Optimization of aerodynamic design for cascade airfoil by means of Boltzmann selection genetic algorithms
Jun Li, Hiroshi Tsukamoto, Nobuyuki Satofuka · 18th Applied Aerodynamics Conference · 2000
Aerodynamic design of cascade airfoil using Genetic Algorithms with single objective and multiple objectives has been presented in this paper. Both inverse and direct design problems are faced. In the first part of this work, Genetic Algorithms based on Boltzmann selection are applied to turbine cascade inverse design through minimizing difference between target pressure distribution and computed pressure distribution. The result shows that the pressure distribution of obtained cascade airfoil is well agreement with the target pressure distribution and so is the geometry shape. In the second part of this work, based on the multi-branch Boltzmann selection and Pareto criteria method, multiobjective Genetic Algorithms have been developed and used for compressor cascade airfoil design. Goal of the compressor cascade design is to search higher pressure rise and lower total pressure loss on the basis of Controlled Diffusion Airfoil(CDA) at the given flow condition. Pareto solution of multiobjective design can supply many design plans for decision maker to select. The cascade with higher pressure rise and lower total pressure loss can be considered to have higher aerodynamic performance than the existed CDA. The optimization results also confirm that the feasibility and robustness of the present Genetic Algorithms based on Boltzmann selection.