Combined Experimental and Numerical Investigation of Geometric Parameters on the Performance of Savonius Vertical Axis Wind Turbine

Abstract

ABSTRACT

Wind energy has become one of the most widely deployed renewable sources because it is scalable, mature, and suitable for both utility-scale and distributed generation applications. This work investigates the aerodynamic development and experimental validation of a Savonius vertical-axis wind turbine (VAWT) intended for low to medium wind-speed applications. A complete workflow was adopted starting from a validated CFD model, followed by systematic geometric development, optimal design selection, additive manufacturing, and wind-tunnel-style testing, then concluding with a direct comparison between numerical and experimental results. The CFD simulations were conducted in ANSYS-CFX using air at  and 1 atm, and the turbulence model . Model validation was first performed at an inlet wind speed of  using the optimal reference configuration reported in the literature (OR0.109BS2BN2). After validation, the baseline turbine was analyzed at  and used as the starting point for parametric development. Three geometric variables were introduced and varied in a controlled one-factor-at-a-time strategy: the upper tilt angle  step), the blade orientation angle, and an arc-edge parameter  (100-500 with stepwise 100 increments). The results showed that  reduced performance across the tested range; therefore,  was retained. In contrast,  had the strongest influence and increased reaching. The effect of  was comparatively weak, producing only minor changes once  was optimized. The optimum turbine was fabricated using 3D printer with Hyper PLA material. A low-medium wind-speed blower test rig was designed and built to generate uniform flow in the range of  . Experimental performance curves were produced at , and the trends matched the CFD predictions: the power coefficient increased with TSR up to an optimum region around , then decreased at higher TSR. The peak experimental value was approximately  near , compared with a corresponding CFD peak near . The remaining differences are attributed to practical effects not fully represented in the numerical model, including mechanical losses, manufacturing tolerances, surface roughness, and measurement uncertainty. The combined CFD-experimental methodology successfully identified an improved Savonius rotor configuration and demonstrated a consistent validation-to-optimization workflow suitable for small-scale wind energy harvesting.