Optimizing propeller performance: a comprehensive constrained multi-objective design approach using blade element theory and evolutionary algorithms

Abstract

Currently, concerns about energy efficiency, sustainability, and the environment are growing, as proposed by the UN through The Global Goals (Goal 9 - Industry, Innovation and Infrastructure), and the search for new, more efficient, and cleaner solutions is notorious in all engineering fields. With the advancement of new manufacturing techniques, discovery and development of new materials, and expansion of computational capabilities, new opportunities for study in the field of aeronautical engineering arise. The present work proposes the elaboration of a new methodology to be used in the design and optimization of propellers, as well as the algorithms and couplings necessary for its accomplishment. During the work, a search was made for aerodynamic profiles, and a database was created with about 1500 of these. Such profiles had their coordinates standardized and refined. All airfoils were analyzed using panel methods through the XFOIL code. Using the evolutionary algorithms AGEMOEA, ARMOEA, MSOPSII, NSGAII, and NSGAIIARSBX present in the PlatEMO optimization platform coupled to the propeller analysis program JAVAPROP, in addition to the coupling of a structural analysis code, optimizations were performed for different objectives. This set of elements was added to PlatEMO as a problem and named OptProp. Initially, analyzes were carried out taking into account only the dimensionless parameters of the propellers, through seven different multi-objective optimization problems for two different powertrain groups. Then, an optimization is performed to minimize the power required for two different flight points and a propeller is selected from the Pareto front obtained. With such a propeller established, operational optimization is sought for a given mission by varying the rotational speed of the system. In all optimizations, geometric constraints are considered, and, in operational optimization, structural constraints through natural frequencies and the Campbell diagram are also used. Energy savings close to 1.4% were found after operational optimization.