Otimização dos perfis aerodinâmicos de uma hélice de turbina eólica

Abstract

Multidisciplinary optimization is essential in the performance of structures used in the aeronautical industry, in the automobile industry, in the industry in the context of renewable energy such as wind farms and so on. In this context, for example, the structural and aerodynamic optimization of propellers, airfoils and turbine blades of wind turbines is illustrated. The optimization of these elements can contemplate several objectives, restrictions and search spaces. It can be parametric or dimensional, shape, topological or even a combination of all these simultaneously. Also, problems can be formulated with a single or multiple objectives. The solution of these optimization problems can be done by deterministic methods, metaheuristics or hybrid methods. This dissertation aims to develop a methodology capable of optimizing the geometry of an aerodynamic profile, aiming at aerodynamic improvement using Genetic Algorithms. To treat the aerodynamic solution, the XFOIL platform, by Mark Drela, widely used in the literature, was used. The goal of optimization is to maximize the lift and drag ratio. A Bézier parameterization was adopted for the generation of the airfoil shape where the coordinates of the Bézier curves are the design variables, which greatly facilitated the description of the airfoil shape. A generational genetic algorithm was used with real coding of the design variables and the simulated binary crossover (SBX) recombination operators and a polynomial mutation were adopted. The MATLAB Platform for Evolutionary Multi-Objective Optimization (PlatEMO) was adopted to solve the optimization problems presented in this dissertation. The optimized profiles are implemented in a case study, a horizontal axis wind turbine produced by NREL. Through the QBlade tool, it is possible to obtain power, torque and thrust information generated by the turbine with the optimized profiles and compare them with the information of the commercial standard turbine. In addition, it was also possible to perform a profile-by-profile comparison between each optimized profile and its associated business profile. It was observed that with the profiles developed in this dissertation, there were aerodynamic improvements in the comparison made, either in the profile-by-profile comparison, or in the comparison via turbine characteristics