Improving the Power Output of Vertical Axis Wind Turbines Using an Optimized Grooved Blade Design

Experimental Setup

To accurately evaluate the aerodynamic performance of the H-Darrieus turbine, experimental testing of the complete three-dimensional turbine model was carried out. During the fabrication stage, the first step involved constructing the primary blade mold using wooden boards (Figure 3). This mold defined the required blade geometry along a 1-meter length. A hot-wire cutter was then used to shape the foam into the NACA 0021 airfoil profile (Figure 4), ensuring that the blade cross-section matched the geometry obtained from the numerical design. To further improve the dimensional accuracy, a 200-W laser cutting machine was used to refine the airfoil outline with high precision, as illustrated in Figure 5. A thin fabric layer impregnated with calcium sulfate and starch was applied to the outer blade surface. After moistening and drying, this coating hardened and provided additional strength and resistance to external environmental effects.

All blade geometric parameters – including airfoil type (NACA 0021), chord length, maximum thickness, blade height, rotor diameter, solidity, and groove dimensions – are summarized in Table 1. The grooved and non-grooved rotor models share the same global geometry, with the groove diameter and position indicated in the CAD drawing presented in Figure 6.

The smooth (non-grooved) rotor has the same global geometry (airfoil, chord, span, diameter, and solidity), but without the grooves. A dimensioned CAD view showing the blade profile and groove locations is provided in Figure 6.

Cork was selected as the core material for its low cost and ease of cutting; however, its outer surface required hardening to improve durability during rotation. Several hardening materials were tested. Commercial hardeners were initially applied, but they reacted chemically with the cork and caused partial dissolution. High-performance coating materials were also evaluated but found to be prohibitively expensive for multiple blades. After testing, a suitable coating was identified: a mixture of calcium carbonate and starch, which forms a rigid shell once sprayed with water and dried for approximately 30 minutes. This coating provided an optimal balance between stiffness, weight, and cost. During assembly, wooden supports were used initially but exhibited insufficient strength and stability, particularly under rotation. They were therefore replaced with lightweight wrought-iron supports to increase rigidity and ensure reliable mounting of the blades. The final groove dimensions were verified manually using calipers. The diameter tolerance was maintained within ±0.5 mm, and the depth tolerance was maintained within ±0.3 mm, ensuring that the manufactured grooves matched the design geometry required for aerodynamic testing.

In the manufacturing process of the fixed structure for the vertical Darrieus turbine, the focus was on designing and building multiple components to achieve the stability and durability required to support the turbine under real operating conditions. The fixed structure (Figure 7, Figure 8), which is directly connected to the ground base, consists of two wooden discs, each with a diameter of 250 mm and a thickness of 2 mm. These discs were manufactured using a laser machine to ensure precision in dimensions and compliance with the engineering design requirements of the turbine. The first disc is connected to the fixed structure, which is considered a vital part of ensuring the stability of the turbine, as it secures the entire structural framework of the turbine to the ground. Meanwhile, the second disc is connected to the generator, which in turn carries the vertical blades of the turbine. This arrangement allows for flexibility of movement, as the second disc rotates with the generator when the blades move due to the wind, resulting in the generation of electrical energy.

The load-bearing part of the vertical blades was constructed using a wooden board measuring 500 mm in length, 30 mm in width, and 2 mm in thickness. This wooden board serves as the main supporting structure for the blades, as it is connected to the rotating disc linked to the generator. This supporting structure ensures the stability of the blades during rotation and allows for the transfer of mechanical energy generated by the wind to the axis of the electric generator. A compact permanent-magnet generator was mechanically coupled to the turbine shaft to obtain the electrical output during rotation. The unit has a nominal mechanical power of approximately 35 W and a rated speed of 3,400 RPM. When operated as a generator, it provides reference voltage levels of about 6 V at 120 RPM, 12 V at 240 RPM, 48 V at 480 RPM, 96 V at 1,000 RPM, and 192 V at 2000 RPM. These values were used only as indicative reference points to understand the expected voltage range during operation, while the turbine speeds used in the experiments were those corresponding to the tip-speed ratios.

Another disc similar to the lower discs in terms of diameter and thickness has been placed at the top of the turbine, with the aim of securing the three manufactured blades and ensuring the complete balance of the turbine structure (Figure 9, Figure 10, Figure 11). This precise process in designing and installing the structure ensures the highest levels of efficiency for the turbine, as the rotational motion caused by the wind is transmitted to the generator with high efficiency, resulting in electricity production when the turbine rotates at various speeds.

2.1.1.

Wind Blower

For the purpose of testing the turbine at medium velocities from 5 m/s to 16 m/s, a blower with a centrifugal fan was used to study the aerodynamics performance of H-Darrieus turbine. The dimension of the tunnel of wind blower was 1 meter high, 1 meter wide, and 2 meters length, as shown in Figure 12. The blowing capacity used is 2 HP. A grid has been manufactured with square perforated holes. The dimensions of each hole are 1 cm × 1 cm to obtain uniform air velocity. The intensity of the flow disturbance was less than 5%, and the regularity of the flow was more than 95%. Figure 13 shows the stages of manufacturing the wind blower tunnel.

Each operating point was measured three times to ensure repeatability (Figure 14, Figure 15). The reported voltage, current, and rotational speed values correspond to the mean of three consecutive runs. Measurement uncertainty was quantified using standard deviation, variations of 0.1–0.3% for voltage, 2–6% for current, and below 1.5% for rotational speed. The resulting uncertainty in electrical and mechanical power was approximately 1–3%, confirming stable and repeatable turbine–generator performance. Wind velocity and rotational speed were measured using two independent instruments to ensure consistent operating conditions during the experiments. The free-stream air velocity at the wind-tunnel outlet was recorded using a HOLDPEAK HP-856A digital anemometer, which was used solely to verify the linear airflow velocity (V∞ = 5 m/s). The device provides a measurement range of 0.001–45 m/s with an accuracy of ±3% ± 0.1 m/s, which is sufficient for confirming the inlet flow condition throughout all tests. The turbine rotational speed was measured using a Hall-effect RPM sensor (Model 5,135ZSB), operating under an AC/DC 8–24 V supply and capable of measuring 10–9,999 RPM with ±0.1% accuracy. The sensor was installed with a fixed 10 mm gap between the magnet and the sensing head to ensure stable signal detection, while avoiding metallic surfaces and electromagnetic interference. Together, these instruments provided reliable measurements for both airflow and rotor speed under the test conditions.

Numerical Simulations

The numerical simulations were performed using ANSYS CFX 2024 R1 in a transient formulation to accurately capture the unsteady aerodynamic behavior of the rotating turbine. A 3D computational domain was used, and the flow was modeled with the SST k–ω turbulence model, which provides reliable prediction of separation and rotational effects. The pressure–velocity coupling was solved using the High-Resolution scheme, while Second-Order spatial discretization was applied for momentum and turbulence transport equations. The turbine blades were modeled with the NACA 0021 profile, using a chord length of 0.206 m, height of 0.824 m, rotor diameter of 1.03 m, and solidity of 1.2. All models are shown in Table 2. The computational domain used for the Computational Fluid Dynamics (CFD) simulations was a rectangular box following the layout adopted in a study by Sobhani [16], with an overall length of 30D and a height of 15D, where D is the rotor diameter (Figure 16). The rotor center was located 5D downstream of the inlet and 25D upstream of the outlet. All outer boundaries (top, bottom, and side walls) were treated as no-slip walls. At the inlet, a uniform velocity of 9 m/s was prescribed with a turbulence intensity of 3% and an automatic turbulence length scale, while the outlet was specified as an average static pressure of 0 Pa. Two fluid sub-domains were defined: an inner rotating domain surrounding the VAWT rotor and an outer stationary domain, coupled through a Frozen Rotor interface. Air was modeled as an ideal gas at a reference pressure of 1 atm and a temperature of 25°C. Both domains were discretized using an unstructured tetrahedral mesh in ANSYS-CFX. The final grids contained approximately 3.1 × 10⁶ cells for the smooth baseline rotor and 5.2 × 10⁶ cells for the grooved rotor. A mesh-convergence study was performed by successively refining the grid until further refinement produced negligible changes in the predicted rotor torque and peak C_p.

The mesh contained approximately 5.2 million elements and 109,628 nodes, as shown in Figure 17. Input parameters were implemented for the groove diameter (d) and the distance (L) between the driving edge and the groove edge, allowing these geometric variables to be modified directly from the interface. The mesh automatically updated after each modification. The turbine torque was defined as an output parameter, enabling automatic extraction of the mechanical performance after every simulation run.

Experimental Analysis of Airfoil With Circular Grooves

The performance of analysis of airfoil with circular grooves of vertical wind turbine (H-Darrieus) was tested after the three blades were manufactured. The first model, which contains grooves designed on the blades, was tested, while the second model had no grooves. The aim of this experiment was to compare the performance of the two models in terms of the Cp and the tip speed ratio λ at different rotational speeds, by varying the load and at a speed of 5 m/s. As shown in Table 3, the extracted values of the Cp were compared at different tip speed ratios λ for the two models:

The results showed that the performance of this model gradually improved with the increase in tip speed ratio (TSR). The value of Cp (Figure 19) was initially very low at 0.04 with a tip speed ratio of λ = 0.93, but with the increase in tip speed, the Cp gradually rise to its maximum value of 0.41 at a tip speed ratio of λ = 5.12. The same applies to the power (Figure 20). This increase in Cp reflects the significant role that the grooves play in improving airflow around the blades, thereby increasing the turbine’s efficiency. For the turbine without the grooves, the results showed that the performance of the CHDR model was slightly lower than that of the model with grooves (GHDR). For example, at a tip speed ratio of λ = 5.12, the Cp value for this model was 0.40 compared to 0.41 for the model with grooves (GHDR). The overall performance was also slightly lower across all tip speed ratios. Experiments have shown that adding grooves to the blades improves airflow around the turbine, leading to an increase in the lift coefficient and consequently an increase in Cp, due to the vertexes generated are small compared to the CHDR, and the high pressure in the pressure side of the airfoil (lower surface). This improvement is clearly evident in the high-Cp values at high tip speed ratios. This can be explained by the fact that the grooves generate small vortices at the leading edge of the blades, which reduces flow separation and improves the aerodynamic efficiency of the blades. Although the performance was not significantly better compared to the model with the grooves, the results show that the removal of the grooves led to a slight decrease in turbine efficiency. This decrease could be due to the absence of the vortex effects produced by the troughs, leading to an increase in flow separation and deterioration in the lift coefficient. The enhancement in turbine performance was quantified by calculating the percentage improvement in the Cp. The improvement was evaluated using the standard relation:

Using this approach, the modified design showed a 54.7% improvement relative to the baseline model. When compared to the model without grooves, the grooved configuration achieved a 10.8% increase in the maximum Cp.

Previous studies on grooved or roughened blades and small-scale vortex generators for VAWTs have shown that surface modifications can enhance aerodynamic performance by delaying flow separation and improving boundary–layer momentum. Researchers such as Sobhani et al. and Sedighi et al. [16, 17] demonstrated that introducing streamwise vortices along the blade surface increases the peak Cp by 8–20%.

Experiments were conducted to measure the impact of grooves on the mechanical torque produced by the vertical wind turbine. As shown in Figure 21, two models were tested: the first contains grooves in the blades, and the second had no grooves. The C_T (Figure 22, Table 4) was calculated at different ratios of the tip speed ratio for each model. At a low tip speed ratio of λ = 0.96, the C_T = 0.284, indicating that the presence of grooves has helped increase the mechanical torque produced by the turbine at these speeds. With the increase in the tip speed ratio to λ = 2.44, the C_T rose to 0.586, indicating that the grooves help improve the airflow around the blades, thereby increasing the resulting torque. At higher speeds, λ = 5.54, the torque decreased slightly to C_T = 0.460, but it is still higher than the model without grooves. As for the model without grooves (Figure 22, Table 5): At λ = 0.96, the C_T was significantly lower (0.038), indicating that the removal of the grooves reduced the turbine’s ability to generate mechanical torque at these speeds. Even with the increase in tip speed ratio to λ = 2.44, the C_T remained lower than the model with grooves, where the value of C_T = 0.080. At high tip speeds, λ = 5.54, the torque C_T = 0.071, which is clearly lower compared to the model with grooves. The model with grooves showed good performance in terms of mechanical torque across all tip speed ratios. This demonstrates the role of the grooves in improving airflow around the blades, which increases the turbine’s efficiency and generates more torque. However, the model without the grooves showed a clear decline in the resulting mechanical torque, as the CT values were lower across all peripheral speeds. This indicates that removing the grooves reduces the blades’ ability to generate the lift necessary to increase mechanical torque.

Visualization of VAWT

In this section, the velocity contours of three different designs of the vertical H-Darrieus wind turbine are analyzed. The velocity contour plots (Figure 23, Figure 24) show wind speed distribution around the turbine blades for the base design, the modified design, and the optimal design with grooves. The velocity contours for the base design, taken from previous research, show clear separation zones and turbulence areas behind the turbine blades. Airflow speed increases significantly around the leading edge of the blades, creating areas of high speeds, while areas of low speeds appear at the rear (the trailing edge). This distribution indicates a significant energy loss in the wake of the blades due to limited aerodynamic performance. The absence of any improvements in the blade design leads to a limited ability to control airflow and maximize energy extraction.

As for the modified design, which includes engineering improvements based on simulation results, it has a more balanced distribution of speeds compared to the base design. The high-speed areas near the leading edge of the blades show greater organization, with reduced low-speed turbulence areas at the rear. These improvements indicate better control of airflow around the blades, leading to increased energy extraction efficiency. The modified design shows a significant reduction in turbulence and flow separation, enhancing aerodynamic performance compared to the baseline model. The optimal design, which includes the addition of two grooves with a diameter of 7 mm and a distance of 10 mm between them, shows a more efficient distribution of speeds among the three models. The grooves contribute to the formation of organized vortices that improve the pressure distribution on the blade surface. This leads to an increase in lift and a reduction in drag. The velocity contours show a more streamlined and uniform flow, with a significant reduction in downstream disturbances and a noticeable improvement in speed recovery behind the turbine. These characteristics confirm that the addition of the grooves enhanced the aerodynamic performance of the turbine, leading to improved energy extraction and reduced losses.

The pressure distribution on the blades and the surrounding field of the turbine shows significant differences between the three designs (basic, modified, and optimal), reflecting the impact of engineering modifications on aerodynamic performance. In the base design (Figure 25a), the pressure distribution shows clear high-pressure zones at the leading edge of the blades, while low-pressure areas appear on the rear side of the blades, indicating significant turbulence and loss of airflow. These results indicate that the basic design suffers from significant flow separation behind the blades, reducing energy extraction efficiency, and the sharp transition between high and low-pressure areas leads to greater mechanical stress on the blades, which may affect the turbine’s sustainability. As for the modified design (Figure 25b), it shows clear improvements in pressure distribution compared to the basic design. The high-pressure areas become more streamlined and less sharp at the leading edge, and the low-pressure area on the rear side shows a more organized distribution, reducing disturbances and improving airflow. These modifications increase lift force and reduce drag forces, leading to improved overall turbine performance and increased efficiency in converting wind energy into mechanical energy. The optimal design, which includes the addition of two grooves with a diameter of 7 mm and a distance of 10 mm, shows the best pressure distribution among the three models (Figure 25c), where the high-pressure areas at the leading edge are more balanced, reducing airflow separation and enhancing dynamic flow. The low-pressure areas on the rear side become less severe and more consistent, reducing wind energy losses. This design shows an excellent balance between high and low pressure, improving the overall efficiency of the turbine and reducing aerodynamic dynamic stresses. Also, the pressure disturbance in the baseline design is clearly visible, while the modified design significantly reduces these disturbances. The optimal design shows less disturbance with a balanced distribution, and the high-pressure effect at the leading edge is more uniform in the modified and optimal designs, which increases blade lift and torque. As for the low-pressure areas behind the blades, they are more stable in the optimal design, which improves airflow and reduces energy loss.