摘要
基于自由变形技术(FFD)和计算流体力学(CFD)建立了离心压缩机叶片形状优化设计的方法,从而避免了对复杂的实体造型本身进行参数化,提高了优化效率。首先基于自由变形技术建立分流叶片外形的参数化表示,根据优化拉丁超立方试验设计方法构建控制点变化的样本空间,接着通过CFD数值仿真获得各样本的性能参数并建立响应面分析法(RSM)模型,以压缩比与等熵效率最大化目标构建多目标优化模型,进一步采用最小偏差法转化为单目标优化模型并进行求解,最后分析比较了形状优化对压缩机性能和流道内流体流动的影响。结果表明形状优化后压缩机的压缩比提升0.46%,等熵效率提升0.84%,同时减少了流道内低速区域,降低了流动损失。
A method for optimizing the shape of centrifugal compressor blade was established based on free form deformation technology(FFD) and computational fluid dynamics(CFD), thereby avoiding parameterization of complex solid model itself and improving optimization efficiency. By using FFD to parameterize the splitter blade, the sample space for the change of control points was constructed according to the Optimal Latin Hypercube method. Then the performance parameters of each sample were obtained through CFD numerical simulation and an response surface methodology(RSM) model was established. Further, a multi-objective optimization model was built with the maximization of the compression ratio and isentropic efficiency as the objective function. And a single objective optimization model was obtained by using the minimum deviation method. Finally, the influences of shape optimization on compressor performance and fluid flow in the flow channel were analyzed. The results showed that the optimized centrifugal compressor model had a compression ratio increase by 0.46% and isentropic efficiency increase by 0.84%, and the low-speed area in flow channel as well as the flow loss was reduced.
作者
邵佳丰
罗晨
周怡君
曹火光
SHAO Jiafeng;LUO Chen;ZHOU Yijun;CAO Huoguang(School of Mechanical Engineering,Southeast University,Nanjing 211189,China;Chengxian College,Southeast University,Nanjing 210088,China;Research and Development Department,Gunai Heavy Industry(Suzhou)Corporation Limited,Suzhou 215400,China)
出处
《航空动力学报》
EI
CAS
CSCD
北大核心
2021年第6期1315-1323,共9页
Journal of Aerospace Power
基金
国家自然科学基金资助项目(51975119,51575107)
江苏省“六大人才高峰”高层次人才选拔培养资助项目(GDZB-002)。
关键词
离心压缩机
分流叶片
FFD技术
响应面模型
最小偏差法
centrifugal compressor
splitter blade
FFD technology
response surface model
minimum deviation method
