基于Co-Kriging代理模型的扇形冷却孔结构优化  

Structural optimization of fan-like cooling holes based on Co-Kriging surrogate model

作　　者：陈萍 左宁 王泽轩 齐婷婷 于明月 CHEN Ping;ZUO Ning;WANG Ze-xuan;QI Ting-ting;YU Ming-yue(College of Civil Aviation,Shenyang Aerospace University,Shenyang 110136,China;College of Autornation,Shenyang Aerospace University,Shenyang 110136,China)

机构地区：[1]沈阳航空航天大学民用航空学院,沈阳110136 [2]沈阳航空航天大学自动化学院,沈阳110136

出　　处：《沈阳航空航天大学学报》2022年第3期63-71,共9页Journal of Shenyang Aerospace University

基　　金：国家自然科学基金(项目编号:51605309);辽宁省教育厅重点攻关项目(项目编号:JYT2020153)。

摘　　要：为进一步提高航空发动机的冷却效率,对扇形气膜冷却孔进行优化研究,优化参数为直孔段孔长L,前向扩展角α和后向扩展角β,优化目标为气膜孔下游内平均气膜冷却效率。首先对优化变量进行拉丁超立方采样,然后对疏、密两种质量的网格进行有限元分析,得到两种可信度数据,进而构建Co-Kriging代理模型,最后引入遗传算法进行全局寻优,寻找最优的扇形孔结构。在吹风比为1.5的工况下,优化后的直孔段孔长、前向扩展角和后向扩展角分别为2.6D、41°、14°,冷却效率提升了45%。研究结果表明,较小的直孔段和较大的后向扩展角可以有效地抑制肾型涡对的产生,气膜冷却效果更好。In order to improve cooling efficiency of aero-engine,optimization of fan-like film cooling holes was studied.Length of straight hole section L,forward expansion angleαand backward expansion angleβwere defined as optimization parameters,while average film cooling efficiency in downstream of the film hole was defined as a optimization objective.Sample points of the parameters were obtained by Latin Hypercube method,and then finite element calculation was carried out in view of two meshes with different qualities to establish the Co-Kriging surrogate model.Genetic algorithm was adopted to find optimal structure of the fan-like holes.When straight hole length,forward and backward expansion angles were 2.6D,41°and 14°respectively,the cooling efficiency increased by 45%under blowing ratio of 1.5.The results indicated that a smaller straight hole section and a larger backward expansion angle could inhibit generation of kidney-like vortex pairs effectively,resulting in a better film cooling effect.

关 键 词：扇形孔 优化设计 Co-Kriging代理模型 遗传算法 冷却效率 

分 类 号：V231.1[航空宇航科学与技术—航空宇航推进理论与工程]