This paper presents an automated optimization design methodology for turbomachinery blades using multiobjective evolutionary algorithms (MOEAs) and Navier-Stokes solver. The multi-branch Tournament selection and Pareto solution conception are used in the presented MOEAs. Elitist method and generation gap are adapted to ensure the optimization performance and decrease the computation expense. The Bezier-curves are utilized to parameterize the designed blade profile and corresponding control points are used as the designed variables. Reynolds-Averaged Navier-Stokes solver is applied to evaluate the aerodynamic performance of the designed candidates. Two design cases of the 2D and 3D compressor blade are used to demonstrate the presented methodology performance. A 2D axial compressor blade is optimized for maximization of the static pressure rise and minimization of the total pressure loss at the fixed flow condition. The Pareto solutions are obtained using the presented optimization algorithm. The detailed analysis between the certain Pareto solutions with the higher static pressure rise and the lower total pressure loss and the initial design are illustrated. A 3D centrifugal compressor impeller blade is optimized as the second case. The maximization of the pressure rise and blade load and minimization of the rotational total pressure loss at the given flow conditions are worked as the design targets. The present method obtains many reasonable Pareto optimal designs that outperform the original centrifugal impeller at the designed condition.

This content is only available via PDF.
You do not currently have access to this content.