Evolutionary optimization of a Savonius rotor with sandeel-inspired blades

Abstract

A possible way to capture the kinetic energy available in the wind and then convert it into a useful form of energy is to use wind turbines. Out of many kinds of wind turbines, a lot of in-depth studies have been conducted to improve the performance of the drag-based vertical-axis wind turbine called Savonius rotor. This kind of vertical-axis wind turbine usually suffers from poor energy conversion efficiency, particularly for its conventional shape (thin half-cylinders). In the current work, a bionic blade shape of the Savonius-type wind turbine rotor inspired by sandeels (also called sandlances) is proposed, combined with an optimization step in order to maximize performance. Sandeels are slim elongated fish that often swim in vast shoals. The bionic blade is described by two principal geometrical parameters, both divided by the blade length (d); the maximum blade camber (fmax) and its location (Xfmax). The optimization relies on evolutionary algorithms to maximize the average power coefficient of the Savonius rotor with sandeel-inspired blades when varying the two design parameters. A total of 128 individuals have been evaluated numerically using a series of two-dimensional transient simulations with the software Star-CCM+. In comparison to the conventional rotor with two arc-type semi-circular blades, the optimal design with sandeel-inspired blades exhibits a considerable increase in turbine efficiency of up to 9.21% at the design tip speed ratio (λ = 0.8). Investigating the performance over the entire operating range results in a maximum power coefficient of 0.255 by the optimal rotor, leading to a gain of about 10.58% with respect to the standard rotor at λ=1.