A Savonius Vertical Axis Wind Turbine is a simple type of wind turbine that works based on drag. The wind pushes against its curved blades, causing it to spin. Because of their simple design, they are reliable and can capture wind from any direction. However, the airflow around the spinning blades is very complex and turbulent. To accurately predict a turbine’s power and efficiency, we must use the correct turbulence model in a CFD simulation.

This project investigates the Turbulence Models Effect On Savonius CFD simulation. We will compare four different turbulence models in ANSYS Fluent to see how each one predicts the airflow. The goal is to understand the strengths of each model for this specific application. Our simulation setup uses assumptions from the reference paper by Talukdar, et al. [1].

• Reference [1]: Talukdar, Parag K., Vinayak Kulkarni, and Ujjwal K. Saha. “Performance Characteristics of Vertical-Axis Off-Shore Savonius Wind and Savonius Hydrokinetic Turbines.” International Conference on Offshore Mechanics and Arctic Engineering. Vol. 51319. American Society of Mechanical Engineers, 2018.
• Reference [2]: Farajyar, Shayan, et al. “CFD investigation and optimization on the aerodynamic performance of a Savonius vertical axis wind turbine and its installation in a hybrid power supply system: a case study in Iran.” Sustainability6 (2023): 5318.

Figure 1: A schematic showing the basic design and operation of a Savonius VAWT [2].

Simulation Process: CFD Modeling of a Rotating Turbine in Fluent

Simulating a rotating machine like a Savonius wind turbine Fluent requires a special setup. We created two separate domains: a smaller, circular domain that rotates with the turbine blades, and a larger, stationary domain that represents the surrounding air. A high-quality mesh was generated for both regions, with special care taken at the interface where they meet.

To handle the rotation, the Sliding Mesh technique was used in ANSYS Fluent. This powerful module allows the inner domain to spin relative to the outer domain, accurately simulating the turbine’s motion. The core of this investigation was to test the performance of four different turbulence models:

• k-omega SST
• SST Transition
• Spalart-Allmaras* Reynolds Stress model (RSM)

This comparative approach allows us to see the direct Turbulence Models Effect On Savonius Fluent predictions.

Figure 2: The computational domain for the Savonius Vertical Axis Wind Turbine CFD simulation, showing the boundary conditions [1].

Post-processing: Comparing Velocity and Wake Structures

The results of the simulation show that the choice of turbulence model has a significant impact on the predicted flow field. Figure 3 shows the velocity contours around the Savonius VAWT for all four models. In all cases, we can see that the velocity is highest near the tips of the advancing blade (the blade moving with the wind), reaching a maximum value of approximately 1.675 m/s. This high-velocity region is critical because it is where the turbine generates the most torque.

However, the most important differences appear in the wake region behind the turbine. The key finding of this study is how each turbulence model predicts the size, shape, and behavior of the low-velocity wake. The wake is a region of highly turbulent, swirling air that has lost energy to the turbine. For example, the Reynolds Stress model, which is the most complex model, predicts a larger and more diffuse wake. In contrast, the simpler Spalart-Allmaras model shows a more contained and defined wake structure. This difference is critical for real-world applications, as the wake from one turbine directly affects the performance of any turbines placed behind it. This comparative analysis demonstrates that an accurate Wind Turbine Performance CFD study must carefully consider the selection of the turbulence model.

Figure 3: Velocity contours from the Savonius Vertical Axis Wind Turbine Fluent simulation, comparing the results from four different turbulence models.