摘要
以单级跨声速风扇为研究对象,采用叶片参数化造型、人工神经网络构建近似函数及遗传算法寻优相结合的方法,在级环境下对转子叶片进行周向积叠的多目标三维气动优化设计。结果表明,该单级风扇在保持质量流量和总压比基本不变的前提下,优化后设计点绝热效率提高了0.88%,优化后所获得的根部反弯、顶部正弯叶片,可有效改善叶根和叶片中上部的流动损失。
Taking the single-stage transonic fan as the research object, the multi-objective three dimen?sional aerodynamic optimization design of circumferential stacking was carried out under stage environmentbased on the methods consisting of blade section parameterization, artificial neural network and genetic al?gorithm. The results show that under the condition of the mass flow and pressure ratio were unchanged, theadiabatic efficiency increased by 0.88%. The optimized blade with negative curve in the hub and positivecurve in the tip effectively improve the flow loss near the hub and the upper middle of the blade.
作者
黄磊
余华蔚
HUANG Lei;YU Hua-wei(China Gas Turbine Establishment,Chengdu 610500,China)
出处
《燃气涡轮试验与研究》
北大核心
2016年第3期30-34,共5页
Gas Turbine Experiment and Research
关键词
航空发动机
跨声速风扇
弯叶片
优化设计
遗传算法
人工神经网络
aero-engine
transonic fan
bow blade
optimization design
genetic algorithm
artificial neural network
