Optimization and parametric analysis of a novel design of Savonius hydrokinetic turbine using artificial neural network

This study focuses on enhancing the efficiency of vertical axis Savonius Hydrokinetic turbines designed for marine applications, historically characterized by a power coefficient below 0.1. Prior efforts aimed at improving rotor performance have primarily involved modifications to blade designs. In this article, a new approach is introduced, incorporating twisted blades inspired by the Archimedes screw turbine. Utilizing a 3D incompressible flow analysis based on the Navier-Stokes equation, this research explores and compares the turbine's effectiveness with varying screw pitches (0.5, 0.75, 1). The system of equations is solved numerically using ANSYS 2020 R2 fluid fluent. The performance assessment involves contrasting each proposed rotor against a pitchless semi-circle rotor. An innovative aspect of this work involves investigating the impact of asymmetry using two different ratios (2:1 and 3:1). Specifically, the lower half of the optimal pitch screw remains constant, while the upper half varies based on these ratios. To understand performance trends, the study employs visualizations of pressure, velocity contours, and streamlines to grasp the flow field and its underlying principles. Turbulent kinetic energy and eddy viscosity are also visualized. The results reveal an 18.25 % improvement in performance with the proposed rotor featuring a pitch screw of 0.5. Notably, the asymmetric rotor with a 2:1 ratio demonstrates the highest performance. According to the ANN, the optimum pitch screw value is determined to be 0.6, achieving a power coefficient of 0.1938. This investigation employs novel design modifications and asymmetrical configurations, offering valuable insights into significantly enhancing the performance of Savonius turbines for marine applications.