The Study Of Aerodynamic Numerical Optimization Of Axial Compressor Blade Section

With the development of CFD and the theory of numerical optimization, aerodynamic numerical optimization of blade section attracts more and more attention. Aerodynamic numerical optimization of blade section is a kind of method that combines simulation of cascades and numerical optimization. It depends on the numerical optimization methods instead of experience of designers to improve a design. Compared with the traditional methods, it is more objective, economical and efficient. The method has been one of the hot subjects of turbo-machinery field.In the process of optimization design, the modeling approach of blade, the selection of design variables and the design of objective function are important factors that impact the optimization results deeply. From the above three aspects, the middle section of the 4th stage vane of a certain type of compressor was selected as the initial blade. The initial blade was fitted by Bezier curves and least squares method, and then the parametric blades were modeled based on the way of arc pasting thickness. The mesh was generated automatically by autogrid. Flow field simulation is based on CFX and SST turbulent model. On the basis of previous, a new objective function was designed which the amount of calculation was greatly reduced and the optimization direction meets the requirement of this paper well. The selection of design variables is based on the sensitivity of modeling parameters to objective function,and the sample database of sensitivity analysis was generated by Latin Hypercube experiment design methods. After comparison, the optimization algorithms which are suitable for blade optimization design were selected. Finally the above process was integrated by Isight, and the methods of aerodynamic numerical optimization of blade section in this paper were established. A detailed comparison from the angle of attack-loss, Mach - Loss and Reynolds-loss was conducted to prove the superiority of optimized blade, and thus verified the aerodynamic numerical optimization of blade section in this paper.