Multi-objective optimum design of propellers using the blade element theory and evolutionary algorithms

A multi-objective optimum design of propellers is presented. Traditionally, the definition of propellers is done by choosing models that already exist on the market. However, with the advancement of technologies such as 3D printing and computer numerical control, the possibility for the designer to seek an optimized propeller for his design is increasingly feasible. In this scenario, this paper aims to propose a methodology that allows the designer to find alternative propeller designs. The methodology comprises the creation, refinement, and extension of an airfoil database available from the literature. The aerodynamic analysis uses the blade element theory and the optimization with evolutionary algorithms. Seven multi-objective constrained optimization problems were formulated, with two, three, and four conflicting objective functions concerning the propeller aerodynamic and geometric information, such as the thrust coefficient, the bending coefficient, the efficiency, and the volume of the propeller. Among them, for instance, two proposed formulations are: (i) maximizing the thrust and minimizing the bending coefficients simultaneously, (ii) maximizing the thrust coefficient, minimizing the bending coefficient, minimizing the volume of the propeller, and minimizing the efficiency simultaneously. The experiments analyzed the optimization of small propellers for two different engines, combustion and electric brushless. Five evolutionary algorithms were used to solve these problems, and solutions were extracted from the Pareto fronts using multi-criteria decision-making according to the decision-maker preferences. Finally, the complete data of the extracted solutions are provided, such as dimensions, angles, shapes, and performance characteristics.