The Savonius wind rotor is a drag-typed vertical-axis wind rotor, developed by S. J. Savonius [3]. Its shape is similar to the character “S”, therefore, it is also called S-rotor. This kind of rotor is spinned by drag force. The drag force difference between concave and convex surfaces drives the rotor, leading to a large starting torque with a relatively low rotational speed. Many improving researches are bringing up after the Savonius’ development.
Blackwell et al. [4] investigated the performances of fifteen configurations of Savonius wind rotors by testing in a low speed wind tunnel. What they investigated included parameters, such as number of blades, wind velocity, wind rotor height, and blade overlap ratio. The results showed that the two-bladed configurations have better performance than the three-bladed ones, except the starting torque. Besides, the performance increases with aspect ratio slightly, and the optimum overlap ratio is between 0.1 and 0.15.
3
Fujisawa [5] studied the performances of two-bladed Savonius wind rotors with different overlap ratios ranged from 0 to 0.5 by experimental investigations.
The results reveal that the performance of Savonius wind rotor reaches a maximum value at an overlap ratio of 0.15. It is due to the fact that the flow through the overlap strengthens the forward movement. However, when the overlap ratio becomes larger, the recirculation zone grows accordingly, causing the performance deterioration.
Gupta et al. [6] studied the performances of a Savonius wind rotor and a Savonius-Darrieus machine with overlap variation by experimental investigations. For the Savonius-Darrieus machine, there was a two-bladed Savonius wind rotor in the upper part and a Darrieus machine in the lower side.
The result showed that Cp with 20% overlap is higher than 16.2% without overlap. They also concluded that the improvement of Cp can be achieved for the Savonius-Darrieus wind machine compared with the general Savonius rotor.
Irabu and Roy [7] introduced a guide-box tunnel to improve the Cp of Savonius wind rotors and prevent the damage by strong wind disaster. The guide-box tunnel was like a rectangular box as wind passage and the test wind rotor was included. It was able to adjust the inlet mass flow rate by its variable area ratio between the inlet and outlet. The experimental results showed that the maximum Cp of the two-bladed wind rotor using the guide-box tunnel is about 1.23 times of that the wind rotor without the guide-box tunnel and 1.5 times of that using a three-bladed wind rotor. Apparently, it verified that the two-bladed wind rotor is better than the three-bladed one for converting wind power.
Saha et al. [8] used a wind tunnel to test and investigate the performances by different number of blades and stages, different geometries of blade and inserting valves on the concave side of blade or not. The results were as follows.
4
First, with an increases of the number of blades, the performance of wind rotor decreases. Second, twisted geometry blade profile has a better performance than the semicircular blade geometry. Third, the Cp of a two-stage Savonius wind rotor is higher than those of single-stage and three-stage wind rotors. Last, the valve-aided Savonius wind rotor with three blades shows a better performance than the conventional wind rotor.
Altan et al. [9] introduced a curtaining arrangement to improve the performance and increase the efficiency of a two-bladed Savonius wind rotor without changing its basic structure. They placed two wind-deflecting plates in front of the wind rotor to prevent the negative torque opposite the wind rotor rotation. The experimental results showed that Cp increase about 38% with an optimum curtain arrangement and it is 16% much higher than that without curtaining.
Antheaume et al. [10] applied the CFD software, Fluent, to investigate the performance of vertical axis Darrieus wind rotor in different working fluids by using k-ε turbulent model under steady-state conditions. They also discussed the average efficiency of several wind rotors connected in parallel pattern. The results showed that increasing the number of wind rotors or decreasing the distance between wind rotors can make the efficiency higher due to the velocity streamlines straightening effect by the configuration. In addition, the performances working in water are much higher than those in air.
Zhao et al. [11] applied the computational fluid dynamics (CFD) software to investigate the performance of new helical Savonius wind rotors. They analyzed the behaviors of the wind rotors with different aspects ratio, number of blades, overlap distance and helical angle. The results showed that three-blade helical wind rotor has lower Cp compared with two-bladed helical wind rotor.
5
And the best overlap ratio, aspect ratio, and helical angle are 0.3, 6.0 and 180°, respectively.
Howell et al. [12] applied Fluent to investigate the performances of one VAWT in 2-D and 3-D simulations and compared the predictions with experimental data. The turbulence model used was RNG k-ε model, by which the applicability in flow fields involves large flow separations. The error bars on experimental data were fixed at ±20% of measured values. The results showed that the performances predicted by 2-D simulations are apparently higher than those by 3-D simulations and experimental measurements due to the effect of the generation of over tip vortices.
Pope et al. [13] applied Fluent to investigate the performances of zephyr VAWT and compared the predictions with experimental data. By the reason that a free spinning turbine cannot be fully simulated, they used constant rotational speeds of the VAWT in simulations and changed the specification of parameters to reveal freely moving turbine blades in experiments. They indicated that determining the performance at constant rotational speed is valuable since any power generation connected to the electricity grid needs to operate at constant speed.
Shigetomi et al. [14] studied the interactive flow field around two Savonius wind rotors by experimental investigation using particle image velocimetry.
They found that there exist power-improvement interactions between two rotating Savonius rotors in appropriate arrangements. The interactions are caused by the Magnus effect to provide the additional rotation of the downstream rotor and the periodic coupling of local flow between two wind rotors. However, they are quite sensitive between two wind rotors to the wind direction so that wind rotors arranged together will lose one of the VAWTs inherent advantages, such as
6
no influence of wind direction.
Akwa et al. [15] investigated the influence of the buckets overlap ratio of a Savonius wind rotor on the averaged moment and power coefficients in the complete cycles of operation by the commercial software, Star-CCMþ.
Numerical simulations of the air flow on the rotor in 40 different situations were carried out. The results showed that the best performance occurs at bucket overlap ratios with values close to 0.15, giving an averaged power coefficient of 0.3161 under the tip speed ratio 1.25. However, in the range of high bucket overlap, the moment and performance of the rotor fall dramatically due to the reduced incidence of air on the concave side of the rotor bucket.
McTavish et al. [16] studied a performance assessment of a novel vertical axis wind turbine by using CFD software, CFdesign 2010. A validation study consisting of steady and rotating simulations was conducted using a Savonius rotor. Fair agreement was obtained by comparing with experimental data. Steady two-dimensional CFD simulations had demonstrated that the new VAWT has the similar average static torque characteristics to existing Savonius rotors.
Three-dimensional simulations were conducted at several tip speed ratios with a free stream speed of 6 m/s. The predicted dynamic torque generated by the rotor decays more rapidly with increasing tip speed ratio than the torque output of Savonius rotors due to its asymmetric design and the curvature of the outer rotor wall.
Akwa et al. [17] presented a review on the performance of Savonius wind turbines. Simple construction, high start up and full operation moment, wind acceptance from any direction, low noise and angular velocity in operation, reducing wear on moving parts, are some advantages of using this type of machine. Savonius rotor performance is affected by operational conditions,
7
geometric and air flow parameters. Each different arrangement of Savonius rotor affects its performance. The range of reported values for maximum averaged power coefficient includes values around 0.05–0.30 for most settings.
Performance gains of up to 50% for tip speed ratio of maximum averaged power coefficient are also reported with the use of stators.
Zhou et al. [18] explored the non-linear two-dimensional unsteady flow over a conventional Savonius-type rotor and a Bach-type rotor, and developed a simulation method for predicting their aerodynamic performance. The simulations were performed using Star-CCMþ. A comparative study of the two types of rotors was carried out, and numerical simulation results were compared with experimental data. The results showed that the Bach-type rotor is demonstrated to have better performance for torque and power coefficient than the conventional Savonius-type rotor. A discussion of the causes of these differences was presented that is based on a detailed study of the respective flow field characteristics, including the behavior of moment coefficients, velocity vectors and pressure distribution. A simulation method for further study of new blades shapes was suggested.
Feng [1] applied the computational fluid dynamics (CFD) software, Star-CD, to investigate the flow field around two-bladed Savonius wind rotors and their corresponding performances. The study mentioned that using the parallel matrix arrangement of Savonius wind rotors can get higher power efficiency. It indicated that the parallel matrix system has an optimum Cp, which is about 1.45 times of Cp of one single Savonius wind rotor in 3-D simulation.
Huang [2] applied the CFD software, Fluent, to investigate a four two-bladed Savonius wind rotors in parallel matrix system and compared the predicted results with experimental data. For the simulation results, the Cp of
8
parallel matrix system with curtain is about 1.03 times of that of the system without curtain. Both of those are occurred at TSR 0.8. Besides, The Cp of parallel matrix system with curtain at TSR 0.6 is about 1.16 times of that of the system without curtain. However, the predicted results are higher than the experiment measurements because the resultant Cps in experiments by the generated power, which needs to consider energy transform loss.