Certificate

Certified that Suraj Yadav(1302940161), Abhishek Jain(1302940010), Prab-

hat Kumar (1302940094) and Bharat Kaushik(1402940909) have carried out the
research work presented in this thesis entitled “Numerical Analysis of Different Blade
Profiles of Wind Turbine” for the award of Bachelor of Technology from Dr. A.P.J.
Abdul Kalam Technical University, Lucknow under my supervision. The thesis em-
bodies results of original work, and studies are carried out by the students themselves
and the contents of the thesis do not form the basis for the award of any other de-
gree to the candidates or to anybody else from this or any other University/Institution.

Mr Gaurav Kumar
Assistant Professor
Department Of Mechanical Engineering
Krishna Institute of Engineering and Technology, Ghaziabad
Date:

i
 Abstract

The performance of the Savonius wind rotor with conventional semicircular blades
and a novel blade configuration was studied by means of 2–D flow computation in An-
sys Fluent 14.0. The variation in dynamic torque and power coefficients with respect
to the tip speed ratio of the rotors have been systematically analyzed. The self starting
capability of the rotor is assessed by studying the static torque exerted on the rotors at
various angles with respect to the wind velocity. The optimal blade profile geometry
was identified by examining the performance characteristics of each of the geometries
under study. The most favorable design was also found to be capable of self starting
at any rotor angle with respect to the wind velocity. The idea behind the new de-
signs was to harvest the incoming wind to generate a wind jet to the concave side of
the advanced blade and prevent the convex side of the returning blade from coming
upwind stream by guiding it through different guide plates, to impinge the concave
side of return blade and hence to eliminate the negative torque. The research reflected
that using structured grid under k-w SST turbulent modeling in transient conditions
for 2-D modeling under dynamic mode, is a method of great accuracy to generate nu-
merical data that follows closely the experimental curve. So the models were studied
in k-w SST transient turbulent model under dynamic conditions. The results showed
that the suggested designs improve the performance of savonius rotor in view of the
power coefficient and the operating range. The performance achieved with one of the
suggested guide plate design in terms of power coefficient is 0.14 for tip speed ratio 0.8.
With the use of suggested designs of guide plates, the model can be operated even at
higher tip speed ratios.

ii
 Acknowledgement

It gives us a great sense of pleasure to present the report of the B. Tech Project
undertaken during B.Tech final year. We owe special debt of gratitude to Assistant
Professor Mr. Gaurav kumar, Department of Mechanical Engineering, Krishna Insti-
tute of Engineering and Technology, Ghaziabad for his constant support and guidance
throughout the course of our work. His sincerity, thoroughness and perseverance have
been a constant source of inspiration for us. It is due to his endless efforts that this
project has come out with great results.

We also take the opportunity to acknowledge the contribution of Dr. K.L.A Khan,
Head of Department of Mechanical Engineering, Krishna Institute of Engineering and
Technology, Ghaziabad for his constant support and help due to which our work could
be fulfilled.

We would also like to thank all faculty members and staff of the department for
their kind assistance and cooperation during the development of our project. And
finally, our thanks are also due to our friends who, at various points kept helping us
by valuable suggestions.

Suraj Yadav
(1302940161)

Abhishek jain
(1302940010)

Prabhat Kumar
(1302940094)

Bharat Kaushik
(1402940909)

iii
Contents

1 INTRODUCTION 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 LITERATURE SURVEY 6
2.1 Numerical Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 Experimental Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

3 PROBLEM FORMULATION 19
3.1 Computational Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.1 For Elliptical Rotor . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.2 For Designs of Guide Plates . . . . . . . . . . . . . . . . . . . . 20
3.2 Details of Domain Discretization . . . . . . . . . . . . . . . . . . . . . . 22
3.3 Computational methodology . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3.1 Simulation Model . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3.2 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3.3 Details of Setup Solver . . . . . . . . . . . . . . . . . . . . . . . 25
3.4 Calculation of performance coefficients . . . . . . . . . . . . . . . . . . 25

4 RESULT AND DISCUSSION 26

4.1 Starting Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.2 Velocity Flow Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.3 Pressure Distribution Analysis . . . . . . . . . . . . . . . . . . . . . . . 32

5 CONCLUSION 36
5.1 Key Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.2 Scope for future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

References 37

iv
List of Figures

1.1 Types of Wind Turbine Rotors . . . . . . . . . . . . . . . . . . . . . . . 2

1.2 Two Bladed Savonius Rotor . . . . . . . . . . . . . . . . . . . . . . . . 3

2.1 Effect of TSR on Power Coefficient . . . . . . . . . . . . . . . . . . . . 6

2.2 Effect of end plates on Cp with variation in TSR . . . . . . . . . . . . 8
2.3 Effect of overlap ratio on the average static torque coefficient for two
bladed VAWT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.4 Performance analysis of the Savonius rotor angle . . . . . . . . . . . . . 9
2.5 Evaluation of the resultant force coefficients developed by rotor A and
rotor B. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.6 Comparison of average co-efficient of power at different TSR to the ex-
perimental results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.7 Cp vs TSR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.8 Cp and Ct vs TSR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.9 Variation of SPL with Frequency . . . . . . . . . . . . . . . . . . . . . 13
2.10 Average starting torque characteristics . . . . . . . . . . . . . . . . . . 14
2.11 Velocity distributions in and around stationary Savonius rotors measure-
ments and numerical calculations . . . . . . . . . . . . . . . . . . . . . 15
2.12 Design of curtain arrangement placed in front of the Savonius wind rotor. 16
2.13 The torque change obtained through experiment for β = 10 and 15
between α = 30 and 60 when rotor position is 45 degree . . . . . . . . 16
2.14 Comparison of the power changes with the rotor without curtain and
the rotor with different curtain types for α = 45 degree and β = 15 degree 17

3.1 Classical semi circular profile of the Savonius rotor . . . . . . . . . . . 20

3.2 Modified rotor geometry under study . . . . . . . . . . . . . . . . . . . 21
3.3 Geometrical parameters of the rotor blade under study . . . . . . . . . 21
3.4 Rotor with Guide plate design A and design B . . . . . . . . . . . . . . 22
3.5 3D Isometric geometry of Design B . . . . . . . . . . . . . . . . . . . . 23
3.6 3D Isometric geometry of Design B . . . . . . . . . . . . . . . . . . . . 23
3.7 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4.1 Variation of Torque Coefficient with Tip Speed Ratio . . . . . . . . . . 27

4.2 Variation of Power Coefficient with Tip Speed Ratio . . . . . . . . . . . 27
4.3 Variation of Static Torque Coefficient with Rotor Angle . . . . . . . . . 28
4.4 Velocity contour for semicircular profile at 90 degree . . . . . . . . . . . 29
4.5 Velocity contour for novel shaped profile at 90 degree . . . . . . . . . . 30
4.6 Velocity contour for semicircular profile . . . . . . . . . . . . . . . . . . 30
4.7 Velocity contour for novel shaped profile . . . . . . . . . . . . . . . . . 31

v
4.8 Velocity contour for novel shaped profile with guide vanes design A . . 31
4.9 Velocity contour for novel shaped profile with guide vanes design B . . 32
4.10 Pressure contour for semicircular profile at 90 degree . . . . . . . . . . 33
4.11 Pressure contour for novel shaped profile at 90 degree . . . . . . . . . . 33
4.12 Pressure contour for semicircular profile . . . . . . . . . . . . . . . . . . 34
4.13 Pressure contour for novel shaped profile . . . . . . . . . . . . . . . . . 34
4.14 Pressure contour for novel shaped profile with guide vanes design A . . 35
4.15 Pressure contour for novel shaped profile with guide vanes design B . . 35

vi
List of Tables

4.1 Variation of Torque Coefficient with Tip Speed Ratio . . . . . . . . . . 26

4.2 Variation of Power Coefficient with Tip Speed Ratio . . . . . . . . . . . 28

vii
Nomenclature

Acronyms

Cp Coefficient of Power

Ct Coefficient of Torque

D Diameter of semi-circular blade (m)

e Overlap Ratio

g Acceleration due to gravity

H Height of the rotor (m)

P Turbine power (kW)

R Radius of Savonius rotor

T Torque of wind turbine blade

U Fluid velocity (m/s)

u, v, w Velocity in x, y and z direction

x, y, z Cartesian coordinates

Greek Symbols

α Aspect ratio

φ Coefficient of viscosity

Subscripts

CFD Computational Fluid Dynamics

SPL Sound pressure level

SSWT Savonius style wind turbine

TSR Tip speed ratio

viii
Chapter 1

INTRODUCTION

1.1 Motivation
In the present era, economical and industrial growth is very vital for develop-
ing countries. For which Energy is of great importance, especially for socio -economic
development. In the present situation, indispensible contribution of energy is being
sensed in everyday needs of mankind, production, transport and communication pro-
cesses. In contrast, the conventional energy resources are depleting at higher rate
resulting the hike in prices. As well as the recent impacts of climate change and global
warming caused by the carbon derivatives and the nuclear plant calamities are a major
concern for the global environment. As an effect of this, the energy scenario of the
world is rapidly changing and governments of many countries are focusing on devel-
oping some alternate energy sources which are clean, safe, reliable and economical,
especially renewable energy so as to reduce the dependence on fossil fuels. Renewable
energy sources are expected to play an important role in solving the energy problems
in decentralized locations, rural and remote areas. Fruitful progress has been achieved
in various renewable technologies that include wind power, hydro power, solar and
geothermal energy.

It is observed that among the non-hydro renewable sources, wind energy has shown
great promise among various renewable energy technologies and can be a feasible al-
ternative to reduce the dependency on conventional energy sources because of its tech-
nological maturity, good infrastructure, abundance in nature and relative cost compet-
itiveness. So in this light, efforts are been made to utilize the wind energy through
development of various wind turbines which is renewable as well as clean as far as
ecology is concerned.

The Global Wind Energy Council released its annual market statistics: the 2016
market was more than 54 GW, bringing total global installed capacity to nearly 487
GW. Led by China, the US, Germany, and India; and with surprisingly strong show-
ings from France, Turkey and the Netherlands, the global market was nonetheless less
than 2015’s record total.

“Wind power continues to grow in double digits; but we can’t expect the industry
to set a new record every single year, said SteveSawyer [15] , GWEC Secretary Gen-

1
 Figure 1.1: Types of Wind Turbine Rotors

eral. “Chinese installations were an impressive 23, 328 MW, although this was less than
2015’s spectacular 30 GW, which was driven by impending feed-in tariff reductions.
Also, Chinese electricity demand growth is slackening, and the grid is unable to handle
the volume of new wind capacity additions; although we expect the market to pick
up again in 2017.” The Chinese offshore market began what many hope is the sector’s
long awaited take-off in 2016, with China passing Denmark to achieve 3rd place in the
global offshore rankings, after the UK and Germany.

US installations (8, 203 MW) were nearly equal to 2015’s strong market, bringing
the US total to more than 82 GW. The US industry now employs more than 100, 000
people and has more than 18 GW under construction or in advanced stages of develop-
ment, a harbinger for a strong market again in 2017. Canada (702 MW) and Mexico
(454 MW) posted solid though modest gains.

Thus, wind power harvesting with the help of wind turbine is a new and an emerg-
ing field in power generation. With the intention of using this abundant source of
wind energy, various designs of wind turbine have been proposed till date. However,
only few have been found to be practically suitable and useful. Depending on their
axis of rotation wind turbines are mainly two types - horizontal axis wind turbine
(HAWT) and vertical axis wind turbine (VAWT). HAWTs are very well-known for
their comparatively high efficiency over VAWTs and inherently have been used for
electricity production. Multi blade type and propeller type are few HAWTs. However,
the VAWTs have significant advantages over HAWTs such as simple in construction,
lower installation and maintenance cost, good self-starting capability, independent of
the wind direction and no need of over speed control etc. These advantages are con-
firming the use of VAWT for small scale applications. VAWTs can be further classified
into various categories such as Savonius and Darrieus rotor etc.

2
 Figure 1.2: Two Bladed Savonius Rotor

Among this VAWTs, Savonius rotors offer an easy way to harness this vast wind
power for small scale applications because of their simple construction and wind direc-
tion independence. This class of turbine is a very useful device for water pumping in
agricultural purposes, especially in rural areas where the water level is within a dis-
tance of 5 m below the ground. The installation and maintenance costs are very low
and it can be installed on rooftops for local electricity production.

The Savonius rotor turbine is designed to be driven by the drag effect of wind on
the turbine blade and that is why known as drag type rotors. The top view of Savo-
nius rotor is somewhat analogous to ’S’ shape. It is simple in structure, have a very
low cut-in speed and accepts wind from any direction (Ross and Altman, 2010)[5].The
conventional Savonius wind rotor is made by cutting a cylinder into two halves, along
its central axis an relocating the two semi-cylindrical airfoils sideways as shown in fig-
1.2.The semi cylindrical airfoils are called blades which are used to convert the kinetic
energy of the wind into electrical energy.

It was originally introduced by a French engineer named (Savonius) in the 1920

s (Paraschivoiu, 2002 [8]) and basically is a modification of the flettner’s rotor used
in ships. It has low noise emissions because of low rotational speed. Its mechanical
and aerodynamic characteristics make it suitable for the high turbulence intensity ex-
perienced in the urban area. A typical conventional Savonius style turbine possesses
a power conversion efficiency of 12-20 percant which is relatively lesser in comparison
with Darrieus style turbine. Nevertheless, it is worth mentioning that if we modified
the design of savonius style turbines, this peak power conversion efficiency can be in-
creased upto 30-35 percent (Shepherd and Zhang ,2011; Abraham et al.2012; Akwa
et al. 2012). The conventional Savonius rotor has a high starting torque and this is
an advantage for particularity this kind of wind turbine (Hayashi et al, 2005), as they
are often applied for starting torques for other VAWTs (Chauvin and Benghrib, 1989
[2]).This devices are insensitive to the wind flow directions, and thus, are very useful
for the specific locations of variable wind directions. The vertical rotational axis allows
them to be positioned in confined space or at top of buildings (Abraham et al. 2012;
Akwa et al. 2012). A Savonius rotor designed with three blades generates favorable

3
torque characteristics. However, it has a poor maximum power coefficient compared to
the VAWT with two blades (Ross and Altman, 2011). So the number of rotor blades
has a direct impact on the rotor performance (Al-Bahadly, 2009 [7]). Menet (2004)[10]
affirmed that a two-stage Savonius rotor develops improved torque and power coeffi-
cients in comparison to the rotor in a single stage (Akwa et al, 2012). In the present
work, a Savonius rotor with two blades and a single stage is modeled for simplicity of
the rotor geometry and because it is the most commonly-studied type of Savonius rotor.

In the wind energy sector the modeling of the motion and behavior of a vertical
axis wind turbine (VAWT) is a technique that is necessary to deepen the knowledge
of the operation, control, structural stability, and performance of this type of wind
turbine. The research effort at this level is conducted in order to have a more reliable
and efficient wind turbine (Ross and Altman, 2011). The modern VAWTs currently in-
tegrated in the urban area are exploited to maximize the security of the energy supply.
Nowadays, the development of the wind energy system dependson the research effort
spent on analytical and numerical methods. The unsteady forces acting on the rotor
will have implications related to the structural design of in-service rotors. Additionally,
the transient lateral lift and longitudinal drag forces have not been explicitly treated
in literature. The relative contributions of these forces throughout a revolution may
help to develop a more complete understanding of the torque generation mechanisms
of Savonius-type VAWTs.

Savonius rotors have many advantages like-

- These turbines have a very low cut-in speed and can operate at low wind speed as it
has high starting torque.
- They are simple machines, so they are easy to build.
- This devices are insensitive to wind direction so very useful for the specific location
of wind directions.
- They are supposed to run even in case of “strong” winds (when most of fast running
wind turbines must be stopped).
- The components which convert the mechanical energy extracted from the wind into
electrical energy can be located at the surface level.
- They are simple machines, so they are easy to build as well as the Maintenance op-
erations are considerably simplified.

The main disadvantages of Savonius rotors are:

- The paddles are running at a speed of the same order as the wind velocity, which
makes it difficult to use them for the production of electric energy, especially when the
current produced is directly transferred on the electricity distribution network.
- This type of turbines are not competitive with other styles of wind turbines in terms
of their aerodynamic performance, and is only application specific because of its low
power conversion efficiency.

4
1.2 Objectives
The objective of the present investigation is set to develop a Savonius style wind
turbine with the intention of improving its performance in terms of power and torque
coefficients, starting characteristics, and also to analyze the flow behavior around this
turbine. This is accomplished by using modified rotors instead of conventional rotor.
The choice of the geometry of the elliptical-bladed Savonius rotor is because, of the
slenderness of the design, which allows the boundary layer attached to the wall of the
rotor to separate at a larger distance than classical Savonius rotor, and thus helps to
reduce the generation of secondary vortices and hence reduces the losses caused by the
rotor profile, leading to increase in its performance.

5
Chapter 2

LITERATURE SURVEY

2.1 Numerical Work

Sheldahl et al. (1978)[11] from the beginning of research on Savonius rotor,
TSR has been matter of interest for investigators. In the analysis, it has been observed
that the power coefficient is optimum at TSR ranging between 0.7-1.0. This study has
been further assured by several investigators and shows the effect of tip speed ratio
on power coefficient. From this fig.2.1 it can be concluded that the power coefficient
obtained is optimum at tip-speed ratios ranging between 0.7-1.0.

Fernando and Modi (1993) [18] have studied a numerical model, based on the
discrete vortex method, to predict performance and wake of a Savonius wind turbine is
discussed, for both stationary and rotating configurations. They found that the Savo-
nius rotor is not a pure drag device. This is logical as, at times, the rotor behaves like a
slender body with lift contributing to the power. They also studied the mathematical
model for viscous separated flow with an unsteady wake, developed here, represents a
versatile tool applicable to a large class of transient fluid dynamics problems. When
applied to a Savonius rotor it is able to predict performance and evolution of the wake
geometry with a reasonable accuracy.

Afungchui et al. (2010) [4] have explored the non-linear two-dimensional un-

Figure 2.1: Effect of TSR on Power Coefficient

6
steady potential flow over a Savonius rotor and developed a code for predicting its
aerodynamics performances. In the model developed, the rotor is represented in a
median plane by two semicircles, displaced along their common diameter. The two
semicircles can be considered to produce lifting effects. The dependence of the numer-
ical results to the time step is investigated. Given that the time step is quite small,
the inverse of the time step has been used. The unsteady pressures at points of the
baseline profileare presented and compared to the measurements of Chauvin et al.
(1989), for one complete turn of the rotor. They concluded form their study that the
validity of the code in predicting the average power and torque coefficients is however,
limited to tip speed ratios only as high as 0.6 due to increasing viscous effects. For
most practical purposes, the rotor runs for tip speed ratios greater than 0.6, thus it
could be recommended that, the implementation of an empirical scheme for viscous
effects can considerably improve the results.

Mohamed et al.(2011) [16] studied the optimal blade shape of a modified Savonius
rotor using an obstacle shielding the returning blade and found that the k-e turbulence
model can be employed for a quantitative analysis of the performance, provided a suf-
ficiently fine grid and sufficiently large computational domain are used. A performance
gain of at least 30 percent is found for the full operating range from 0.36 to 61.4. This
positive effect was also observed for the torque coefficient.

Akwa et al.(2012) [12] have studied the effects of different parameters on the perfor-
mance of Savonius rotors and have concluded the addition of end plates on a Savonius
turbine can greatly increase the maximum averaged power coefficient, CP. The turbine
also operates more efficiently at higher tip speed ratios. Also Savonius rotors with high
aspect ratios have low losses due to the effect of the tips of the buckets. The aspect
ratio growth of a Savonius rotor therefore has an effect similar to adding end plate. The
best value is 2.0. They also added that with increase of the rotor buckets number, it
decreases the ranges of values, for angular positions of the advancing bucket, where the
rotor moment is low since the probability of a rotor bucket to be in a position favorable
to the extraction of momentum from air flow increases and about the bucket spacing
they found that the null buckets spacing gives the best performance of a Savonius wind
rotor with semicircular profile buckets.

Roy and Saha(2012) [13] have undergone unsteady two-dimensional computa-

tional study in order to observe the effect of overlap ratios on static torque characteris-
tics of a vertical axis windturbine (VAWT).It was found that the increase in the static
torque with the increase of overlap ratio is mainly triggered by the increased pressure
on the concave side of the returning blade owing to the flow through the overlap. How-
ever, it is realized that as the overlap ratio increases beyond an optimum value, the
effective pressure on the concave side of the advancing blade reduces. Thus, at an
optimum overlap ratio of 0.20, the peak static torque characteristic of the turbine is
observed.

It was also observed that with the increase in the blade overlap, positive static
torque coefficients are recognized in the complete rotational cycle of this turbine. As
the overlap ratio increases at a value of 0.20, the large variation of the static torques at

7
 Figure 2.2: Effect of end plates on Cp with variation in TSR

Figure 2.3: Effect of overlap ratio on the average static torque coefficient for two bladed
VAWT

various angular positions is also reduced. An optimum mean static torque coefficient of
0.224 is obtained with D=0.20 at U= 10.44 m/s. The increase of the static torque with
the increase of overlap ratio is mainly due to increased pressure on the concave side of
the turbine returning blade. Moreover, it is seen that with an increase of overlap ratio
beyond 0.20, the mean static torque coefficient value decreases, although a positive
value of static torque coefficient is observed over the complete 360 degree rotation.
These reductions beyond D = 0.20 are mainly due to reduced pressure on the concave
side of the rotor advancing blade.

Zullaha and Lee(2011) [9] have studied a numerical modeling devoted to inves-
tigate the effect of wave on the performance and internal flow of the Savonius rotor
in the components of an oscillating water column (OWC) system used for the wave
energy capture. In that paper, the flow behavior is modeled by using the commercial
code ANSYS CFX. The numerical flow models have been elaborated and simulated
independently with different inlet geometries of a water chamber. The flow is assumed
to be two-dimensional (2-D), viscous, turbulent and unsteady. The commercial CFD
code is used with a solver of the coupled conservation equations of mass, momentum

8
 Figure 2.4: Performance analysis of the Savonius rotor angle

and energy, with an implicit time scheme. Turbulence is modeled with the k-e model.
Simulation result shows that the water velocity, differential pressure, and the best ef-
ficiency vary considerably depending on the Savonius rotor angle. CFD simulation
studies has shown the potential of the Savonius rotor with 20 degree blade angle in
terms of smooth running,higher efficiency and self-starting capability as compared to
the other Savonius rotor angles. The principal observations of the present findings can
be briefly stated as that for the higher blade curvature, the maximum force moves
towards to the tip of the blade because of the curve in the blade. Due to the change
inflow at the blade area, a higher curvature rotor gets a higher value of net positive
torque. The maximum torque is obtained at blade angle of 20 degree which is 3.85
kNm. This increase is due to the fact that the blade is able to capture or extract the
energy more effectively from the returning flow when water flows out of the rear cham-
ber. This in turn increases the average net positive torque and hence the performance
of the turbine. Decreasing the blade angle further has an unfavorable impact on the
output power and also by increasing the blade curvature. The simulated wave energy
flux is 13.77 kW/m. The highest total efficiency is 18.58 percent. Therefore the 20
degree rotor angle Savonius rotor converts 2.22 kW/m of the total energy flux.

Zhou and Rempfer(2015) [21] have studied under the realizable k-e model for
turbulence. Numerical simulation results were compared with experiment data. Sepa-
ration of the flow at the blade tips is well resolved. They concluded that the conven-
tional semi-circular shape is not the best option for power generation. In contrast to
the reputation of these turbines as ’drag-based devices’, they found that in fact a large
portion of the power they produce comes from the circulation inherent in the attached
flow on their blades.

Jaohindy et al.(2013) [14] have analyzed the transient forces of a Savonius verti-
cal axis wind turbine (VAWT) using computational fluid dynamics (CFD). Unsteady,
Reynolds Averaged Navier-Stokes CFD simulations were conducted using two different
rotors and two simulation methodologies. The first rotor had an aspect ratio of 1.1 and
simulations were conducted by incorporating the dynamic equations of the rigid-rotor
motion using a one degree-of-freedom (1-DOF) module to evaluate the rotor’s free

9
Figure 2.5: Evaluation of the resultant force coefficients developed by rotor A and rotor
B.

motion. The second rotor had an aspect ratio of 0.7 and simulations were conducted
by specifying affixed rotational velocity. The predicted torque andpower curves were
validated using experimental data. The transient forces from the two rotor blades, the
resultant forces, and the longitudinal drag and lateral lift coefficients were evaluated
using the two complementary data sets. The lateral force coefficient had a large con-
tribution to the rotor torque at low azimuth angles and at low tip speed ratios. At
tip-speed ratios above 0.6, the effect of the longitudinal drag force on the rotor oper-
ation increased and the lateral lift force contribution decreased. The resultant force
angle was shown to become more acute due to the increase in the longitudinal force
with increasing TSR. The numerical simulations allowed the transient nature of the
forces acting on a Savonius rotor to be studied. The simulations with rotor permitted
the forces acting on each rotor blade to be analyzed. The resultant force generated
by one of the two rotors was compared to the resultant force generated by the other.
The resultant force angle was calculated and was shown to become shallower with
increasing TSR. The intensity of the longitudinal drag and lateral lift coefficients de-
veloped by the Savonius rotor are dependent on the azimuth angle and the tip speed
ratio of the rotor. The contribution of the lateral coefficient has a similar magnitude
to the longitudinal coefficient at low tip speed ratios and begins to reduce after the
peak power condition of the rotor. This demonstrates that a Savonius rotor does not
generate its torque purely from drag effects at low to moderate tip speed ratios. At
high tip speed ratios, the relative effect of the lateral lift coefficient is reduced and the
longitudinal drag coefficient is the dominant driver of the rotor motion. It is suggested
that a Savonius rotor should not be referred to solely as a drag-driven rotor, but as a
VAWT whose torque is generated by the combined effects of lateral and longitudinal
forces.

Kacprzak et al.(2013) [20] studied about the performance of a Savonius wind tur-
bine with constant cross-sections and numerically examined by means of quasi 2-Dflow
predictions executed in ANSYS. They found that all examined rotors attain maximum

10
Figure 2.6: Comparison of average co-efficient of power at different TSR to the exper-
imental results

power at a tip speed ratio of 0.8 , in terms of the power coefficient the Bach-type rotor
is superior to the other tested geometriesand at the same time the elliptical Savonius
rotor exhibits better power characteristics than the Classical one, in the range of tip
speed ratios between 0.2 and 0.4 the elliptical-bladed Savonius rotor performs better
than the Bach-type and the Classical Savonius rotor, the wake produced by the tur-
bines seems to be limited, when the overlap is introduced. That is why, in case of the
Bach-type rotor, where the overlap was not present, the evoked wake is spread over
a larger area and its energy is dissipated over a larger frequency range, the tip speed
ratio parameter can be used in order to predict the frequency of fluctuations, which
appear in the wake. This is very important information when multiple turbines are
supposed to operate in wind farms. Their angular velocities and the phase orientation
can be adjusted with respect to each other to maximize the overall efficiency.

Roy and Saha(2013) studied the performance parameters of the Savonius rotors
and found that the performance of the Savonius rotor mainly depends upon various
parameters such as tip speed ratio, overlap ratio, aspect ratio, blade angle, blade
shape, and number of blades. Moreover, the performance can also be improved by in-
troducing various augmentation techniques such as curtain design, obstacle shielding,
multi-staging, and nozzles. Their studies revealed:
-The hybrid DES/k–w SST turbulence model has shown the benefits of capturing the
unsteady effects more precisely. However, 3-D k-w SST turbulence model also shown
a close prediction capability of the rotor performance.
-The 2D computational study always over-predicts the performance, whereas the 3D
study shows a better prediction capability.
-The finite volume method appears to be superior as compared to finite difference and
finite element methods in solving the complex flow phenomena over the Savonius rotor.

Roy and Saha(2014) [1] analysis the performance and starting characteristics of
the newly developed two bladed savonius rotor turbine. This novel shape of the tur-
bine blade is evolved from a series of experiments and unsteady simulations on classical
Bach type SSWT by varying the overlap distance from 0 to 0.5d (d is the blade chord)

11
 Figure 2.7: Cp vs TSR

and blade arc angle from 90 to 165. The simulation studies show an improved design
of the Bach type SSWT at an overlap distance of 0.4d and a blade arc angle of 135.
They test edit in an open type test section and its performance is assessed in terms
of power and torque coefficients. Experiments have also been conducted with other
standard blades such as semi-circular, semi-elliptic, Benesh and Bach types in order
to have a direct comparison. After the comparison they found the power conversion
efficiency gains of 3.3 percent, 6.9 percent, 19.2 percent and 34.8 percent over modified
Bach, Benesh, semi-elliptical, and conventional savonius rotor respectively. With the
increase of Reynolds number the dynamic performance index i.e. the power coeffi-
cient increases up to a certain limit of Re = 1.2*105, beyond which it again decreases.
In contrary, with the increase of Reynolds number, the starting performance index i.e.
the static torque coefficient increases irrespective of the magnitude of Reynolds number.

Banerjee et al.(2014) [1] studied the unsteady flow analysis around an Elliptical
bladed Savonius style wind turbine. They used SST k-w turbulence model and per-
form 2D simulation on elliptical blade in ANSYS fluent. The power and the Torque is
calculated at the rotating condition, and the result obtained are validated with the ex-
perimental data. From these unsteady simulations, they found that the elliptical bladed
SSWT shows a performance improvement of 10.7 percent over classical semi-circular
shaped SSWT. However, from the experimental study, a 17.4 percent performance im-
provement has been recorded. The maximum value of Cp is found to be 31 percent
at a tip speed ratio of 0.8.They also observed that the trend of Ct is decreasing as tip
speed ratio increases.

Sanghyeon Kim and Cheolung Cheong(2015) [21] have studied the aerody-
namic noise characteristics of Savonius wind turbines using hybrid computational aero-
acoustics techniques, and proposed a low-noise designs based on the understanding of
the noise generation mechanism. They analyzed the flow field around the turbine in
detail by solving threedimensional unsteady incompressible Reynolds-averaged Navier
Stokes equations using computational fluid dynamics techniques. The aerodynamic
noise radiating from the wind turbine was predicted using the Ffowcs Williams and
Hawkings equation with the obtained flow field information and identifed two distinct

12
 Figure 2.8: Cp and Ct vs TSR

Figure 2.9: Variation of SPL with Frequency

harmonic noise components, firstly the blade passing frequency (BPF) and secondly
the harmonics with a fundamental frequency that is much higher than the BPF were
identified in the predicted noise spectrum. The origin of the higher harmonic compo-
nents was found to be related to vortex shedding from the rotating turbine. Based on
this finding, they proposed a low-noise design for Savonius wind turbines uses S-shaped
blades and find the reduction in noise by upto 2.7dB.

2.2 Experimental Work

Fernando and Modi(1989) [17] suggested that the Savonius rotor is not a pure
drag device, lift forces also contribute to the power. They also found Presence of a
vortex filament at the center of the rotor and its shedding which is predicted by the
analytical model is also confirmed through the flow visualization study. Also the wake
of the stationary blade shows periodic behavior over the range beta = 60-150 degree.
The vortex shedding frequency corresponds to a Strouhal number of approx. 0.12. The
theoretically predicted peak starting torque coefficient occurs at beta approx 30 degree
which is confirmed by the experimental results. Starting torque becomes negative over

13
 Figure 2.10: Average starting torque characteristics

the range of beta = 130 - 180 degree. For rotating blades they found existence of the
central vortex and its shedding is evident for tip-speed ratio less than one. However,
for TSR greater than 1 theoretical results do not show presence of the central vortex.

Fujisawa(1996) [19] worked on the flow fields in and around Savonius rotors at
various overlap ratios. He has investigated to clarify the effect of overlap on the flow
mechanisms. Measurements of phase-averaged velocity distributions were carried out
using particle imaging velocimetry with a conditional sampling technique, and the re-
sults were compared with numerical calculations by a discrete vortex method. The
measured velocity distributions indicate clearly the effect of the overlap both on the
flow through the overlap and on the formation region of the vortices downstream of the
rotor. The measured velocity distributions for a stationary rotor indicate an increase
in flow rate through the overlap from the advancing side to the returning side with
increase of the overlap ratio. These phenomena cause a decrease in the wake width of
the rotor, and an increase in the presence on the concave side of the returning blade,
which results in an improvement in the torque performance. Although numerical cal-
culations can predict the basic features of the flow variations with rotor angle, the flow
field inside and downstream of the rotors is poorly predicted for a stationary rotor,
especially at small rotor angles, which is caused bythe numerical assumption of flow
separation at the tips of the blades, and the three dimensional nature of the flow. For a
rotating rotor, the agreement with the measurement is much improved compared with
that for a stationary rotor, due to the better matching of the flow with the numerical
assumptions.

Altan and Atılgan(2008) [3] designed a curtain to increase the low performance
of the Savonius wind rotor, a type of vertical-axis wind rotor, and the effect of this
curtain on the static rotor performance has been analyzed both experimentally and nu-
merically. Designed to prevent the torque that occurs on the convex blade of the rotor
in the negative direction, this curtain has been placed in front of the rotor. Experi-
mental measurements and numerical analysis have been conducted when the Savonius
wind rotor is with and without curtain. The static torque values of the rotor have been
measured by experiments and calculated by numerical analysis, and finally they have
been compared. The best results have been obtained by means of the rotor with cur-
tain. Low static torque values have been obtained with the short curtain dimensions,

14
Figure 2.11: Velocity distributions in and around stationary Savonius rotors measure-
ments and numerical calculations

while a considerable increase has been acquired in the static torque values with the
long curtain dimensions. When the rotor position is 45, 60 and 90 degree the static po-
sition of the Savonius wind rotor, the static torques have been found for the rotor with
and without curtain through the experimental and numerical analysis made by using
experimental measurements and Fluent 6.0 program. Then they have been compared
through Figures. It is seen both from the experiment and the numerical analysis that
the static torque values at 45 degree are the highest. The major outcome of the study
is that, this has shown that when the rotor is static for the same positions, the rotor
with curtain can provide a better performance than the rotor without curtain. In the
rotor with 16 curtain, the best torque values have been obtained with the curtain 1,
which is the longest one. On the other hand, it has been indicated by experiments and
numerical analysis that the best performance for beta=15 degree and alpha= 45 degree
has been obtained with curtain 1 when rotor blade angle 60 degree. When the rotor
position is 90 degree, most of the fluid guided by the curtain escapes from over the
convex blade and thus creates an adverse torque effect. This is the main reason why the
lowest torque values have been obtained at 90 degree in the present study. However,
more fluid escapes from between the curtain-end and rotor blade, thus leading to a
decrease in the value of torque exerted. When the rotor position 45 degree, more fluid
escapes from between curtain-end and rotor blade, and the amount of torque that is
exerted decreases in this α case.

Altan and Atilgan(2010) [6] designed the system to prevent the negative torque
on the convex blade of the rotor, this curtain has been placed in front of the rotor, and
performance experiments have been carried out when the rotor is with and without
curtain. It has been determined fromhere that a significant increase can be achieved in

15
Figure 2.12: Design of curtain arrangement placed in front of the Savonius wind rotor.

Figure 2.13: The torque change obtained through experiment for β = 10 and 15 between
α = 30 and 60 when rotor position is 45 degree

16
Figure 2.14: Comparison of the power changes with the rotor without curtain and the
rotor with different curtain types for α = 45 degree and β = 15 degree

the rotor performance by means of the curtain design. Experiments have been made at
different curtain angles (α,β) of three different curtain arrangements and the optimum
curtain angle has been found as =45 degree and β=15. When the rotor is in static
state, curtain 1, which is the longest curtain has yielded to the best torque values.
Moreover, the torque values obtained from the rotor with each of the three curtains
are higher than the ones obtained from the rotor without curtain.

Most of the fluid guided by the curtain when the rotor’s position is 90 degree escapes
from over the convex blade and thus leads to an adverse negative torque effect. This
is the essential reason why it is when the rotor’s position is 90 degree that the lowest
torque value is obtained. When the rotor’s position is 45 degree, however, more fluid
escapes from between the end of the curtain and blade and the value of the applied
torque decreases.

D’Alessandro et al. (2010) [23] have studied the complex flow field developing
around a Savonius wind rotor and evaluated its performance. A mathematical model
of the interaction between the flow field and the rotor blades was developed and val-
idated by comparing its results with data obtained at Environmental Wind Tunnel
(EWT) laboratory of the ’Polytechnic University of Marche’. Major findings were the
use of dimensionless quantities allowed to verify the Reynolds number effect on the
Savonius rotor’s performance. A computational methodology able to calculate the flow
field around the rotor was also developed. RANS equations were solved in order to
obtain accurate information about the flow field. The rotor blades were treated as rigid
bodies and their behavior were modeled by means of the Second Cardinal Equation
of Dynamic.The experimental data were used to validate the developed computational
methodology. The comparison of performance data obtained by numerical simulations
and experimental measurements shows very good agreement with the practical results

17
obtained.

Banerjee et al.(2014) [1]studied the unsteady flow analysis around an Elliptical

bladed savonius style wind turbine. They performed the wind tunnel experiments with
a elliptical profile blade with the inlet velocity of around 6.2 m/s and found that the
maximum power conversion efficiency is found to be 0.23 at a tip speed ratio of 0.71.
They used SST k-w turbulence model and perform 2D simulation on elliptical blade in
ANSYS fluent. The power and the Torque is calculated at the rotating condition, and
the result obtained are validated with the experimental data. From these simulations,
they found that the elliptical bladed SSWT shows a performance improvement of 0.7
percent over semi-circular shaped SSWT. From the experimental study, a 17.4 percent
performance improvement has been recorded.

Roy et al.(2014) [1] have studied and analysis the performance and starting char-
acteristics of the newly developed two-bladed Savonius rotor turbine. This novel shape
of the turbine blade is evolved from a series of experiments and unsteady simulations
on classical Bach type SSWT by varying the overlap distance from 0 to 0.5d (d is the
blade chord) and blade arc angle from 90degree to 165degree. The simulation studies
show an improved design of the Bach type SSWT at an overlap distance of 0.4d and
a blade arc angle of 135degree. They tested it in an open type test section and its
performance is assessed in terms of power and torque coefficients. Experiments have
also been conducted with other standard blades such as semi-circular, semi-elliptic,
Benesh and Bach types in order to have a direct comparison. After the comparison
they found the power conversion efficiency gains of 3.3, 6.9, 19.2 and 34.8 over modified
Bach, Benesh, semi-elliptical, and conventional rotor respectively. With the increase of
Reynolds number the dynamic performance index i.e. the power coefficient increases
up to a certain limit of Re = 1.2 * 105, beyond which it again decreases. In contrary,
with the increase of Reynolds number, the starting performance index i.e. the static
torque coefficient increases irrespective of the magnitude of Reynolds number.

2.3 Summary
After reviewing the literature it is found that, many researchers have carried
out numerical and experimental investigations to analyze and improve the design and
performance of Savonius rotors. Improvements in the performance and static torque
characteristics were achieved with the use of deflectors, guide vanes, or curtains placed
upstream to the turbines blades. However, these studies mostly revolved around semi-
circular blades. A few investigators have carried the study on helical and twisted blade
(Kamoji M A, Kedare S B, Prabhu S V, 2009). Although noticeable gains in perfor-
mance were achieved, the designs reported seemed to have been more complex. In this
aspect, studies with modified blade shapes like elliptical, batch type etc. were also
attempted. Computational studies using different turbulence models is also reported
to predict the performance and flow behavior. Elliptical style Savonius blade provides
better result as compared to conventional Savonius blade as per the work done by
Banerjee et.al. (2014), so more emphasis on elliptical blade is required.

18
Chapter 3

PROBLEM FORMULATION

The problem was developed to study the aerodynamic characteristics of the Var-
ious elliptical Savonius rotors by slicing the ellipse at various angles and compare the
result obtained for 2-D models using unsteady analysis. Also the research was taken
ahead on the optimal blade profile by using two newly designed guide plates and the
aerodynamic characteristics were studied under the same conditions.
The basic mathematical formulations required to study the aerodynamics of Savonius
rotors under unsteady conditions are:-

Average Coefficient of Power,

Tw
Cp = (3.0.1)
0.5pAU 3

Average Coefficient of Torque,

4T
Ct = (3.0.2)
pU 2 D2 H
Tip Speed Ratio, λ
wR
λ= (3.0.3)
U
Aspect Ratio, α
H
α= (3.0.4)
D

The governing equations that were solved during the simulations are:

(a) Continuity Equation:

∂p ∂(pu) ∂(pv) ∂(pw)
+ + + =0 (3.0.5)
∂t ∂x ∂y ∂z

The first term in the left hand side of the equation is zero here, because we have
assumed the density of the fluid (air) is constant in the entire simulation.

19
 Figure 3.1: Classical semi circular profile of the Savonius rotor

(b) Momentum Equation:

Du ∂p ∂ 2u ∂ 2v ∂ 2w
p = pg − + µ( 2 + 2 + ) (3.0.6)
Dt ∂x ∂x ∂y ∂z 2

Dv ∂p ∂ 2u ∂ 2v ∂ 2w
p = pg − + µ( 2 + 2 + ) (3.0.7)
Dt ∂y ∂x ∂y ∂z 2
Dw ∂p ∂ 2u ∂ 2v ∂ 2w
p = pg − + µ( 2 + 2 + ) (3.0.8)
Dt ∂z ∂x ∂y ∂z 2

3.1 Computational Domain

3.1.1 For Elliptical Rotor
In the present study, the classical semi circular blade configuration is depicted in
Fig.3.1 describing the important geometrical factors. The modified blade configuration
(Fig.3.2) is featured by spline connecting a set of three characteristic points[22]. The
terminal points of the blade viz., A(0.03m, 0m) and C(-0.2m, 0m) remain unaltered.
The third characteristic point B, which is chosen on the tangent to the semi- circle
at a point where the height from the diameter is the radius and the coordinates are
B(-0.14m,0.1m). The common geometrical parameters of the rotor geometry under
study have been presented in fig.3.3

3.1.2 For Designs of Guide Plates

For further study, two designs of guide plates are proposed as shown in fig.3.4.
In both the designs, the upper plate is inclined with 15 degree to the horizontal as

20
 Figure 3.2: Modified rotor geometry under study

Figure 3.3: Geometrical parameters of the rotor blade under study

21
 Figure 3.4: Rotor with Guide plate design A and design B

suggested by (Altan and Atilgan 2010). The length of the upper plate is kept as 4R
in order to give sufficient developing length and hence reducing the entrance effects.
The upper plate is aimed to collect more amount of wind and generate a jet to the
concave side of the advanced blade. The lower part of the design B has collection area
greater than the design A. The exit jet from the lower passage has width of 0.5R for
both the designs. The position of the exit jet center has a distance 0.75R for design
A and 1.25R for design B. The exit jet position is selected to prevent the convex side
of returning plate from generating negative torque at rotor angle between 90 and 170
degree. For clear observation the 3D isometric geometry of design B is also included.

3.2 Details of Domain Discretization

The computational domain is discretized using unstructured domain by using
quad cells to solve the equations using finite volume method. An overview of the com-
putational domain and its discretization is shown in fig. It consist of two parts: the
first part is fixed mesh and the other is rotating mesh as shown in fig. .The rotating
mesh is bounded by the interface of both the parts. The rotating region has dia 0.8
m and is placed at the center of the domain. The dimension of the domain is 7 m *
4 m. The mesh generation is done in ANSYS fluent software only and is shown in fig.
Near the blades, fine layer are adjusted in order to capture the flow properties near the
blade surfaces. The minimum size of each cell is fixed as 0.001 m for all the models.
There is uniform concentration of cells at the interface of stator and rotor.
A grid independence study is carried in between triangular cells and quadrilateral cells
to approach the realizable grid distances and good resolution that produces reasonable
computational results and found quadrilateral cells as more competitive as shown in
Fig. and Fig.. Mesh refinement was done to model the geometry of the Savonius ro-
tors and no noticeable variations are noticed in the final solutions with increasing the
number of cells.

22
Figure 3.5: 3D Isometric geometry of Design B

Figure 3.6: 3D Isometric geometry of Design B

23
 Figure 3.7: Boundary Conditions

3.3 Computational methodology

3.3.1 Simulation Model
The flow around the SSWTs is assumed to be fully unsteady and turbulent, op-
erating at a free stream wind speed of 6 m/s. The dynamic simulations are carried out
by assigning a certain rotational rate (w in rad/s) to the rotating zone. To predict the
performance of the turbine in terms of torque coefficient (Ct ) and power coefficient (Cp )
with respect to TSR. The SST k-w turbulence model is found to be suitable for analysis
as compared to other turbulence models. It is due to fact that the SST k-w turbulence
model is a blended model of k-w and k-e turbulence models. The reason for the selec-
tion of this model was, its better accuracy with practical results as is mentioned in the
literature (Afungchui et al., 2010: D’Alessandro et al., 2010; Akwa et al., 2012). SST
k-w model is also known to give highly accurate predictions of flow separation under
adverse pressure by an implementation of the transport effects on the formulation of
the eddy viscosity (Sargolzai and Kainifer, 2009). So, it is utilized whenever a high
boundary layer accuracy is required, which is the case in these investigations.

3.3.2 Boundary Conditions

Once the meshing is accomplished, the boundary conditions are assigned to the
domain. The inlet is given as velocity inlet (V= 6 m/s), outlet as pressure outlet (at-
mospheric pressure), turbine models as no-slip boundary condition, and the rotating
zone is assigned as a sliding mesh at a certain rotational speed with respect to TSR.
The upper and bottom side of the domain is assumed to be symmetry boundary con-
ditions. As shown in Fig.3.7.

24
3.3.3 Details of Setup Solver
A pressure based transient FVM solver, ANSYS Fluent is used to discretize the
equations. The spatial discretization of the conservative equations is treated with
2nd order upwind scheme and the temporal terms of the conservative equations are
discretized using 1st order fully implicit temporal scheme. Good solution stability is
ensured through the pressure- velocity coupling with the SIMPLE method (Semi Im-
plicit Linked Equations). For the solution iteration, the time step size and the number
of iterations per time step are taken as 0.01 and 20, respectively.

3.4 Calculation of performance coefficients

As the simulation is carried out with the unsteady flow assumptions, the perfor-
mance coefficients are averaged to capture a more accurate time averaged value. In
further report it is shown that the variations of torque coefficients at different time
intervals. It is observed that after the initial period, it follows almost a cyclic path,
and hence, averaged to give a more accurate value.

25
Chapter 4

RESULT AND DISCUSSION

With the SST k-w turbulence model, at 6 m/s, 2D unsteady simulations were
carried out on all the elliptical blades. After finding the optimal blade profile, using dif-
ferent guide plates unsteady simulation were again carried out. Fig.4.1 and shows the
torque coefficient respectively at different TSR for Elliptical savonius rotor and the new
designs also. It indicates a gradual decrease in Ct as rotational speed increases. From
the figures, it is clear that design B has the highest value torque coefficient compared
with elliptical blade design. The static torque coefficient increases with high values in
the range of rotor angle of 90-135 for both the guide plates because the generated wind
jet in the new designs impinge into the concave side of the return blades.

The variation of the coefficient of power has also been studied for all the 2-D ellipti-
cal style rotors as well as for rotor with new guide plates by varying the tip speed ratio
under transient dynamic mode and unsteady k-w turbulent model. The comparison of
the numerical values of coefficient of power of elliptical rotors along with the both the
guide plates with respect to tip speed ratio, has been shown in Fig.4.2. The result shows
that k-w SST turbulent model in transient method gives similar values for structured
grid. From the Fig, it is clear that the rotor using guide plate design B has highest
value of power coefficient when compared to other rotors. The directed wind harvested
in the passage of design B includes a smooth turning and hence, a reduction of the
lost energy compared with the design A. Also, the momentum forces directed to the
concave sides of advanced and return blades in design B is greater than that in design A.

Table 4.1: Variation of Torque Coefficient with Tip Speed Ratio

TSR Classical Rotor Modified Rotor Rotor with Guide Plate A Rotor with Guide Plate B
0.2 0.235 0.2 0.25 0.28
0.4 0.18 0.24 0.23 0.25
0.6 0.15 0.15 0.18 0.2
0.8 0.14 0.17 0.19 0.2
1.0 0.07 0.12 0.14 0.148
1.2 0.04 0.1 0.13 0.131

26
Figure 4.1: Variation of Torque Coefficient with Tip Speed Ratio

Figure 4.2: Variation of Power Coefficient with Tip Speed Ratio

27
 Table 4.2: Variation of Power Coefficient with Tip Speed Ratio
TSR Classical Rotor Modified Rotor Rotor with Guide Plate A Rotor with Guide Plate B
0.2 0.046 0.044 0.051 0.051
0.4 0.075 0.086 0.087 0.091
0.6 0.08 0.095 0.103 0.109
0.8 0.087 0.11 0.13 0.14
1.0 0.064 0.1 0.12 0.129
1.2 0.044 0.089 0.108 0.11

Figure 4.3: Variation of Static Torque Coefficient with Rotor Angle

4.1 Starting Characteristics

The Classical rotor shows a wide range of variation in the static torque coef-
ficient. The major disadvantage of this configuration is that it experiences negative
torque coefficient in the range of 60 degree – 75 degree i.e., no self starting is possible
when the rotor is aligned in any angle in this range with respect to the wind velocity.
The modified rotor is capable of self starting from any angular position with respect to
the wind direction. This is because the negative torque region completely disappears
for this configuration as depicted in fig.4.3

28
 Figure 4.4: Velocity contour for semicircular profile at 90 degree

4.2 Velocity Flow Field

The velocity distributions at the time when the turbines reached their peak torque
with TSR = 0.8 has been presented in following Figures. The asymmetrical nature of
the flow above and below the rotors are evident from the velocity plots which is due to
the rotation of the turbine. Another important observation is the recirculation flow on
the convex side of the advancing blade. The recirculation flow plays a vital role in the
performance of the turbine. Moving the recirculation zone downstream enhances the
performance of the rotor i.e., the flow needs to be attached to the walls of the rotor
blades as long as possible in order to get the best performance. Another important
flow phenomena is the overlap flow. It is the flow through the overlap gap ratio of the
rotor. The velocity of the overlap flow is responsible for pushing the returning blade
in the advancing direction especially when the rotor is experiencing minimum torque.

After study the optimal blade profile, velocity flow field of savonius rotor using
guide plates designs were also studied. Fig4.8 and Fig4.9 shows the velocity contours
around the rotor with the design A and design B respectively. It can be noticed that
strong wakes behind the rotor are generated. The upstream wind enters the designs
from two stream; one generates the first wind jet to the concave side of advancing blade
and the other generates the second wind jet to the concave side of returning blade. The
flow direction guiding of such design is of course a subsidizing tool to generate a strong
active momentum which is responsible for enhancing the generated torque from such
new design. When comparing both the designs, one can observe the design B generates
asymmetric large wakes with strong vortex shedding.

29
Figure 4.5: Velocity contour for novel shaped profile at 90 degree

Figure 4.6: Velocity contour for semicircular profile

30
 Figure 4.7: Velocity contour for novel shaped profile

Figure 4.8: Velocity contour for novel shaped profile with guide vanes design A

31
 Figure 4.9: Velocity contour for novel shaped profile with guide vanes design B

4.3 Pressure Distribution Analysis

From the pressure contour, it is found that the low pressure region is formed
near the advancing blade and a high pressure region is found near the returning blade.
The high pressure on the returning blades causes the negative torque generation in the
turbine, which adversely affects the power generation capability of the turbine. The
negative impact of this pressure is restored to some extent by the returning flow behind
the advancing blade which impacts the concave side of the returning blade.

In order to reduce the negative effect of pressure on the returning blade total pres-
sure contours for rotor with guide plate designs were also studied in fig4.15 and fig4.15.
From the figure, it is clear that pressure difference between the concave and convex
sides of blade participates, with the most part of the generated torque. It can be seen
that the pressure difference for the design B is higher than design A. The pressure
increment on the concave side of returning blade is due to the presence of wind jet
which results in higher torque.

32
Figure 4.10: Pressure contour for semicircular profile at 90 degree

Figure 4.11: Pressure contour for novel shaped profile at 90 degree

33
Figure 4.12: Pressure contour for semicircular profile

Figure 4.13: Pressure contour for novel shaped profile

34
Figure 4.14: Pressure contour for novel shaped profile with guide vanes design A

Figure 4.15: Pressure contour for novel shaped profile with guide vanes design B

35
Chapter 5

CONCLUSION

5.1 Key Observations

The present study attempts to improve the performance of savonius style wind
turbine in terms of power and torque coefficient, starting characteristics, and also to
analyze the flow behavior around the turbine blades. Numerical studies have been car-
ried out on savonius rotor with different modifications to enhance the turbine power.
The outcomes of the study have been listed below :
- The modified savonius rotor is more efficient as compared to classical blade profile
having same chord length.
- The maximum CP of modified blade profile without using any guide plates is 0.11
and the value achieved is in the TSR 0.7-0.9.
- The use of guide plates helps in reducing the negative torque of the returning blade
which results in the increment of the power coefficient.
- The result showed that a new design with curved passage shape has the best perfor-
mance, with noticeable effects of the wind speed on its performance.
- The use of designed guide plates also results in better performance of SSWT even at
higher tip speed ratios.
- With the use of guide plates, it has been possible to increase the speed of the wind
entering the rotor, without changing the original shape of the rotor.
- The strong vortex shedding and wide wake generated around and behind the new
designs needs more attentions from the point of view of noise generation and dynamic
loads in the design of wind turbine farm.

5.2 Scope for future work

The models were simulated numerically, so to validate the results experimental
study can be performed in the wind tunnel. Unstructured grids were used in the present
study. So implementation of the structured grids can be studied to have a more accurate
analysis of the flow behavior. The other parameters like aspect ratio of the rotor, no.
of buckets of the rotors, the spacing between the buckets can be studied and the results
can be compared to get the better designing performance. Other blade profiles using
the designed guide plates can be developed in order to get better efficiency.

36
References

[1] Banerjee A, Roy S, Mukherjee P, and Saha UK. Unsteady flow analysis around an
elliptic-bladed savonius-style wind turbines. ASME Gas Turbine India Conference,
pages 2014–814, 2014.

[2] Chauvin A and Benghrib D. Drag and lift coefficient evolution of a savonius rotor.
Experiments in Fluids, 8:118–120, 1989.

[3] Altan BD and Atilgan M. An experimental and numerical study on the im-
provement of the performance of savonius wind rotor. Energy Conversion and
Management, 49:3425–3432, 2008.

[4] Afungchui D, Kamoun B, and Helali A. The unsteady pressure field and the
aerodynamic performances of a savonius rotor based on the discrete vortex method.
Renewable Energy, 35:307–313, 2010.

[5] Altan B D and Atilgan M. The use of a curtain design to increase the performance
level of a savonius wind rotors. Renewable Energy, 35:821–829, 2010.

[6] Altan B D and Atilgan M. The use of a curtain design to increase the performance
level of a savonius wind rotors. Renewable Energy, 35:821–829, 2010.

[7] Al-Bahadly I. Building a wind turbine for rural home. Energy for Sustainable
Development, 13:159–165, 2009.

[8] Paraschivoiu I. Wind turbine design: with emphasis on darrieus concept. Presses
Inter Polytechnique, Canada, 2002.

[9] Ross I and Altman A. Wind tunnel blockage corrections: review and application to
savonius vertical-axis wind turbines. Journal of Wind Engineering and Industrial
Aerodynamics, 99:523–38, 2011.

[10] Menet JL. A double-step savonius rotor for local production of electricity: a design
study. Renewable Energy, 29(11):1843–62, 2004.

[11] Abraham Jp, Plourde BD, and Mowry GS. Summary of savonius wind turbine
development and future applications for small- scale power generation. Journal of
Renewable and Sustainable Energy, 4:1–21, 2012.

[12] Akwa JV, Vielmo HA, and Petry AP. A review on the performance of savonius
wind turbines. Renewable and Sustainable Energy Reviews, 16:3054–3064, 2012.

37
[13] Akwa JV, Junior GA S, and Petry AP. Discussion on the verification of the
overlap ratio influence on performance coefficients of a savonius wind rotor using
computational fluid dynamics. Renewable Energy, 38:141–189, 2012.

[14] Kacprazak K, Liskeiwicz G, and Sobczak K. Numerical investigation of conven-

tional and modified savonius wind turbines. Renewable Energy, 60:578–588, 2013.

[15] Kamoji MA, Kedare SB, and Prabhu V. Performance test on helical savonius
rotor. Renewable Energy, 34:521–529, 2009.

[16] Mohamed MH, Janiga G, and Thevenin D. Optimal blade shape of a modified
savonius rotor using an obstacle shielding the returning blade. Energy Conversion
and Management, 52:236–242, 2011.

[17] Fernando MSUK and Modi VJ. On the performance of the savonius wind turbine.
ASME Journal of Solar Energy Engineering, 111:71–81, 1989.

[18] Fernando MSUK and Modi VJ. Unsteady aerodynamics and wake of the savonius
wind turbine: a numerical study. Journal of Wind Engineering and Industrial
Aerodynamics, 46:811–816, 1993.

[19] Fijusawa N. Velocity measurements and numerical calculations of flow fields in

and around savonius rotors. Journal of Wind Engineering and Industrial Aerody-
namics, 59:39–50, 1996.

[20] Jaohindy P, Mctavish S, and François Grade. An analysis of the transient forces
acting on savonius rotors with different aspect ratios. Renewable Energy, 55:286–
295, 2013.

[21] Hayashi T and Hara Y, Li Y. Wind tunnel tests on a different phase three-stage
savonius rotor. JSME International Journal Series B, 48:9–16, 2005.

[22] Nilavarasan T and Dheenadhayalan J. Numerical investigation of savonius wind

turbines with modified blade configurations. International Conference on Energy
Systems and Developments, 21, 2015.

[23] D’Alessandro V, Montelpare S, Ricci R, and Secchiaroli A. aerodynamics of a

savonius wind rotor: a new computational approach for the simulation of energy
performance. Energy, 35:3349–63, 2010.

38