optimum design configuration of savonius rotor through wind tunnel experiments

ARTICLE IN PRESS

Contents lists available at ScienceDirect

Journal of Wind Engineeringand Industrial Aerodynamics

Journal of Wind Engineering and Industrial Aerodynamics 96 (2008) 1359– 1375

0167-61

doi:10.1

� Corr

E-m

journal homepage: www.elsevier.com/locate/jweia

Optimum design configuration of Savonius rotorthrough wind tunnel experiments

U.K. Saha a,�, S. Thotla a, D. Maity b

a Department of Mechanical Engineering, Indian Institute of Technology Guwahati, Guwahati 781 039, Indiab Department of Civil Engineering, Indian Institute of Technology Guwahati, Guwahati 781 039, India

a r t i c l e i n f o

Article history:

Received 30 July 2006

Received in revised form

27 February 2008

Accepted 14 March 2008Available online 7 May 2008

Keywords:

Savonius rotor

Twisted blade

Wind tunnel

Power coefficient

Rotational speed

Valves

05/$ - see front matter & 2008 Elsevier Ltd

016/j.jweia.2008.03.005

esponding author. Tel.: +91361 2582663;

ail address: [email protected] (U.K. Saha).

a b s t r a c t

Wind tunnel tests were conducted to assess the aerodynamic

performance of single-, two- and three-stage Savonius rotor

systems. Both semicircular and twisted blades have been used

in either case. A family of rotor systems has been manufactured

with identical stage aspect ratio keeping the identical projected

area of each rotor. Experiments were carried out to optimize the

different parameters like number of stages, number of blades (two

and three) and geometry of the blade (semicircular and twisted).

A further attempt was made to investigate the performance of

two-stage rotor system by inserting valves on the concave side of

blade.

& 2008 Elsevier Ltd. All rights reserved.

1. Introduction

In recent years, an interest in wind energy has been growing and many researchers haveattempted the development to introduce cost-effective, reliable wind energy conversion systems allover the world. In practice, however, there are many difficulties, to introduce the wind turbine intothe community because of less wind energy source, profitability, noise emission, etc. Therefore, thedecentralization or local clusterization of renewable energy plant made it attractive not only todeveloping area, where a lot of people do not yet have access to conventional electricity service, butalso to an urban area where one can make better living space for future generation (Shikha et al.,2003; Grinspan et al., 2004; Menet, 2004).

This project was undertaken to optimize the design configuration of Savonius rotors with theexpectation that this inherently simple vertical axis machines could be manufactured at low cost,

. All rights reserved.

fax: +913612690762.

www.sciencedirect.com/science/journal/indaer

www.elsevier.com/locate/jweia

dx.doi.org/10.1016/j.jweia.2008.03.005

mailto:[email protected]

ARTICLE IN PRESS

Nomenclature

A projected area of the rotor (m2)AR aspect ratio of blade (h/d)c chord of the bladed diameter of cylinder constituting

the paddles (chord)Cp power coefficientDf, Do diameter of the end plate (m)D diameter of rotor (m)e gap between the two blades: main

overlape0 gap between the two blades: sec-

ond overlaph height of the blade (m)N rotational speed of the rotor (RPM)Ps shaft power (W)V wind velocity (m/s)a twist angle (in degrees)b overlap ratio (e/d)r density of air (kg/m3)1-2sc single-stage two-bladed semicircu-

lar rotor1-2tw single-stage two-bladed twisted ro-

tor

1-3sc single-stage three-bladed semicir-cular rotor

1-3tw single-stage three-bladed twistedrotor

2-2sc two-stage two-bladed semicircularrotor

2-2tw two-stage two-bladed twisted rotor2-3sc two-stage three-bladed semicircu-

lar rotor2-3tw two-stage three-bladed twisted ro-

tor3-2sc three-stage two-bladed semicircu-

lar rotor3-2tw three-stage two-bladed twisted ro-

tor3-3sc three-stage three-bladed semicircu-

lar rotor3-3tw three-stage three-bladed twisted

rotor2-3sc (wv) two-stage three-bladed semicir-

cular rotor with valves2-3tw (wv) two-stage three-bladed twisted

rotor with valves

U.K. Saha et al. / J. Wind Eng. Ind. Aerodyn. 96 (2008) 1359–13751360

leading to their widespread use. The research proposal noted that small units could be manufactured fordistributed generation of electricity in residential and commercial locations. The units would be gridconnected to take advantage of net metering and would provide pollution-free generation of electricityusing a renewable resource at a cost competitive with power supplied by the grid. The operation ofSavonius wind turbine rotor is based on the difference of the drag of its semicircular vanes, dependingon whether the wind is striking the convex or the concave part of the vane. The advantage of this type ofrotor is that it is self-starting and relatively independent of the wind direction. It is simple to design andhas relatively low construction cost. However, it has a low efficiency.

It is a known fact that accessories like end plates, shielding and guide vanes (flat, curved) usuallyincrease the Savonius rotor performance; however, all of these increase the complexity of the rotor(Huda et al., 1992, Rajkumar, 2004). The rotor can develop a relatively high torque at low rotationalspeeds and is cheap to build, but it harnesses only a small fraction of the wind energy incident uponit. An attractive proposition for augmenting its harnessing effectiveness is to keep non-return valvesplaced inside the concave side of the blades. The valve opens automatically as a result of windpressure when the blade advances towards the wind thereby experiencing lower flow-resistance. Thecentrifugal force automatically closes this valve during the power-harnessing part of the cycle. Valve-aided rotor is the mechanism to make direction independent and is the effective way of increasingpower capability without unduly affecting the simplicity of rotor (Rajkumar and Saha, 2006). Inaddition to this, damages to turbine at higher velocities will be reduced with the valve mechanism.

2. Project objective

In recent times, a double-step or two-stage Savonius rotor has been investigated to find itsfeasibility for local production of electricity (Menet 2002, 2004). The challenge was to design,

ARTICLE IN PRESSU.K. Saha et al. / J. Wind Eng. Ind. Aerodyn. 96 (2008) 1359–1375 1361

develop and ultimately build a prototype of such a rotor, which was considered as a completeelectromechanical system. An optimum configuration was chosen for the geometry of theprototype. The building data were calculated on the basis of the nominal wind velocity of 10 m/s.The whole design of the prototype has confirmed the high efficiency and the low technicalityof the Savonius rotors for local production of electricity. It has been suggested to test this prototypein a wind tunnel to verify the design performances. It is important to emphasize here thataerodynamic performance studies of double-step or two-stage Savonius rotor have not so far beenreported in open literature. In view of this, the present work attempts to investigate thewind tunnel tests of a two-stage Savonius rotor with semicircular and twisted blades. In orderto assess its performance, tests are also planned for single- and three-stage rotor systems.Aerodynamic performances of the rotors have been evaluated on the basis of power coefficient (Cp),and no-load speed (RPM) at various airspeeds. The sequential order of the present investigation ismentioned below.

�
Performance study of two-stage rotor with two- and three-bladed system using both semicircularand twisted blades. � Comparison of the above two-stage system with single- and three-stage systems. � Optimization of rotor configuration in terms of number of stages, number of blades and blade
geometry (semicircular/twisted).
� Performance of valve-aided Savonius rotor with semicircular and twisted blades.
3. Energy in the wind

For an airstream flowing through an area A, the mass flow rate is rAV, and therefore the power

P ¼ rAV12V2¼ 1

2rAV3

where r is the air density (kg/m3), V is the wind speed (m/s) and P is the power (watts). The power isalso known as the energy flux or power density of the air (Walker and Jenkins, 1997; Bansal et al.,2002; Menet, 2004). The ratio of shaft power (Ps) to the power available in the wind (P) is known asthe power coefficient (Cp), and this indicates the efficiency of conversion. Thus

Cp ¼Ps

P

In the present investigation, A is the projected area of rotor (m2), and V is the airspeed (m/s) at thetunnel exit. The shaft power (Ps) is calculated from brake torque and rotational speed (RPM).

4. Blade design and fabrication

Savonius rotor made out of half cylinders (nominal diameter d, height H) is a very simple conceptwhere the whole rotor turns around a vertical axis. There are a number of geometrical parametersthat affects the efficiency of Savonius rotor (Alexander and Holownia, 1978; Mojola, 1985; Ushiyamaand Nagai, 1988; Modi and Fernando, 1989; Islam et al., 1993; Coton et al., 1996). Among thoseparameters, the aspect ratio (AR) plays an important role in the aerodynamic performances of aSavonius rotor. Globally high values of AR should greatly improve this efficiency. Values of AR around4.0 seem to lead to the best power coefficient for a conventional Savonius rotor. It is further knownthat end plates lead to better aerodynamic performances. The influence of the diameter Df of theseend plates relatively to the diameter D of the rotor has been experimentally studied. The higher valueof the power coefficient is obtained for a value of Df around 10% more than D, whatever the velocitycoefficient (Menet, 2004). The influence of the overlap ratio b( ¼ e/d) has also been studied(Ushiyama and Nagai, 1988; Fujisawa, 1992). The best efficiencies are obtained for values of bbetween 20% and 30%. It is not necessary to create another separation gap e0 between the paddles,which would consist of removing the chord of the paddles from the diameter of the rotor (Fig. 1); on

ARTICLE IN PRESSU.K. Saha et al. / J. Wind Eng. Ind. Aerodyn. 96 (2008) 1359–13751362

the contrary, the power coefficient and the torque coefficient decrease when the separation gap e0 issuperior to zero (Sheldahl et al., 1978; Ushiyama and Nagai, 1988).

In the present work, both two- and three-bladed systems have been studied in single-, two- andthree-stage rotor systems using semicircular and twisted blades. The top view of two- and three-bladed systems are shown in Figs. 1 and 2. An attempt has been made to study a variety of rotor

Fig. 1. Schematic of the two-bladed rotor.

Fig. 2. Schematic of the three-bladed rotor.

ARTICLE IN PRESS

Fig

.3

.S

che

ma

tic

dia

gra

mo

fse

mic

ircu

lar

an

dtw

iste

db

lad

es.

U.K. Saha et al. / J. Wind Eng. Ind. Aerodyn. 96 (2008) 1359–1375 1363

ARTICLE IN PRESS

Fig. 4. Solid models of single-, two- and three-stage rotor systems.


configurations with identical stage aspect ratio. Blades, manufactured from galvanized iron sheets,are brazed on the sleeve 1801 apart if the number of blades is two, and 1201 apart if the number ofblades is three. The schematic diagram of semicircular and twisted blades is shown in Fig. 3. Twistedblades are manufactured with optimum twist angle of a ¼ 12.51. The twist angle of the blade wasoptimized from a series of investigations (Grinspan et al., 2004; Rajkumar, 2004; Saha and Rajkumar,2006). The semicircular and twisted shapes were made on a rolling machine. In a two-stage system,top and bottom rotors are placed orthogonal to each other. The solid models for all the three differentstage systems are shown in Fig. 4.

5. End plate design and fabrication

It is known that end plates lead to better aerodynamic performances. The influence of diameter Do

of these end plates relatively to the diameter D of the rotor has been experimentally studied(Fujisawa, 1992). The higher value of the power coefficient Cp is obtained for a value of Do around 10%more than D, whatever be the velocity coefficient. In the present investigations, three end plates areused in a two-stage system, one at the top, one at the middle and one at the bottom of the rotor.

6. Valve design and fabrication

In this mechanism, a small raxine-type cover is pasted in the concave side of the blade, which ispurposely holed (Thotla, 2006). When wind is facing the concave side, the raxine cover will beattached to the blade; else it would allow air to flow from convex side to concave side therebyreducing the pressure deference on both sides, as it is the form drag that contributes to the powermechanism of rotor. The static torque performance of the rotor, especially of the returning blade, canbe improved by this mechanism without affecting the simplicity of rotor. This is mainly caused by theincreased pressure on the concave side of the returning blade, and due to the flow through the valve.


This would also reduce the negative torque of the rotor considerably. The schematic diagram of thevalve-aided mechanism is shown in Fig. 5.

7. Test facility

To study the performance of Savonius wind turbine a low-speed wind tunnel with an open testsection facility has been designed, developed, fabricated (Grinspan, 2002; Grinspan et al., 2003) andshown in Fig. 6. The rotor axis is placed at a distance of 205 mm from the tunnel exit having a cross-section area of 375 mm�375 mm. By changing the input voltage with the help of variac, the windtunnel exit air velocity can be changed. The entire tests have been conducted in the range of airvelocity of 6–11 m/s. A thermal velocity probe anemometer was used to measure the airspeed withan accuracy of 70.1 m/s, while the rotational speed (RPM) of the rotor was measured with a digitaltachometer. A brake dynamometer measures the static and the dynamic torques.

Fig. 6. Schematic diagram of the low-speed wind tunnel.

Fig. 5. Valve-aided semicircular and twisted blades.

ARTICLE IN PRESSU.K. Saha et al. / J. Wind Eng. Ind. Aerodyn. 96 (2008) 1359–13751366

8. Discussion of results

The performance of the Savonius rotor depends on the different parameters like number of blades,number of stages and geometry of the blades. Till now, there is no exact theoretical procedure to assessthe performance the Savonius rotor. The best way of optimizing the various parameters is to carryoutnumber of experiments on the different types of rotors in a low-speed open test section wind tunnel.Here, experiments have been conducted with different types of rotors by varying number of blades,number of stages and blade geometry or shape of the blades to optimize the Savonius rotor. A total of14 different types of Savonius rotors have been manufactured with identical stage aspect ratio ( ¼ 1.58)keeping the projected area of each rotor to be same ( ¼ 0.0377 m2). Prior to the conduct of the tests, thetunnel was calibrated by means of flow visualization and velocity measurements.

8.1. Optimum number of stages

The power coefficient (Cp) of the Savonius rotor also depends on the number of stages. Tests havebeen conducted by varying the number of stages (from one to three), by varying the number ofblades (two and three), and the blade geometry (semicircular and twisted shapes) to optimize thenumber of stages. The variation of power coefficient (Cp) with velocity for the three rotors (1-2sc,2-2sc and 3-2sc) is shown in Fig. 7. In this case, the number of blades and geometry of the blades arekept fixed. When the number of stages increased from one to two, the value of Cp increasedconsiderably. But, when the number of stages increased from two to three, the performance isdecreased because of the increase in inertia of the rotor. The same experiments have been repeatedwith the twisted blades and the results are shown in Fig. 8. For the twisted blades also, two-stagerotor shows a better performance characteristics when compared the three-stage rotor. As thenumber of stages was increased, the inertia of the rotor was found to increase thereby reducing itsperformance, which is independent on the blade geometry for this case.

In the above experiments, there are two blades in each stage, i.e., a two-stage system has fourblades altogether. Experiments have also been conducted with three blades in each stage, i.e., a two-stage system would have six blades altogether. Results obtained are shown in Figs. 9 and 10. With anincrease of number of blades from two to three, the rotor performance seemed to have decreased inall the three cases, showing a better performance coefficient for the two-stage system. Further, fromthe above plots, it can be concluded that when the number of stages got increased from one to two,the rotor shows a better performance characteristics, however, the performance gets degraded whenthe number of stages becomes three. This may be due to an increase in inertia of the rotor. Fromthese experiments, it is clear that the optimum number of stages for the Savonius rotor is two.

8.2. Optimum number of blades

The effect of number of blades on the performance of Savonius rotor is also studied. Experimentshave been conducted by varying the number of blades from two to three in single-, two- and three-stage rotor systems. Both semicircular and twisted-bladed rotors have been tested, and resultsobtained are shown in Figs. 11 and 12. It is observed that the power coefficient of the rotor decreaseswhen the number of blades got increased from two to three. When the number of blades is increasedto three, the air which strikes on one blade get reflected back on the following blade so that thefollowing blade rotates in negative direction as compared to the succeeding blade. Hence, with anincrease of number of blades, the rotor performance decreases. It can be concluded from theexperimental evidence that a two-bladed system gives optimum performance. Again, in a two-bladed system, the performance of twisted-bladed rotor is superior to the semicircular-bladed rotor.

8.3. Optimum geometry of the blade

One of the important parameter that affects the performance of the Savonius rotor is geometry ofthe blade. Both semicircular (twist angle of 01) and twisted (twist angle of 12.51)-bladed geometry

ARTICLE IN PRESS

1-2tw2-2tw3-2tw

Velocity (m/s)

Pow

er C

oeffi

cien

t

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

0 0

0.03 0.03

0.06 0.06

0.09 0.09

0.12 0.12

0.15 0.15

0.18 0.18

0.21 0.21

0.24 0.24

0.27 0.27

0.3 0.3

0.33 0.33

Fig. 8. Variation power coefficient with velocity for two-bladed twisted Savonius rotor system.

1-2sc2-2sc3-2sc

Velocity (m/s)

Pow

er C

oeffi

cien

t

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

0 0

0.03 0.03

0.06 0.06

0.09 0.09

0.12 0.12

0.15 0.15

0.18 0.18

0.21 0.21

0.24 0.24

0.27 0.27

0.3 0.3

0.33 0.33

Fig. 7. Variation of power coefficient with velocity for two-bladed semicircular Savonius rotor system.


have been tested and the results are shown in Fig. 13. It is observed that twisted-bladed rotor shows agood performance when compared to the semicircular-bladed rotor in all the cases of single-, two-,and three-stage systems. In case of semicircular rotor, the maximum force acts centrally (curvaturecenter) and vertically; whereas for the twisted blade, the maximum force moves towards to the tip ofthe blade because of the twist in the blade. Due to these changes, a twisted blade gets a longermoment arm, and hence a higher value of power coefficient.

ARTICLE IN PRESS

1-3sc2-3sc3-3sc

Velocity (m/s)

Pow

er C

oeffi

cien

t

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

0 0

0.03 0.03

0.06 0.06

0.09 0.09

0.12 0.12

0.15 0.15

0.18 0.18

0.21 0.21

0.24 0.24

0.27 0.27

0.3 0.3

0.33 0.33

Fig. 9. Variation of power coefficient with velocity for three-bladed semicircular Savonius rotor system.

Velocity (m/s)

Pow

er C

oeffi

cien

t

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

0 0

0.03 0.03

0.06 0.06

0.09 0.09

0.12 0.12

0.15 0.15

0.18 0.18

0.21 0.21

0.24 0.24

0.27 0.27

0.3 0.3

0.33 0.331-3tw2-3tw3-3tw

Fig. 10. Variation of power coefficient with velocity for three-bladed twisted Savonius rotor system.


8.4. Results of valve-aided rotor

Experiments were carried out with valve-aided Savonius rotor, and the results are compared withconventional Savonius rotors (Figs. 14 and 15). In case of valve-aided rotor, a hole is made on theblade and a rexine cover is pasted on the concave side of the blade so that it can act as a non-return

ARTICLE IN PRESS

Velocity (m/s)

Pow

er C

oeffi

cien

t

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

0 0

0.03 0.03

0.06 0.06

0.09 0.09

0.12 0.12

0.15 0.15

0.18 0.18

0.21 0.21

0.24 0.24

0.27 0.27

0.3 0.3

0.33 0.331-2sc1-3sc2-2sc2-3sc3-2sc3-3sc

Fig. 11. Variation of power coefficient with velocity for semicircular Savonius rotor system.

Velocity (m/s)

Pow

er C

oeffi

cien

t

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

0 0

0.03 0.03

0.06 0.06

0.09 0.09

0.12 0.12

0.15 0.15

0.18 0.18

0.21 0.21

0.24 0.24

0.27 0.27

0.3 0.3

0.33 0.331-2tw1-3tw2-2tw2-3tw3-2tw3-3tw

Fig. 12. Variation of power coefficient with velocity for twisted Savonius rotor system.


valve. Generally, a three-bladed Savonius rotor system had low power coefficient when compared to atwo-bladed system. But, when valves are provided on the three-bladed system, the power coefficienthas exceeded the two-bladed system whether blade geometry is semicircular or twisted. In absence ofvalves, the air strikes on the incoming blade and rotor rotates in the negative direction. With valves inopen position, the air passing through it reduces the negative pressure on incoming blade.

ARTICLE IN PRESS

Velocity (m/s)

Pow

er C

oeffi

cien

t

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

0 0

0.03 0.03

0.06 0.06

0.09 0.09

0.12 0.12

0.15 0.15

0.18 0.18

0.21 0.21

0.24 0.24

0.27 0.27

0.3 0.3

0.33 0.331-2sc1-2tw2-2sc2-2tw3-2sc3-2tw

Fig 13. Variation of power coefficient with velocity for two-bladed Savonius rotor system.

Velocity (m/s)

Pow

er C

oeffi

cien

t

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

0 0

0.03 0.03

0.06 0.06

0.09 0.09

0.12 0.12

0.15 0.15

0.18 0.18

0.21 0.21

0.24 0.24

0.27 0.27

0.3 0.3

0.33 0.33

2-2sc2-3sc(wv)2-2tw2-3tw(wv)

Fig. 14. Variation of power coefficient with velocity for two-stage three-blade valve-aided Savonius rotor system.


8.5. Variation of RPM with velocity

Initially, the rotor is kept stationary, and then the airspeed at the inlet to the rotor is increasedslowly with the help of a variac. At different airspeed, the RPM of the rotor is found with a digitaltachometer. The range of air velocity was varied from 6.02 to 10.17 m/s. Initially, experiments have

ARTICLE IN PRESS

Velocity (m/s)

Pow

er C

oeffi

cien

t

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

0 0

0.03 0.03

0.06 0.06

0.09 0.09

0.12 0.12

0.15 0.15

0.18 0.18

0.21 0.21

0.24 0.24

0.27 0.27

0.3 0.3

0.33 0.33

2-3sc(wv)2-3sc2-3tw(wv)2-3tw

Fig. 15. Variation of power coefficient with velocity for two-stage three-bladed valve-aided Savonius rotor system.


been done on the single-stage with two and three blades for both semicircular and twisted-bladedsystem. It is observed that the increase in RPM with increase in velocity is more for two-bladedtwisted Savonius rotor (Fig. 16). As the number of blades increased from two to three, the RPM of therotor was found to decrease because of the interference of the air with the blades. This happened forboth the cases of semicircular and twisted rotor systems.

Similar types of experiments have been carried out with two-stage and three-stageSavonius rotor system with semicircular and twisted blades (Figs. 17 and 18). In the two-stagesystem, two-bladed rotor shows an increase in RPM as compared to the three-bladed rotorsystem. However, the overall increase in RPM is more in the two-stage two-bladed twisted rotorsystem.

The variation of RPM of the valve-aided Savonius rotor is studied and compared with the two-stage three-bladed Savonius rotor (Fig. 19). As the number of blades got increased from two to threein a two-stage rotor system, the RPM of the rotor decreased. However, with the insertion of the valve,the three-bladed system showed an improved rotational performance due to a reduction of itsnegative torque.

9. Conclusions

Due to slow rotational speed and low power production, the Savonius rotor is unsuitable forelectricity generation. Therefore, not enough work has been progressed in the area of this verticalaxis wind turbine as opposed to its horizontal axis counterpart. However, for a small-scale powerrequirement, Savonius rotors are quite useful. Therefore, it has become necessary to go through itsvarious prospects so that its performance can be improved to a greater extent. In the presentinvestigation, optimization of blade configuration for single-, two- and three-stage Savonius rotorshas been made through wind tunnel testing. All the tests have been conducted in the range 6–11 m/s.Tests were also carried out with valve-aided Savonius rotor in a two-stage system. The principal

ARTICLE IN PRESS

Velocity (m/s)

RPM

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

350 350

400 400

450 450

500 500

550 550

600 600

650 650

700 700

750 750

800 800

850 850

2-3sc2-2sc2-3tw2-2tw

Fig. 17. Variation of RPM with velocity for two-stage Savonius rotor system.

Velocity (m/s)

RPM

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

300 300

350 350

400 400

450 450

500 500

550 550

600 600

650 650


Fig. 16. Variation of RPM with velocity for single-stage Savonius rotor system.


observations of the present investigation have been summarized below. Some of the salient data ofthe present investigation are shown in Tables 1 and 2.

�
Optimum number of blades is two for the Savonius rotor whether it is single-, two- or three-stage. � Twisted geometry blade profile had good performance as compared to the semicircular blade
geometry.

ARTICLE IN PRESS

Velocity (m/s)

RPM

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

350 350

400 400

450 450

500 500

550 550

600 600

650 650

700 700

750 750

800 800

850 850

2-3sc2-3sc(wv)2-3tw2-3tw(wv)

Fig. 19. Variation of RPM with velocity for valve-aided Savonius rotor system.

Velocity (m/s)

RPM

5

5

6

6

7

7

8

8

9

9

10

10

11

11

12

12

350 350

400 400

450 450

500 500

550 550

600 600

650 650

700 700

750 750


Fig. 18. Variation of RPM with velocity for three-stage Savonius rotor system.


�
Two-stage Savonius rotor had better power coefficient as compared to the single- and three-stagerotors. � Valve-aided Savonius rotor with three blades shows better performance coefficient as compared
to the conventional three-bladed rotor.

ARTICLE IN PRESS

Table 2Performance of valve-aided two-stage three-bladed Savonius rotor system

Rotor system Blade shape Blade

height

(m)

Blade

chord

(m)

Aspect

ratio

Projected area

(m2)

Free stream

velocity (m/s)

Max. power

coefficient

(Cp)

Two-stage Semicircular 0.122 0.077 1.58 0.0377 7.30 0.31

Twisted 0.32

Table 1Performance of Savonius rotor system

Rotor

system

No. of

blades

Blade shape Blade

height

(m)

Blade

chord

(m)

Aspect

ratio

Projected

area (m2)

Free stream

velocity

(m/s)

Max. power

coefficient

(Cp)

Single-

stage

2 Semicircular 0.173 0.109 1.58 0.0377 8.23 0.18

Twisted 0.19

3 Semicircular 0.15

Twisted 0.16

Two-

stage

2 Semicircular 0.122 0.077 1.58 0.0377 7.30 0.29

Twisted 0.31

3 Semicircular 0.26

Twisted 0.28

Three-

stage

2 Semicircular 0.100 0.063 1.58 0.0377 8.23 0.23

Twisted 0.24

3 Semicircular 0.20

Twisted 0.21


References

Alexander, A.J., Holownia, B.P., 1978. Wind tunnel test on a Savonius rotor. J. Ind. Aerodyn. 3, 343–351.Bansal, R.C., Bhatti, T.S., Kothari, D.P., 2002. On some design aspects of wind energy conversion systems. Energy Convers.

Manage. 43, 2175–2187.Coton, F.N., Galbraith, R.A.McD., Jiang, D., 1996. The influence detailed blade design on the aerodynamic of straight-bladed

vertical axis wind turbines. Proc. Inst. Mech. Eng. 210, 65–74.Fujisawa, N., 1992. On the torque mechanism of Savonius rotors. J. Wind Eng. Ind. Aerodyn. 40, 227–292.Grinspan, A.S., 2002. Development of a low speed wind tunnel and testing of savonius wind turbine rotor with twisted blades.

M.Tech. Thesis, Department of Mechanical Engineering, IIT Guwahati, India.Grinspan, A.S., Mahanta, P., Saha, U.K., 2003. Design, Development and calibration of a low speed wind tunnel. In: Seventh

Triennial International Symposium on Fluid Control, Measurement and Visualization, FLUCOME, August 25–28, Sorrento,Italy.

Grinspan, A.S., Saha, U.K., Mahanta, P., 2004. Experimental investigation of twisted bladed Savonius wind turbine rotor. RERICInt. Energy J. 5 (1), 1–9.

Huda, M.D., Selim, M.A., Islam, A.K.M.S., Islam, M.Q., 1992. The performance of an s-shaped Savonius rotor with a deflectingplate. RERIC Int. Energy J. 14, 25–32.

Islam, A.K.M.S., Islam, M.Q., Mandal, A.C., Razzaque, M.M., 1993. Aerodynamic Characteristics of a Stationary Savonius Rotor.RERIC Int. Energy J. 15, 125–136.

Menet, J.L, 2002. Local production of electricity with a small Savonius rotor. In: Proceedings of the 2002 Global Wind PowerConference, April 2–5, Paris, France.

Menet, J.L., 2004. A double-step Savonius rotor for local production of electricity: a design study. Renew. Energy 29,1843–1862.

Modi, V.J., Fernando, M.S.U.K., 1989. On the performance of the Savonius wind turbine. ASME J. Sol. Eng. 111, 71–76.Mojola, O.O., 1985. On the aerodynamic design of the Savonius windmill rotor. J. Wind Eng. Ind. Aerodyn. 15, 223–231.Rajkumar, M.J., 2004. Experimental investigation and flow simulation of Savonius rotor, M.Tech. Thesis, Department of

Mechanical Engineering, IIT Guwahati, India.


Rajkumar, M.J., Saha, U.K., 2006. Valve aided twisted Savonius rotor. J. Wind Eng. 30 (3), 243–253.Saha, U.K., Rajkumar, M.J., 2006. On the performance analysis of Savonius rotor with twisted blades. Renew. Energy 31 (11),

1776–1788.Sheldahl, R.E., Blackwell, B.F., Feltz, L.V., 1978. Wind tunnel performance data for two and three bucket Savonius rotor. J. Energy

2, 160–164.Shikha, Bhatti, T.S., Kothari, D.P., 2003. Wind energy conversion systems as a distributed source of generation. ASCE J. Energy

Eng. 169, 69–80.Thotla, S., 2006. Optimum design configuration of Savonius rotor through wind tunnel testing. M.Tech. Thesis, Department of

Mechanical Engineering, IIT Guwahati, India.Ushiyama, I., Nagai, H., 1988. Optimum design configurations and performance of Savonius rotors. Wind Eng. 2, 59–75.Walker, J.F., Jenkins, N., 1997. Wind Energy Technology. Wiley.

optimum design configuration of savonius rotor through wind tunnel experiments

Documents