为拓展叶型设计空间，实现跨声速涡轮叶型设计中的精细化型面控制，并在叶型优化时考虑叶身气膜冷气出流的影响，提出了基于Bezier曲线构型的“12+7参数化造型方法”，叶型吸力面进口段、出口段、压力面和前缘均采用Bezier曲线，通过集成自主涡轮叶型造型程序、MATLAB人工神经网络工具箱及NUMECA商用软件，搭建了可考虑冷气的涡轮叶型数值优化平台，在考虑冷气掺混的条件下，对典型重燃透平一级导叶进行了气动优化研究，优化变量包括安装角、后弯角、前缘半楔角以及吸力面Bezier曲线控制系数LLS1和LLS2，优化目标为总压损失系数最小。结果表明：经过优化后的叶型模拟与实验拟合良好，总压损失系数减小了7.23%(从0.06519减小为0.06048)，能量损失系数减小了7.76%(从0.0518减小为0.04778)，气动性能得到增强。

In order to expand the design space and realize the precise control of blade surfaces in transonic blade design, and to consider the influence of film cooling during airfoil optimization, a "12+7 parametric modeling method" based on Bezier curves is proposed. Bezier curves are applied to the suction surface inlet, outlet, pressure surface, and leading edge of the blade. A numerical optimization platform for turbine blade design considering film cooling is constructed by integrating an autonomous turbine blade modeling program, MATLAB artificial neural network toolbox, and the commercial software NUMECA. The aerodynamic optimization with coolant ejection is performed on the first-stage guide vane of a typical heavy-duty gas turbine. The optimization variables include the stagger angle, trailing edge angle, leading edge 1/2 wedge angle, suction surface Bezier curve control coefficients LLS1 and LLS2, and the optimization objective is to minimize the total pressure loss coefficient. The results show that the simulation of optimized blade profile fits well with experimental result, with a decrease in total pressure loss coefficient by 7.23%(from 0.06519 to 0.06048) and a decrease in energy loss coefficient by 7.76%(from 0.0518 to 0.04778), indicating improved aerodynamic performance.

TK221

国家科技重大专项(J2019-II-0010-0030)