基于叶片振动的航空发动机喘振预警及识别方法  

Surge Warning and Identification Method of Aero-engine Based on Blade Vibration

作　　者：王维民[1,2] 刘延振 林昱隆 李天晴 WANG Weimin;LIU Yanzhen;LIN Yulong;LI Tianqing(State Key Laboratory of High-end Compressor and System Technology,Beijing University of Chemical Technology,Beijing 100029;Beijing Key Laboratory of Health Monitoring and Self-Recovery for High-End Mechanical Equipment,Beijing University of Chemical Technology,Beijing 100029)

机构地区：[1]北京化工大学高端压缩机及系统技术全国重点实验室,北京100029 [2]北京化工大学高端机械装备健康监控与自愈化北京市重点实验室,北京100029

出　　处：《机械工程学报》2024年第21期122-131,共10页Journal of Mechanical Engineering

基　　金：国家自然科学基金重点资助项目(92160203)。

摘　　要：喘振是发动机的一种典型故障,危害极大,尤其是加力状态下的喘振,因此及时诊断预警尤为关键。传统喘振预警方法以压力脉动信号为主,传感器安装难度大且识别范围有限。叶尖定时技术在透平机械的状态监测上受关注较广,传统的叶片异步辨识算法对欠采样问题分析不充分,辨识精度受频率混叠影响较大,通过叶片的频幅特征变化,提出了一种喘振预警及叶片频率识别方法。建立了考虑失谐及耦合的叶片集总参数模型,通过全相位和传统傅里叶变换相结合修正了频率差值,经相位遍历得到倍频值,引入倍频准确度并提出频差明显度作为判据,构成修正后的频率识别算法。采用数值仿真模型模拟BTT传感器采集叶片多频共振时的位移欠采样数据,对比两种算法的辨识结果,所提修正算法的差值和频率识别误差最大分别为0.845%和0.053%,远低于传统算法的1.556%和0.097%,验证了所提算法的精确度。在大风扇叶片实验台上进行实验,研究了喘振阶段的叶片频幅特征,根据幅值有效值和报警阈值实现喘振实时预警,使用所提算法辨识叶片的振动频率,实验结果表明,喘振时叶片幅值报警阈值的设置与位移结果吻合较好,异步振动频率均值为1 261.2 Hz,最大偏差为4.8%,该方法能够实现对透平机械实时非接触喘振预警,为透平旋转叶片的异步频率辨识及损伤监测提供了技术支持。Surge is a typical fault of engine,which is extremely harmful,especially under the state of acceleration,so timely diagnosis and early warning is especially critical.The traditional early warning method of surge is based on pressure pulsation signals,which are difficult to install and has a limited identification range of the sensors.The blade tip timing technology is widely concerned in the condition monitoring of turbomachinery.The traditional blade asynchronous identification algorithm does not adequately analyze the under-sampling problem,and the identification accuracy is greatly affected by the frequency aliasing.So through the change of the frequency and amplitude characteristics of the blades,a method is proposed for surge warning and blade frequency identification.The lumped parameter model of blades considering detuning and coupling is established,the frequency difference is corrected by the combination of APFFT and traditional FFT,the natural number are obtained by the apfft and the traversal algorithms.The accuracy of natural number is introduced and the obviousness of frequency difference is proposed as a criterion to obtain an accurate asynchronous recognition algorithm.A numerical simulation model is used to simulate the displacement under-sampling data of the blade collected by the BTT sensor during multi-frequency resonance,and comparing the identification results of the two algorithms,the maximum error of the proposed modified algorithm is 0.845%and 0.053%for the difference and frequency identification,respectively,which is much lower than that of the traditional algorithm,which is 1.556%and 0.097%,and the accuracy of the proposed algorithm is verified.Experiments are carried out on the large fan blade test bench,the frequency amplitude characteristics of blade in the stage of surge are investigated,real-time warning of surge is realized according to the amplitude effective value and alarm threshold,and the proposed algorithm is used to identify the blade asynchronous vibration frequency.The exp

关 键 词：叶尖定时 喘振预警 异步振动 频率辨识 

分 类 号：TK14[动力工程及工程热物理—热能工程] V211[航空宇航科学与技术—航空宇航推进理论与工程]