Investigation Of 3D Turbomachinery Aerodynamic Optimization Design Based On NUMECA And ISIGHT

Cascades as the thermal and power conversion basic unit of turbomachinery, the research of its aerodynamic theory and design methods,both at home and abroad has been the key subject in the area of turbomachinery aerothermodynamics.With the rapid development of computational fluid dynamics and modern optimization calculation,the optimization design technology of turbomachinery has developed rapidly.But turbomachinery optimization aerodynamic design is still a challenging area of research. In this paper, a three-dimensional mechanical design optimization system,which is applicable to the turbomachinery is developed based on the commercial software NUMECA and iSIGHT-a optimization platform.A three-dimensional aerodynamic design optimization study to a certain power turbine rotor blade is carried out.Mainly in the following several aspects of the work:First, Reference Blade is fitted by Autoblade module and Autoblade_fitting module of the software package NUMECA, and it istransformed into parametric blade. Blade stacking line will be fitted with two angle control curve. Five blade sections are selected as key sections to be fitted. Each section is defined by the type of used with end line, the camber line is construeted in the form of a 4 degree Bezier Curve of 4 control points. Both pressure and suction sides are construeted in the form of B-spline curves. These provide optimized design variables for the follow-up Optimization workSecondly, the integration research between NUMECA and iSIGHT is studied,three-dimensional fluid dynamics calculation software and optimization platform are united. A three-dimensional aerodynamic optimization system is constructed. The system includes three parts: geometric model updating module, automatic mesh generation module, and numerical calculation module, the module through the three repeated iterative calculation leaves the optimization of the complete revision.Third,a combinatorial optimization method applicable to the design of three-dimensional turbomachinery is proposed and implemented, The combinatorial optimization method is divided into four phases: Orthogonal matrix method to experiment design (DOE) analysis phase, adaptive simulated annealing algorithm (ASA) global search phase, response surface method(RSM) approximate modeling stage, RSM and sequential quadratic programming (NLPQL) combination of local optimization stage. Through the use of DOE method, it offers a in-depth study of how the changes in design variables affect the objective function and constraints, accesses to the nature of the design space and seize the principal contradiction of optimization problems, reduces the aerodynamic optimization leaves the scale of the problem; The turbomachinery aerodynamic optimization is such a nonlinear multi-optimization problem, simply using some conventional numerical optimization algorithm can not be effective for the global optimal solution, through the use of ASA method of global search by the initial approximation is an effective way to the overall solution. RSM fits the complicated relationship between the response, smoothes the design space "noise" to prevent numerical optimization methods into local maximum points so as to gain a good robustness and adaptability. Local optimization phase in the use of the three-dimensional numerical simulation program to correct the similar model, NLPQL method is used for partial optimization to approximate model,not only ensures the accuracy of the analysis of the aerodynamic performance, and greatly reduces the actual calculation cost of numerical simulation.Finally, with the highest efficiency as the objective function,the three-dimensional aerodynamic optimization system is used to design a certain power turbine blade. The results show that the aerodynamic performances of optimization blade have significantly improved. It shows that the aerodynamic design optimization method of this paper is an effective way to access highly efficient and low-loss performance blade.