基于响应面法的柱塞泵迷宫密封优化设计  

Optimization design of labyrinth seal for piston pump based on response surface method

作　　者：杨煜兵 张春光 刘光恒[1] 李上青[1] 张俊峰 安磊 褚帅 YANG Yubing;ZHANG Chunguang;LIU Guangheng;LI Shangqing;ZHANG Junfeng;An Lei;Chu Shuai(Research and Development Department,Shenyang Academy of Instrumentation Science Co.,Ltd.,Shenyang 110043,China)

机构地区：[1]沈阳仪表科学研究院有限公司研究开发部,沈阳110043

出　　处：《流体机械》2021年第6期44-49,93,共7页Fluid Machinery

基　　金：辽宁省自然科学基金项目(2019-MS-309);沈阳市“双百工程”计划项目(1802869)。

摘　　要：为了优化柱塞泵的迷宫密封结构减少泄漏量,选取密封槽角度、高度、宽度等参数为优化变量,以泄漏量最小为优化目标,采用神经网络响应面对迷宫密封结构建立了优化模型,并分析了各设计参数对泄漏量的影响;基于建立的响应面模型,以泄漏量最小为优化目标,应用遗传算法进行了优化计算,优化后的迷宫密封结构的泄漏量比原型降低了97.3%,并对优化后的结构应用计算流体力学(Computational Fluid Dynamics,CFD)进行了验证,结果表明该优化结果比较准确。In order to reduce the leakage by optimizing the labyrinth seal structure of the piston pump,by choosing angle,height,and width,etc.as the design parameter and the minimum leakage as the goal of optimization,a neural network response surface model is established.The effects of various design parameters on the leakage are analyzed;By taking the minimum leakage as the optimization goal,the genetic algorithm is used to perform the optimization calculation based on the established response surface model.The optimization calculation results show that the leakage of the optimized labyrinth seal structure is reduced by 97.3%compared with the original design;The optimized structure is proved accurate by CFD verification.

关 键 词：柱塞泵 迷宫密封 响应面法 遗传算法 CFD 

分 类 号：TH12[机械工程—机械设计及理论]