﻿ 基于定制可控扩散叶型的压气机多工况优化
 舰船科学技术  2022, Vol. 44 Issue (23): 92-98    DOI: 10.3404/j.issn.1672-7649.2022.23.018 PDF

1. 海军装备部驻哈尔滨地区第三军事代表室，黑龙江 哈尔滨 150078;
2. 中国舰船研究院，北京 100192;
3. 航天时代飞鸿技术有限公司，北京 100097

Multi-condition optimization of compressor based on customized controllable diffusion profile
YU Yang1, REN Jian-yong2, XIAO Nan-nan3, CHENG Yuan2
1. The Third Military Representative Office of the Naval Equipment in Harbin, Harbin 150078, China;
2. China Ship Research and Development Academy, Beijing 100192, China;
3. Aerospace Era Feihong Technology Co., Ltd., Beijing 100097, China
Abstract: In this paper, a multi-condition optimization design method of compressor based on customized controllable diffusion profile is established. A customized controllable diffusion profile design method combining aerodynamic performance of blade with profile geometry parameterization is proposed by using double arc middle arc, multi-segment arc thickness distribution and ellipse leading edge parameterization method Based on the existing experiments, the aerodynamic analysis under multiple operating conditions is compared and verified. Based on the DOE experimental design, the approximate proxy model is established, and the objective function considering the variable weight coefficient under design and off-design conditions is put forward. The multi-island genetic algorithm is used to optimize the design, and the loss of controllable diffusion airfoil at positive angle of attack is reduced by 4%.
Key words: customized blade profile     blade profile design     multi-angle of attack optimization
0 引　言

1 基于双圆弧的可控扩散叶型设计方法及分析 1.1 叶型设计方法

 图 1 叶型设计方法步骤图 Fig. 1 Blade design method step diagram

1）中弧线造型设计

2）厚度分布曲线造型设计

3）椭圆前缘造型设计

4）厚度分布叠加

1.2 几何模型

 图 2 单通道计算域 Fig. 2 Single channel computing domain
1.3 网格划分

1.4 边界条件和湍流模型

1.5 数值方法验证

 图 3 叶型气动性能曲线对比 Fig. 3 Comparison of airfoil aerodynamic performance

 图 4 不同攻角下表面马赫数分布 Fig. 4 Mach number distribution on the surface under different angles of attack

1.6 叶型设计对比

 图 5 改型设计叶型对比图 Fig. 5 Blade shape modification design comparison

 图 6 改型设计前后总压损失系数对比 Fig. 6 Comparison of Total pressure loss coefficient before and after modification design

 图 7 马赫数云图对比 Fig. 7 Mach number cloud map comparison
2 基于双圆弧的可控扩散叶型设计方法及分析 2.1 设计变量

2.2 多工况目标函数

 ${ f = \frac{{5{i_{ - 6}}{\omega _{ - 6}} + 4{i_{ - 4}}{\omega _{ - 4}} + 2{i_{ - 2}}{\omega _{ - 2}} + 0.5{i_0}{\omega _0} + 2{i_2}{\omega _2} + 3{i_4}{\omega _4} + 4{i_6}{\omega _6}}}{{5{i_{ - 6}} + 4{i_{ - 4}} + 2{i_{ - 2}} + 0.5{i_0} + 2{i_2} + 3{i_4} + 4{i_6}}}}。$

2.3 试验设计及代理模型

 图 8 马赫数拟合曲面 Fig. 8 Mach number fitting surface

 图 9 损失系数拟合曲面 Fig. 9 Loss Coefficient fitting surface
2.4 多工况设计优化流程

 图 10 优化流程图 Fig. 10 Optimization flow chart

1）基于本文提出的叶型的参数化设计方法，建立流场分析的集成迭代，获得对设计参数下的响应；

2）确定设计参数与随机变量，利用DOE获取初始样本点，构建Kriging近似代理模型；

3）利用优化算法进行叶型的多工况点稳健性优化体系。

2.5 优化结果 2.5.1 优化前后叶型对比

 图 11 CDA MAN GHH 1-S1叶型优化前后二维叶型对比 Fig. 11 Comparison two- dimensional blade profile before and after CDA MAN GHH 1-S1 blade profile optimization

2.5.2 优化前后性能对比

2.5.3 优化前后流场对比分析

 图 12 CDA MAN GHH 1-S1叶型优化前后总压损失随攻角变化对比图 Fig. 12 Comparison of total pressure loss with angle of attack before and after CDA MAN GHH 1-S1 blade profile optimization

 图 13 CDA MAN GHH 1-S1叶型优化前后绝对马赫数等值线分布图 Fig. 13 Distribution map of absolute Mach number contours before and after CDA MAN blade profile optimization
3 结　语

1）采用双圆弧中弧线、多段圆弧厚度分布以及椭圆前缘参数化方法，能快速生成CDA叶型，建立数值计算方法并进行试验对比验证。

2）基于叶型参数化设计、DOE试验设计、目标函数和近似代理模型建立压气机叶型多工况稳健性设计优化体系，寻优计算获得在多攻角下的宽工况叶型。为考虑结果状态的影响，将权重系数与设计变量相关联，将多目标优化转化成单目标优化，从而大大减少计算工作量，减少优化耗时，实现多工况点的优化。

3）用所建立的优化设计体系对CDA MAN GHH 1-S1叶型优化设计之后，非设计点总压损失系数与优化前相比有了比较明显的下降并使正攻角损失降低4%，实现压气机叶型在非设计工况下气动性能明显提升的优化设计。