Ten Savonius Wind Rotors in Different Tip-Speed Ratio 62

The torque curves of the parallel matrix system with ten Savonius wind rotors and one single Savonius wind rotor are shown in Figure 4.17. As shown in Figure 4.17, it can be seen that for rotors No.1 to 10, each has higher performance than one single rotor. The ten wind rotors rotate in the same direction (counterclockwise). The wind departed from rotor No.1 enhances the rotation of No. 2, and so on. When the wind passes through the wind rotor, it causes a low pressure that contributes extra rotation power to adjacent wind rotor. However, there is no wind rotor to enhance the rotation of No.1. But the performance of rotor No.1 is still higher than one single rotor. The static pressure field and velocity vector distribution around parallel matrix system are demonstrated in Figs. 4.18 and 4.19, respectively.

For the purpose to analyze flow field more clearly, Fig. 4.20 shows the streamline distribution derived from 2-D simulation, which has same parameters with present case. The difference between each adjacent stream functions

62

represents the volume flow rate. Therefore, the thinner streamlines has higher velocity and wider streamlines has lower velocity. The higher velocity indicates that wind passes through the wind rotor easily. On the other hand, the lower velocity indicates the wind rotor to be able to absorb more wind work that lets the velocity goes down. Thus, wind rotors produce an asymmetrical flow field that the dense streamlines take place around wind rotor with lower torque; wider streamlines around wind rotor with higher torque.

For this case, the maximum value of Cp is 0.438 at TSR 0.7, given in Fig.

4.21. It also reveals that parallel matrix system apparently has higher performance than one single rotor.

4.3.3 Ten Savonius Wind Rotors in Different Wind Direction

Feng [1] numerically studied a parallel matrix system, which includes three Savonius wind rotors with the same angular speed, the specified wind direction and the fixed distance on the parallel system. He found that such disposition can cause constructive interference that improves performance of wind rotor and the performance curve of three Savonius wind rotors with different wind directions as shown in Fig. 4.22. These conclusions can be supported by the numerical simulation of Shigetomi et al. [14], illustrated in Fig. 4.9. Therefore, present study will discuss the effect in different wind directions on the parallel matrix system, which is consisting of ten two-bladed Savonius wind rotors. The different wind directions are due to different arrangement of Savonius wind rotor.

According to the research of Feng [1], the best phase angle difference in such system is 90° and the poorest one is 135°. The suitability of phase angle difference for more energy gained from wind is based on the shape of the two

63

semicircular blades of these Savonius wind rotors. The higher performance is resulted from the positive interaction between these Savonius wind rotors, and the flow fluctuation plays the major role in contributing to this effect. According to the research of Howell et al. [12], the fluctuation is caused by reasons, such as the potential disturbances around the rotating blades and the large vortex shedding due to the flow separation from the wind rotor’s blade. The influence of the fluctuation velocity on the power output is explained by separating the inflow velocity into time average and fluctuation components as following:

W = Cu³

where C represents the constant to u³ with all the other factors fixed, and the over bar indicates time averaging. Therefore, it can be concluded that the time average power output will increase with the fluctuation of velocity.

This positive interaction by connecting ten Savonius wind rotors in parallel may gain apparently the higher performance, but it might be sensitive to the direction of wind. Therefore, the influence of wind direction on the parallel system is studied now. The system with phase angle difference 90° are chosen with the wind velocity 7 m/s and tip-speed ratio 0.7. The angles of wind direction are 0°, 37°, 53° and 90°. The results are shown in Fig. 3.5 and Fig.

4.23, and a comparison with a single Savonius wind rotor is given as well. In the figure, the change of wind direction will clearly affect and higher the Cp of the parallel matrix system at 0°. When θ = 0°, the C^p is even lower than that of a

64

single one.

Figs. 4.24 (a)、(b)、(c) and (d) show the velocity vector distribution around the ten Savonius wind rotors with a wind direction 0°, 37°, 53° and 90°, respectively. And Figs. 4.25 (a)、(b)、(c) and (d) show the static pressure field around the ten Savonius wind rotors with a wind direction 0°, 37°, 53° and 90°, respectively. As shown in these figures, the above mentioned effect that the pressure difference on the retuning blade is decreased by the effect of the lower wind rotor is reduced due to the changed arrangement relative to the wind direction. The lower wind rotor could not affect the fluid flowing to the region behind the retuning blade of the upper one to reduce the vortex. Therefore, a negative torque is increased and then causes a lower performance. It indicates that the parallel matrix system is strongly influenced by the change of wind direction, representing that one of the advantages in VAWTs is lost.

The maximums of average power output of the parallel matrix system with ten Savonius wind rotors are calculated and the results are listed in Table 4.10.

Table 4.10 The maximum of average power output of the parallel matrix system with ten Savonius wind rotors

Condition Wind Direction

Rotors

θ=53° 7 57.3

14 520.24

θ=90° 7 82.87

14 700.46

4.4 Comparison between One Single Savonius Wind Rotor and