In the past decades, attention of researchers was
mainly focused on HAWTs due to their high efficiency in steady winds, which makes HAWTs widely
recognized and applied. But, there has been growing interests in VAWTs which are regarded more
suitable for urban and offshore applications compared to HAWTs. In urban areas where the wind
environment is very complex, small-scale VAWTs are considered to have better performance than
their HAWT counterparts because of their ability to work in varied wind directions, higher
adaptability to turbulent or skewed flows and reduced level of noise35-36. For offshore applications,
due to the lower center of gravity, easier manufacturing, reduced installation & maintenance cost and
higher scalability to large sizes, VAWTs are regarded to be more promising than HAWTs.
4.2.2 Types of VAWTs and Their Characteristics
Common VAWTs mainly include the following basic types: H-Rotor, Darrieus, and Savonius types.
H-rotor type, derived from the traditional Darrieus type, is equipped with straight blades rather than
curved blades. Straight blades are easier to be manufactured and installed, resulting in lower costs .
Therefore, manufacturers of VAWTs in developed countries try to bring H-rotor type to commercial
maturity. Besides, a H-Rotor type VAWT is driven by lift force and can regulate itself well in various
wind velocities . It's also chosen to be the research object in this paper. The distribution of the chord
33 Ning Ma, Hang Lei , Zhaolong Han, Dai Zhou , Yan Bao , Kai Zhang , Lei Zhou, Caiyong Chen, “Airfoil optimization
to improve power performance of a high-solidity vertical axis wind turbine at a moderate tip speed ratio”, Energy
150 (2018) 236-252.
34 Same as above.
35 Zamani M, Nazari S, Moshizi SA, Maghrebi MJ. Three dimensional simulation of J-shaped Darrieus vertical axis
wind turbine. Energy 2016.

36 Balduzzi F, Bianchini A, Carnevale EA, Ferrari L, Magnani S. Feasibility analysis of a Darrieus vertical-axis wind
turbine installation in the rooftop of a building. Applied Energy.

94
length and the airfoil profile along the blade span remains unchanged for the H-Rotor type VAWT.
Once the airfoil profile, chord length and the blade span are determined, the 3D shape of the blade is
also determined. This characteristic of the H-rotor type VAWT makes the design and manufacturing
of it much easier than HAWTs. Although VAWTs have attracted much attention, the power coefficient
of them is relatively low compared to that of HAWTs, which is a bottleneck restricting the utilization
of VAWTs. There are different ways of improving the power performance of VAWTs, including:
➢ Changing the geometrical or operational parameters related to aerodynamics;
➢ Developing specific airfoils for VAWTs.
4.2.3 Literature Survey and Procedure
Previous researchers have conducted many studies on the impacts of geometrical and operational
parameters. [Rezaeiha et al.]37 investigated the effects of pitch angle on the loads, angle of attack,
vorticity distribution and flow separation with (CFD) method. Their research suggests that dynamic
pitching could be an effective approach for power performance improvement. [Subramanian et al.]38
used FLUENT to study the behavior of Darrieus VAWTs with different solidity and found that
turbines with lower solidity have better performance at high tip speed ratios (TSRs). In the work of
[Bachant and Wosnik]39, experiments were conducted to study effects of Reynolds number on energy
conversion of a high-solidity VAWT. The power coefficient improves rapidly with growing Reynolds
number (Re) when Re is smaller than a threshold value and becomes Rein dependent after that
threshold.
[Wang and Zhuang]40 adopted serration design on the leading edge of blades, which demonstrates
stronger capability of wind power extraction at low TSRs compared to the model without serrations.
Another effective solution to improve the power performance of VAWTs is to develop specific airfoil
profiles for VAWTs as they can largely affect the capture ratio of wind power and the flow field
around the blade41. Capture ratio means the ratio of the wind power obtained by the wind turbine
to the total amount of power carried by the air flowing through the turbine. Plenty of airfoil families
have been created and applied for aircrafts and HAWTs during the past decades. Although those
airfoil families are not customized for VAWTs, they can be used as original templates and then
optimized for VAWTs, which is much easier than designing brand new airfoil families for VAWTs. A
complete profile optimization process generally contains parametric modelling, calculation and
optimization parts. The widely used traditional methods for calculating the aerodynamics of VAWTs
include the Blade Element Momentum (BEM) theory, Vortex method and Cascade model. Although
assumptions and simplifications involved in these methods once brought much convenience, they
also result in intrinsic limitations which may cause large deviations between experimental outcomes
and the calculated ones. CFD method, supported by computer technologies and favored by
researchers nowadays, has the following advantages compared to traditional methods. On one hand,
it is a more accurate approach and the generated results are usually in good agreement with the
experimental data. On the other, it can reveal the phenomenon of dynamic stall better and even
37 Rezaeiha A, Kalkman I, Blocken B. Effect of pitch angle on power performance and aerodynamics of a vertical
axis wind turbine. Applied Energy 2017.
38 Subramanian A, Yogesh SA, Sivanandan H, et al. Effect of airfoil and solidity on performance of small scale
vertical axis wind turbine using three dimensional CFD model. Energy 2017.
39 Bachant P, Wosnik M. Effects of Reynolds number on the energy conversion and near-wake dynamics of a high
solidity Vertical-Axis cross-flow turbine. Energies 2016.

40 Wang Z, Zhuang M. Leading-edge serrations for performance improvement on a vertical-axis wind turbine at
low tip-speed-ratios. Applied Energy 2017.

41 Wang Q, Chen J, Pang X, Li S, Guo X. A new direct design method for the medium thickness wind turbine airfoil.
J Fluid Structure 2013.

95
visualize detailed information of the flow field including streamlines and vorticity evolution42. When
it comes to airfoil optimization, there has been some research using different algorithms. [Ribeiro et
al]43 utilized the Non-dominated Sorting GA-II (NSGA-II) in the airfoil optimization work for wind
turbines. CFD method combined with artificial neural networks could save about 50 % computational
time when compared with conventional CFD methods. To develop an automatic airfoil profile
optimization process for improving the power performance of a high solidity VAWT at a moderate
TSR. Specific objectives are as follows:
➢ Establishing an appropriate numerical model of a three bladed VAWT and proving its validity;
➢ Building an optimization system based on 3D CFD simulation to improve the power
coefficient of the VAWT.
To achieve these objectives, the present study involves the work of numerical modelling, verification
& validation of CFD model and the development of an optimization system. Findings are shown with
indicators like power coefficient, pressure coefficient and vorticity distribution.
4.2.4 Numerical Model and Selection of CFD Model
3D CFD model is adopted in the present study as it can generate more accurate results than 2D model
when calculating the aerodynamics of VAWTs. 2D simulation cannot take into consideration the
energy loss induced by the vortex near the blade tip which could be simulated in the 3D simulation.
This drawback of 2D simulation will cause the over-prediction of power performance of VAWTs,
which has been verified by the work of [Siddiqui et al.]44. His team once conducted a study comparing
the results of 2D and 3D simulation with a 2.5m diameter VAWT and found that 2D approximation of
the VAWT could lead to an over-prediction of the power performance by 32% in their research. As
the present study requires tens of simulations during the optimization process, RANS method is
selected because it takes less computational cost than other three methods and could still produce
satisfying results. Of all the RANS turbulence models, k-ω model has good performance in dealing
with inner boundary layers while k-ε model does well in free stream zones. Shear Stress Transport
(SST) k-ω model incorporates the mentioned features of k-ω model and k-ε model. Another point
that needs clarifying is that the blades are regarded as rigid in the simulation. But the absence of aero
elasticity has very small effect on the precision of generated power coefficients for the case used in
our paper, which is a 0.8m-diameter small-scale wind turbine operating under low wind speeds. The
effect of the aero elasticity is usually obvious for large, multi-megawatt and flexible wind turbines,
especially under high wind speeds. Based on the considerations above, we chose to perform the
simulation in the absence of aero elasticity. This could also effectively control the complexity of our
work.
4.2.5 Parametric Modelling
As the airfoils are studied as components of a VAWT, it's necessary to define basic parameters of the
VAWT. In the following research, only the airfoil profile will be modified while the other parameters
of the VAWT will remain the same. The schematic diagram of the model adopted in the present study
is shown in Figure 4.1. The chord length of the airfoil c, the rotor diameter D, the blade span h and
the number of blades N are all constants in this paper and their values are shown in Table 4.5. In
the simulations, support arms and the center shaft are ignored because of their small effects. This is
also validated by others in which a maximum power loss of 5.5% was observed when shaft-to-turbine
diameter ratio equals 16% compared to the case without shaft. The similar model and corresponding
42 Lei H, Zhou D, Lu J, Chen C, Han Z, Bao Y. The impact of pitch motion of a platform on the aerodynamic
performance of a floating vertical axis wind turbine. Energy 2017.
43 Ribeiro AFP, Awruch AM, Gomes HM. An airfoil optimization technique for wind turbines. 2012.
44 Siddiqui MS, Durrani N, Akhtar I. Quantification of the effects of geometric approximations on the performance
of a vertical axis wind turbine. Renew Energy 2015.

96
Figure 4.1 Schematic diagram of an H-rotor type VAWT: (a) 3D view; (b) top view - [Courtesy of Ma
et al)
data in the study of [Elkhoury et al.]45 will be used for comparison and verification.
The parameterization of the airfoil profile is the first step of optimization as it can use variables
modified by the algorithm to generate new geometries. Five rules that an effective parameterization
method should satisfy have
been mentioned in46. In this Parameter Value
paper, Eq. 4.1 is used to Chord length c 220 mm
represent the upper or
Diameter of the rotor D 880 mm
lower profile of the airfoil.
Of the two terms, the former Blade span h 880 mm
one px (c - b) can be seen as
a b Number of blades N 3
the mean camber line (MCL) Blade aspect ratio h/c 4
while the latter one qxe(c – Solidity s = Nc/D 0.75
b)f can be seen as the Chord-based Re at inlet 1.02 x 105
thickness distribution. Such
Freestream turbulence intensity 1%
an equation has sufficient
flexibility to cover a large Rotational speed ω 8 - 30 rad/s
search space which allows Tip speed ratio λ 0.4 - 1.5
the appearance of non-
routine profiles. Besides, Table 4.5 Geometrical and Operational Parameters of the VAWT –
this equation is highly [Courtesy of Ma et al. )
adaptive and suitable for
both symmetric and cambered airfoil profiles.
45 Elkhoury M, Kiwata T, Aoun E. Experimental and numerical investigation of a

three-dimensional vertical-axis
wind turbine with variable-pitch. 2015..
46 Giannakoglou KC. Design of optimal aerodynamic shapes using stochastic optimization methods and
computational intelligence. Progress Aero Science2002.

97
p x a1 (c − x)b1 + q1 x e1 (c − x)f1 upper surface

y = { 1 a2
p2 x (c − x)b2 + q2 x e2 (c − x)f2 lower surface
Eq. 4.1
Here a1, a2, b1, b2, e1, e2, f1, f2, p1, p2, q1, q2 are 12 design variables and c is the chord length (c =
0.2 m). The value of x is between 0 and 0.2 m. The variables p and q have global influence on the
profile from the leading edge to the trailing edge. The variables a and e tend to have more impacts on
Table 4.6 Ranges of variables and specific values of variables for NACA0018 airfoil - [Courtesy of
Ma et al.)
the front end while b and f have more effects on the rear end of the airfoil. Curve Fitting Toolbox in
MATLAB® can be used to determine the values of design variables for certain airfoil profile.
NACA0018 airfoil is selected to be optimized in the present study and corresponding values of
variables are listed in Table 4.6.
The value ranges of design variables during the optimization are also set according to the initial
values of variables for NACA0018. For example, the initial value of a1 for NACA0018 is 0.5112, then
the range of a1 is set
to be 0.5112 ± s. The
parameter s is
related to the scope
of the search space of
the optimization
system. It would be
well if 0.2 is assigned
to s and the range of
a1 would be
approximately 0.31 -
0.71 in this case. ‘s =
0.2’ is a moderate
value which can not
only ensure a
sufficient scope of
Figure 4.2 Plan Views of the Computational Domain and Boundary Conditions
search space but also - [Courtesy of Ma et al.)
save much
computational resources. It also needs to be noted that all the variables are greater than zero except
q2.
98
4.2.6 Computational Domain & Meshing

The computational domain can be divided into an outer zone and a rotating inner zone. All the sizes
with respect to rotor diameter D and blade span h as well as the boundary conditions are presented
in Figure 4.2. The determination of these relative sizes is based on the work of previous
researchers. Other studies also provide useful guidelines on settings of domain size despite that their
advice is mainly for 2D/2.5D simulations. In the present study, the center of the inner zone is situated
at 5 rotor diameters downstream of the inlet boundary and 15 rotor diameters upstream of the outlet
boundary, which can allow for sufficient space for the wake generation. The width of the outer zone
is 10 times the rotor diameter while its height is 3 times the blade span. For the inner zone, its
diameter is 2 times the rotor diameter and its height is 1.5 times the blade span. The mesh setup is
completed in STAR-CCM+ which
contains components of meshing,
calculating and post-processing.
The rotational speed of the inner
zone could be adjusted in order
to investigate the power
performance at different Tip
Speed Ratios (TSRs). Figure 4.3
shows the grids of the
computational domain.
Unstructured trimmed cells with
varied sizes are adopted in
different zones. Prismatic
boundary layer cells are used on
the surface of the blades. The
thickness of the boundary layer is
0.01m with a growth ratio of
1.25. The ‘all-y+ wall treatment’
is selected in STARCCM which is a
hybrid treatment and adaptable
to both coarse and fine grids. The
maximum value of y+ on the
blade surfaces is 0.7 which
indicates acceptable resolution of
boundary layer cells. There are
roughly 50 cells along the chord
length and 180 cells along the
span wise direction.
Figure 4.3 Mesh Setup (a) Topology; & BCs. (b) Grids at the Mid-
A uniform velocity profile (V∞ = 8 Height Plane - [Courtesy of Ma et al.)
m/s) is set as the inlet condition
and the outlet boundary is
considered as pressure outlet
with zero relative pressure. The turbulence intensity is set as 1% and the value of turbulent length
scale is 0.014mat the inlet. The slip wall condition is assigned to the remaining boundaries of the
outer zone while the non-slip wall condition is assigned to blade surfaces. Two sub-domains are
linked with interface boundary condition at their contact surfaces. Besides, the grid size of two
subdomains near the interface should be consistent enough to ensure the effective and smooth data
exchange. The settings mentioned above are used in the auto-meshing for each model generated
during the optimization process. In this way, the number of grids and the grid size around the blades
are consistent, which ensures the reliability of the comparative analysis later.
99
4.2.7 Solver Setting

The Mach number in the field of wind turbines is below 0.3, which means the flows can be treated as
incompressible. Unsteady implicit segregated flow model is adopted and the second order difference
formula is employed to solve the discrete differential equations. The Semi-Implicit Method for
Pressure Linked Equations (SIMPLE) is chosen to deal with pressure-velocity coupling and Shear
Stress Transport (SST) k-ω turbulence model. This turbulence model is used for its flexibility in the
treatment of wall boundaries and high reliable results. Solutions are deemed as convergent when
the residuals are equal to or lower than 10-4. For unsteady calculation, settings about the time step
and total physical time need careful consideration. Detailed information about the time step can be
seen in ‘Sensitivity test of time step’ and only the total physical time is discussed here. The peak
values of the torque are almost the same after 4e5 rotations of the rotor, indicating that the
aerodynamics and the flow field can be regarded as stable after 5 rotations. That means choosing 6
rotation periods as the total physical time can not only ensure the accuracy and reliability of results,
but also save computational cost.
4.2.8 Verification and Validation of CFD Model
4.2.8.1 Grid Independence Study
The verification process starts with the grid sensitivity test as the grid quality usually has profound
influence on the simulation results. Here three sets of grids are generated: a coarse set, a medium set
and a fine set. The grid refinement is not uniformly performed in the entire domain. As the wind
environment in the inner zone is more complex than that in the outer zone, a smaller refinement
factor for the inner zone is adopted compared to that for the outer zone. The refinement factor for
the inner zone changes from 0.8 (coarse to medium set) to 0.85 (medium to fine set) while that for
the outer zone remains 0.9. The power coefficient CP at the TSR = 1.0 is selected as the evaluation
criterion. Here power coefficient CP is defined as:
Qω
Cp = 3
0.5ρDhV∞
Eq. 4.2
where Q, u, r D, h and V∞
are torque of the VAWT, Relative change of
Mesh Predicted
angular velocity of the Mesh Set CP with respect to
Number CP
rotor, air density, rotor the fine Mesh set
diameter, blade span and Coarse 1,529,457 0.136 -15.00%
freestream velocity, Medium 2,616,348 0.158 -1.25%
respectively. The number Fine 3,835,735 0.160 l
of grids and predicted
value of CP for each case Table 4.7 Sensitivity test of grid size at TSR = 1.0 - [Courtesy of Ma et
are shown in Table 4.7. al.)
The relative change of CP
from the medium grid set to the fine grid set is 1.25%, which indicates that further refinement of
grids will only lead to negligible change in solution. Considering the accuracy of results and
computational cost, the medium grid set is chosen for the subsequent simulations in the present
study.
4.2.8.2 Sensitivity W.R.T. Time Step
For the unsteady simulation, determining a suitable time step (∆t) is important as it influences the
precision and computational cost. In the present study, three values of the time step are chosen
which are corresponding to the time that the VAWT rotates by 2p/ 90, 2p/180 and 2p/360,
respectively. The test is conducted at TSR = 1.0 which means the VAWT rotates with a rotational
100
speed of 20 rad/s. In such a condition, ∆t1 = T/90 = 0.003491 s, ∆t2 = T/ 180 =0.001745 s and ∆t3 =
T/360 = 0.000873 s. Figure 4.4 shows the time-varying power coefficients of the VAWT during one
Figure 4.4 Time-varying power coefficients in a rotational period when adopting different time steps
- [Courtesy of Ma et al.)
rotational period in three cases with different time steps. It is observed that Curve2 (∆t2) and Curve3
(∆t3) are almost identical while Curve1 (∆t1) is in a lower position. In fact, the relative change of
average CP during one revolution when adopting ∆t2 and ∆t3 is less than 1%, indicating that the
change in solution will be negligible if smaller time step is employed. Therefore, ∆t2 is selected to be
the suitable time step in
the present study for the
trade-off between
precision and
computational cost.
4.2.9 Comparison with
Experimental
Results
This section is aimed to
validate the accuracy and
reliability of the model and
simulation methods by
comparing the simulation
results with experimental
data at different TSRs.
Here the power coefficient
is selected for comparison
Figure 4.5 Comparison of average power coefficients between the
as it is a significant index
experiment and simulation (NACA0018, freestream velocity V∞ = 8 m/s)
to reflect the utilization as well as relative errors of them - [Courtesy of Ma et al.)
ratio of wind power.
101
Although using this single indicator cannot fully prove the advantage of the present model, it actually
grabs the core problem and can prove the suitability of the current CFD model to a large degree. The
experimental data of the three-bladed VAWT is extracted from the work of [Elkhoury et al.]47 from
which the model used in this paper originates. Their experiments were conducted in an open-circuit
type wind tunnel and three freestream velocity values of 6 m/s, 8 m/s and 10 m/s were employed
for a fixed-pitch NACA0018 airfoil. The present study only uses experimental data related to the
freestream velocity value of 8 m/s.
Figure 4.5 illustrates the comparison results and the relative error between numerical and
experimental data. The largest error occurs at TSR = 0.5 with the value of 17.15 % and the second
largest error appears with the value of 14.99 % at TSR = 0.6. The variation range and the absolute
value of angle of attack are both large when TSR is very low, which means the blades will get into the
state of deep dynamic stall easily. However, SST k-u model is a bit weak in predicting the
phenomenon of deep dynamic stall at high angles of attack, which leads to the relatively high error
at low TSRs48,49. Besides, as the support structures of the VAWT and the effect of aero elasticity are
ignored in the numerical simulation, certain error should be expected. at low TSRs50,51. Besides, as
the support structures of the VAWT and the effect of aero elasticity are ignored in the numerical
simulation, certain error should be expected. Overall, the numerical results are consistent with
experimental data as the relative errors are under 10% in most cases. Such a verification outcome
could be accepted and the subsequent simulations are regarded as valid and reasonable.
4.2.10 Optimization System
When it comes to optimization approaches, they are classified into Gradient Search Methods (GSMs)
and heuristic algorithms. GSMs have been utilized to optimize multi-element wing sections and
turbo-machinery airfoils52,53. In spite of the speed of GSMs, there is doubt about their robustness
because outcomes are easily affected by initial conditions. Of all the heuristic algorithms, genetic
algorithms (GAs) are the most widely used which are derived from the nature and follow the rule of
‘Survival of the fittest’ 54. Though GAs usually take much more computational time, they can be used
to seek global optima in complex non-linear situations with higher reliability than GSMs. In the
present study, Multi-Island Genetic Algorithm (MIGA) is chosen to be the core optimization method
which can preserve the diversity of solutions and reduce the probability of premature phenomenon.
It separates the population into subgroups and standard GA is operated on each island respectively.
Every few generations, individuals migrate between the islands periodically to realize the exchange
of genetic material. The migration is controlled by two parameters: Rate of Migration and Interval of
Migration. The “tournament selection” scheme of MIGA chooses best individuals from a subgroup of
randomly chosen individuals. In such a scheme, a larger relative tournament size will lead to more
duplicate best individuals in the filial population. The corresponding settings of MIGA are listed in
Table 4.8. Here the population size and number of generations are relatively small because 3D
47 Elkhoury M, Kiwata T, Aoun E. Experimental and numerical investigation of a three-dimensional vertical-axis

wind turbine with variable-pitch. J Wind Eng Ind Aerod 2015;139:111e23.
48 Lei H, Zhou D, Bao Y, Li Y, Han Z. Three-dimensional improved delayed detached Eddy simulation of a two-
bladed vertical axis wind turbine. Energy Convers Manag 2017;133:235e48.

49 Wang S, Ingham DB, Ma L, Pourkashanian M, Tao Z. Numerical investigations on dynamic stall of low Reynolds
number flow around oscillating airfoils. Comput Fluids 2010.

50 Lei H, Zhou D, Bao Y, Li Y, Han Z. Three-dimensional improved delayed detached Eddy simulation of a two-
bladed vertical axis wind turbine. Energy Convers Manag 2017;133:235e48.

51 Wang S, Ingham DB, Ma L, Pourkashanian M, Tao Z. Numerical investigations on dynamic stall of low Reynolds
number flow around oscillating airfoils. Comput Fluids 2010.

52 Nemec M, Zingg DW, Pulliam TH. Multipoint and multi-objective aerodynamic shape optimization. AIAA 2004.
53 Rai MM, Madavan NK. Aerodynamic design using neural networks. AIAA J 2000.
54 Chehouri A, Younes R, Ilinca A, Perron J. Review of performance optimization techniques applied to wind
turbines. Appl Energy 2015.

102
simulation is adopted in the present study which consumes large computational resources. For the
simulation of one individual, it takes about 8 h using 40 threads in parallel. Two individuals are
simulated at the same time as there
are 88 threads in total. The single
objective is to maximize the power
coefficient CP. To clarify the scope of
this optimization problem, certain
constraints should be imposed.
Firstly, to effectively control the
search space, stabilize the calculation
and speed up the convergence, upper
and lower limits of the design
variables are defined which have
been stated in. Secondly, the
maximum relative thickness Table 4.8 Settings of multi-island genetic algorithm
(thickness-to-chord ratio) should not [Courtesy of Ma et al.)
-
be too small as it is a simple indicator
of the structural performance. Combined with the consideration of controlling the search space, the
maximum relative thickness is set in the range of [0.175, 0.205]. Thirdly, to ensure the geometric
compatibility, the location of the maximum relative thickness is restricted in the range [0.25c, 0.32c].
4.2.11 Optimization System Procedure
The optimization process is operated at TSR = 1.0 and the overall aim of the entire system is to
provide an optimized airfoil profile which can help improve the aerodynamic performance of the
VAWT. This process is carried out with an iterative loop which contains tools for modelling, meshing
& calculation, data processing and an optimization algorithm. ISIGHT (Villacoublay, France) is
employed to integrate all the mentioned sub-processes which is famous for its ability to manage data
flows between different programs and to automate the operation of simulation process. As there are
no standard rules for integrating function components in ISIGHT, each optimization system is
designed according to special needs of the corresponding study. The complete scheme of the loop is
illustrated in Figure 4.6.
Figure 4.6 The scheme of the optimization system - [Courtesy of Ma et al.)
The modelling component firstly uses design variables (a1, a2, b1, b2, e1, e2, f1, f2, p1, p2, q1, q2)
and other constant parameters (c, D, h) to generate airfoil profiles with the help of NX (Siemens PLM
Software, Texas, United States). An initial population is needed to start the whole optimization
process. Apart from NACA0018 airfoil, other individuals of the initial population are randomly
103
generated within the scope which is defined by the ranges of variables shown in Table 4.6. Then
the geometry is imported into STARCCM where the meshing & calculation work is done. After that,
values of time varying dependent variables (e.g. the torque of the VAWT) are exported to MATLAB
for post processing in which the values of objectives and constraints are calculated for the
optimization. The running of the entire system is controlled by macros which can call applications
and issue the orders. After the input of the initial population, the process automatically operates until
the ultimate results are outputted.
4.2.12 Results of Geometric Profile Evolution
The optimization system smoothly runs and the final results are generated after the evolution
involving hybridization, mutation and migration. Table 5 shows the final values of variables for the
optimized profile. To visualize these values, Figure 4.7 illustrates the difference between the initial
and optimized profiles as well as the mean camber line (MCL) of the optimized one. At first glance,
the profile of the outer side rises especially in the middle part. The front part of the profile of the
inner side falls while the rear part rises.
Here the inner side represents the side facing the shaft of the VAWT. The MCL is not aligned to the
chord anymore and it intersects the chord line at the position of two ends and 0.38c. The offset of
MCL reaches the
maximum downward at
the position of 0.08c and
upward at the position of
0.78c. shows the relative
thickness distribution of
the initial and optimized
profiles. The initial
profile NACA0018 is
symmetric with the
maximum thickness of
0.18c at the position of
0.3c while the maximum
thickness of the
optimized profile
increases to 0.203c with
the position moving to Figure 4.7 Profiles of the initial airfoil and optimized airfoil
0.275c. The radius of the
leading edge of the
optimized profile
sensibly increases while
the angle of the trailing
edge only decreases a
little compared to the
initial profile. Such an
evolution of profile is
beneficial to the
structural performance
to some degree as the
overall thickness is
increased. The pressure
distribution on the inner Figure 4.8 Relative thickness distribution of the initial and optimized
side may change greater airfoil profiles - [Courtesy of Ma et al.)
104
than that on the outer side in the optimized case. Here the profile changes of airfoils from other
studies are listed for comparison. The profile changes are different as the optimization methods and
objectives are different in diverse studies. In the work of [Ribeiro et al.]55, optimization of airfoil was
conducted using 2D CFD simulations coupled with NSGA-II algorithm. The study focused on the
characteristics like lift coefficient, drag coefficient and lift-to-drag ratio of the airfoil. The optimized
airfoil has larger leading edge radius and the mean camber line becomes more curved.
Archive based Micro Genetic
Algorithm was employed and the
numerical analysis was
conducted using XFOIL. The
overall thickness and leading
edge radius of the optimized
airfoil both increase while the
mean camber line doesn't
change significantly. [Goçmen et
al.]56 investigated six widely-
used airfoils in the field of small
scale wind turbines in order to
improve both the acoustic and
aerodynamic performance. After
optimization, the profile of the
pressure side of each airfoil Figure 4.9 Comparison of average power coefficients and the
becomes closer to the chord line. growth rate of CP between the initial and optimized airfoils at
The angle of trailing edge of different TSRs
some airfoils becomes smaller.
4.2.13 Power Performance
This section firstly discusses about the average power performance improvement at different TSRs
as shown in . In terms of the absolute growth value of CP, the maximum growth occurs at TSR = 0.9
with a value of 0.035 and the minimum growth is 0.007 occurring at TSR = 0.4. The growth values
all exceed 0.03 when TSR falls into the interval [0.9, 1.2]. In terms of the growth rate, it is obvious
that the power performance is largely improved with the maximum growth rate of 26.82% at TSR =
0.9 and the minimum growth rate of 9.02% at TSR = 1.5. The growth rates are all above 20% when
the TSR is no larger than 1.0. The growth rate is relatively low at relatively high TSRs, which is mainly
because the power coefficients are much higher at larger TSRs. Even if the absolute growth value of
CP is the same, the growth rate at larger TSRs will be lower than that at smaller TSRs. The reverse
point where the power coefficient starts to fall off also changes. The reverse point of the initial case
is at TSR = 1.3 while that of the optimized case is at TSR = 1.2, which means the appropriate angular
speed to generate the maximum power output has dropped down when the freestream velocity
equals 8 m/s.
As the power coefficient CP is derived from the torque values, a comparison of the time-varying
torque values during one rotational period is conducted, which is illustrated in Figure 4.10 with the
solid line representing the optimized case and the dashed line representing the initial case. Here the
analysis is divided into two parts. The first part is about the variation trend at a specific TSR and TSR
= 0.9 is selected at which the largest growth rate of the power coefficient CP occurs. The analysis at
other two TSRs is much alike and will not be discussed in detail. The second part investigates the
effects of TSR on the torque values and involves all the six diagrams in Figure 4.10. At TSR = 0.9,
55 RibeiroAFP, Awruch AM, Gomes HM. An airfoil optimization technique for wind turbines. Applied Math, 2012.
56 Goçmen T, Ozerdem B. Airfoil optimization for noise emission problem and aerodynamic performance criterion
on small scale wind turbines. Energy 2012.
105
the peak torque value of one blade with the optimized profile is 3.87 Nm appearing at ϴ = 82 degrees
while that with the initial profile is 3.45 Nm occurring at ϴ = 76 degrees.
The maximum difference of the torque values takes place at ϴ = 90 degrees with a difference value
of 1.09 Nm. Besides, the torque value of one blade with either the optimized or the initial profile
keeps negative in most parts of the downstream area due to the increase of the drag force. In Figure
4.10 (b), the solid line is obviously above the dashed line which means the power performance has
been improved at every second during
the operation of the VAWT. The peak (a) TSR = 0.6
torque values of the VAWT equipped
with three optimized blades during one
rotational period appear at ϴ= 82, 202
and 322 degrees with an approximate
value of 3.7 Nm while that of the VAWT
equipped with initial blades are
approximately 3.2 Nm occurring at ϴ =
76, 196 and 316 degrees. The maximum
differences of the torque values of the
VAWT appear at ϴ = 90, 210 and 330
degrees. Besides, the torque value is
always above zero for the VAWT with
either optimized or initial blades. In fact, (b) TSR = 0.9
by comparing Figure 4.10 (b), it is
obvious that the peak values of the
VAWT appear at the same time when the
peak value of every single blade occurs.
When comparing the torque curves at
different TSRs, the curves of the
optimized cases are focused on and
there are two changing rules worth
attention. On one hand, the position
where the peak value appears moves
with the change of TSR. For the torque
curve of one blade, the positions of peak
value are ϴ = 60 degrees at TSR = 0.6, ϴ
= 82 degree at TSR = 0.9 and ϴ = 94 (c) TSR=1.2
degree at TSR = 1.2. The peak value
occurs later during one rotational period
when the TSR is larger. On the other, the
interval in which the relatively high
torque values appear becomes larger
with the increase of TSR. It can be seen
that the azimuthal interval is 44 degree
at TSR = 0.6, or 52 deg at TSR = 0.9, or 60
degree at TSR = 1.2 during 1/3 rotational
period. This can also explain why the
peak value at TSR = 0.9 is larger than that
at TSR = 1.2 while the average power
Figure 4.10 Instantaneous torque values during one
coefficient at TSR = 0.9 is less than that rotational period - [Courtesy of Ma et al.)
at TSR = 1.2.
106
4.2.14 Comparison of Pressure Distribution

Pressure distribution on the entire surfaces of one blade is discussed. For the sake of brevity, this
paper only mentions the cases at the position of ϴ =90 where the difference of the instantaneous
torque value of one blade with the initial airfoil and that with the optimized airfoil reaches the
maximum. When ϴ = 90, the suction side of the blade is the leeward side while the pressure side is
upwind. On the suction side of the blade with the initial profile, the pressure always keeps negative.
The absolute value of pressure increases from 200 Pa at the leading edge to 350 Pa at the position.
[see Ma et al]57.
4.2.15 Discussion
This section discusses the findings from figures mentioned above, which involves discussions on the
flow regime around blades, airfoil profile change and pressure distribution on the blade. They could
be referred to when designing airfoils of VAWTs in real applications. Specific airfoil profile could
reduce the energy loss caused by vortices (related to the flow regime), thus improving the power
performance. Besides, understanding the change in pressure distribution induced by the airfoil
profile evolution could help improve the structural performance of blades.
A small vortex forms near the trailing edge because of the circulation of flow leaving blade surfaces
at ϴ = 0. This vortex moves along the inner side of blade and induces a reversed pressure wave
transmitting towards the leading edge. In fact, the adverse pressure gradient could easily cause the
formation of bubbles on the inner side of blade and drives them towards the leading edge. This
phenomenon is also reflected by the pressure coefficient change on the inner side from ϴ = 0 to 60.
During such an azimuthal range, the maximum negative pressure on the inner side keeps increasing
and its position continues to move towards the leading edge. From ϴ = 90 to 120, a large vortex is
detected as there is an obvious trough of the negative pressure coefficient on the inner side, which is
in accordance with the vorticity distribution in Figure 4.10.
This vortex has started to shed from the inner side when ϴ = 90 degree and its moving speed is
smaller than that of the blade as the center of the trough of pressure coefficient changes from 0.22c
to 0.4c. Besides, the vortex center keeps moving away from the inner surface because the maximum
negative pressure coefficient decreases a lot from q = 90 to q= 120 degrees. When ϴ= 150 , this large
vortex has lost contact with the blade surface (Figure 4.10(b)) and consequently has little impact
on pressure distribution.
It needs to be noted that the pressure distribution is obviously different between the initial and
optimized cases when ϴ = 90 degree. There isn't a trough of pressure coefficient on the inner side for
the optimized case, which is caused by the shape change of airfoil profile. As is stated before, the
radius of the leading edge increases and the slope of inner side profile also enlarges (Figure 4.7) for
the optimized case. The increased leading edge radius might delay the movement of leading edge
vortex towards the trailing edge at the beginning. But the larger slope of the inner side profile would
cause the vortex to move a bit faster towards the trailing edge later on. The key parameters like the
radius of the leading edge, angle of the trailing edge and slope of the profile should be determined
with care as they have impacts on the movement of vortices and pressure distribution.
Leading edge vortices could be easily found in Figure 4.11 (a)-(d), which also correspond to the
small troughs of pressure coefficient near the leading edge on the blade inner side. Such leading-
edge vortices are usually induced by the laminar-turbulent transition on the blade inner side in the
upstream area. In the downstream area, the flow reattaches to the inner side while small laminar
separation bubbles are induced on the outer side. The scale of such bubbles is relatively small
57 Ning Ma, Hang Lei , Zhaolong Han, Dai Zhou , Yan Bao , Kai Zhang , Lei Zhou, Caiyong Chen, “Airfoil optimization
to improve power performance of a high-solidity vertical axis wind turbine at a moderate tip speed ratio”, Energy
150 (2018) 236-252.
107
compared to those forming in the upstream area. They keep growing and moving towards the trailing
edge as is shown from ϴ =210 degree to ϴ = 330 degrees in Figure 4.11.
Figure 4.11 Instantaneous Contours of Vorticity Magnitude at TSR = 0.9 for the VAWT with
Optimized Airfoils - [Courtesy of Ma et al.)
4.2.16 Limitations of Current Study

The results and discussion in the present study are only valid under the following limitations:
➢ The study focuses on the improvement of power performance which means the optimized
airfoil profile might not be the optimized choice for performance in other aspects like
structural performance.
108
➢ The research object is a high-solidity VAWT with the solidity of 0.75. The findings in this
paper may not be applicable to VAWTs with low or moderate solidity. But the optimization
system developed in this study could be used for VAWTs with various levels of solidity.
➢ The research is conducted at a moderate tip speed ratio (TSR =1.0). This means the
optimization results may change when adopting higher or lower TSRs. When the system is
used for actual design, the TSR could be adjusted to the real value and the system will also
work.
4.2.17 Conclusion
This work establishes a comprehensive optimization system to improve the power performance of a
three-bladed high-solidity VAWT at a moderate TSR. The system is highly adaptive featuring the
combination of 3D CFD simulation, Multi-Island GA and auto operation. Besides, the airfoil is studied
as part of a three-bladed VAWT rather than as a single isolated body. Following conclusions are
obtained:
➢ The maximum thickness of the optimized profile increases by 12.8% compared to NACA0018
profile and the position of the maximum thickness moves from 0.3c to 0.275c. The mean
camber line of the optimized airfoil is not aligned to the chord anymore and the offset of it
reaches the maximum downward at the position of 0.08c and upward at the position of 0.78c.
➢ The radius of the leading edge of the optimized profile enlarges while the angle of the trailing
edge shrinks compared to the initial profile. The increased leading edge radius might delay
the movement of leading edge vortex towards the trailing edge. But the larger slope of the
inner side profile would cause the vortex to move a bit faster towards the trailing edge later
on.
➢ The power performance of the VAWT equipped with the optimized blades improves sensibly.
In terms of the power coefficient, the maximum growth rate of it occurs at TSR = 0.9 with a
value of 26.82% and the growth rate keeps over 20% when the TSR is no larger than 1.0. The
power performance improvement mainly benefits from the increase of torque value in the
azimuthal angle range [60 degree, 120 deg.].
109
4.3 Case Study 3 - Winglet Design for Vertical Axis Wind Turbines Based on a Design
of Experiment (DoE) and CFD
Authors : [Tian-tian Zhang, Mohamed Elsakka, Wei Huang, Zhen-guo Wang, Derek B. Ingham, Lin Ma,
Mohamed Pourkashaniana] 58
Article In : Energy Conversion and Management
Journal Homepage : www.elsevier.com/locate/enconman
Citation : Tian-tian Zhang, Mohamed Elsakka, Wei Huang, Zhen-guo Wang, Derek B. Ingham, Lin Ma,
Mohamed Pourkashanian, Winglet design for vertical axis wind turbines based on a design of
experiment and CFD approach, Energy Conversion and Management, Volume 195, 2019, Pages 712-726,
ISSN 0196-8904, https://doi.org/10.1016/j.enconman.2019.05.055.
Vertical axis wind turbines (VAWTs) have been attracting an increasing attention in recent years
because of their potential for effectively using wind energy. The tip vortices from the VAWT blades
have a negative impact on the power efficiency. Since a winglet has been proved to be effective in
decreasing the tip vortex in the aerospace field, this paper numerically studies the aerodynamic effect
of appending a winglet on the blade of a VAWT.
Based on the theoretical motion pattern of the VAWT blade, this paper simplifies the three-
dimensional full-scale rotor simulation to a one-blade oscillating problem in order to reduce the
computational cost. The full rotor model simulation is also used in validating the result. The
numerical approach has been validated by the experimental data that is available in the open
literature. Six parameters are applied in defining the configuration of the winglet. The orthogonal
experimental design (OED) approach is adopted in this paper to determine the significance of the
design parameters that affect the rotor’s power coefficient. The OED results show that the twist angle
of the winglet is the most significant factor that affects the winglet’s performance.
A range analysis of the OED results produces an optimal variable arrangement in the current scope,
and the winglet’s performance in this variable arrangement is compared with the blade without a
winglet. For the single blade study, the comparison result shows that the optimal winglet can
decrease the tip vortices and improve the blade’s power performance by up to 31% at a tip speed
ratio of 2.29. However, for the full VAWT case, the relative enhancement in the power coefficient is
about 10.5, 6.7, and 10.0% for TSRs of 1.85, 2.29, and 2.52, respectively. The winglet assists in
maintain the pressure difference between the two sides of the blade, thus weakening the tip vortex
and improving the aerodynamic efficiency of the surface near the blade tip.
4.3.1 Introduction
Wind is widely recognized as being a source of clean energy, and it will never be exhausted. Using
wind energy to generate electricity, instead of fossil fuels, can not only save the conventional energy
resources but also protect the environment [1]. Nowadays, many countries are investigating an
increasing amount of funds into establishing wind turbines [2]. Horizontal axis wind turbines
(HAWTs) dominate the wind turbine market, and it is effective in large electricity plants.
However, the high cost in manufacturing the blades and installing the tall towers limits the use of
wind energy. In comparison with HAWTs, the Darrieus type vertical axis wind turbines (VAWTs)
are simpler in structure because they can work in any flow direction without the assistance of a yaw
control mechanism. Normally, VAWTs have straight blades, which are easier to manufacture than the
twisted blades in HAWTs. In addition, the structure of VAWTs make it possible to place the generator
and gearbox on the ground and this can reduce the installation and maintenance complexities. All the
characteristics in VAWTs can reduce the cost of VAWTs in comparison with HAWTs [2,3]. In addition,
58Tian-tian Zhang, Mohamed Elsakka, Wei Huang, Zhen-guo Wang, Derek B. Inghama, Lin Ma, Mohamed
Pourkashanian, “Winglet design for vertical axis wind turbines based on a design of experiment and CFD
approach”, Energy Conversion and Management, 2019.
110
VAWTs are attractive for distributed small scale electricity facilities in urban environments where
there are unstable wind conditions, as well as large scale offshore floating turbine for deep-sea
installations [5].
4.3.2 Current Study on VAWTs
The Darrieus type rotor is a lift based VAWT, and it uses blades with airfoil shaped cross sections [6].
The study on Darrieus type VAWTs has many aspects, such as the pitching of the blade, the roughness
of the blade surface, the blade thickness, the blade tip speed ratios, etc. Experimental studies on
VAWTs are direct and effective. [Li et al.][7-12] implemented numerous experiments in a wind
tunnel and obtained many valuable results. They concluded that the power coefficient decreases with
an increase in the number of blades and that two blades have a higher annual generating capacity in
high wind velocity regions.
Through change in the pitch angle of the blade, they found that an optimum pitch angle (6°) can
effectively improve the power efficiency of their turbine design. In addition, the Reynolds number
and solidity of the rotor also affects the performance of the VAWTs . Although wind tunnel
experiments are ideal in the study of VAWTs, the expense in building wind tunnels and
manufacturing the rotor has limited the development. Also the results show some differences with
the field experimental results.
The rapid development of computational power makes computational fluid dynamics (CFD) ever
increasingly more popular for use in wind turbine aerodynamic design. Numerical simulations of the
performance of VAWTs are cheaper and faster than experiments. Therefore, numerous research
papers on VAWTs have been published based on CFD analyses. For example, [Lanzafame et al.][13]
studied the CFD simulation process of VAWTs using the Ansys Fluent software. 2D unsteady CFD
models of Darrieus VAWTs using the sliding mesh method is adopted, and the results have been
compared with the available experimental data. Based on the CFD approach, [Rezaeiha et al.][14]
investigated the impact of the operational parameters (e.g. the tip speed ratio, the Reynolds number
and the turbulence intensity) on the performance of VAWTs.
Guidelines for the azimuthal increment, domain size and convergence are proposed for the accurate
CFD study of the VAWT [15]. [Rezaeiha et al.][16] also studied the influence of a fixed pitch angle on
the power coefficient of VAWTs. They conclude that a negative pitch angle can improve the power
performance of a VAWT, and dynamic pitching might be a more promising approach than fixing the
pitching angle. Following [Rezaeiha]’s expectation, [Li et al.] [17] optimized the blade pitch using a
genetic algorithm to maximize the turbine power performance. It is noticed that the optimized pitch
angle changes continuously as the rotor rotates and the pitch angle curves differ with different tip
speed ratios. Furthermore, in order to determine the optimum airfoil shape for the blades of VAWTs,
the VAWTs with the blade airfoils of NACA0018, NACA0021 and NACA0025 have been simulated
using a 2D CFD approach by [Shukla, et al.][18]. [Jafaryar et al.] [19] studied the influence of the
shape of the airfoil on the performance of VAWTs based on the response surface model and the
central composite design technique.
The results obtained confirm that the optimum shape of the airfoil is very similar to those obtained
using symmetrical blades. In addition to the RANS approach that is normally used in VAWTs
simulations, the ALM-LES coupled with the URANS and RANS-LES approaches have been developed
in order to obtain more accurate CFD results [20,21].
The main weakness of a VAWT is its low efficiency in power generation. For this reason, many devices
are proposed to enhance the performance of VAWTs. For example, [Wong et al.] [22,23] studied the
influence of a flat plate deflector, which is placed upstream, on the performance of VAWTs. Due to
the flat plate, the free flow accelerated and was deflected by the deflector before impinging on the
turbine and thus the rotor’s power coefficient is enhanced. [Hashem et al.] [24] proposed the wind-
lens concept which consists of a diffuser and flanges to collect the wind energy. This mechanism
needs yawing devices to ensure that the wind-lens faces towards the wind direction.
111
4.3.3 Influence of 3D Effects on the Performance of VAWTs

Because of the straight character of the VAWTs blade, most CFD works have simplified the model into
a 2D problem since 2D simulations are easier and faster than 3D simulations. However, 2D
simulations often over predict the performance of VAWTs because of the neglect of the vertical flow
and the vortices from the trailing edge and the tip end of the blade. [Siddiqui et al.] [25] investigated
the effect of the 3D flow characteristics on the VAWT’s simulation results and compared 5 cases in
order to study the effect of the tip vortices and the supporting arm. The results obtained show that
2D simulations can over predict the performance by up to 32% in comparison with real time 3D CFD
simulations. This result does not only reflect the inaccuracy in the 2D simulations but also illustrates
that the efforts in reducing the 3D impact has a promising potential in improving the VAWT’s
aerodynamic performance.
Among all the 3D effects, the tip vortices have a great impact on the performance of lift-based blades
and the scope does not limit to VAWT. Taking the aircraft’s wing as an example, the lift is generated
because of the pressure difference on the upper and lower surfaces. However, at the tip end of the
wing, the high-pressure air is transported from the lower surface upwards to the upper surface, thus
causing a reduction in the lift and generating the so-called tip vortex. The flow in the blade’s span
wise direction caused by the transportation in the blade tip also decreases the efficiency of the wing.
The span wise flow on the VAWT’s blade is similar to the span wise flow on an aircraft wing. As the
VAWT’s rotor rotates, both the suction and pressure sides of the blade change their roles continually.
This results in more complicated tip vortices than the static wing. Mie-type tip vane and end plate
have been proved to be effective in weakening the tip vortices and enhance the power performance
of both HAWTs and VAWTs [26–30]. However, the vane results in increasing the fatigue loading and
decreasing the lifetime for the VAWT [31]. Among these devices, the winglet performs better in terms
of improving the power coefficient [31,32]. Although these tip devices have been proposed, more
detailed studies are still limited in comparison with the study on the HAWT.
4.3.4 The Novelty in the Design of the VAWT’s Winglet
A winglet is now commonly observed in most of the commercial long range airplanes and this reduces
the drag and hence saves fuel [33]. This device helps to isolate the suction and pressure sides of the
blade, thus decreasing the tip vortices and improves the blade’s performance. During the study of a
winglet on an aircraft, the influences of the cant angle, sweep angle, twist angle as well as the winglet
direction have been numerically investigated [33–37]. In addition, on using a multi-tipped winglet
to decrease the strength of the tip vortices and improve the blade’s performance, it has been proved
to be an effective way to further enhance the performance of the wing [38].
Inspired by the winglet design in the aeronautic field, engineers are attempting to improve wind
turbines’ performance in a similar way because these blades also encounter tip vortex problems. For
example, [Gupta and Amano][39] found that a winglet with a cant angle of 45° performs better than
that of the vertical winglet on the blade of the HAWT. [Johansen and Sorensen][40] found that the
vertical winglet on the blade of the HAWT has an optimal twist angle when all the other parameters
are kept constant. [Farhan et al.][41] combined the winglet planform with the airfoil design using a
CFD approach.
The result suggests that a smaller tip length configuration may have a better performance. However,
the flow patterns are very different among the static wing of the aircraft, the blade on the HAWT and
the blade on the VAWT. To be more specific, the blade on the HAWT experiences an unevenly
distributed incident velocity in the span wise direction and the tip velocity is the largest. Although
the blade of the VAWT has a consistent incident velocity in the span wise direction, the velocity value
changes continuously as the blade rotates. In addition, the suction side and the pressure side of the
blade of the VAWT convert with each other as the rotation proceeds. This results in time-varying tip
vortices in the flow field. As a result, the results obtained from the aircraft winglet design and the
HAWT winglet design cannot be applied directly to the winglet design of VAWTs.
112
Although the wake and tip vortices also have a very strong impact on the VAWT’s performance [42],
only a few investigations have studied the winglet design for a VAWT. In order to reduce the tip
vortices’ influence on the efficiency of the VAWT, [Rohmann-Shaw][32] performed 3D CFD
simulations on a VAWT and investigated the blade tip design. Through a comparison of the end plate
and winglet, the author found that the winglet is more effective in improving the aerodynamic
efficiency of the VAWT. [Tao et al.][43] used the Non-uniform rational basis spline (NURBS) based
free form deformation approach in the parametric method in order to design the winglet. This
parametric
approach assists in the studying of the winglets’ configuration more comprehensively [43,44].
Although previous studies have illustrated that the winglet can improve the aerodynamic
performance of the VAWT, they failed to give an optimal approach for the winglet configuration
design. It is still not clear which parameters are significant in the winglet design process.
This paper defines 6 parameters to parameterize the winglet’s configuration [39] and studies their
importance in influencing the blade’s power coefficient using the orthogonal experimental design
(OED) approach [45].
The aim is to determine the most influential design parameters of the winglet of the VAWT and obtain
a good design through the OED approach. Through comparing the aerodynamic performance of
VAWTs with and without a winglet, we aim to obtain the performance enhancement by adding a
winglet on the blade tip and determine the working mechanism of the winglet.
4.3.5 Physical and Numerical Model
4.3.5.1 Physical Model Introduction
The NACA 0015 is commonly used as the airfoil for VAWT’s blade in the experimental investigations
[6]. It is a symmetric airfoil with a good power coefficient performance [24]. [Li et al. ][12] performed
many wind tunnel experiments including using this airfoil and the results are available in the open
literature. Based on the work of [Li], this paper investigates three-dimensional numerical simulations
on a 2-bladed VAWT with a NACA 0015 airfoil for the blades. The chord length (C) of the blade is
0.225 m, the height (H) is 1.02 m, and because of the symmetric character in the span wise direction,
only half the model is studied
numerically in this paper. The
radius of the rotor (R) is 0.85m
and the blade’s pitching angle is
set to be 6° according to the
optimization study of [Li et
al.][7,12]. The geometrical
features of the rotor is listed in Table 4.9 The Operational Conditions of the
Table 4.10. VAWT Model
Figure 4.12 illustrates a
schematic of the VAWT’s full-
model geometry studied in this
paper. The freestream wind
speed is V∞(7m/s) flowing from
the left-hand-side across the
turbine, with a Reynolds number
Re = 2.89×105, which is typical
for the VAWT. It is expected that Figure 4.12 Schematic of the VAWT Geometry
the results obtained using this
turbine will give useful
information for VAWTs using other types of airfoils. The operational conditions of the reference rotor
in listed in
113
Table 4.10 The Geometrical Features of the VAWT Model
Figure 4.13 illustrates the relationships among the

velocities and force coefficients in different coordinate
systems. Vreal is the incident velocity of the blade, and it
is composed of the free flow velocity V∞ and the rotation
speed ω R. CL and CD are the lift and drag coefficients in
the velocity coordinate system whose X-axis points to
the free flow direction. The azimuthal angle is θ, and the
blade’s angle of attack is α. Ct and Cn are the tangential
and normal force coefficients, respectively, relative to
the rotation of the blade. Cx and Cy are the horizontal and
vertical force coefficients respectively in the absolute
coordinate system that fixed on the shaft. The Tip Speed
Ratio (TSR) defines the ratio between the speed of the
blade and the free flow velocity, namely
ωR
TSR =
V∞ Figure 4.13 Relationships Among
Eq. 4.3 Different Coordinate Systems on the Blade
According to the sine rule theorem, the relationship of VAWT
between the different velocities is as follows:
ωR V∞
=
sin (θ − α) sinα
Eq. 4.4
Therefore, the TSR can be calculated as follows:
sinθcosα − sincosθ
TSR = = sinθcotα − cosθ
sinα
Eq. 4.5
As a result, the relationship between the azimuthal angle (θ) and the attack angle (α) is given by
sinθ
α = atan ( )
TSR + cosθ
Eq. 4.6
In this paper, an optimal pitch angle β = 6° [11] is adopted so that the VAWT operates in the optimal
condition. In this case, the real attack angle of the blade is giving by
114
sinθ
α = atan ( )−β
TSR + cosθ
Eq. 4.7
The torque of a single blade and the power of the rotor are calculated as follows:
R 2π NR 2π
Q= ∫ F(θ) dθ , P= ∫ F(θ) dθ = NωQ
2π 0 (2π/ω) 0
Eq. 4.8
where FT is the transient tangential force when the azimuthal angle is θ, and N is the number of blades
on the VAWT. The wind turbine performance is normally characterized by its power coefficient (CP),
and the relationship between CP and the torque coefficient CQ is given as follows:
1 2π
∫0 F(θ) dθ NCω c
C̅T = 2π , CP = C̅ , CQ = C̅
(1/2)ρcHU02 2U0 T 2R T
Eq. 4.9
where c and H are the chord length and height of the rotor, respectively. The tangential force and
tangential force coefficient can be obtained directly in the simulations. When the blades rotate, it is
easy to acquire the horizontal and vertical force coefficients namely Cx and Cy. The tangential and
normal force coefficients are given as follows:
Ct = Cx cosθ − Cy sinθ , Cn = Cx sinθ + Cy cosθ

Eq. 4.10
4.3.5.2 Modeling Strategy
Full size simulations of VAWTs can produce accurate results but the meshing and calculation process
can be very time consuming. The grid around the blade, as well as the winglet, occupies a large
portion of the total grid number. This can greatly increase the dependence on the computer memory
and time. According to the pattern of the motion of the VAWT’s blades, the theoretical incident
velocity and the attack angle of the blade are functions of the azimuthal angle. Through the
relationship shown in Figure 4.13, the incident velocity can be calculated, using the cosine theorem,
as follows:
Vreal = √ω2 R2 + V∞
2 + 2ωRV cosθ
∞
Eq. 4.11
In order to simplify the simulations, we only consider one blade, and we employ a coordinate system
that is fixed on the blade of the turbine, and in this relative coordinate system, the blade is performing
an oscillating motion around its central axis with an upstream wind speed and angle of attack
described by Eq. 4.11 and Eq. 4.7, respectively. Although this motion pattern is different from the
full rotor rotation process, the flow conditions for a single blade are similar for both cases. Therefore,
without the need to mesh around every blade and winglet, the computation cost can be dramatically
reduced, and it is also easier to implement.
Figure 4.14 schematically shows the computational domain for the one blade model as employed in
this paper. The domain size is defined according to [Ma et al.][51]. The blade is rotating around its
central axis at a frequency defined according to the rotational speed of the turbine.
This is realized by establishing a cylindrical sub-domain around the blade and applying the sliding
mesh approach to make this sub-domain rotate. The non-conformal interface boundary condition is
used to connect the sub-domain and the stationary flow field domain. Further, a symmetry condition
115
Figure 4.14 Schematic of the computational domain for the one blade model (dimensions are not to
scale)
is adopted on the bottom surface so that only half of the whole blade has to be simulated. Also the
slip wall condition is used on the side and top of the outfield to avoid the influence of the wall
boundary layer on the blade. The velocity inlet condition is employed on the inlet of the flow field
domain and the velocity value is calculated by using Eq. 4.11. At the outlet of the flow field, the
pressure outlet condition is used.
For a single blade simulation, the incident velocity’s direction is always horizontal. Therefore, the
dynamic force coefficients obtained in the Ansys Fluent software are the drag and lift force
coefficients, namely CD and CL. According to the coordinate relationship in Figure 4.13, the
tangential and normal force coefficients can be acquired using the following formulae:
Ct = CD cosθ − CL sinθ , Cn = CD sinθ + CL cosθ

Eq. 4.12
where α is the angle between the body coordinate system and the velocity coordinate system shown
in Figure 4.13 and it is normally calculated using Eq. 4.6. Since the blade has a fixed pitch angle (β)
in the current work, α should be calculated using Eq. 4.7.
Using one blade to simulate the motion of the VAWT saves computational cost but it cannot take the
blade-blade interaction and blade shaft interaction into consideration [52]. The virtual incidence
effect is also neglected [53]. These may cause slight inconsistences between the numerical results
and the real situation. Nevertheless, it can be used in the primary design of the blade and the
validation using a full-scale rotor simulation is discussed in the following sections. In order to make
the design in this work more reliable, a full 3D model simulation of the VAWT will be performed in
order to validate the numerical model and verify the efficiency of the final design.
4.3.6 Code Validation and Grid Independency Analysis
Figure 4.15 shows the structured mesh topologies around a blade and an adjacent arm for the mesh
in the full-scale 3D model simulation. The grid is clustered towards the surface of the blade, the
winglet and the supporting arms in order to capture the boundary layer more precisely. In order to
choose a mesh that is not only precise but also time-saving, three different meshes are used to check
the mesh independency. The grid independency analysis is based on the full-scale model. The
geometry, as well as the mesh, in the flow field is generated in the workbench in ANSYS 18.2. The
domain is discretized using a fully structured shaped mesh, except in the blade tip region and the
winglet region, when applicable.
116
Figure 4.15 An illustration of the baseline computational mesh, showing the mesh topologies around a
blade and an adjacent arm on different vertical and horizontal sectional planes
Table 4.11 shows three different attributes of the meshes. While the number of the nodes around
the blade profile changes for the different meshes, the non-dimensional wall distance, y+, is kept
constant in order to adhere to the
recommended value for the
enhanced wall treatment that is
used by the transition turbulence
model. In these meshes, there are
43 span wise mesh divisions along
the blade half length. The
geometry, as well as the mesh, in
the flow field is generated in the
workbench in ANSYS 18.2. The Table 4.11 Three Sets of Meshes
domain is discretized using a fully
structured shaped mesh, except in
the blade tip region and the winglet region, when applicable. The flow around the VAWT is modelled
using a commercial CFD solver, ANSYS FLUENT [54]. The three-dimensional Reynolds-averaged
Navier-Stokes equations and the transition SST k-ω turbulence model have been utilized in the
simulation. The double precision pressure based coupled solver algorithm is utilized in the transient
mode and the second-order scheme is used for the spatial and temporal discretization. Using the
coupled scheme, a temporal resolution of 540-time steps per cycle is chosen according to [Elsakka et
al.][52] and this corresponds to an azimuthal increment of 0.67° which slightly coarser than the 0.5
value recommended by [Rezaeiha et al.][15] using the SIMPLE scheme.
The results are considered at the fifth cycle and this is found to be sufficient to reduce the differences
between the results at successive cycles to 0.3% under the current setup using the coupled solver.
However, some other studies that utilize the SIMPLE scheme recommend significantly larger number
of cycles, namely between 20 and 30 turbine cycles. The relative motion between the turbine and the
absolute reference frame is modelled using a sliding mesh interface.
117
Figure 4.16 shows the

CFD predictions of both
the power coefficients
from the turbine mid-span
section and the overall
turbine power coefficients
at a TSR of 2.29. The
relative difference
between the predicted
values based on the
baseline mesh and fine
meshes is about 1%, which
means that further
increasing of the mesh size
has a very small influence
on the numerical result.
Therefore, the settings in Figure 4.16 The prediction of both the power coefficient contribution
the baseline mesh are of the turbine mid-span section and the turbine overall power
chosen for the further coefficients for the different
tests. meshes at a TSR of 2.29.
4.3.7 3D VAWT CFD Validation Using Experimental Data

The baseline mesh has been chosen to simulate the motion of the VAWT and the results obtained is
validated by the experimental data of [Li et al. [12]. Both the experiment and the simulation use the
same geometrical and operational characteristics as listed and
Table 4.9. The experiment in [Li et al.][12] is implemented in an open circular type wind tunnel with
the six-component balance installed on the basement of wind turbine to measure the force and the
moment.
The high frequency pressure scanners are located at different positions along the spanwise directions
of the rotor blade. Figure 4.17
shows a comparison between
the CFD predictions of the
instantaneous torque
coefficient and the
corresponding experimental
data obtained by the
integration of the pressure
distribution from the pressure
scanner at the blade mid-span
[12]. This comparison shows a
good agreement between the
CFD prediction and the
experimental data, which
confirms that the numerical
setup and the mesh are Figure 4.17 A comparison between the experimental and the
convincing in the numerical numerical data of the torque coefficient contribution at the turbine
study of the motion of the mid-span at a TSR of 2.29
VAWT. Therefore, the current
setup is used to study the effect
of the proposed winglet design on the performance and the flow behavior around the turbine. As
118
mentioned in Section 4.3.5.2, the winglet design process used the one blade model because this
model can save the meshing and simulation cost.
Figure 8 of [Zhang, et al.]59 shows the mesh in the flow field and the surface mesh on the blade for the
one blade model simulation. This mesh is generated based on the baseline mesh used in the full-scale
simulations. The Reynold number, the meshing strategy and the settings in Fluent are all the same
with the full-scale simulation. In this one blade simulation domain, the flow field is divided into 3
parts, namely the outfield, the rotation field around the straight blade and the rotational field around
the winglet. It should be noticed that only the third part adopts an unstructured mesh because this
can improve the mesh generation efficiency around a morphing winglet. The node number along the
profile of the winglet is the same as that on the blade (172 nodes).
Since an unstructured mesh is adopted in the subdomain around the winglet, the nodes number is
relative with the winglet configuration. The maximum face size in the winglet subdomain is limited
to 0.01m and the maximum dimensionless wall distance, y+, around the surface is less than 1.2. The
grid in the flow field is clustered towards the rotating region in order to capture the flow
characteristics more precisely. The encrypted set up is the same with the baseline mesh validated in
the full-scale model. Since the one blade simulation has to consider the blade’s motion and the
transient incident velocity, the User-Defined-File (UDF) file and the sliding mesh are adopted based
on Ansys Fluent. In the current simulation, the free flow velocity is V∞ = 7m/s and the TSR is 2.29
based on [Li]’s experiment [12] so that the rotation speed is ω=18.86/s.
The relative incident velocity condition is applied at the velocity inlet, and the incident velocity is
defined according to Eq. 4.11. The rotating part around the blade and the winglet employs a sliding
mesh approach and the blade’s motion is realized by rotating the mesh around it according to Eq.
4.7, in which θ = ω · t. Both the transient incident velocity condition and the rotating mesh are
realized by UDF files, which allow users to customize ANSYS FLUENT.
4.3.8 Orthogonal Experimental Design (OED)
4.3.8.1 A Brief Introduction of the OED Approach
Orthogonality is an important index in the Design of Experiments (DOE) because it reflects the
independence among the design variables. An experimental analysis of an orthogonal design is
usually straightforward in that each main effect and the independency of the parameters can be
estimated. The OED assists in the sampling among the parameter combinations with multi levels and
shows the impact of the parameters when employing only a few experiments. Designers can choose
the most influential variables and simplify the optimization process through an orthogonal
experiment design.
This approach is now commonly used in the energy and aerospace fields, especially in those
circumstances when the single experiment or simulation is costly [45–47]. For example, [Wang et al.]
[46] combined the OED method and the CFD approach in optimizing the performance of a three-
bladed VAWT with leading-edge serrations and helical blades. [Liu et al.][47] applied the OED
approach to study the effects of many factors on the optical-thermal efficiency of a parabolic trough
solar collector. Since the samples in the OED have orthogonality, they can also be used to establish a
surrogate model before optimization [48–50]. Due to the continual study of OED, different orthogonal
arrays that are suitable for different numbers and levels of the variables are now available.
This paper adopts the OED method in studying the importance of the design variables and the optimal
variable value arrangement in the pre-set level arrays. There are 6 parameters that define the
geometry of the winglet, and these are shown in Figure 4.18. This figure shows that there is a dashed
line inside of the winglet, and the winglet is generated by sweeping the airfoil along the dashed line.
119
This dashed line composes of an arc

and its tangent line. The radius and
the radian of the arc, namely Rcant
and Ωcant, define the radius and
radian of the winglet. The length of
the tangent line (Ltip) defines the
length of the winglet, and Ctip is the
chord length of the airfoil at the tip
of the winglet. σ is the twist angle of
the winglet after the sweep
operation, and its positive direction
is noted in the figure. Θ sweep is the
sweep angle of the winglet, and it is
controlled by the moving of the
dashed line along the blade’s chord
line. In the scope of this paper, the
winglet bends towards the inside
direction of the rotor of the VAWT. Figure 4.18 Schematic Diagram of the Parameters that Define
Further, the winglet’s airfoil in the Winglet Geometry
every section is maintained to be
the same as that of the blade’s airfoil.
4.3.8.2 Design Variables, Levels and Objective of the Winglet Design
The design variables in the winglet design are mainly based on the design parameters shown in
Figure 4.18 but with some modifications. For example, Θ sweep is controlled mainly by moving the
dashed line along the blade’s chord line but also C tip has an impact on Θ sweep. Since C tip is already
defined as a design variable, in this paper we use the sweep distance (ϑ sweep) as the design variable.
When ϑ sweep = −0.057, the corresponding sweep angle is 0. A larger ϑ sweep value than −0.057
means a positive Θ sweep while a smaller value means a negative Θ sweep (See to Figure 4.18). The
chord length of the winglet’s tip (Ctip) is dimensioned by the blade’s chord length (Cb), and the ratio
of Ctip and Cb is defined by γtip.
In the OED design of the winglet, the aim is to determine the most significant design variable and the
optimal arrangement of the variables. The design variables can be set to be in a wide range but a
distorted configuration has a negative impact on the meshing and convergence. As a result, this paper
balances the range of the variables and the convergence requirements and divides every parameters
into 5 levels. The design variables and the design levels are listed in Table 4.13. The design levels
and ranges of the design variables are chosen to cover as many configurations as possible. However,
considering the regularity of the grid and the convergence of the simulations, we have limited these
levels to be in an acceptable range. Examples of other suitable designs are conducted and available
by [Malla et al.]60.
4.3.8.3 L25(56) Orthogonal Array
The combination of the variable values listed in Table 4.13 results in Figure 4.18. Schematic
diagram of the parameters that define the winglet geometry. 56 = 15625 cases to be investigated.
However, the Orthogonal Array uses only a few of these combinations to determine the influence and
significance of each variable. In this case, the L25(56) orthogonal array is adopted to design the
samples, where ‘L’ stands for the orthogonal array and 25 is the total samples needed in the array
[56]. Using 56 means that there are 6 design variables in total and each variable has 5 levels. The
60 Anamika Malla, Zhaolong Han, and Dai Zhou, “Effect of a winglet on the Power Augmentation of Straight Bladed
Darrieus Wind Turbine”, 6th International Conference on Environment and Renewable Energy, 2020.
120
Table 4.13 Design Parameters and Design Levels of the Winglet
orthogonal array used in this paper is listed in Table 4.12. Cp is the power coefficient obtained in
Table 4.12 The L25(56) Standard Orthogonal Array for the Parametric Study of the Winglet and the
Power Coefficients
121
each case and Cp0 is the power coefficient of the blade without a winglet. The value of Cp0 is 0.0917.
The last column in Table 4.12 reflects the effect of using the winglet in each case. Different
configurations have different shadow areas but this effect is considered when presenting the results
in the following sections.
4.3.9.1 Analysis of the OED Results
Since the power coefficients of the 25 cases investigated are listed in Table 4.12, an analysis of the
range is based on these results to assist in the identifying of the significance of the design variables.
Table 4.14 lists the mean power coefficient (Cp) distribution as a function of the different
parameters and different levels. R(Cp) is the range between the maximum and minimum values of a
specific variable at different levels. It is clear that σ (the twist angle) has the largest R(Cp), which
Table 4.14 Mean Cp Distribution as a Function of the Different Design Parameters and Different
Parameter Levels
means that the variance of σ has the most significant impact on Cp. The value in the dark shade is the
maximum values in each column. These values mean that the power coefficient can be relatively high
when the corresponding variables adopt the corresponding levels. Error! Reference source not f
ound. is a graphic display of Table 4.14, and it reflects the influence of the parameters more clearly.
Overall, Table 4.14 and Error! Reference source not found. show the importance in the order of
the 6 parameters is σ > ϑ sweep > Ltip > Ωcant > γtip > Rcant and their optimal levels are L1, L4, L2, L3,
L4, L3, respectively. The corresponding optimal value for the design variables in the current range
are as follows: σ =−14.4°, ϑ sweep=0 m, Ltip=0.04 m, Ωcant=60°, γtip=0.45, Rcant= 0.05 m.
4.3.10 Analysis of the Variance (ANOVA) of the OED Results
The ANOVA is used to estimate the relative significance of each parameter in terms of the percentage
contribution to the overall response. It reflects the design parameters that have a significant impact
on the results. The sum of the squares of the deviations that are specific to one variable is divided
into the sum of the squares within the group (SSW) and the sum of the squares between the groups
(SSB) [57,58]. The degree of freedom within each group (DoFW) is the value of that observation
number minus the group number. The degree of freedom between the groups (DoFB) is the value of
that group number minus 1. Fvalue is determined as follows:
SSB / DoFB
Fvalue =
SSW/ DoFW
Eq. 4.13
122
Table 4.15 The Data Employed in the Analysis Steps in the Study of the Parameter Significance
The standard Fvalue is determined from an F-table for

a given statistical level of the significance. In this
paper, the standard Fvalue at the 5% significance level
is 2.8661 when the DoFW and DoFB are 20 and 4,
respectively [59]. When the Fvalue of a specific
variable is greater than the standard value F0.05,
then this means that this variable is significant. In
Table 4.15, the Fvalue of the twist angle (σ) is
17.75056, which is larger than F0.05. The Fvalues for the
other variables are all smaller than F0.05. Therefore,
the ANOVA result reflects that σ is the only
significant factor among all the chosen design Figure 4.20 The designed winglet using the
variables. This result is consistent with the ANORA optimal variable level arrangement
result. This means that when the optimization of the
winglet’s configuration is required in the following work, the twist angle should be the most
important parameter to consider.
4.3.11 Working Mechanism
Analysis of the Winglet
As discussed in Section 4.3.9, the
best arrangement for the design
variables in the current range are
as follows: σ = −14.4°, ϑ sweep= 0
m, L = 0.04 m, Ωcant = 60°, γtip = 0.45,
Rcant = 0.05 m. Figure 4.20 shows
the winglet configuration in the
optimal data set. The aerodynamic
performance for this configuration
is calculated and compared with
the non-winglet blade (reference
blade) in the following sections.
4.3.12 Comparison of the Best
Result and the Reference
Blade Figure 4.19 Comparison of the tangential and normal force
coefficients as a function of the azimuthal angle between the
Figure 4.19 shows a comparison
reference blade and the blade with the optimized winglet.
of the tangential and normal force
123
coefficients as a function of the azimuthal angle between the reference blade and the blade with the
optimized winglet. The corresponding angle of attack can be calculated using Eq. 4.4. The figure
shows that both the tangential and normal force coefficients of the blade with the optimal winglet
have higher maximum values than the reference value.
Figure 4.21 16. The flow field characteristics when the azimuthal angle is 180°. (Left side: the
reference blade, right side: the blade with the optimal winglet).
Further, their values are almost the same when the azimuthal angle is 0°and 180° but the difference
becomes larger when the azimuthal angle approaches 90° and 270°. The same characteristics appear
in the torque coefficient comparison.
The results above can be explained by Figure 4.22 and Figure 4.21 in which the flow field
characteristics are compared separately when the azimuthal angle is 90° and 180°, respectively. In
Figure 4.22, the near-tip streamlines derived from the upstream of the reference blade transfer
from the pressure side to the suction side at the blade tip. This phenomenon results in the pressure
difference between the two sides of the blade becoming smaller and the blade cannot effectively use
the pressure difference to generate thrust. The winglet separates both sides of the blade so that most
Figure 4.22 The flow field characteristics when the azimuthal angle is 90°. (Left side: the reference
blade, right side: the blade with the optimal winglet).
124
of the streamlines cannot cross the blade tip, which maintains the pressure difference between the
two sides of the blade and this improves the aerodynamic efficiency of the surface near the blade tip.
In Figure 4.21, the near-tip streamlines of the reference blade do not cross the blade tip so that the
winglet’s function is not obvious. The reason is that the attack angle of the blade is small when the
azimuthal angle is 180° and the pressure difference of both sides is not apparent.
As a result, the winglet improves the blade’s aerodynamic performance when the azimuthal angle is
90° but the improvement is not obvious when the azimuthal angle is 180°. The reason why the
winglet improves
the blade’s performance at the azimuthal angle around 300° is similar to that when the azimuthal
angle is 90°. Since the one-blade model does not consider the wake of the rotor, the flow contours
when the azimuthal angle ranges from 180° to 360° is not discussed in this section. Following section,
where full rotor model simulations have been implemented, will further compare the results.
Fig. 14 [see Zhang, et al.]61 also shows that the torque coefficient of the optimized configuration has
a higher maximum value and average value than the reference blade. As a result, the power
coefficient for the designed blade is larger than the reference value. In the current study, the power
coefficient of the optimized configuration is 0.1207, which is 31.6% higher than the power coefficient
of the reference blade (0.0917). This means that the winglet can greatly improve the power efficiency
of the VAWT’s blade. Fig. 17 [see Zhang, et al.]62 shows a comparison of the iso-surface contours
between the optimized configuration and the reference blade when the Q-criterion level is 0.005.
The Q-criterion is a method to describe the instantaneous vortex structures, which reflects the 3D
structures of the flow field [60].
The iso-surface is colored by the norm of the resultant velocity of the flow field. In the same flow
condition, a longer vortex structure means a stronger vortex strength and a more revere pressure
interaction at the blade tip. Since the pressure interaction at the blade tip reduces the pressure
difference between two sides of the blade, and thus decrease the power efficiency, a reduced vortex
Figure 4.23 The contours of the turbulence kinetic energy in normal and lateral planes when θ=90°
61 Tian-tian Zhang, Mohamed Elsakka, Wei Huang, Zhen-guo Wang, Derek B. Inghama, Lin Ma, Mohamed
62 See Previous.
125
length can visually reflect a weaker pressure interaction phenomenon and an improvement in the
power efficiency.
The first column is the result for the reference blade and the second column is the result for the blade
with the optimal winglet. Fig. 17 [see Zhang, et al.]63 shows that when the azimuthal angle is 90°, the
optimized configuration’s vortex area is smaller than the reference blade’s vortex area. Also the same
phenomenon appears when the azimuthal angle is 270° and 360°. The contours for both
configurations are similar when the azimuthal angle is 180°.
Figure 4.23 compares the contours of the vorticities in the normal and lateral planes when the
azimuthal angle is 90°. The contours are sliced in the same position for both the reference blade flow
filed and the blade with winglet flow field. These figures show that the winglet can change the
distribution of the vorticity field and decrease the intensity and the influence area of the tip vortex.
The results presented in both Fig. 17 (not shown here) and Figure 4.23 show that the tip vortex of
the blade can be weakened by appending the optimized winglet on the blade during most part of the
rotation cycle. This means that a drop in the pressure interaction appears at the end of the blade
when the winglet is adopted. Therefore, the power efficiency of the blade is improved.
4.3.12.1 Full Rotor Model Simulation of the Winglet in the OED Result
The winglet design process and the result comparison above are all based on the one blade simulation
model. We have mentioned in Section 4.3.5.1 that the one-blade model is a simplification of the
full-scale VAWT model, and it does not consider the wake of the rotor. In order to validate the
practical influence of the proposed winglet on the performance of the turbine, the full-rotor VAWT
model with the proposed winglet amended to the turbine is simulated in this section. The
computational domain and the full-rotor mesh
have been shown in Figure 4.15 and the
accuracy of this full-rotor VAWT model has been
validated as shown in Figure 4.17.
It is found that amended winglet increases the
height of the blade by about 14.6% and therefore
this changes the swept area of the rotor. In
calculating the aerodynamic coefficients of the
turbine, the modified swept area of the model is
shown in Figure 4.24 and is bounded by the
blade profile trace line. Figure 4.26 is the
numerical results based on the full-rotor model
and the simulations by considering the blade-
blade interaction and blade-shaft interaction,
thus making it more in line with the actual
situation. Figure 4.24 An illustration of the swept area of
the turbine with the selected winglet
126
Figure 4.26 Comparison Between the Torque Coefficient for Different Spanwise Locations at a TSR of
2.29 with and without the Proposed Winglet.
More enhancement is achieved near the

blade tip because the pressure transection
in the blade tip has more severe impact on
the blade surface close to the tip. The
winglet separates the pressure side and
the suction side of the blade in the tip, and
the tip vortex is weakened. Therefore, the
pressure distribution difference on the
blade along the span wise direction is
decreased and the closer to the blade’s tip,
the more obvious the improvement
appears. In addition, Figure 4.26 also
shows that the performance improvement
appears mainly at the upstream stage
(when the azimuthal angle ranges from 0°
to 180°) of the rotor and the difference
among the torque coefficients at the Figure 4.25 The influence of the winglet on the turbine
power coefficient at different TSRs.
downstream stage is not obvious.
Figure 4.25 shows the enhancement of
the overall power coefficient due to the use of the winglet for three different TSRs. The relative power
coefficient enhancements of the blade with a winglet are about 10.5%, 6.7% and 10.0% for TSRs of
1.85, 2.29 and 2.52, respectively. This means that although the winglet is designed under the
condition that the TSR is 2.29, it can improve the VAWT’s power coefficient under other TSR
127
conditions. As a result, adding the designed winglet on the blade tip is a good choice to enhance the
overall performance of the VAWT.
4.3.13 Discussion of the Results
Here, we have used six parameters for the design variables in defining the winglet’s configuration.
Although this paper is not designed to optimize the winglet configuration, the adopted OED approach
discovered the most influential design parameter and obtained a winglet with a recommended
parameter arrangement. Since the winglet is a 3D configuration, it requires many parameters in the
design process. The results in the OED analysis show that the twist angle is the most influential
parameter so that special attention should be paid on the twist angle in the primary optimization of
the winglet configuration.
The designed winglet has been proved to be beneficial for both the one blade model and the full-scale
VAWT rotor. According to the full scale rotor simulation results, the winglet enhances the rotor’s
power coefficient by 10.5%, 6.7% and 10.0% for TSRs of 1.85, 2.29 and 2.52, respectively. Although
the design process is implemented in the condition that the TSR is 2.29, the obtained winglet
improves the power coefficients at other TSR conditions. This means that the turbine’s efficiency is
enhanced in different free flow conditions by installing the winglet on the tip.
The development of the tip vortex can directly reflect the pressure interaction at the blade tip.
Through a flow contour comparison, we find that the winglet enhances the turbine’s power efficiency
by separating the pressure side and the suction side of the blade and maintaining the pressure
difference on the blade. The power performance comparison at different locations along the span
wise direction of the blade reflects that the performance enhancement is increasingly obvious as it
approaches the blade tip.
On adding the designed winglet in this work onto the blade tip will increase the blade volume by
11.7% and of course this will increase the cost of the blade. The real cost increment will depend on
the manufacturing technique employed. For example, the 3D print technique can be used to produce
winglets for small-scale turbines, and for large-scale turbines, the manufacturing of the winglet can
refer to the turbine blade manufacturing technique.
These technologies are very mature now and will not increase the cost too much. When considering
all the other parts of a VAWT system then the percentage increase in the cost will be very small. In
addition, the winglet can be designed and manufactured separately from the blade and be appended
on the blade tip during installation. This makes it easy to produce, maintain and replace.
Further, winglets have already been very successfully used on the tip of aircraft wings and the blades
of some HAWTs. This means that there are available techniques to consult in manufacturing and
installing the winglet onto the blade tip of the VAWT. Since the blade of the VAWT is even simpler
than the aircraft wing and the working environment is less harsh than that of the aircraft, it is feasible
to use a winglet.
4.3.14 Conclusion
This paper numerically studies the configuration of the winglet on the power performance of VAWTs.
Based on the OED approach, the significance of the design variables have been investigated, and the
optimal variable level arrangement is obtained in order to generate an optimal configuration to
compare with the non-winglet blade. From the results presented in this paper, we draw the following
conclusions:
➢ The twist angle is the most significant design parameter in the current parametric approach
with an F value of 17.75 while the F values for the other design variables are all smaller than
F0.05. It is seen that the power performance of the winglet increases with the absolute value
of negative twist angle.
➢ The winglet with the optimal design variable arrangement can decrease the magnitude of the
tip vortex of the blade by maintaining the pressure difference between two sides of the blade
128
and thus improve the power coefficient. The power output of the VAWT, as well as the power
improvement by the winglet, mainly occurs in the upstream stage because the blade-blade
interaction and the blade shaft interaction of the VAWT make the downstream performance
incomparable with the upstream stage.
➢ The improvement in the aerodynamic performance of VAWTs by the winglet becomes
increasingly more obvious when it gets closer to the blade tip along the span wise direction.
The specific improvement of the power coefficient increases from 1.3% to 16% from the mid-
span profile to the profile at z=0.8*H/2. This is because the pressure interaction at the blade
tip region has an influence mainly on the near tip region and adding the winglet onto the
blade tip can reduce the pressure interaction.
➢ An improvement in efficiency of about 6.7% could be achieved by the designed winglet at the
optimal TSR while up to 10% of improvement could be achieved in off-design TSRs. Thus
adding the winglet onto the blade’s tip is beneficial for enhancing the efficiency of the VAWT
in different wind conditions.
4.3.15 References
[1] Meng H, Wang M, Olumayegun O, Luo X, Liu X. Process design, operation and economic evaluation
of compressed air energy storage (CAES) for wind power through modelling and simulation. Renew
Energy 2019.
[2] Ghasemian M, Ashrafi ZN, Sedaghat A. A review on computational fluid dynamic simulation
techniques for Darrieus vertical axis wind turbines. Energy Convers Manage 2017;149:87–100.
[3] MacPhee D, Beyene A. Recent advances in rotor design of vertical axis wind turbines. Wind Eng.
2012.
[4] Aslam Bhutta MM, Hayat N, Farooq AU, Ali Z, Jamil SR, Hussain Z. Vertical axis wind turbine – a
review of various configurations and design techniques. Renew Sustain Energy Rev 2012;16:1926–39.
[5] Maître T, Amet E, Pellone C. Modeling of the flow in a Darrieus water turbine: wall grid refinement
analysis and comparison with experiments. Renew Energy 2013.
[6] Tummala A, Velamati RK, Sinha DK, Indraja V, Krishna VH. A review on small scale wind turbines.
Renew Sustain Energy Rev 2016.
[7] Li Q, Maeda T, Kamada Y, Murata J, Furukawa K, Yamamoto M. Effect of number of blades on
aerodynamic forces on a straight-bladed Vertical Axis Wind Turbine. Energy 2015;90:784–95.
[8] Li Q, Maeda T, Kamada Y, Murata J, Shimizu K, Ogasawara T, et al. Effect of solidity on aerodynamic
forces around straight-bladed vertical axis wind turbine by wind tunnel experiments (depending on
number of blades). Renew Energy 2016.
[9] Li Q, Maeda T, Kamada Y, Murata J, Yamamoto M, Ogasawara T, et al. Study on Fig. 22. The influence
of the winglet on the turbine power coefficient at different TSRs. T.-t. Zhang, et al. Energy Conversion
and Management 195 (2019). Also, power performance for straight-bladed vertical axis wind turbine
by field and wind tunnel test. Renew Energy 2016.
[10] Li Q, Maeda T, Kamada Y, Ogasawara T, Nakai A, Kasuya T. Investigation of power performance
and wake on a straight-bladed vertical axis wind turbine with field experiments. Energy 2017.
[11] Yang Y, Guo Z, Song Q, Zhang Y, Li Q. Effect of blade pitch angle on the aerodynamic characteristics
of a straight-bladed vertical axis wind turbine based on experiments and simulations. Energies 2018.
[12] Li QA, Maeda T, Kamada Y, Murata J, Kawabata T, Shimizu K, et al. Wind tunnel and numerical
study of a straight-bladed vertical axis wind turbine in three-dimensional analysis (Part I: For predicting
aerodynamic loads and performance). Energy. 2016.
[13] Lanzafame R, Mauro S, Messina M. 2D CFD modeling of H-Darrieus wind turbines using a
transition turbulence model. Energy Procedia 2014.
[14] Rezaeiha A, Montazeri H, Blocken B. Characterization of aerodynamic performance of vertical
axis wind turbines: impact of operational parameters. Energy Convers Manage 2018.
129
[15] Rezaeiha A, Montazeri H, Blocken B. Towards accurate CFD simulations of vertical axis wind
turbines at different tip speed ratios and solidities: guidelines for azimuthal increment, domain size and
convergence. Energy Convers Manage 2018.
[16] Rezaeiha A, Kalkman I, Blocken B. Effect of pitch angle on power performance and aerodynamics
of a vertical axis wind turbine. Appl Energy 2017.
[17] Li C, Xiao Y, Xu YL, Peng YX, Hu G, Zhu S. Optimization of blade pitch in H-rotor vertical axis wind
turbines through computational fluid dynamics simulations. Appl Energy 2018.
[18] Shukla DL, Mehta AU, Modi KV. Dynamic overset 2D CFD numerical simulation of a small vertical
axis wind turbine. Int J Ambient Energy 2018.
[19] Jafaryar M, Kamrani R, Gorji-Bandpy M, Hatami M, Ganji DD. Numerical optimization of the
asymmetric blades mounted on a vertical axis cross-flow wind turbine. Int Common Heat Mass Transfer
2016.
[20] Thé J, Yu H. A critical review on the simulations of wind turbine aerodynamics focusing on hybrid
RANS-LES methods. Energy 2017.
[21] Hezaveh SH, Bou-Zeid E, Lohry MW, Martinelli L. Simulation and wake analysis of a single vertical
axis wind turbine. Wind Energy 2017.
[22] Wong KH, Chong WT, Poh SC, Shiah YC, Sukiman NL, Wang CT. 3D CFD simulation and parametric
study of a flat plate deflector for vertical axis wind turbine. Renew Energy 2018.
[23] Wong KH, Chong WT, Sukiman NL, Shiah YC, Poh SC, Sopian K, et al. Experimental and simulation
investigation into the effects of a flat plate deflector on vertical axis wind turbine. Energy Convers
Manage 2018;160:109–25.
[24] Hashem I, Mohamed MH. Aerodynamic performance enhancements of H-rotor Darrieus wind
turbine. Energy 2018;142:531–45.
[25] Siddiqui MS, Durrani N, Akhtar I. Quantification of the effects of geometric approximations on the
performance of a vertical axis wind turbine. Renew Energy 2015;74:661–70.
[26] Abdulrahim A, Anik E, Uzol O. Effects of Mie Vanes and tip injection on the performance and wake
characteristics of a HAWT. 34th Wind Energy Symposium. 2016. p. 0519.
[27] Dong Z. Effects of tip vanes on aerodynamic performance of H-Type VAWT. IOP Conference Series:
Earth and Environmental Science. IOP Publishing; 2018. p. 042138.
[28] Islam M, Fartaj A, Carriveau R. Analysis of the design parameters related to a fixed pitch straight-
bladed vertical axis wind turbine. Wind Eng. 2008;32:491–507.
[29] Shimizu Y, Imamura H, Matsumura S, Maeda T, Van Bussel G. Power augmentation of a horizontal
axis wind turbine using a Mie type tip vane: velocity distribution around the tip of a HAWT blade with
and without a Mie type tip vane. J Sol Energy Eng. 1995;117:297–303.
[30] Shimizu Y, Ismaili E, Kamada Y, Maeda T. Power augmentation of a HAWT by Mie type tip vanes,
considering wind tunnel flow visualization, blade-aspect ratios and Reynolds number. Wind Eng.
2003;27:183–94.
[31] Shoutu L, Yin W, Congxin Y, Ye L. Numerical investigation the effect of the different tip vanes on
the loading of an H-VAWT. E3S web of conferences. EDP Sci 2018:03041.
[32] Rohmann-Shaw C. The Effects of Blade Tip Design on the Aerodynamic Performance of the Vertical-
Axis Wind Turbine Blade. Sheffield: Mechanical Engineering. The University of Sheffield; 2016.
[33] Abdelghany ES, Khalil EE, Abdelatif O, Elharriry G. Aircraft winglet design and performance: cant
angle effect. J Robot Mech Eng Res 2016;1:28–34.
[34] Pooladsanj S, Tadifar M. Numerical study of winglet at low Reynolds numbers. Proceedings of the
ASME 2013 Fluids Engineering Summer Meeting, Nevada, USA. 2013.
[35] Elham A, van Tooren MJL. Winglet multi-objective shape optimization. Aero. Sci. Tech. 2014.
[36] Krebs AT, Bramesfeld BDG. Using an optimization process for sailplane winglet design. Aeronaut
J 2016.
[37] Yadav R, Kumar S, Lakshminarayana S, Rathlamwala I, Rathlamwala B. Computational analysis
for random winglet designs on light aircraft. J Adv Res Mech Eng. Technol 2018.
130
[38] Narayan G, John B. Effect of winglets induced tip vortex structure on the performance of subsonic
wings. Aerospace Sci Technol 2016.
[39] Gupta A, Amano RS. CFD analysis of wind turbine blade with winglets. Proceedings of the ASME
2012 International Design Engineering Technical Conferences & Computers and Information in
Engineering Conference. 2012.
[40] Johansen J, Sørensen NN. Aerodynamic Investigation of Winglets on Wind Turbine Blades using
CFD. Roskilde, Denmark: Risø National Laboratory; 2006.
[41] Farhan A, Hassanpour A, Burns A, Motlagh YG. Numerical study of effect of winglet planform and
airfoil on a horizontal axis wind turbine performance. Renew Energy 2019.
[42] Shamsoddin S, Porté-Agel F. A Large-Eddy simulation study of vertical axis wind turbine wakes in
the atmospheric boundary layer. Energies 2016.
[43] Tao J, Sun G, Si J, Wang Z. A robust design for a winglet based on NURBS-FFD method and PSO
algorithm. Aerospace Sci Technol 2017.
[44] Zhang T-T, Wang Z-G, Huang W, Yan L. A review of parametric approaches specific to aerodynamic
design process. Acta Astronaut 2018.
[45] Tang J, Gong G, Su H, Wu F, Herman C. Performance evaluation of a novel method of frost
prevention and retardation for air source heat pumps using the orthogonal experiment design method.
Appl Energy 2016.
[46] Wang Z, Wang Y, Zhuang M. Improvement of the aerodynamic performance of vertical axis wind
turbines with leading-edge serrations and helical blades using CFD and Taguchi method. Energy
Convers Manage 2018.
[47] Liu X, Huang J, Mao Q. Sensitive analysis for the efficiency of a parabolic trough solar collector
based on orthogonal experiment. Int J Photoenergy.
[48] Yan L, Huang W, Zhang T-T, Li H, Yan X-T. Numerical investigation of the nonreacting and reacting
flow fields in a transverse gaseous injection channel with different species. Acta Astronaut 2014.
[49] Ou M, Yan L, Huang W, Zhang T-T. Design exploration of combinational spike and opposing jet
concept in hypersonic flows based on CFD calculation and surrogate model. Acta Astronaut 2019.
[50] Huang W, Yang J, Yan L. Multi-objective design optimization of the transverse gaseous jet in
supersonic flows. Acta Astronaut 2014.
[51] Ma N, Lei H, Han Z, Zhou D, Bao Y, Zhang K, et al. Airfoil optimization to improve power
performance of a high-solidity vertical axis wind turbine at a moderate tip speed ratio. Energy 2018.
[52] Elsakka MM, Ingham DB, Ma L, Pourkashanian M. CFD analysis of the angle of attack for a vertical
axis wind turbine blade. Energy Convers Manage 2019.
[53] Bianchini A, Balduzzi F, Ferrara G, Ferrari L. Virtual incidence effect on rotating airfoils in Darrieus
wind turbines. Energy Convers Manage 2016.
[54] Fluent. ANSYS. 18.2 User’s Guide. Ansys Inc.; 2017.
[55] Roache PJ. Quantification of uncertainty in computational fluid dynamics. Ann. Rev. Fluid Mech
1997.
[56] Huang W, Pourkashanian M, Ma L, Ingham DB, Luo S-B, Wang Z-G. Effect of geometric parameters
on the drag of the cavity flame holder based on the variance analysis method. Aerospace Sci Technol
2012.
[57] Huang W, Wang Z-G, Ingham DB, Ma L, Pourkashanian M. Design exploration for a single
expansion ramp nozzle (SERN) using data mining. Acta Astronaut 2013.
[58] Huang W. Design exploration of three-dimensional transverse jet in a supersonic crossflow based
on data mining and multi-objective design optimization approaches. Int J Hydrogen Energy 2014.
[59] Roy RK. A primer on the Taguchi method, competitive manufacturing series. New York; 1990.
[60] Lei H, Zhou D, Bao Y, Li Y, Han Z. Three-dimensional Improved Delayed Detached Eddy Simulation
of a two-bladed vertical axis wind turbine. Energy Convers Manage 2017. Also, T.-t. Zhang, et al.
Energy Conversion and Management 195 (2019).
131
4.4 Case Study 4 - Flow Characteristics And Dynamic Responses of a Parked Straight
bladed Vertical Axis Wind Turbine (VAWT)
Authors : Limin Kuang | Jie Su | Yaoran Chen | Zhaolong Han | Dai Zhou | Yongsheng Zhao | Zhiyu Jiang
| Yan Bao
Appeared in : Energy Sci Eng. 2019;00:1–17. May 2019.
Source : DOI: 10.1002/ese3.389
With the development of urbanization and the application of renewable energy, wind turbine is
becoming an important approach for wind energy reservation and utilization [Kuang et al.]64. This
study provides a numerical investigation on understanding the surface pressure distribution, flow
characteristics and dynamic responses of a parked straight‐bladed vertical axis wind turbine
(VAWT), which is helpful for its design. Together with the two‐way coupling method between
simulation platforms such as STAR‐CCM+ and ABAQUS, the SST k‐ω turbulence model is used to obtain
the surface pressure and surrounding flow of the VAWT, and the finite element method is used to
obtain the dynamic responses of its structural components.
The results show that the contours of the pressure distribution on the windward surface of the VAWT
are similar even under a few different conditions, and the deformation of the VAWT can lead to
changes in surface pressure; the turbulent flow characteristics and the wake effect become more
obvious as the wind velocity increases; the blades and support arms of the VAWT need to be
reinforced during the design, and the effect of the parked condition on the dynamic responses of the
VAWT can be neglected. The two‐way coupling method as well as the numerical simulation results is
expected to provide references for the design of VAWTs subjected to coming wind action.
4.4.1 Introduction
With the depletion of fossil fuel resources, the energy issue has become a key factor affecting the
sustainable development of a country [1]. Therefore, countries around the world have made the
development of renewable energy as an important strategic target. The renewable energy sources
that researchers mainly concerned with include biomass, solar, geothermal, hydroelectric, and wind
[2]. Among them, wind energy has the advantages of clean and pollution‐free, wide distribution
range and large reserves, hence has been gradually becoming one of the most important choices for
countries to develop and utilize renewable energy. As a wind energy conversion equipment, wind
turbines have been continuously improved and widely used.
According to the position of the rotating shaft, wind turbines can be classified into horizontal axis
wind turbine (HAWT) and vertical axis wind turbine (VAWT) [3,4]. Due to the higher wind energy
conversion efficiency of HAWTs, they have been the focus of researchers for the past few decades
[5,6]. However, with the rapid development of urbanization and relying on unique advantages,
VAWTs have obtained increasing attention recently. Compared with HAWTs, the following
characteristics of VAWTs are found to be more suitable for urban areas where winds are unsteady
and gusty [2,7-10].
1. VAWTs can capture incoming wind from any direction without yaw mechanisms, which reduces
design and manufacturing costs.
2. The inertial and gravity forces of VAWTs remain unchanged, and the load on the wind turbine is
relatively constant, helping to reduce structural vibration and fatigue damage.
3. VAWTs produce less aerodynamic noise and have better performance in turbulent wind
conditions. Vertical axis wind turbines can mainly be divided into Darrieus type, Savonius type, and
Giromill type [11].
64Limin Kuang | Jie Su | Yaoran Chen | Zhaolong Han | Dai Zhou | Yongsheng Zhao | Zhiyu Jiang | Yan Bao, “Flow
Characteristics And Dynamic Responses of a Parked Straight bladed Vertical Axis Wind Turbine”, Energy Sci Eng.
2019;00:1–17. May 2019.
132
Among them, the Darrieus‐type wind turbines can convert the highest amount of energy due to a
higher tip‐speed ratio [12]. The straight‐bladed VAWTs developed from traditional Darrieus type
wind turbines have simple blade shapes and wind wheel structures, which are convenient to design
and install. Besides, for numerical modeling, the regular blade shapes can reduce the workload and
computational cost during the simulation process.
Although VAWTs have broad application prospects in urban areas, structural safety still restricts
their further development. The support towers of VAWTs are usually slender and are susceptible to
wind loads, especially in areas with strong winds. Such a characteristic poses great challenges to the
normal operation of wind turbine systems. In urban areas with dense traffic, damaged wind turbines
can also pose a safety hazard to pedestrians. Therefore, more and more attention has been paid to
the dynamic responses of wind turbines under wind loads, both experimentally and numerically.
Several laboratory vibration tests were conducted to study the dynamic behavior of VAWTs. [Wang
et al] investigated the ambient dynamic responses of a rooftop VAWT and found that the wind
direction has little effect on its vibration amplitude. [Mclaren et al] showed that the vibration of the
VAWT support tower may be influenced by the wind turbine's operation.
Nowadays, the finite element method is widely used to analyze the dynamic responses of VAWTs
numerically [14] [Rebelo et al] used a numerical model of support tower setup by finite element
technique to compare with the measurement results and verified its feasibility. [Avila et al] solved
the vibration equation of a wind turbine tower and got a valuable result. [Li et al] analyzed the
dynamic behaviors of a VAWT and discovered that the influence of gyroscopic moment on the
dynamic characteristics of the support tower during the rotation can be ignored. [Feliciano et al]
demonstrated that the mass imbalance due to the rotational effect of the blades has a minor effect on
the tower deflection through finite element analysis (FEA). However, so far, there are still many
problems that need to be further analyzed and solved. For example, [Avila et al] simplified the wind
action to a concentrated load, which affected the creditability of the computational results. And the
dynamic responses of the wind turbine under different wind loads are rather different, which were
not considered in the research of [Li et al]. In addition, most of the existing researches mainly focus
on the dynamic responses of the wind turbine support tower, while the researches on other
structures such as blades and support arms are relatively few.
However, the damage probability of the blades and support arms is much higher than the support
tower, which deserves more attention [19]. On the other hand, in the above studies, the
characteristics of flow around the wind turbine have not been described, which may have important
influence on dynamic behaviors of the wind turbine structural system.
Computational fluid dynamics (CFD) method can describe the loading process of wind to structure
more accurately and overcome some limitations of wind tunnel tests such as high cost and complex
test environment [20]. It is often utilized in the studies of aerodynamic performance of wind
turbines, however, rarely be used in dynamic responses analysis [21]. Hence, in order to simulate
the flow around the wind turbine and better reflect the wind loads, the two‐way coupling method
combining the CFD prediction and FEA technique is proposed.
The main work of the present study is to analyze the surface pressure distribution, flow
characteristics, and dynamic responses of a straight‐bladed VAWT under different steady wind
conditions, using the two‐way coupling method. Compared with HAWTs, since the wind direction
and operating state have relatively small influence on the vibration of straight‐bladed VAWTs,
[12,18,22] and in order to reduce the workload of the numerical simulation, only the parked
conditions of the VAWT at different rotation angles are considered. The following points make this
study different from previous ones:
(a) the two‐way coupling method is adopted;
(b) the dynamic responses of the whole structural system of the VAWT including the blades and
support arms are calculated;
(c) the variations of the flow around the VAWT under different wind conditions are revealed;
133
(d) the effects of different parked conditions on the surface pressure distribution, flow
characteristics, and dynamic responses of the VAWT are analyzed.
4.4.2 Numerical Simulation Method
Under the commercial CFD platform STAR‐CCM+, the SST k‐ω turbulence model is used to obtain the
surface pressure and surrounding flow of the VAWT in different steady wind conditions and different
parked conditions. The time‐domain dynamic responses of the VAWT are calculated by the finite
element software ABAQUS, using the traction loads extracted from STAR‐CCM+. The two‐way
coupling procedure between the two solvers is reached through an iterative process to obtain a full
solution by means of exchanging results information through the fluid‐structure interface. The
specific simulation methods are shown in the following sections.
4.4.2.1 CFD
The CFD simulation methods of turbulence mainly include direct numerical simulation (DNS), large
eddy simulation (LES), detached eddy simulation (DES), and Reynolds‐averaged Navier‐Stokes
(RANS) method [5]. Compared with other three methods, RANS method takes less computational
cost and can still produce satisfying results. Among RANS models, the SST k‐ω turbulence model
proposed by Menter [23] is found to have good prediction in the simulation of straight bladed VAWTs,
as verified by [Ma et al] Therefore, it is chosen in the present study. In SST k‐ω turbulence model, the
k and ω equations can be expressed as
∂ ∂ ∂ ∂k
(ρk) + (ρkui ) = [ Γk ] + Gk − Yk
∂t ∂xj ∂xj ∂xj
∂ ∂ ∂ ∂ω
(ρω) + (ρωui ) = [ Γω ] + Gω − Yω + D𝜔
∂t ∂xj ∂xj ∂xj
Eq. 4.14
where ρ, k, ω, ui (uj), and t are the density, turbulent kinetic energy, turbulent dissipation rate,
velocity, and time, respectively. Γk, Gk, and Yk are the convection term, production, and effective
diffusion term of k, respectively. Γω, Gω, Yω, and Dω are the convection term, production, effective
diffusion term, and cross convection term of ω, respectively.
4.4.2.2 Dynamic Responses By Finite Element Analysis
A VAWT is like a vertical cantilever structure. When calculating its dynamic responses, the kinetic
equation is given as
[𝐌]{ẍ (t)} + [𝐂]{𝑥̇ (t)} + [𝐊]{𝑥(t)} = {𝐅(t)}

Eq. 4.15
where {x¨(t)}, {ẋ (t)} , and {x(t)} are acceleration, velocity, and displacement vectors, respectively.
[M], [C] , and [K] are mass, damping, and stiffness matrices, respectively. {F(t)} is the wind load
extracted from STAR‐CCM+. Afterward, the implicit dynamic algorithm of Newmark method [24] is
chosen to obtain the dynamic responses of the VAWT in ABAQUS.
4.4.2.3 Two‐Way Coupling Procedure
The algorithm describing the two‐way coupling procedure between STAR‐CCM+ and ABAQUS is
featured in Figure 4.27. First, the surface pressure and surrounding flow of the VAWT are obtained
from STAR‐CCM+ with SST k‐ω turbulence model (step 1). Then, the traction loads are applied to the
finite element model in ABAQUS, to simulate dynamic responses of the VAWT in an iterative way
(steps 2 and 3). Afterward, the displacements are sent back to STAR‐CCM+ to update the mesh
morphing (step 6). The results information is exchanged at frequent intervals until the maximum
134
Figure 4.27 Flowchart illustrating the Two‐Way Coupling Between STAR‐CCM+ and ABAQUS
physical time is reached (step 5). For every circular simulation, the results are stored in step 4 for
further analysis.
4.4.3 Validation of the Two Way
4.4.3.1 Coupling Method
In this section, the oscillating plate model from the research conducted by Gluck et al25 is used to
verify the feasibility and accuracy of the two‐way coupling method. In this research, the dynamic
behaviors of a vertical plate under a steady fluid flow are analyzed by CFD‐CSD method. The sizes
and mechanical properties of the plate are shown in
Table 4.16, and the bottom of the plate is fixed. The Parameter Value
fluid has a density of 1 kg/m3, and its dynamic Length 400 mm
viscosity is 0.2 Pa s. The computational model Width 10 mm
constructed in STAR‐CCM+ and ABAQUS is shown in Height 1000 mm
Figure 4.28. The length, width, and height of the Young's modulus 3.5e9 N/m2
computational domain are 6000 mm, 6000 mm, and Poisson's ratio 0.32
2000 mm, respectively, as shown in Figure 4.28-A. In Density 1200 kg/m3
order to be consistent with the settings of the VAWT
numerical simulation, the unstructured trimmed cells Table 4.16 Oscillating plate model
with varied sizes are adopted in this validation parameters
(Figure 4.28-B,C), and there are roughly 600 000
grids in the whole computational domain. The C3D8R element is used in ABAQUS to generate a total
of 32 000 grids as shown in Figure 4.28-D. The inlet velocity is 10 m/s, and the outlet pressure is
0.0 Pa. The total simulation time is 2.5 seconds, and the constant coupling time step is 0.01 seconds.
The displacement of the free edge of the plate in the x‐direction is selected as the comparison object,
135
Figure 4.28 Computational model: A, computational domain; B, over view of the grids; C, grids around
the plate; D, 3D view of the finite element model
and the specific position is indicated by a red point which is shown in Figure 4.28-A. The numerical
results of the validation are shown in Figure 4.29, and the two‐way coupling simulation results fit
well with those of the literature [25]. The vibration amplitude and frequency of the oscillating plate
are basically coincident under the two computation conditions. Hence, the two‐way coupling method
can accurately predict the dynamic
behaviors of the vertical plate,
which can be used in subsequent
computations of the VAWT.
4.4.4 Numerical Model
To start up a two‐way coupling
simulation, separate numerical
models are prepared in STAR‐CCM+
and ABAQUS, respectively. The
boundary conditions and physics
models are defined in STAR‐CCM+,
and the material properties are set
up in ABAQUS. The specific
numerical models are given in the
following sections.
Figure 4.29 Validation of the two‐way coupling method

between the simulation and literature25
136
4.4.4.1 Wind Turbine Model

A straight‐bladed VAWT is selected as the research object in this study. Its three‐dimensional (3D)
model is built in the structural design software Unigraphics NX as shown in Figure 4.30-A and then
is import into STAR‐CCM+ and ABAQUS for numerical simulation. The major structural components
of this model include blades, support arms, and support tower. The airfoil section of the blades is
NACA0021 airfoil with a chord length of c = 925.6 mm (Figure 4.30-B). The blade span is h = 6300
mm, and the pitching angle is β = 8°. The support tower consists of two shafts with different
diameters. The lower part has a diameter of d1 = 650 mm and a height of H1 = 5400 mm. The upper
part has a diameter of d2 = 700 mm and a height of H2 = 4200 mm. The junction of the two parts is
Figure 4.30 Straight‐bladed VAWT model: A, 3D view; B, top view (the first parked condition, θ = 0°); C,
top view (the second parked condition, θ = 90°); D, top view (the third parked condition, θ = 135°)
137
defined as the variable section (Figure 4.30-A). The support arms are d3 = 200 mm in diameter, and
the diameter of the rotor is D = 7000 mm (Figure 4.30-B).
In this study, the parked conditions of the VAWT at three different rotation angles are investigated.
According to the relative position of the blades and the support tower, and considering the
periodicity during the rotation of the three‐blade wind turbine system, the state when the angle
between blade 1 and y‐axis is θ = 0° (Figure 4.30-B) is taken as the first parked condition, and the
angles of 90° and 135° are set as the second and third parked conditions, respectively (Figure 4.30-
C,D).
4.4.4.2 Computational Fluid Dynamics Model
An appropriate 3D computational domain is constructed in STAR‐CCM+ as shown in Figure 4.31-A.
The first parked condition is used to indicate the specific situation. As the current study considers
the parked conditions of the VAWT, the
computational domain only contains the
stationary zone and a block is set to
improve the accuracy of the simulation.
Considering the wake generation and flow
field blocking rate requirements, the
length, width, and height of the
computational domain are 70000 mm, 30
000 mm, and 20000 mm, respectively, as
shown in Figure 4.31-B,C. The types of
the inlet and outlet are set as velocity inlet
(V1 = 5 m/s, V2 = 10 m/s, and V3 = 20 m/s)
and pressure outlet (P = 0.0 Pa),
respectively. The freestream turbulence
intensity is set as 1%. The surfaces of the
VAWT and the bottom side of the
stationary zone are set as no‐slip walls.
The other three sides of the stationary
zone are set as symmetry walls.
The mesh topology is also completed in
STAR‐CCM+ with automatic meshing
technique, and unstructured trimmed
cells are generated in the computational
domain as shown in Figure 4.33-A. In
order to investigate the flow
characteristics of the VAWT, the mesh
around the VAWT is refined as shown in
Figure 4.33-B. The boundary layer
conditions of the support tower and blade
are shown in Figure 4.33-C,D. The
prismatic boundary layer grids are set Figure 4.31 Computational domain and boundary
along the normal direction of the support conditions: A, 3D view; B, top view; C, front view
tower and blades. The total thickness of
the boundary layer is 0.00294 m with 30
layers and a growth ratio of 1.1. Based on the inlet velocity (5 m/s, 10 m/s and 20 m/s) and the chord
length (0.9256 m), the Re (Reynolds number) of the present study is about 3.09e5 ‐1.23e6. In order to
meet the requirements of the SST k‐ω turbulence model, the thickness of the first prismatic layer grid
is set as 1.79e−5 which ensures y+ < 1.
138
4.4.4.3 Finite Element Model

The finite element model of the VAWT is
built in ABAQUS for two‐way coupling as
shown in Figure 4.32. The bottom of the
VAWT support tower is fixed without
displacement and rotation. Due to the
difficulty of meshing the trailing edge of
the blades and the support arm joints, the
Tet (tetrahedral) element is utilized in the
current study to divide the grid. The
number of grids in the finite element
model is basically consistent with the CFD
model. The material of the support tower
and support arms is steel, and its density
and Young's modulus are set to 7850
kg/m3 and 2.06e11 N/m2, respectively.
The blades are made with aluminum alloy,
and its density and elastic modulus are
taken as 2700 kg/m3 and 7e10 N/m2, Figure 4.32 3D view of the finite element model of the
respectively. The Poisson's ratio of the two VAWT
materials is equal to 0.3.
Figure 4.33 Mesh topology of the computational domain: A, over view of the grids; B, mesh topology
around the VAWT; C, boundary layer grids of the support tower; D, boundary layer grids of the blade
139
4.4.4.4 Mesh Independence Test

The mesh quantity and quality have an important impact on the computational accuracy and
efficiency of the numerical simulation. Therefore, the mesh independence of the numerical model
needs to be tested before the formal calculations. In the present mesh independence test, three types
of mesh topologies are considered and compared, which are called “Coarse,” “Medium,” and “Fine,”
respectively. Since the flow characteristics and wind‐induced dynamic responses of the VAWT are
mainly affected by the mesh quantity around the VAWT, the difference between the three types of
mesh topologies is the grid density along the surfaces of the blades and support structure. The grid
sizes of the blades and support structure are set to 0.1 m, 0.04 m, and 0.02 m in the three types of
mesh topologies, respectively. In the first parked condition, the representative dynamic responses
data of the top section (z = 9.6 m) and variable section (z = 5.4 m) of the VAWT support tower at the
wind velocity of 10 m/s are selected as the comparison objects, including the maximum
displacement, average displacement, and maximum velocity. The specific locations of the two
sections are shown in Figure 4.30-A. Due to lack of experimental data about the wind induced
dynamic responses, the computational results of the “Fine” mesh topology are taken as the reference
for error relative calculation.
Variable Section
Table 4.17 Mesh Independence Test Based on The Relative Errors of Dynamic Responses of The Top
Section
The results are shown in Table 4.17 and Table 4.18. It can be seen that the “Coarse” mesh topology
produces the largest relative error, approximately 20%, and its results are unreliable The
computational results of the “Medium” and “Fine” mesh topologies are relatively close. However, a
larger number of grids do not always mean more accurate results which it may affect the mesh
quality, as verified by [Lei et al] and [Su et al] Therefore, considering the computational cost and
accuracy, the “Medium” mesh topology is selected to accomplish the formal numerical simulation.
4.4.4.5 Time Step Test
Time step test is also an important part to ensure reliability of numerical simulation results.
Therefore, the current study sets three different time steps for comparison to assess the
computational cost and precision, which are 0.05 seconds, 0.01 seconds, and 0.005 seconds,
respectively. As with the mesh independence test, the representative dynamic responses data with
the wind velocity of 10 m/s at critical sections of the VAWT support tower in the first parked
condition are chosen as the comparison objects. The results are shown in Table 4.20 and Table
140
4.19. These are shown that the largest relative error occurs at the time step of 0.05 seconds and the
peak reaches 30%. The results are nearly coincident when the time steps are 0.01 seconds and 0.005
seconds. Combining with the computational cost, the time step 0.01 seconds is selected for all
simulations.
Table 4.20 Time step test based on the relative errors of dynamic responses of the top section
Table 4.19 Time step test based on the relative errors of dynamic responses of the variable section
4.4.4.6 Two‐Way Coupling Simulation Setup

For each numerical model, a set of two‐way coupling simulation parameters are specified. In STAR‐
CCM+, the VAWT makes up the fluid‐structure interface. The pressure and wall shear stress are
transmitted to ABAQUS with a constant coupling time step. In ABAQUS, the nodal displacement is sent
back to STAR‐CCM+ and the increment size of the implicit dynamic solution is consistent with the
constant coupling time step to ensure the smooth progress of the two‐way coupling simulation.
4.4.4.7 Solver Settings
For the simulation in STAR‐CCM+, the second‐order upwind scheme is selected for the discretization
of convection terms and the diffusion terms discretization is accomplished with the second‐order
central‐differencing scheme. The implicit unsteady segregated flow method is used for calculating
discretization equations, and the AMG (algebraic multi‐grid) technique combined with Gauss‐Seidel
iterative method is adopted. Meanwhile, the pressure‐correction approach is utilized to solve the
continuity and momentum equations. The coupling between the pressure‐velocity equation and SST
k‐ω turbulence model is achieved with the SIMPLE algorithm. In the calculation of ABAQUS, the
implicit dynamic algorithm of Newmark method is utilized to solve the kinetic equations of the
VAWT. The coupling time step is set as 0.01 seconds, and 10 iterations are set in one time step, which
can achieve the convergence of the numerical simulation with less computational costs. Due to the
good convergence performance in the above mesh independence test and time step test, the under‐
relaxation factor is consistent with the default setting of STARCCM+, which is taken as 0.8. All
simulations are performed in parallel on a Small‐Scale Server with two Intel (R) Xeon (R) CPUs (E5‐
2630 v3), and the calculation of one time step requires about 260 seconds.
Due to the large computational cost of the two‐way coupling between STAR‐CCM+ and ABAQUS, the
results within 20 seconds are given in this study which can illustrate the surface pressure
distribution, flow characteristics, and dynamic responses of the VAWT. The analysis is mainly for the
first parked condition and the other two conditions are taken as comparison objects.
141
4.4.5.1 Surface Pressure Analysis

This section discusses the surface pressure of the VAWT in three steady wind conditions with V1 = 5
m/s, V2 = 10 m/s, and V3 = 20 m/s, and the effects of different parked conditions on the pressure
distribution of the structure are studied. The pressure distributions on the windward surface of the
VAWT in the first parked condition at 20 seconds are presented in Figure 4.34. The results show
that the contours of the pressure distribution are similar in different wind conditions and the
pressure increases significantly as the wind velocity increases. It can be seen that the contours of the
support tower are basically symmetrical along the z‐axis which gradually change from positive
pressure area to negative pressure area. For blade 2, the larger positive pressure occurs in the mid‐
span parts near the trailing edge and the pressure decreases to 0.000 Pa or negative values at the
other three edges; for blade 3, the larger positive pressure occurs at the two ends of the trailing edge
and a smaller positive pressure occurs at the middle of the trailing edge; for blade 1, the maximum
negative pressures appear at the right edge in all three conditions, which are −38.888 Pa, −175.390
Pa, and −729.210 Pa, respectively.
It is also found from Figure 4.34 that the contours of the pressure distribution on the blade 2 are
relatively regular while they are disorderly on blade 3. This is because blade 2 first interacts with the
incoming flow, which affects the surface pressure of the subsequent structure while disturbing the
flow field. Meanwhile, there are some differences in the pressure distributions between the three
wind conditions. As shown in Figure 4.34-A,B, when the wind velocity is V1 = 5 m/s and pressure
appears in the middle of the two parts of the support tower, which is 16.891 Pa and 61.121 Pa,
respectively. However, the maximum positive pressure only exists in the middle of the lower part of
the support tower when the wind velocity is V3 = 20 m/s (Figure 4.34-C), which is 269.47 Pa.
Besides, among the three wind conditions, when the wind velocity is 10 m/s, the area of the larger
positive pressure on blade 2 is the biggest, while blade 3 is the smallest. The reason can be that the
deformation of the VAWT gradually leads to changes in surface pressure.
Figure 4.34 Contours of the instantaneous pressure distribution on the windward surface of the VAWT
in the first parked condition: A, V1 = 5 m/s; B, V2 = 10 m/s; C, V3 = 20 m/s
Figure 4.35 shows the comparison of the surface pressure of the VAWT in three parked conditions
at the velocity of 10 m/s. It can be seen that under the three parked conditions, the maximum and
minimum values of the surface wind pressure are not much different. For the pressure distribution
on blade 2 in the third parked condition (Figure 4.35-C), the same result is found on blade 3 in the
first parked condition (Figure 4.35-A). However, there are subtle differences in the pressure
142
distribution on the blades. For blade 1 in the second parked condition (Figure 4.35-B), the larger
positive pressure occurs in the mid‐span parts near the leading edge, rather than the trailing edge,
which is different from the phenomenon that occurs in blade 2 under the first parked condition
(Figure 4.35-A).
Figure 4.35 Contours of the instantaneous pressure distribution on the windward surface of the VAWT
in three parked conditions at V2 = 10 m/s: A, the first parked condition, θ = 0°; B, the second parked
condition, θ = 90°; C, the third parked condition, θ = 135°
4.4.5.2 Flow Characteristics Analysis

The flow characteristics of the VAWT at different wind velocities with V1 = 5 m/s, V2 = 10 m/s, and V3
= 20 m/s are discussed in this section. The differences between the flow field in three parked
condition are analyzed. Figure 4.36 shows the vorticity magnitudes of the horizontal section (z =
7.5 m) and vertical section of the VAWT in the first parked condition at 20 seconds. The specific
locations of the two sections are shown in Figure 4.32-A. As shown in Figure 4.36-A,C,F, the
vorticity increases significantly as the height of the support tower increases, and the three‐
dimensional motion characteristics of the turbulent flow become more obvious. It also shows that as
the wind velocity increases, the disturbance range of the wake flow increases continuously which the
same phenomenon is obtained in Figure 4.36-B,E,G, the vorticity near the inner boundary layers of
the support tower increases obviously while the vorticity does not change significantly in other areas
of the flow field, and the vortices at the end of the flow field generate in an asymmetry way. It can be
seen from Figure 4.36-B,E,G that the large vortex is not formed around the blade 1, the flow
attaches to its surface and the flow separation does not appear. It is also clearly seen that the vortex
structures around the blade 2 and blade 3 become very different when the wind velocity increases,
and the vortices substantially disappear, where only two banded vortex structures are formed on the
trailing
edge and leading edge of the airfoil. This is because the flow velocity of the flow field increases
rapidly, and the vortex structure around the blade is taken away as soon as it develops. The specific
process can be seen in Figure 4.36-E,F. When the flow field has not been fully developed, a series
of different magnitudes of vortices around the airfoil can be observed, the Karmen vortex street
phenomenon behind the support tower is obvious, and its shedding vortices interact with the blade
3. However, the above phenomena no longer exist as the flow field develops completely.
Figure 4.37 shows the vorticity magnitudes of the horizontal and vertical sections of the VAWT in
three parked conditions at the velocity of 10 m/s. From Figure 4.37-A,C it can be seen that there is
almost no difference in the flow field between the two conditions. In the second parked condition,
143
the vorticity intensity near the inner boundary layers of the upper part support tower is significantly
reduced due to the blades and support arms interfering with the incoming flow. It is found in Figure
4.37-E that the phenomenon of the Karmen vortex street does not appear behind the support tower
due to the fully developed wake disturbance of blade 1. In addition, the vortex structures around the
airfoil in the second and third parked conditions are more abundant than the first one, indicating that
the slight difference in the position of the blade relative to the incoming flow will result in a large
change in the flow field.
Figure 4.36 Instantaneous contours of vorticity magnitude of the VAWT at different wind velocities: A,
B, V1 = 5 m/s; C, D, E, V2 = 10 m/s; F, G, V3 = 20 m/s
144
4.4.5.3 Dynamic Responses Analysis

In this section, the dynamic responses of the VAWT in three steady wind conditions are discussed.
The effects of different parked conditions on the dynamic responses of the structure are also
analyzed. Besides, the dynamic behaviors of some critical sections of the support tower, blades, and
support arms are compared. The top section (z = 9.6 m) and variable section (z = 5.4 m) of the support
tower, the top section (z = 10.65 m) and bottom section (z = 4.35 m) of the blades, and the mid‐span
part (z = 5.925 m, 9.075 m) of the support arms are selected as the research objects. The specific
locations of the critical sections are shown in Figure 4.30-A. The time‐domain displacement and
velocity responses of the above structural components in the x‐direction (Figure 4.30) under three
parked conditions are calculated at different wind velocities, where V1 = 5 m/s, V2 = 10 m/s and V3 =
20 m/s.
Figure 4.37 Instantaneous contours of vorticity magnitude of the VAWT in three parked conditions at V
2 = 10 m/s: A, D, the first parked condition, θ = 0°; B, E, the second parked condition, θ = 90°; C, F, the
third parked condition, θ = 135°
Figure 4.38 and Figure 4.39 show the displacement and velocity responses of the support tower
in the first parked condition, respectively. The representative dynamic responses data are given in
Table 4.21. It can be seen that in all three wind conditions, the displacements have certain
periodicity and fluctuate around different average values, and the velocity curves also fluctuate
steadily around 0.000 m/s. The similar result for the displacement is obtained in the research of
[Wang et al]. It shows that the support tower reaches a new forced vibration equilibrium state from
the free vibration equilibrium state under wind loads. It is also found in Table 4.21 that the
displacement and velocity at the top section are relatively 2.5 times than those of the variable section,
which is consistent with the theory of structural mechanics. Meanwhile, it is obvious that as the wind
145
velocity increases, the dynamic responses of the support tower increase significantly, affecting the
structural safety of the VAWT.
Table 4.21 Representative dynamic responses data of the support tower in the first parked condition
The dynamic responses of the blades and support arms in the first parked condition at V3 = 20 m/s
are given in [Kuang et al.]65, respectively. It can be seen from the two figures that the displacement
and velocity of the blades and support arms are substantially larger than that of the support tower.
For the blades, the reason can be that the material of the blade is aluminum alloy, and its density and
elastic modulus are rather smaller than the steel. As for the support arms, its diameter is small,
resulting in less rigidity.
Figure 4.38 Displacement responses in the x‐direction at critical sections of the support tower in the
first parked condition: A, D, V1 = 5 m/s; B, E, V2 = 10 m/s; C, F, V3 = 20 m/s
65Limin Kuang | Jie Su | Yaoran Chen | Zhaolong Han | Dai Zhou | Yongsheng Zhao | Zhiyu Jiang | Yan Bao, “Flow
2019;00:1–17. May 2019.
146
Besides, the contact area between the two components is small, and the structural performance is
not ideal. These indicate that the blades and support arms deserve more attention whether its
deformation meets the regular service requirements of the VAWT. (See Figure 14 [Kuang et al.]66).
shows that the dynamic responses of the blade top section are more obvious than that of the bottom
section and blade 1 vibrates most intensely. It is also found in (see Figure 15 )67 that the deformation
of the upper arm is larger than the lower arm, and arm 1 has the most severe vibration. Combined
with the flow characteristics and surface pressure distribution of the VAWT in the first parked
condition, the reason can be that blade 1 and arm 1 first interact with the incoming flow and disturb
the flow field, so that the surface pressure of blade 3 and arm 3 is reduced, resulting in a decrease in
the dynamic responses.
Figure 4.39 Velocity responses in the x‐direction at critical sections of the support tower in the first
parked condition: A, D, V1 = 5 m/s; B, E, V2 = 10 m/s; C, F, V3 = 20 m/s
Table 4.22 shows the comparison of the dynamic responses of the VAWT support tower under three
parked conditions at the velocity of 10 m/s. It can be seen that the average displacement and
maximum velocity of the support tower are smallest in the first parked condition. For the third
parked condition, the average displacement of the support tower is significantly larger than that of
the first and second conditions. However, considering that the deformations of the support tower in
the three conditions are much smaller than the limiting value under regular service, the effect of the
parked condition on the dynamic responses of the VAWT can be neglected.
66 Limin Kuang | Jie Su | Yaoran Chen | Zhaolong Han | Dai Zhou | Yongsheng Zhao | Zhiyu Jiang | Yan Bao, “Flow
2019;00:1–17. May 2019.
67 Same Source as abone.
147
Table 4.22 Comparison of the representative dynamic responses data of the support tower in three
parked conditions under V2 = 10 m/s
4.4.6 Conclusions
This study mainly investigates the surface pressure distribution, flow characteristics, and dynamic
responses of a straight‐bladed VAWT in different steady wind conditions and different parked
conditions by combining CFD and FEA techniques. Although the computational cost of the two‐way
coupling between STAR‐CCM+ and ABAQUS is high, it can capture the deformation and surrounding
flow of the VAWT simultaneously and accurately. The main conclusions are as follows:
1. The contours of the pressure distribution on the windward surface of the VAWT are similar under
a few different conditions and the pressure increases significantly as the wind velocity increases. The
distribution of the surface pressure on blades shows a certain regularity. The interaction between
the blade and flows affects the pressure distribution of the subsequent structure. The deformation
of the VAWT leads to changes in the maximum positive pressure distribution on the support tower.
The maximum and minimum values of the surface wind pressure are not much different in different
parked conditions.
2. The three‐dimensional motion characteristics of the turbulent flow and the turbulent effect of the
wake flow become more obvious as the wind velocity increases. Meanwhile, the vortex structures of
the airfoils change significantly which may have an impact on the dynamic responses of the VAWT.
The slight difference in the position of the blade relative to the incoming flow will result in a large
change in the flow field.
3. The VAWT works normally at a certain position under wind action, accompanied by the vibration.
Due to the large displacement and velocity, the blades and support arms deserve more attention
during the design to ensure the structural safety of the VAWT. The effects of the parked condition on
the dynamic responses of the VAWT can be neglected.
4.4.7 References
1. Tummala A, Velamati RK, Sinha DK, Indraja V, Krishna VH. A review on small scale wind turbines.
Renew Sustain Energy Rev. 2016;56:1351‐1371.
2. Bhutta M, Hayat N, Farooq AU, Ali Z, Jamil SR, Hussain Z. Vertical axis wind turbine‐a review of
various configurations and design techniques. Renew Sustain Energy Rev. 2012;16:1926‐1939.
3. Lei H, Zhou D, Bao Y, Li Y, Han Z. Three‐dimensional improved delayed detached eddy simulation of
a two‐bladed vertical axis wind turbine. Energy Convers Manage. 2017;133:235‐248.
4. Ismail MF, Vijayaraghavan K. The effects of airfoil profile modification on a vertical axis wind turbine
performance. Energy. 2015;80:20‐31.
5. Ma N, Lei H, Han Z, et al. Airfoil optimization to improve power performance of a high‐solidity
vertical axis wind turbine at a moderate tip speed ratio. Energy. 2018;150:236‐252.
148
6. Jacob J, Chatterjee D. Design methodology of hybrid turbine towards better extraction of wind
energy. Renewable Energy. 2019;131:625‐643.
7. Lin J, Leung L, Xu Y, Zhan S, Zhu S. Field measurement, model updating, and response prediction of
a large‐scale straight‐bladed vertical axis wind turbine structure. Measurement. 2018;130:57‐70.
8. Nguyen L, Metzger M. Optimization of a vertical axis wind turbine for application in an
urban/suburban area. J Renew Sustain Energy. 2017;9(4):043302.
9. Tjiu W, Marnoto T, Mat S, Ruslan MH, Sopian K. Darrieus vertical axis wind turbine for power
generation Ⅱ: challenges in HAWT and the opportunity of multi‐megawatt Darrieus VAWT
development. Renewable Energy. 2015;75:560‐571.
10. Lei H, Zhou D, Bao Y, Chen C, Ma N, Han Z. Numerical simulations of the unsteady aerodynamics of
a floating vertical axis wind turbine in surge motion. Energy. 2017;127:1‐17.
11. Li Q, Maeda T, Kamada Y, Murata J, Furukawa K, Yamamoto M. Effect of number of blades on
aerodynamic forces on a straight bladed Vertical Axis Wind Turbine. Energy. 2015;90:784‐795.
12. Wang Y, Lu W, Dai K, Yuan M, Chen S. Dynamic study of a rooftop vertical axis wind turbine tower
based on an automated vibration data processing algorithm. Energies. 2018;11(11):3135.
13. Mclaren K, Tullis S, Ziada S. Measurement of high solidity vertical axis wind turbine aerodynamic
loads under high vibration response conditions. J Fluids Struct. 2012;32:12‐26.
14. Mabrouk IB, Hami EA. Dynamic response analysis of Darrieus wind turbine geared transmission
system with unsteady wind inflow. Renewable Energy. 2018;43:482‐493.
15. Rebelo C, Veljkovic M, da Silva LS, Simoes R, Henriques J. Structural monitoring of a wind turbine
steel tower‐Part I: system description and calibration. Wind Struct. 2012;15:285‐299.
16. Avila SM, Shzu M, Pereira WM, Morais M, Prado Z. Numerical modeling of the dynamic behavior
of a wind turbine tower. J Vib Eng Technol. 2016;4(3):249‐257.
17. Li C, Yu A, Li J, Xiang B, Peec-Org C. Dynamic analysis of vertical‐axis wind turbines under the
rotation. Power and Energy Engineering Conference. 2010;2010:634‐638.
18. Feliciano J, Cortina G, Spear A, Calaf M. Generalized analytical displacement model for wind
turbine towers under aerodynamic loading. J Wind Eng Ind Aerodyn. 2018;176:120‐130.
19. Shi F, Wang Z, Zhang J, Gong Z, Guo L. Influences of wind and rotating speed on the fluid‐structure
interaction vibration for the offshore wind turbine blade. J Vibroeng. 2019;21(2):483‐497.
20. Sorensen JN. Aerodynamic aspects of wind energy conversion. Annual Rev Fluid Mech. 2011.
21. Posa A, Parker CM, Leftwich MC, Balaras E. Wake structure of a single vertical axis wind turbine.
Int J Heat Fluid Flow. 2016;61:75‐84.
22. Bazilevs Y, Korobenko A, Deng X, Yan J, Kinzel M, Dabiri J O. Fluid‐structure interaction modeling
of vertical‐axis wind turbines. J Appl Mech‐Trans ASME. 2014;81(8):081006.
23. Menter FR. Two‐equation eddy‐viscosity turbulence models for engineering applications. AIAA J.
1994;32(8):1598‐1605.
24 . Newmark NM. A method of computation for structural dynamics. J Eng Mech, ASCE. 1959;85:67‐
25. Gluck M, Breuer M, Durst F, Halfmann A, Rank E. Computation of fluid‐structure interaction on
lightweight structures. J Wind Eng Ind Aerodyn. 2001;89:1351‐1368.
26. Su J, Lei H, Zhou D, et al. Aerodynamic noise assessment for a vertical axis wind turbine using
Improved Delayed Detached Eddy Simulation. Renewable Energy. 2019;141:559‐569.
27. Wang Z, Zhao Y, Li F, Jiang J. Extreme dynamic responses of MW‐level wind turbine tower in the
strong typhoon considering wind‐rain loads. Math Problems Eng. 2013;2013:1‐13.
149
4.5 Case Study 5 - Mesh Sensitivity of Vertical Axis Turbine Wakes For Farm
Simulations
Authors : P.-L. Delafin, S. Guillou, J. Sommeria, T. Maître
Title : Mesh Sensitivity of Vertical Axis Turbine Wakes For Farm Simulations
Appeared in : 24ème Congrès Français de Mécanique , August 2019
Source : https://www.researchgate.net/publication/338920770
Citation : Pierre-Luc Delafin, Sylvain Guillou, J. Sommeria, Thierry Maître. Mesh sensitivity of vertical
axis turbine wakes for farm simulations. 24ème Congrès Français de Mécanique, Aug 2019, Brest,
France. hal-03132153
This work uses 2D URANS simulations of an isolated vertical axis turbine (one single turbine with a
very low blockage ratio) in order to study the mesh sensitivity of the wake prediction. To that end,
four grids are tested and the comparison of the average velocity and turbulence kinetic energy
profiles shows that it is better to use a minimum refinement ∆x = ∆y = D/20, with D the turbine
diameter. Furthermore, the comparison of velocity and turbulence kinetic energy profiles up to 12
diameters downstream the turbine with experimental data shows the limitations of a 2D model if
farms applications are considered.
4.5.1 Introduction
The wind turbine industry is one of the fastest growing renewable energy industries and wind farms
are Now being built offshore with projects of operating floating turbines. Vertical axis turbines (VAT)
present some advantages over the classical horizontal axis turbines when considering floating
applications. They do not need any yawing system since they are insensitive to the wind direction
and the drive train and generator may be placed close to the sea surface which makes maintenance
easier. Also, the lower altitude of the drive train components compared to a horizontal axis
configuration increases the stability of the turbine which in turn reduces the cost of the floating
platform. These advantages are also valid for tidal or river flows applications.
Within a farm, turbines interact with each other and it is therefore important to have a good
understanding of the wake and its recovery. The wake of horizontal axis turbines has already been
studied a lot, both experimentally and numerically. However, the wake of VAT has started to be
studied more recently. Especially, studies covering both near and far wakes have been carried out
recently [1, 2, 3]. Simulations of VAT generally base their grid convergence study on the variation of
the coefficient of power of the rotor. When focusing on the turbine wake, this criteria is not sufficient
because the grid resolution may change significantly between the rotor region and the far wake
region. Furthermore, the far wake needs more time than the coefficient of power of the turbine to
reach a stationary (or periodic) state.
In this work, we use 2D URANS simulations to study the sensitivity of a VAT wake to the mesh
resolution. Both near and far wakes are addressed (up to 12 diameters downstream the rotor) and
simulations results are compared to experimental data recently published [2]. The aim is to provide
guidelines regarding the minimum mesh resolution to use for an efficient wake description and to
give information about the level of accuracy that can be expected from a 2D simulation, when
compared to results of an actual 3D turbine.
4.5.2 Methods
4.5.2.1 Turbine And Operating Conditions
The turbine used in this study is the same as the one tested in [2] (Figure 4.40 a). It is a 3-bladed
vertical axis turbine with straight blades. The turbine diameter is D = 0.175 m and the blade span is
l = 0.175 m which gives a rotor swept area S = 0.0306 m2. Blades are NACA 0018 sections whose mean
lines are projected on the circle along which they travel (Figure 4.40 b). The chord length c = 0.032
150
m is constant over the span. The turbine solidity, as defined in [1], is therefore σ = nc/πD = 0.175,
with n the number of blades.
The freestream water speed is U∞ = 2.3 m/s as in the experiments and water has the standard
characteristics: ρ = 998.2 kg/m3 and υ = 1.0048 x 10-6 m2/s. For most of the work presented here,
the turbine is operating at a tip speed ratio (TSR = ωR/U∞, with ω the turbine rotational speed and R
the turbine radius) equal to 2 which corresponds to its optimal operating condition. The Reynolds
number based on the relative flow speed (W) and the chord length (c), Rec = Wc/ , varies between 7
x104 and 2.2 x105 over a revolution.
Figure 4.40 Picture of the experimental turbine (a) and schematic of a horizontal section of the rotor
showing the 3 blades and the boundaries (interfaces) with the two other sub-domains (b).
The experiments were run in a water tunnel with a width w = 0.6 m and a height h = 0.55 m. The
blockage ratio is then ∊ = S/(w x h) = 9.3%. The turbulence intensity at the turbine location was
measured at 5.5% [2].
4.5.2.2 Numerical Model
4.5.2.2.1 Mesh
The 2D computational domain is
meshed with ICEM CFD, following a
multi-block structured approach
(Figure 4.41) and is divided into
three sub-domains (Figure 4.40
b):
➢ An outer sub-domain
extending 15D upstream,
40D downstream and 30D
on each side.
➢ A rotating ring containing Figure 4.41 Close view of the mesh in the rotor region
the three blades and
151
extending from r = 0.7D to r = 1.3D, with r the

radius from the turbine center. (102 x 103 cells)
➢ An inner sub-domain (with no shaft). (25 x 103 cells)
The boundaries of the outer domain are located away from the turbine to limit their effect on the
solution as much as possible. The blockage ratio is here: ∊ = 1.7%, which is lower than the
experimental value. It was deemed better to study a purely isolated turbine configuration
considering that it would have been difficult to reproduce in two-dimensions the blockage ratio of
the experiments (3D).
Each blade is represented by 314 mesh nodes and the size of the first row of cells, perpendicular to
the blade surface, is 10-5 m
at the leading edge and 3 x
10-5 m at the trailing edge so
that y+ max < 5 at any time
during a revolution. A time
step corresponding to a
variation of the azimuthal
angle of the turbine ∆ϴ = 1◦
is used in all cases. Although
not shown here, the
sensitivity to y+ and time
step have been studied at
TSR = 2 and showed
convergence for the values
aforementioned. The
computational domain size
and the number of mesh
nodes on each blade are
beyond the recommended
values [4, 5, 6].
Figure 4.42 D10 mesh around the turbine and in the near wake.
To study the sensitivity of Water flows from left to right.
the turbine wake to mesh
resolution, four outer sub-
domain grids are tested (Table 4.23). The strategy adopted is to have a constant grid spacing (∆x =
∆y) downstream of the turbine (Figure 4.42) and until the outlet boundary. Therefore, within the
O-grid surrounding the rotor
sub-domain, the cells’ size
increases smoothly from the
small cells observed in the rotor
(Figure 4.41) to the target cells’
size described in Table 4.23. The
region where this strategy is
applied covers 4D in the y
direction (Figure 4.42). We
made an exception to this method Table 4.23 Details of the four mesh resolutions used for the
with the D80 grid for which the outer domain
characteristics ∆x = ∆y is only
valid until 20D downstream of
the turbine. Further downstream, ∆y remains the same but ∆x is loosened in order to save cells.
152
4.5.2.2.2 Turbulence Model and Numerical Procedure

Incompressible Unsteady Reynolds-Averaged Navier-Stokes (URANS) equations are solved using
Open-FOAM v1812. The k - ω SST (Shear Stress Transport) turbulence model [7] is selected to model
the Reynolds stresses. The numerical schemes used in the simulations are summarized in Table 4.24
and the Pimple FOAM solver is used with 20 iterations per time step. A case with 35 iterations per
time step was run and showed no difference with the corresponding case using 20 iterations per time
step.
A constant velocity Ux =
U∞ = 2.3 m/s is defined
at the inlet boundary
together with a
turbulence intensity
equal to 5% (k = 0.0199
m2/s-2). At the outlet
boundary, a constant
pressure P = 0 Pa is
imposed. Blades are
considered as walls with
a movingWallVelocity
type for U, a Table 4.24 Main numerical schemes used for every simulation in this
zeroGradient condition study
on the pressure, a
kqRWallFunction on k
and the omegaWallFunction on ω. Note that due to the fine mesh used close to the blades, the
boundary layers are resolved and the wall function treatment is not used. The side boundaries of the
outer domain are considered as symmetryPlanes. Finally, the interfaces between sub-domains are set
as cyclicAMI boundaries.
Figure 4.43 Comparison of measured and simulated coefficients of power (left) and coefficients of
thrust (right).

4.5.3.1 Turbine Performance
Before focusing on the turbine wake, we run simulations at TSR = 1.5, 2 and 2.5 with the grid D20 in
153
order to compare the turbine performance (coefficient of power, CP = CΩ/0.5ρSU3∞ and coefficient
of thrust CT = Fx/0.5ρSU2∞ , with C the torque generated by the rotor, the rotor rotational speed, and
Fx the rotor thrust force) with experimental data [2]. Figure 4.43 shows that the URANS 2D
simulations overpredict the coefficient of power but follow the same trend as the measurements.
Especially, the best operating point is found to be TSR = 2 with both 2D simulations and experiments.
The higher coefficients of power predicted by the 2D simulations can easily be explained by the
absence of 3D effects like tip vortices, arm/blade junction losses and the interaction with the wake
of the main shaft.
The simulations results shown in Figure 4.43 are average values of CP and CT calculated over the
10th revolution (TSR = 1.5) and over the 15th revolution (TSR = 2 and 2.5). The convergence criteria
used in these cases was a decrease in the average coefficient of power by less than 1% between two
consecutive revolutions. This convergence criteria is often used even though it is a little coarse. If
those simulations had been run for a longer time (e.g. for 80 revolutions as in the next section), the
coefficient of power would decrease further by a few percent’s. Figure 4.43 (right) shows that the
2D simulations underpredict the thrust coefficient. Similar explanations as with the CP values can be
made plus the fact that the thrust measured in the experiments includes a contribution from the arms
and the shaft connecting the turbine blades to the generator. The arms and the shaft are not taken
into account in the 2D simulations.
4.5.3.2 Mesh Sensitivity of The Turbine Wake
This section focuses on the sensitivity of the simulation at TSR = 2 to the mesh resolution in the
turbine wake. The four grids presented in Table 4.23 are used to this end. Before looking at the
wake, Table 4.25 shows that the resolution of the mesh in the outer sub-domain does not affect
significantly the prediction of both CP and
CT after 80 rotor revolutions. Especially,
grid D20 leads to only 0.5% and 0.3%
increase in CP and CT , respectively,
compared to the finest grid tested, D80. It
should be noted that when increasing the
mesh resolution in the outer sub-domain,
cells are refined both upstream and
downstream of the rotor (Figure 4.42).
Table 4.25 Coefficients of Power (CP) and Thrust (CT )
It is therefore difficult to determine
Obtained with Different Grid Resolutions in the Turbine
whether the difference in CP and CT Wake. Values are Averaged Over the 80th Revolution.
comes from the better resolved wake or
upstream flow.
Figure 4.44 presents a qualitative analysis of the flow field in the near wake of the turbine (up to 6D
downstream). Simulations using grids D20 to D80 show very similar flow patterns although the
resolution of small flow structures (e.g. just downstream the turbine) improves with the number of
cells. Grid D10 leads to a flow pattern in the near wake which is similar to the other grids. However,
6D downstream the turbine, where the wake seems to transition to a bluff-body regime, as described
in [1], the grid D10 leads to a different behavior (or induces at least a phase shift in the wake
oscillation).
A comparison of both average streamwise velocity and average turbulence kinetic energy profiles
obtained with the grids D10, D20, D40 and D80 is presented in Figure 4.45. To produce this figure,
simulations have been run for 80 turbine revolutions and results averaged over the last 20
revolutions (60 to 80). In the far wake, 20 turbine revolutions correspond to 5 periods of velocity
fluctuation (which corresponds to a Strouhal number St = fD/U∞= 0.16).
Figure 4.45 shows that grids D20, D40 and D80 lead to almost exactly the same velocity profiles
whatever the distance downstream the turbine. However, grid D10 leads to noticeable, even if still
154
small, differences in the velocity profiles. This is especially true at x = 8D where D10 predicts a slightly
faster wake recovery than the other two grids. The turbulence kinetic energy (k) profiles are a little
more scattered than the velocity profiles. Especially at x = 2D, it is clear that the finer the mesh, the
larger the peaks of k. Further downstream, D20, D40 and D80 curves superimpose almost perfectly
while the D10 curves show some significant differences with the two others.
The results presented in Table 4.25, Figure 4.44 and Figure 4.45 show that although the mesh
resolution in the outer sub-domain does not affect significantly the turbine performance and the
prediction of its wake, it is better to use a resolution equivalent to ∆x = ∆y = D/20 or finer in order to
obtain a mesh insensitive solution. Lower mesh resolutions should be avoided.
Figure 4.44 Instantaneous streamwise velocity (left) and turbulence kinetic energy (right) fields after
40 revolutions for grids D10 to D80. The line plotted on the right side of the figures is located at x = 6D.
4.5.3.3 Comparison to Experimental Data

In this section, we compare the simulation results obtained with gridD40 to measurements [2].
Figure 4.46 a emphasizes the great difference between 2D simulations and actual 3D experiments.
155
Figure 4.45 Comparison of non-dimensioned streamwise velocity profiles (a) and turbulence kinetic
energy profiles (b) for grids D10, D20, D40 and D80 at five locations downstream of the turbine
The 2D simulation predicts a wider wake at x = 2D and significantly underpredicts the wake recovery
from x = 4D to x = 12D. The agreement between the 2D simulation and the measurements at x = 12D
is better than at lower x values but a 0.15 difference in non-dimensioned streamwise velocity
Figure 4.46 Comparison of measured and simulated (with gridD40) non-dimensioned streamwise
velocity profiles (a) and turbulence kinetic energy profiles (b) at five locations downstream the turbine.
156
remains. Figure 4.46 b shows that the 2D simulation underpredicts k at x = 2D and x = 4D. The peak
values are predicted further away from the turbine center than in the experiment. At x = 6D and
beyond, the 2D simulation significantly overpredicts k and the recovery is longer than in the
experiment, as already observed on the streamwise velocity.
Note that due to the long distance between the upstream boundary and the turbine, the ambient
turbulence intensity around the rotor is not 5% as set at the inlet. Instead, this value has decreased
to almost 0. This is explained by the absence of velocity gradient upstream of the turbine which
results in only the destruction term to be active in the k equation of the turbulence model. This is not
considered to have a significant effect on the turbine performance but it could affect the wake
development and mixing. In particular, it can explain why the minimum value of k at x = 2D is very
close to zero. However, some studies have shown that the wake of a vertical axis turbine is not much
affected by the ambient turbulence. Therefore, the main explanation for the differences observed in
this work is assumed to be the 2D versus 3D approach.
4.5.4 Conclusions
2D URANS simulations of a vertical axis tidal turbine have been run with OpenFOAM using four grids
with different mesh resolutions in the wake. The simulations have been compared between
themselves in order to find the minimum mesh resolution in the wake above which further
refinements do not improve significantly the solution. It was found that a mesh resolution equal to
∆x = ∆y = D/20 is enough to capture accurately the average velocity profiles in the wake of the
turbine. This guideline, based on 2D simulations, is believed to be valid for 3D simulations as well.
The 2D simulation results have also been compared to already published experimental data [2]. The
comparison gives valuable information about the differences between 2D simulations and 3D
experiments and the consequences on the wake prediction. It was found that the 2D simulations
significantly underpredict the wake recovery. This is of particular interest as it means that 2D
simulations should not be used to study farms of high solidity turbines. 3D simulations using the
same configuration as the one used in this study are planned in the near future to confirm this
observation.
4.5.5 References
[1] D.B. Araya, T. Colonius, and J.O. Dabiri, Transition to bluff-body dynamics in the wake of vertical
axis wind turbines, Journal of Fluid Mechanics, 813 (2017) 346-381.
[2] V. Clary, T. Oudart, T. Maître et al., A simple 3D river/tidal turbine model for farm computation -
Comparison with experiments, Sixth International Conference on Estuaries and Coasts (ICEC), Caen,
France (2018).
[3] H.F. Lam, H.Y. Peng, Study of wake characteristics of a vertical axis wind turbine by two- and
three-dimensional computational fluid dynamics, Renewable Energy, 90 (2016) 386-398.
[4] A. Rezaeiha, I. Kalkman, B. Blocken, CFD simulation of a vertical axis wind turbine operating at a
moderate tip speed ratio: Guidelines for minimum domain size and azimuthal increment, Renewable
Energy, 107 (2017) 373-385.
[5] F. Balduzzi, A. Bianchini, R. Maleci, G. Ferrara, L. Ferrari, Critical issues in the CFD simulation of
Darrieus wind turbines, Renewable Energy, 85 (2016) 419-435.
[6] T. Maître, E. Amet, C. Pellone, Modeling of the flow in a Darrieus water turbine: wall grid
refinement analysis and comparison with experiments, Renewable Energy, 51 (2013) 497-512.
[7] F. Menter, Two-equation eddy-viscosity turbulence models for engineering applications, AIAA
Journal, 32 (1994) 1598-1604.
157
4.6 Case Study 6 - Numerical Investigations of the Savonius Turbine with Deformable
Blades
Authors : Krzysztof Sobczak, Damian Obidowski, Piotr Reorowicz and Emil Marchewka
Affiliation : Institute of Turbomachinery, Lodz University of Technology, 90-924 Lodz, Poland;
Title : Numerical Investigations of the Savonius Turbine with Deformable Blades
Original Appearance : Energies , MDPI
Citation : Sobczak K, Obidowski D, Reorowicz P, Marchewka E. Numerical Investigations of the Savonius
Turbine with Deformable Blades. Energies. 2020; 13(14):3717. https://doi.org/10.3390/en13143717
Savonius wind turbines are characterized by various advantages such as simple design,
independence of wind direction, and low noise emission, but they suffer from low efficiency68.
Numerous investigations were carried out to face this problem [Sobczak et al.]. In the present paper,
a new idea of the Savonius turbine with a variable geometry of blades is proposed. Its blades, made
of elastic material, were continuously deformed during the rotor revolution to increase a positive
torque of the advancing blade and to decrease a negative torque of the returning blade. In order to
assess the turbine aerodynamic performance, a two-dimensional numerical model was developed.
The fluid-structure interaction (FSI) method was applied where blade deformations were defined by
computational solid mechanics (CSM) simulations, whereas computational fluid dynamics (CFD)
simulations allowed for transient flow prediction. The influence of the deformation magnitude and
the position of maximally deformed blades with respect to the incoming wind direction were studied.
The aerodynamic performance increased with an increase in the deformation magnitude. The power
coefficient exceeded Cp = 0.30 for the eccentricity magnitude of 10% and reached 0.39 for the highest
magnitude under study. It corresponded to 90% improvement in comparison to Cp = 0.21 in the case
of the fixed-shape Savonius turbine.
Keywords: vertical axis wind turbine (VAWT); Savonius Turbine; Deformable Blades; Power
Coefficient Cp; Blade Load; Fluid-Structure Interaction (FSI).
4.6.1 Introduction
Vertical axis wind turbines (VAWTs) are typically characterized by lower wind energy-conversion
efficiency than commonly used horizontal axis wind turbines (HAWTs). However, they are often
favored in micro power generation due to their simple design, a possibility to locate a generator near
the ground, and to accept the wind blowing from different directions [1,2]. They are quieter and safer
than small-scale HAWTs and, thus, suitable for applications in urbanized areas [3–5].
A vertical axis wind turbine, referred to as the Savonius turbine, was invented by S.J. Savonius [6].
In the top view, it resembles the letter “S” with two typically semi-cylindrical blades, often slightly
overlapping. Its primary advantage lies in simple and, thus, cheap and robust design [7,8]. Similar to
other VAWTs, Savonius turbines are independent of the wind direction. However, contrary to
Darrieus wind turbines, they are characterized by a high starting torque for selected rotor positions.
They are classified as drag-driven turbines and operate at low rotational speeds, with tip speed ratios
not exceeding 2, which makes them safer than HAWTs at strong winds [9]. They perform well at low
wind speeds most often encountered close to the ground and they are characterized by a low level of
noise emission. Thus, Savonius turbines are suitable for application in urbanized areas [3].
Unfortunately, the primary disadvantage of Savonius turbines is their low efficiency.
Typically reported values of the power coefficient for designs with semi-cylindrical blades fall l1
within the range Cp = 0.15–0.20 [7,10,11]. Therefore, this kind of turbine in its basic configuration is
not usually a reasonable alternative when compared to other types of wind turbine. Nonetheless,
owing to their advantages, Savonius turbines were subject to numerous investigations aimed at
68Krzysztof Sobczak, Damian Obidowski, Piotr Reorowicz and Emil Marchewka, “Numerical Investigations of
the Savonius Turbine with Deformable Blades”, Energies , MDPI, 2020.
158
increasing their efficiency. Many works were focused on the search for optimal dimensions of
geometrical parameters of rotors. The influence of the number of blades, overlap ratio, aspect ratio,
end plates and other factors was studied both with experimental and numerical methods. Results of
those investigations were presented in numerous papers, which further were summarized in
thorough reviews [7,12].
Many studies were focused on modifications of the blade shape or an application of additional
elements to direct the air flow towards blades. The replacement of conventional semi-cylindrical
blades by more sophisticated shapes allowed one to increase the Savonius turbine performance
significantly, with maximal values of the power coefficient up to Cp = 0.25–0.30. A substantial part
of the research concentrates on two-dimensional (2D) thin blade configurations, i.e.,: Bach, Benesh,
elliptical and spline [11,13,14]. Airfoil shape blades were studied in [15,16]. Optimization methods
were also applied in order to search for an optimal blade shape in [16–19], taking advantage of 2D
simplifications in numerical simulations. Three-dimensional (3D) blade arrangements with twisted
or helical blades were tested as well. In this case, it was possible to reduce static and dynamic torque
variations for different angular positions of the rotor with respect to the incoming wind, however, no
significant improvement was reported as far as the turbine performance is considered [12,14].
Different augmentation systems can be used to change the wind flow path around and in the Savonius
rotor. Its power output was increased by 20% up to 50% if the turbine rotor was equipped with flat
plate deflectors [18,20,21], v-shaped deflectors [22] or a combination of flat and circular deflectors
[23], shielding the returning blade and reducing its negative moment. A similar effect was achieved
if the wind was directed towards the advancing blade with a curtain-deflector system [24], self-
adjusting conveyor-deflector curtains [15], a system of adjustable shielding plates for twin rotors [9]
or even a rectangular guide-box tunnel surrounding the rotor [25]. A comprehensive summary of
different augmentation systems with the power coefficient exceeding considerably Cp = 0.3 can be
found in [11,26]. However, a disadvantage of such approaches consists in larger dimensions of the
turbine, an increase in the complexity of its geometry and dependence on the wind direction.
An idea of the Savonius turbine with a variable geometry of blades is proposed in order to enlarge
the projected area of the advancing blade (increase the positive moment) and, at the same time, to
diminish the area of the returning blade (decrease the negative moment). Elaborate two-dimensional
(2D) computational fluid dynamics (CFD) simulations were performed to assess the output power
gain for different arrangements of blade deformations.
The deformations of blades in this case were determined with a structural solver and then the
geometry was transferred to a fluid solver for the aerodynamic analysis. This approach is referred to
as the fluid-structure interaction (FSI). The FSI is a very wide concept of solvers coupling in order to
obtain high-fidelity numerical solutions. Solvers can be one-way coupled once the data, i.e., loads
from the fluid acting on the wall, are transferred to the structural solver where stress and strains are
determined [27,28]. Another example of one-way coupling takes place where the deformation of the
structure influences the flow structure and loads determined in the fluid solver. The most advanced
method, called the two-way FSI, requires co-simulation between computational fluid dynamics and
structural mechanics. Both CFD and structural solvers are coupled and synchronized to attain
converged solutions. The two-way FSI is applied to highly dynamical systems as in the case of the
aeroelastic response analysis [29] or whenever the structure is flabby [30]. The FSI strategy is
typically used in horizontal axis wind turbines (HAWTs). It is also used in the analysis of vertical axis
wind turbines (VAWTs), e.g., in the case of an H-rotor [31,32], but no reference reporting an analysis
of the Savonius-based turbine is known to the authors. It is due to much lower blade loads than in
the case of lift-driven HAWTs or VAWTs.
4.6.2 Savonius Rotor with Deformable Blades
An idea of the novel turbine with a variable geometry of blades is shown in Figure 4.47 for
subsequent phases of the rotor revolution [33]. Shapes of the blades made of a flexible material
159
change constantly during its rotation. This is achieved by guiding the outer edges of the blades (tips
marked as dots in the color of the blades) along the guide ring (red) placed eccentrically with respect
to the rotor shaft. Rods attached to the outer edges of the blades can move linearly with respect to
sliders, which are fixed to the shaft. The rode slider mechanisms, marked by a red dashed line, are
applied at the top and bottom of the turbine and they transfer the torque generated by the blades to
the shaft. The inner edges of the blades are also attached to the rotor shaft.
Figure 4.47 Principle Of The Turbine Operation.
The phases 90° and 270° presented in Figure 4.47 illustrate the maximal deformation of the blades.
If the extended, advancing blade is located in such a way that its concave side is exposed to the wind
and, simultaneously, the wind blows at the convex side of the contracted, returning blade, one can
expect that the turbine will be driven with the wind energy more efficiently than in the case of both
blades having the same, fixed shape. This implies that the turbine needs to be properly located with
respect to the incoming wind. Therefore, the guide ring needs to change its position by rotation
around the axis of the base of the turbine, which is coaxial with the rotor shaft axis, marked as a black
X in Figure 4.47. The guide ring mechanism has to be equipped with an aerodynamic or mechanical
160
system in order to adjust its position with respect to the wind direction. The turbine generator is
fixed to the frame and a gear or a transmission has to be applied to transfer the mechanical energy
from the turbine blades. The additional mechanisms make the design of the proposed turbine more
complex than the original Savonius. However, they consume a part of the energy generated by the
turbine. Thus, all elements need to be carefully designed to be resistant and efficient. The present
study focuses on the aerodynamic performance of a novel, efficient design but neither mechanical
losses due to friction
4.6.3 Materials and Methods
The numerical simulation of the Savonius turbine with deformable blades was performed within the
FSI method available in AnsysWorkbench v19.2. The geometry and position of the rotor blades change
continuously and, thus, transient simulations were performed both for a structural analysis with the
ANSYS® solver and a fluid flow analysis with the ANSYS fluent solver. Full 3D simulations of the
turbine were not possible due to enormous requirements of the coupled fluid flow and structural
solvers. Therefore, it was decided to solve the problem as the 2D one, actually quasi-2D, as the
structural solver demanded a three-dimensional model to be applied. Despite its limitations as 3D
effects at the blade ends are disregarded, a 2D approach is a frequently applied simplification, which
allows one to learn about performance of the turbine configuration. It is especially useful to observe
changes in the performance if different configurations of the blades are compared. However, one can
keep in mind that significant differences can be obtained comparing a 2D prediction with the full
three-dimensional one, especially when the aspect ratio of Savonius turbines is low [34 , 35].
The geometry of the blades, the guidance system and the fluid domain were prepared in SolidWorks®.
In general, the guide ring can be of any arbitrary smooth shape, preferably elliptical. However, in this
first study the guide ring of the constant diameter D = 1 m was selected. It was also the value of the
diameter of the reference turbine rotor with fixed-shape blades, where the eccentricity was equal to
zero. The blades of that reference rotor had a semi-circular shape. In the case of the rotor with
eccentricity, the arc length of the blades was the same as in the reference rotor, but the distance
between outer edges of the deformable blades was variable during the rotor revolution. In order to
simplify the numerical model, it was decided to disregard the rotor shaft. Its impact on the flow
around the rotor blades is rather limited and it can be assumed to be similar in all configurations.
The blade overlap can have a positive effect on the turbine performance. However, in those
investigations, it was decided not to overlap the blades.
As one can see in Figure 4.48, the fluid domain was divided into two regions. The internal domain
of the diameter 1.5D including turbine blades was surrounded by the external one. The total length
of the domain was 60D in the flow direction, with the turbine axis located 20D from its inlet and in
the middle of the domain height, which was set to 40D. The domain blockage was similar to our
previous studies [34] or in [18] and it did not affect comparisons between different turbine
configurations. Due to the complexity of blade deformations, combined with their rotation, a tool
outside the fluid flow solver was needed to define the instantaneous rotor geometry. It was decided
to use the structural ANSYS® solver in order to take advantage of the FSI method implemented in the
AnsysWorkbench®.
A one-way system coupling was defined between simulation components, where the deformation of
blades obtained in the structural analysis was transferred to the fluid flow solver. Because the
pressure variation around the blade for the considered wind speed (v = 4 m/s) was less than 100 Pa,
the two-way coupling, where the pressure load on blades would be transferred from the flow to the
structural analysis, was disregarded. The one-way system coupling method was successfully used
and presented in [27].
161
Figure 4.48 Computational domain scheme (external domain dimensions do not correspond to the
scale)
Due to the complexity of blade deformations, combined with their rotation, a tool outside the fluid
flow solver was needed to define the instantaneous rotor geometry. It was decided to use the
structural ANSYS solver in order to take advantage of the FSI method implemented in the Ansys
Workbench. A one-way system coupling was defined between simulation components, where the
deformation of blades obtained in the structural analysis was transferred to the fluid flow solver.
Because the pressure variation around the blade for the considered wind speed (v = 4 m/s) was less
than 100 Pa, the two-way coupling, where the pressure load on blades would be transferred from the
flow to the structural analysis, was disregarded. The one-way system coupling method was
successfully used and presented in [27].
In order to satisfy a good coupling of the structural and flow parts of the problem, the same timestep
of transient simulations was applied to both of them. The value of the timestep was selected on the
basis of solution stability tests performed for the rotors with the highest magnitude of blade
deformation. The typically accepted timestep corresponding to the revolution of the rotor by 1° was
selected initially in the Savonius rotor simulations. However, due to solution instabilities during the
remeshing procedure, it was successively reduced. In the case of the timestep equal to 0.001 s, the
numerical errors resulting from the mesh deformation and the remeshing algorithm were very
limited and the computations were successful. The simulations were carried out for the tip speed
ratio TSR = 0.8, for which Savonius rotors of typical aspect ratios (AR = H/D = 0.8–1.5) reach the
maximal value of the power coefficient Cp [34, 36]. For this TSR, the angular velocity of the turbine
was 6.4 rad/s. Thus, the selected timestep of 0.001 s corresponded to a revolution of the rotor by
approximately 0.367°, resulting in 982 steps per one revolution, which is sufficiently low as far as the
time discretization is concerned [16]. The total time of simulations was 10s, which corresponds to
more than 10 full revolutions, thus eliminating an influence of the initial conditions onto the
simulation results.
162
The tip speed ratio TSR Eq. 4.16 (left) and the power coefficient Cp Eq. 4.16 (right), i.e., the energy
extracted by the turbine to the available wind energy, were defined as follows:
ωR Tω
TSR = , Cp =
v′ 1 3 ′
ρv A
2
Eq. 4.16
Where ω - angular velocity [rad/s], R - turbine radius, v - wind speed [m/s], T - output torque [Nm],
Ρ - air density [kg/m3], A - projected area of the rotor (DH) [m2], D - rotor diameter (2R) [m], H -
rotor height [m].
4.6.3.1 Definition of Structural Simulations
As mentioned above, the structural solver was used to determine the deformation of the blades for
each timestep. One structural model, whose scheme is presented in Figure 4.49, was designed to
perform a series of simulations. Different eccentricity magnitudes and various angular positions of
the eccentricity line with respect to the incoming wind were attained by changing the position of the
guide velocity Inlet Deformable turbine rotor 20D Interface Symmetry 100D(m) ring center (red X)
with respect to the rotor shaft (black X) as indicated by white arrows. That position was fixed in
particular simulations.
Figure 4.49 Turbine Motion Constrains In The Structural Simulation

163
The kinematics of the blades (blue and green arcs) was restricted by several constraints indicated
schematically in Figure 4.49. The turbine shaft rotated around the axis indicated by a black X,
where the blades were tangent. Their outer tips marked with green and blue points were constrained
by frictionless movement along the guide ring (red circle) as indicated by green and blue arrows. The
torque generated by the blades was transferred to the shaft by rode–slider mechanisms. The rode
marked with a purple dashed line was fixed to the turbine shaft. The sliders, marked as blue and
green cylinders, were connected to the blade tips. The rode–slider mechanisms allowed for
frictionless linear movement of the blade outer tips with respect to the axis of rotation, but the
mechanisms kept them in same relative angular position during the turbine rotation. Additionally,
the blade outer tips could rotate without friction with respect to the rods as indicated by yellow
arrows.
The blade was made of structural steel from the standard ANSYS material library, with the Young
modulus and Poisson’s ratio equal to 210 GPa and 0.3, respectively. The material was selected to
limit deformation to the elastic range to ensure stability of the structural solution for selected
eccentricity magnitudes.
4.6.3.2 Definition of Flow Simulations
In comparison to typical simulations of Savonius turbines with blades of fixed shapes (where the
internal domain rotates), deformations of the blades enforced an application of an advanced meshing
approach in that fluid flow analysis. In the case under study, both flow domains are in the stationary
frame of reference.
The computational mesh was
generated in ANSYS Meshing.
Following the mesh
dependence test, described in
the next subsection, a mesh
composed of 1.16 million
control volumes was used in
the simulations. The external
domain did not change during
the computational campaign,
thus, it was meshed with a
single layer of hexahedral
elements through the slice.
Refinements were applied
around the interface with the
internal domain and in the
wake downstream of the
turbine (Figure 4.50).
The internal domain consisted
of two regions marked in
green and in red. Due to the Figure 4.50 Computational mesh in the external and internal
requirements of the ANSYS domains with details of its refinement in the blade vicinity presented
Fluent concerning the in the successive magnifications
dynamic mesh options of
deformation and remeshing, tetrahedral elements were used in those regions. In order to ensure a
high-quality mesh in the region around the blades (marked in red), a highly refined mesh with 12
layers of prismatic elements at the wall was generated. That number of layers was sufficient to satisfy
the condition y+ < 1 of the first mesh element at the wall for almost all the simulations. The value was
exceeded only for very limited regions at the blade tips. Thus, the mesh was sufficient to solve fully
164
the flow in the boundary layer, which is very important as the flow is characterized by numerous
boundary layer separations, especially at convex sides of the blades. Refinement of the mesh around
the blades (especially at their tips) caused the number of elements through its width to be increased
to two and even four elements at the blade tips in this region of the domain. In order to avoid
numerical errors, the sizes of elements at the interface between the outer and inner domains did not
differ significantly.
The problem of rotation of turbine blades with a constant angular velocity as well as their
deformation due to eccentricity was addressed by deforming, remeshing and smoothing algorithms
of the dynamic mesh offered by ANSYS Fluent [37]. The instantaneous blade geometry was
transferred from the structural analysis every timestep. In the region marked in red, the mesh was
following the blade and the elements were deforming as the blade was changing its shape. In the
region marked in green, the continuous motion of the blade caused deformation of mesh elements
and remeshing was launched wherever they degenerated and the quality measures were not
satisfied, namely, the minimal and maximal length scale (1 and 5 mm) with the maximum face
skewness of 0.6 and maximum cell skewness of 0.8 were set. The highest skewness of the grid cells
was in the region of the deformed and re-meshed grid, with the maximal value of 0.84. In the
remaining regions, the skewness was below 0.6. The maximal cell aspect ratio at the blades was equal
to 68 and was almost constant during the solution as the deformation of the prismatic elements was
limited.
The flow simulations were performed in a transient mode with the pressure-based ANSYS Fluent
solver. Reynolds-averaged Navier–Stokes (RANS) equations were calculated with the k-! SST (Shear
Stress Transport) turbulence model of Menter [38], which is one of the most frequently used in
Savonius turbine simulations [16, 39, 40], due to its good performance in the adverse pressure
gradient and separated flows. Despite the fact that changes in air density are negligible, the fluid was
defined as compressible according to the ideal gas law, in order to compensate for numerical
instabilities resulting from mesh deformation and remeshing.
A set of the boundary conditions applied in the simulations is indicated in Figure 4.48. At the inlet,
the velocity of 4 m/s was specified with 1% of the turbulence intensity and the turbulent viscosity
ratio equal to 1. This velocity corresponds to the Reynolds number based on the turbine diameter
equal to 2.5 x 105. The pressure outlet condition (absolute pressure of 101,325 Pa) was defined at
the opposite end of the domain. At the upper and lower surfaces of the domain in Figure 4.48, a
free-slip wall condition was imposed. As the task was solved in a quasi-2D way, the symmetry
condition was defined on the side surfaces, perpendicular to the axis of turbine rotation. The motion
of the blades as governed in the structural simulation and was transferred to the Fluent solver by
system coupling. The blades of the turbine were defined as non-slip and smooth walls. The interface
between both domains was applied to exchange data. The initial conditions were determined on the
basis of the boundary conditions.
The SIMPLE (semi-implicit method for pressure-linked equations) pressure-velocity coupling
method was used. The second-order spatial discretization schemes were applied for mass,
momentum and energy equations. A first-order implicit transient scheme was also used. Maximally
20 iteration loops were solved in every timestep of transient simulations. Typically, it was enough
to reach the residual target of 1e-3, selected as the convergence criterion for all equations solved in
Fluent. This criterion was not reached only for a few timesteps, due to numerical problems resulting
from mesh deformation or remeshing.
4.6.3.3 Validation of the Numerical Procedure
In order to assess the precision of the numerical method with mesh deformation and remeshing used
in this investigation, it was decided to compare its results with the results of a typical simulation of
the Savonius turbine, applied, e.g., in [13, 16, 17]. It was possible to do it for the eccentricity
magnitude equal to zero, i.e., when rotor blades were not deformed. In this case, the flow was
165
additionally simulated with a fixed mesh, where continuous changes of the rotor position were
obtained by rotation of the circular inner domain and exchange of data at the sliding mesh interface
between the stationary and rotational domains. The rest of the simulation conditions was preserved.
Taking advantage of much faster simulations for such a task arrangement, the mesh size dependence
was also verified by means of Richardson’s extrapolation, similarly as in [41]. The same task was
solved on three different grids with a mesh refinement ratio of 2. The numbers of control volumes
of the grids and the average power coefficient values obtained for them are presented in Table 4.26.
Table 4.26 Mesh Size Dependence Study
Figure 4.51 Comparison of the power coefficient Cp for one revolution of the non-deformable rotor
from simulations with the rotating inner domain and the deformed/re-meshed one. Subsequent positions
of the turbine illustrated in the schemes above the graph correspond to the turbine revolution angle.
According to the Richardson’s extrapolation procedure described in [35], the extrapolated value
(RE), the apparent order p, the ratio of error R and the fine-grid convergence index (FGCI) were
166
determined. The negative value of R indicates an oscillatory convergence. Only a slight variation of
the average power coefficient was observed (1.2%), however, the y+ < 1 condition was fulfilled for
the mesh composed of 1.16 M69 control volumes shown in Figure 4.50. The uncertainty due to
discretization determined for the average Cp on the basis of the Richardson’s extrapolation procedure
described in [42] for this mesh was low, i.e., 1.7%. Thus, it was decided to use it in the further
numerical investigations.
In the cases of typical 2D simulations of fixed-shape Savonius rotors within the RANS method [16,17],
after simulations of a few or more of rotor revolutions, one can reach the limit cycle of the blade load.
In this case, a change in the period-averaged value of the power coefficient was lower than 1% of its
value after 7 revolutions. The same criterion for this coefficient for particular instants of the rotor
revolution (limit cycle) was reached after 10 revolutions. This limit cycle is marked in Figure 4.51
with a dashed line.
In the case of the simulations with mesh deformation and remeshing applied in the further
investigations, such a limit cycle was not obtained, mostly due to numerical instabilities. They
resulted from the continuous deformation of mesh control volumes, which successively deteriorated
in quality up to the moment when the specific regions were re-meshed. Remeshing introduced some
further numerical errors due to the data interpolation onto a new mesh. Thus, slight random
fluctuations of the blade load were present. Therefore, the data presented in Figure 4.51 with a solid
line were averaged for the last 5 out of 15 rotor revolutions. This number of periods was sufficient to
reach no significant changes of the data averaged for the same angular positions during the rotor
revolution. The period-averaged values of the power coefficient changes were lower than 1% after
10 revolutions.
As one can see, no significant differences can be observed for two parts of the period for the fixed-
mesh simulations (dashed line in Figure 4.51). In the case of simulations with mesh deformation
and remeshing (solid line), differences for two parts of the period are significant. They are observed
especially for the revolution angles where the rotor performance is the lowest (75-120∘ and 255–
300∘).
In these particular ranges, random fluctuations of the blade loads were the highest. The most
probable explanation is a very complex flow pattern with numerous vortex structures at the convex
side of the advancing blade. The vortex in particular at its tip is very strong with high velocity
gradients and, thus, it can be vulnerable to some fluctuations due to the mesh quality deterioration
or the remeshing procedure. The poorer flow prediction in those ranges caused that the period-
averaged value of the power coefficient was 4.6% higher with respect to the fixed mesh (Cp = 0.206
for re-meshed, Cp = 0.197 for fixed-mesh simulations). This difference is definitely not negligible, but
it is an order of magnitude lower than the pressure coefficient increase predicted for rotors with
deformable blades. Thus, despite its drawbacks, the method with mesh deformation and remeshing
can be considered as sufficiently reliable for the needs of these investigations.
The influence of the deformation magnitude and the position of maximally deformed blades with
respect to the incoming wind direction was studied. A detailed analysis was performed for the
turbine with deformable blades in comparison to the non-deformable (fixed-shape) Savonius rotor.
The presentation and discussion of the results was divided into three sections. In the first section, the
turbine performance was analyzed for different values of the deformation magnitude in a full range
of the position of maximal deformation with respect to the incoming wind. The favorable range of the
position was identified. In the second section, changes in the power coefficient during the rotor
revolution were presented for the rotor with deformable blades in the optimal position and
compared to the non-deformable Savonius rotor. The parts of the revolution cycle where the power
69 M = million
167
output of the deformable rotor was significantly increased were identified and contributions of the
individual blades were revealed. In the third section, a detailed analysis of blade loads was performed
for three instants of the rotor revolution to display the main differences between the rotors with
deformable and non-deformable blades.
Note – Readers should consult the [Sobczak et al.] for discussion regarding the Influence of
Eccentricity of Deformed Blades on the Turbine Performance and Changes in the Power Coefficient
during Rotor Revolution.
Figure 4.52 Comparison of the averaged values of the power coefficient Cp for one revolution of the
rotors with non-deformable and deformable blades. Subsequent positions of the turbine illustrated in
the schemes above the graph correspond to the turbine revolution angle
4.6.4.1 Blade Loading and Torque Generation

In order to better understand the contribution of the particular blade loading to the torque
generation, torque distributions along the blades (from the axis towards the blade tips) are shown in
Figure 4.54 and pressure fields in Error! Reference source not found., correspondingly. The dT/dL
derivative per 1 m of the blade span (actually it is DT/DL as it is based on the finite volume
simulations) is shown in Figure 4.54, thus, in order to obtain the torque value, this parameter needs
to be integrated along the blade. Three particular instants of the rotor revolution were selected (15∘,
55∘ and 105∘), which is sufficient to display the main differences between the rotors with deformable
and non-deformable blades. The data were presented for a half of the cycle, because no significant
differences can be distinguished for its second part. The data in Figure 4.54 and Error! Reference
168
source not found. are attained for the 10th rotor revolution, thus there is no perfect agreement with
the results in Figure 4.53 obtained from averaging over 5 revolutions.
Figure 4.53 Comparison of the averaged values of the power coefficient Cp for each rotor blade for one
revolution of the rotors with non-deformable and deformable blades. Subsequent positions of
the turbine illustrated in the schemes above the graph correspond to the turbine revolution
angle.
In the 15∘ (also 195∘) instant of the revolution cycle, the non-deformable rotor reaches the maximum
of its power output (Figure 4.52). In this case, one can notice in Error! Reference source not found.
that a high pressure at the concave side of the advancing blade (A) and a low pressure at its convex
side (due to flow acceleration) result in the highest positive torque generated by this blade. It is
diminished by the retarding contribution of the returning blade (B), mostly due to a pressure
difference for its part near
the axis of the turbine. In this instant of the revolution cycle, the blades of the deformable rotor have
a semi-circular shape like the blades of the non-deformable rotor. No significant increase in the
power coefficient is observed at this moment for the rotor in reference to the non-deformable one
(Figure 4.52).
The data presented in Figure 4.53 reveal slight positive contributions both of the advancing (A) and
returning (B) blades. The positive contribution of the advancing blade (blue lines) can be
distinguished also in Figure 4.54. An increase of the torque can be seen for a substantial portion of
the blade starting from the rotor axis (internal part), whereas the tip (external) part has slightly
lower loading. In the case of the returning blade (red lines), positive and negative contributions of
the blade deformation can be observed locally for the internal part of the blade, whereas a positive
effect is clearly visible for the external part.
169
Figure 4.54 Torque derivative (dT/dL) variation along the blade for selected instants ( 55° and 105°)
of the rotor revolution with non-deformable and deformable blades
The highest positive effect of the blade deformation with respect to the non-deformable turbine is
observed for the 55∘ (also 235∘) instant of the revolution cycle . In this position the deformable rotor
reaches the highest power output, whereas it is already significantly diminished for the non-
deformable one. In Figure 4.53 it can be seen that both the blades contribute to the power gain. It is
due to the blade deformation, i.e., expansion of the advancing blade (A) and contraction of the
170
returning one (B), as shown in Error! Reference source not found.. The torque of the advancing
blade is significantly higher at its external part (Figure 4.54). It results from higher velocity at the
convex side in a very limited way. The main reason of the higher torque is just a higher arm (radius)
due to the blade expansion. Despite the fact that the pressure build-up (stagnation zone) at the
convex side of the returning blade
is higher than for the non-deformable blade, its contraction and higher pressure at the concave side
do not reduce further the torque at its central part. On the other hand, much higher fluid acceleration
171
at the convex side of the deformed blade for its external part results in lower pressure and locally
changes its torque to the positive one. Thus, the retarding effect of the contracted returning blade is
reduced significantly.
Figure 4.55 Pressure fields for selected instants (15∘, 55∘ and 105∘) of one revolution of the rotors
with non-deformable and deformable blade
A significantly positive effect of the blade deformation can be observed also for the 105∘ (and 285∘
as well) instant of the revolution cycle (Figure 4.52). In this position both the deformable and non-
deformable rotors reach the lowest power output. The blade deformation in this case is the highest
as presented in Error! Reference source not found..
172
In Figure 4.53 one can see that the deformation of the advancing blade (A) decreases the pressure
coefficient, nevertheless, it remains positive. A beneficial effect of the higher arm in the torque
definition due to blade expansion can be noticed for the external part of this blade (Figure 4.54). It
is lowered by higher pressure at its convex side due to lower intensity of the vortex structure at the
blade tip in comparison to the non-deformable rotor (Figure 4.55). This positive effect does not
compensate for a decrease in the torque in the internal part of the blade. This decrease is mainly due
to much lower pressure at the concave side of the blade, which is an effect of the flow acceleration in
this region. Deformation of the returning blade (B) improves considerably its performance almost
along its whole length, excluding the tip only (Figure 4.54). It is partially due to a much lower torque
arm of the contracted blade. Additionally, the higher curvature of the deformed blade yields higher
flow acceleration and a significant reduction in the high pressure stagnation zone at its convex side
in comparison to the non-deformable rotor (Figure 4.55). The blade contraction is also followed by
a pressure increase at the convex side of the returning blade. Thus, the pressure difference between
the convex and concave sides of the blade is significantly diminished. All these aspects contribute to
a reduction in the retarding torque of the deformable blade almost to zero (Figure 4.53).
4.6.5 Summary and Conclusions
An idea of the Savonius turbine with a variable geometry of blades was proposed. Blades made of an
elastic material were continuously deformed during the rotor revolution in order to increase a
positive torque of the advancing blade, and, at the same time, to decrease a negative moment of the
returning blade. The main outcomes of the performed investigations are outlined below:
➢ An elaborate two-dimensional numerical model was developed to simulate a transient flow
in the variable-geometry rotor in order to assess its aerodynamic performance. The shape
and position of the rotor blades were subject to continuous changes according to the
constraints defined in the structural analysis. The rotational motion and deformations of the
blades were transferred to the fluid flow (CFD) analysis, where deformations of grid elements
and remeshing options were applied. This method yielded a satisfactory agreement with a
typical method of simulations, which consists in rotation of the internal domain surrounding
the fixed-shape Savonius rotor.
➢ An improvement of the aerodynamic turbine performance in comparison to the non-
deformable Savonius turbine was obtained. It was achieved when the blades were maximally
deformed in the range of angular position 45∘180∘ with respect to the direction of the
incoming wind. The maximum of Cp was attained at the angle of 105∘, i.e., when the blade
chords were almost perpendicular to the wind direction.
➢ An increase in the blade deformation increased the rotor performance. For the eccentricity
position of 105∘, the power coefficient increased by 37%, 66% and 90% for 5%, 10% and
15% of the eccentricity magnitude, respectively. However, the range of angular positions for
which a gain with respect to the non-deformable Savonius turbine occurred decreased.
➢ A detailed flow analysis for the rotor with eccentricity in the 105∘ position showed an
increase in the power coefficient of the deformable blade rotor with respect to the one having
the fixed shape during all their revolutions. The radial expansion of the advancing blade
increased the positive torque and the contraction of the returning blade decreased the
negative one, with slight changes in the flow patterns.
A significant increase in the aerodynamic performance of the Savonius turbine with continuously
deformed blades was confirmed. The power coefficient exceeded Cp = 0.3 and it reached almost Cp =
0.4 for the highest eccentricity magnitude. Thus, this design is at least comparable to Savonius
turbines equipped with augmentation systems presented in the literature [11 , 26]. However,
additional mechanisms applied to deform blades and locate the rotor in the proper position with
respect to the incoming wind are required. They will make the design of the turbine more complex
173
than the original Savonius and they will consume a part of the energy generated by the turbine. Thus,
all turbine elements need to be carefully designed to be resistant and efficient. Additionally, as the
turbine performance depends considerably on the blade deformation, it is necessary to select an
easily deformable material with high fatigue resistance. The numerical model of the deformable
Savonius rotor needs to be further developed to limit the numerical instabilities during the solution.
The mechanical losses due to friction and blade deformations are to be included. Also the problem of
time consuming simulations has to be addressed.
Supplementary Materials: The following are available online at http://www.mdpi.com/1996-

1073/13/14/3717/s1: Video S1: NonDeformable_Savonius.wmv; Video S2: Deformable_Savonius.wmv.
Author Contributions: Conceptualization, K.S. and D.O.; methodology, K.S., D.O., P.R.; investigation, K.S.,
D.O., P.R., E.M.; writing—original draft preparation, K.S. and D.O.; All authors have read and agreed to
the published version of the manuscript.
Funding: The investigations have been financially supported by the research project POWR.03.02.00-
00-I042/16-00 of the National Centre for Research and Development and Innovation Incubator 2.0
project MNISW/2019/157/DIR of the Ministry of Science and Higher Education.
Acknowledgments: We would like to thank Malgorzata Jozwik for her significant linguistic help during
the preparation of the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.
4.6.6 References
1. Tummala, A.; Velamati, R.K.; Sinha, D.K.; Indraja, V.; Krishna, V.H. A review on small scale wind
turbines. Renew. Sustain. Energy Rev. 2016, 56, 1351–1371.
2. Aslam Bhutta, M.M.; Hayat, N.; Farooq, A.U.; Ali, Z.; Jamil, S.R.; Hussain, Z. Vertical axis wind
turbine—A review of various configurations and design techniques. Renew. Sustain. Energy Rev.
2012, 16, 1926–1939.
3. Kumar, R.; Raahemifar, K.; Fung, A.S. A critical review of vertical axis wind turbines for urban
applications. Renew. Sustain. Energy Rev. 2018, 89, 281–291. Energies 2020.
4. Toja-Silva, F.; Colmenar-Santos, A.; Castro-Gil, M. Urban wind energy exploitation systems:
Behaviour under multidirectional flow conditions—Opportunities and challenges. Renew. Sustain.
Energy Rev. 2013, 24, 364–378.
5. KC, A.; Whale, J.; Urmee, T. Urban wind conditions and small wind turbines in the built
environment: A review. Renew. Energy 2019, 131, 268–283.
6. Savonius, S.J. The S-rotor and its applications. Mech. Eng. 1931, 53, 333–338.
7. Akwa, J.V.; Vielmo, H.A.; Petry, A.P. A review on the performance of Savonius wind turbines. Renew.
Sustain. Energy Rev. 2012, 16, 3054–3064.
8. Roy, S.; Saha, U.K. Review on the numerical investigations into the design and development of
Savonius wind rotors. Renew. Sustain. Energy Rev. 2013, 24, 73–83.
9. Doer_er, P.; Doer_er, K.; Ochrymiuk, T.; Telega, J. Variable Size Twin-Rotor Wind Turbine. Energies
2019, 12, 2543.
10. Abraham, J.P.; Plourde, B.D.; Mowry, G.S.; Minkowycz, W.J.; Sparrow, E.M. Summary of Savonius
wind turbine development and future applications for small-scale power generation. J. Renew.
Sustain. Energy 2012, 4, 042703.
11. Alom, N.; Saha, U.K. Four Decades of Research into the Augmentation Techniques of
SavoniusWind Turbine Rotor. J. Energy Resour. Technol. Trans. ASME 2018, 140, 1–14.
12. Kumar, A.; Saini, R.P. Performance parameters of Savonius type hydrokinetic turbine-A Review.
Renew. Sustain. Energy Rev. 2016, 64, 289–310.
13. Kacprzak, K.; Liskiewicz, G.; Sobczak, K. Numerical investigation of conventional and modified
Savonius wind turbines. Renew. Energy 2013, 60, 578–585.
174
14. Chen, L.; Chen, J.; Zhang, Z. Review of the Savonius rotor’s blade profile and its performance. J.
Renew. Sustain. Energy 2018, 10, 013306.
15. Tartuferi, M.; D’Alessandro, V.; Montelpare, S.; Ricci, R. Enhancement of savonius wind rotor
aerodynamic performance: A computational study of new blade shapes and curtain systems. Energy
2015, 79, 371–384.
16. Kerikous, E.; Thévenin, D. Optimal shape of thick blades for a hydraulic Savonius turbine. Renew.
Energy 2019, 134, 629–638.
17. Tian,W.; Mao, Z.; Zhang, B.; Li, Y. Shape optimization of a Savonius wind rotor with different
convex and concave sides. Renew. Energy 2018, 117, 287–299.
18. Mohamed, M.H.; Janiga, G.; Pap, E.; Thévenin, D. Optimal blade shape of a modified Savonius
turbine using an obstacle shielding the returning blade. Energy Convers. Manag. 2011, 52, 236–242.
19. Zhang, B.; Song, B.; Mao, Z.; Tian, W.; Li, B.; Li, B. A Novel Parametric Modeling Method and Optimal
Design for Savonius Wind Turbines. Energies 2017, 10, 301.
20. Ogawa, T.; Yoshida, H. The e_ects of a deflecting plate and rotor end plates on performances of
Savonius-type wind turbine. Bull. JSME 1986, 29, 2115–2121.
21. Golecha, K.; Eldho, T.I.; Prabhu, S.V. Influence of the deflector plate on the performance of
modified Savonius water turbine. Appl. Energy 2011, 88, 3207–3217.
22. Shaughnessy, B.M.; Probert, S.D. Partially-blocked savonius rotor. Appl. Energy 1992.
23. Alexander, A.J.; Holownia, B.P. Wind tunnel tests on a savonius rotor. J. Wind Eng. Ind. Aerodyn.
1978, 3, 343–351.
24. Altan, B.D.; Atılgan, M. The use of a curtain design to increase the performance level of a Savonius
wind rotors. Renew. Energy 2010, 35, 821–829.
25. Irabu, K.; Roy, J.N. Characteristics of wind power on Savonius rotor using a guide-box tunnel. Exp.
Therm. Fluid Sci. 2007, 32, 580–586.
26. Wong, K.H.; Chong, W.T.; Sukiman, N.L.; Poh, S.C.; Shiah, Y.C.; Wang, C.T. Performance
enhancements on vertical axis wind turbines using flow augmentation systems: A review. Renew.
Sustain. Energy Rev. 2017, 73, 904–921.
27. Lipian, M.; Czapski, P.; Obidowski, D. Fluid–Structure Interaction Numerical Analysis of a Small,
Urban Wind Turbine Blade. Energies 2020, 13, 1832.
28. Karczewski, M.; Sobczak, K.; Lipian, M.; Jozwik, K. Numerical and experimental tools for small
wind turbine load analysis. In Structural Control and Fault Detection of Wind Turbine Systems;
Institution of Engineering and Technology: London, UK, 2018; pp. 45–79, ISBN 9781785613944.
Energies 2020.
29. Shkara, Y.; Cardaun, M.; Schelenz, R.; Jacobs, G. Aeroelastic response of a multi-megawatt upwind
horizontal axis wind turbine (HAWT) based on fluid–structure interaction simulation. Wind Energy
Sci. 2020, 5, 141–154.
30. Löhner, R.; Haug, E.; Michalski, A.; Muhammad, B.; Drego, A.; Nanjundaiah, R.; Zarfam, R. Recent
advances in computational wind engineering and fluid-structure interaction. J. Wind Eng. Ind.
Aerodyn. 2015.
31. Bazilevs, Y.; Korobenko, A.; Deng, X.; Yan, J.; Kinzel, M.; Dabiri, J.O. Fluid–Structure Interaction
Modeling of Vertical-AxisWind Turbines. J. Appl. Mech. 2014, 81.
32. MacPhee, D.W.; Beyene, A. Fluid-structure interaction analysis of a morphing vertical axis wind
turbine. J. Fluids Struct. 2016.
33. Obidowski, D.; Sobczak, K.; Jozwik, K.; Reorowicz, P. Vertical Axis Wind Turbine with a Variable
Geometry of Blades. European Patent Application 19199085.2, 24 September 2019.
34. Sobczak, K. Numerical investigations of an influence of the aspect ratio on the Savonius rotor
performance. J. Phys. Conf. Ser. 2018, 1101, 012034.
35. Kacprzak, K.; Sobczak, K. Numerical analysis of the flow around the Bach-type Savonius wind
turbine. J. Phys. Conf. Ser. 2014, 530, 012063.
175
36. Kamoji, M.A.; Kedare, S.B.; Prabhu, S.V. Experimental investigations on single stage modified
Savonius rotor. Appl. Energy 2009.
37. ANSYS. ANSYS Fluent 19.2 Theory Guide; Ansys Inc.: Canonsburg, PA, USA, 2018.
38. Menter, F.R. Two-equation eddy-viscosity turbulence models for engineering applications. AIAA
J. 1994, 32, 1598–1605.
39. Ferrari, G.; Federici, D.; Schito, P.; Inzoli, F.; Mereu, R. CFD study of Savonius wind turbine: 3D
model validation and parametric analysis. Renew. Energy 2017, 105, 722–734.
40. Kacprzak, K.; Sobczak, K. Computational assessment of the influence of the overlap ratio on the
power characteristics of a Classical Savonius wind turbine. Open Eng. 2015, 5, 314–322.
41. Aramendia, I.; Fernandez-Gamiz, U.; Zulueta, E.; Saenz-Aguirre, A.; Teso-Fz-Betoño, D. Parametric
Study of a Gurney Flap Implementation in a DU91W(2)250 Airfoil. Energies 2019, 12, 294.
42. Celik, I.B.; Ghia, U.; Roache, P.J.; Freitas, C.J.; Coleman, H.; Raad, P.E. Procedure for Estimation and
Reporting of Uncertainty Due to Discretization in CFD Applications. J. Fluids Eng. 2008, 130, 078001.

Wind Turbines

Uploaded by

Copyright:

Available Formats

Wind Turbines

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Wind Turbines

Uploaded by

Copyright:

Available Formats

CFD Open Series

Off Shore Wind

2 Wind Turbine Design and Working Principle................................................................. 19

3 Case Studies Involving Horizontal Axis Wind Turbine (HAWTs)............................. 51

3.1 Case Study 1 - Flow Around a Wind Turbine........................................................................................... 51

4 Case Studies Involving Vertical Axis Wind Turbines (VAWTs) ................................. 87

4.3.14 Conclusion............................................................................................................................................... 127

1 Enrgy.Gov , Office Of Energy Efficiency & Renewable Energy

The first electricity-generating wind turbine

2 Wind Turbine Design and Working Principle

2.1 Airfoils and Blade Design

2.1.1 Blade Twist

2.1.4 Wind Turbine Aerodynamics

Figure 2.8 Actuator Disk Concept for Wind Turbine Rotor

2.1.5 Actuator Disk Concept

4 Albert Betz, “Betz Law for Wind Turbines”, http://en.wikipedia.org/wiki/Betz%27s_law.

P = 2ρv0 a(1 − a)AR

(c) The control (d) The swept-

Figure 2.11 Scheme of Axial Type Wind-Turbine

2.2.3 Theoretical Analysis

2.2.4 Results and Discussion

2.3 Case Study 2 - Blade Thickness Effect on The Aerodynamic Performance of An

2.3.3 Computational Settings

flow physics and the turbine performance in a trustworthy manner.

A significant over estimation of turbine

simulation done in the present work,

2.3.4 Results and Discussion

flow separation, which transforms into

With increase in blade thickness, a

2.4 Wind Profile6

2.5 Wind Turbine Cost Figures7

Figure 2.31 Capital cost breakdown of an offshore wind farm [16]

3 Case Studies Involving Horizontal Axis Wind Turbine (HAWTs)

Turbulent Momentum Flux = ⏟̅̅̅̅̅

Lateral Direction Vertical Direction

3.1.3 Summary & Conclusions

5. Simley E, Angelou N, Mikkelsen T, Sjöholm M, Mann J, Pao LY (2016) Characterization of wind

that is based on a SERI 820 airfoil at 95% of the tip radius.

3.2.2 Design Variables and Design Space

Table 3.1 Design Space

𝐗11 𝐗12 ⋯ 𝐗1Nvar → y1

Nvar Nvar Nvar −1 Nvar

𝐲̂ = ̂f(𝐱, 𝛃) = β1 + ∑ βi xi + ∑ βjj xj + ∑ ∑ βij xi xj

Figure 3.10 Rotating Body in the Inertial Reference Frame

M elements) were used for obtaining the

Blade AOC 15/50 Atlantic Orient Corporation

Table 3.2 Table of CFD Solver Settings

Table 3.3 Table Showing Optimized Candidate Point for Routine 1

Table 3.4 Table Showing Optimized Candidate Point for Routine 2

3.3 Case Study 3 - Multi-Element Ducts for Ducted Wind Turbines

Aerospace Research and Development NEUILLY-SUR-SEINE, France, 1979.

of Aeronautics, Cranfield, 1956.

multi-element duct–AD model with CT;AD = 0.550 (Model C (ii) flap).

A typical converged RANS solution

3.3.5.2 Duct Force Coefficient

(a) Panel method (b) RANS method

3.3.5.3 Power Augmentation

(a) Panel method (b) RANS method

4 Case Studies Involving Vertical Axis Wind Turbines (VAWTs)

Table 4.1 Design Optimization Summary