摘要
目前,航空发动机管路系统的卡箍布局优化设计大多数为确定性优化,没有考虑卡箍刚度和人工装配误差等不确定因素,这些不确定因素的波动性可能会导致管路系统的输出响应与预计值有较大偏差,从而引起系统故障。为了在改善航空发动机管路系统振动性能的同时保证管路系统具有较强的可靠性与稳健性,以Z型管路为研究对象,提出一种以避开激振频率和降低振动频率分散度为目标的卡箍布局稳健优化设计方法;该方法首先采用壳单元和弹簧单元分别对管体和卡箍进行有限元建模;然后对设计参数进行拉丁超立方抽样,通过有限元模型计算样本的固有频率,并构造频率响应的高精度神经网络代理模型;最后采用蒙特卡罗模拟和改进型非支配遗传算法,展开了卡箍支撑位置的稳健优化设计。结果表明,该方法能在保证管路系统避开激振频率的设计要求基础上,有效地提高其动力学特性的稳健性。
At present,the clamp layout optimization design of an aero-engine piping system is mostly deterministic optimization,without considering the uncertainties such as clamp stiffness and manual assembly error.The fluctuation of these uncertainties may cause the output response of the piping system to have a large deviation from the expected value,thus causing system failure.In order to improve the vibration performance of an aero-engine piping system and ensure the piping system has strong reliability and robustness,a robust optimization design method for clamp layout was proposed in this paper,taking Z-type piping as the research object,which aims to avoid the excitation frequency and reduce the dispersion of vibration frequency.Firstly,shell element and spring element are used to model the tube body and clamp,respectively.Then,the Latin hypercube sampling of the design parameters is carried out,the natural frequency of the samples is calculated by finite element model,and the high precision neural network proxy model of frequency response is constructed.Finally,Monte Carlo simulation and an improved non-dominated genetic algorithm were used to carry out robust optimization design of clamp support position.Results show that the proposed method can effectively improve the robustness of dynamic characteristics on the basis of ensuring that the piping system avoids the design requirement of excitation frequency.
作者
孙一冰
王晓伟
SUN Yibing;WANG Xiaowei(College of Energy and Power Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2023年第24期194-203,共10页
Journal of Vibration and Shock
基金
国家科技重大专项经费(J2019-I-0008-0008)。
关键词
Z型管路
有限元建模
神经网络
遗传算法
稳健设计
Z-type pipeline
finite element modeling
neural network
genetic algorithm
robust design
