Process Integration and Automated Multi-Objective Optimization Supporting Aerodynamic Compressor Design


Abstract

Nowadays industrial aerodynamic compressor design is based on mature computer programs developed during several decades. State of the art is to split the complex design process into subsequent design subtasks which are solved by different experts via time-consuming parameter studies. Isolated design of subproblems based on human intuition, however, will result in sub-optimal solutions only. Due to the increasing demand on higher aero engine performance and design cycle time reduction the aspects of process integration and automation as well as numerical optimization become more and more important in today’s aerodynamic compressor design. The intention of this work is to show how process integration and optimization can be used efficiently to support engineering design work in optimal solution finding. Since the aerodynamic compressor design is characterized by many design parameters, multiple constraints and contradicting objectives, multi-objective optimization is used to find Pareto-optimal solutions from which the design engineer can choose trade-offs for his particular design problem. The improvements in terms of process acceleration and design optimization are demonstrated for three selected, but typical industrial engineering design tasks required in three different design phases of the aerodynamic compressor design process, namely preliminary design, throughflow off-design, and blading procedure.