Access Official Technical PDF

Share Summary:LinkedInEmail

A modified Savonius wind turbine blade can improve rotor performance without adding external deflectors or other components. A computational fluid dynamics (CFD) study compared a modified rotor with a conventional semicircular design, examining power coefficient, torque and airflow. For engineers and technical decision-makers assessing vertical axis wind turbines (VAWTs), the results illustrate how blade geometry can influence energy capture and the braking effect that limits Savonius performance.

Why Savonius rotor performance is a design challenge

Savonius turbines are drag-driven VAWTs. Their two offset, half-cylinder blades form an S-shaped cross-section. Wind produces a pressure difference across the concave and convex blade surfaces, creating rotation. The configuration is mechanically simple, can accept wind from different directions and is suited to operation at relatively low wind speeds.

The same drag principle also creates a performance trade-off. The advancing blade generates positive torque, while the returning blade experiences forces that can act against rotation. This negative torque reduces the net torque available to drive a load. Conventional Savonius rotors are consequently less efficient than horizontal-axis turbines, with the paper describing typical power coefficient values in the range of 0.1 to 0.25.

Past approaches to improving output have included obstacle shields, curtains and conveyor-deflector systems, as well as changes to blade shape and rotor geometry. Added structures can make a rotor more complex and may increase its dependence on wind direction. The study instead examines whether reshaping the blades can improve performance while retaining a comparatively simple arrangement.

What changes in the modified Savonius wind turbine?

The modified blade combines an elliptical profile with a straight trailing edge. The study’s design objective is to improve the torque balance between the advancing and returning blades. A longer effective moment arm on the advancing blade can increase positive torque, while the trailing-edge geometry is intended to reduce the returning blade’s negative contribution.

To make the comparison meaningful, the researchers kept the rotor radius, swept area and blade thickness consistent between the modified and conventional models. The conventional design used an overlap ratio of 0.2, which the paper identifies as an optimum configuration based on prior work; the same overlap was used for the modified rotor. These shared parameters help focus the comparison on the change in blade shape.

How CFD was used to compare the designs

The team used transient CFD in ANSYS Fluent to model the flow around each rotor. The calculations applied the finite volume method to solve the fluid-flow equations, with a k-ω turbulence model. The computational domain included a rotating region containing the blades and a larger stationary region, separated by a sliding interface. The models used unstructured meshes.

For the simulations, the inlet wind speed was set to 6 m/s. The study compared numerical results for the conventional rotor against experimental data from earlier work. The simulated power and torque coefficient curves followed similar trends to the experimental results, with the paper reporting errors below 1% over a wide range of tip speed ratios and a maximum error of about 8% at a tip speed ratio of 1.2. The authors considered this agreement adequate for using the numerical method to compare rotor designs.

Two measures help interpret the results:

  • Tip speed ratio (TSR) compares the speed of the blade tip with the wind speed. It provides a way to assess turbine performance as rotor speed changes.
  • Power coefficient (Cp) expresses the fraction of the wind’s available power extracted by the turbine. Torque coefficient (Ct) describes torque in relation to the study’s performance measures and TSR.

Performance and airflow findings

Both rotor models reached their maximum power coefficient at a TSR of 0.9. The modified design achieved a maximum Cp of 0.335, compared with 0.26 for the conventional semicircular blades. The modified rotor also showed a higher torque coefficient, particularly at low TSR values; above a TSR of 1, the values for the two designs moved closer together.

The flow-field results offer a physical explanation for the performance difference. The researchers examined velocity and pressure distributions at a rotor position of 85 degrees and a TSR of 0.9. In the modified model, the pressure on the concave side of the advancing blade was higher, while the pressure pattern and flow around the returning blade were associated with a reduced negative-torque effect overall.

The simulations also showed stronger suction vortices and a different distribution of flow near the blade tips for the modified rotor. The study reports a peak tip-vortex velocity of 22.7 m/s for the modified model and 16.7 m/s for the conventional one under the examined conditions. In the authors’ interpretation, the altered flow helps maintain useful forces on the rotor and moves the stagnation zone toward the blade extremity, increasing the effective drag moment arm.

What the results mean for turbine evaluation

For turbine designers and system evaluators, the study highlights the value of considering blade-level aerodynamics alongside overall output metrics. A change in profile can affect the positive torque from the advancing blade and the negative torque from the returning blade at the same time. Reviewing Cp alone may show that performance changed; examining pressure, velocity and Ct can help explain why.

The results are based on two-dimensional numerical simulations and a specific set of model conditions. They support a comparative assessment of the designs, but do not by themselves establish performance for every rotor scale, wind environment or operating installation. Application decisions should account for the intended operating conditions and the limits of the simulation approach.

Conclusion

The modified Savonius rotor pairs an elliptical blade shape with a straight trailing edge to address a central limitation of the conventional design: negative torque from the returning blade. In the reported CFD comparison, it produced a higher peak power coefficient and torque coefficient under the examined conditions. For engineering teams, the practical takeaway is that a relatively simple blade-geometry change can merit careful evaluation, while validated simulations and context-specific testing remain important before drawing conclusions about a particular turbine application.

Access Official Technical PDF

About The Author