节点文献
声发射技术在旋转机械碰摩故障诊断中的应用
Application of Acoustic Emission Technique in Diagnostics of Rubbing Fault
【作者】 印欣运;
【导师】 何永勇;
【作者基本信息】 清华大学 , 机械工程, 2005, 硕士
【摘要】 动静件间发生碰摩是旋转机械中常见典型故障,其对旋转机械的安全稳定运行有着极大的危害性。因此,对碰摩故障进行及时有效的识别和诊断对于保证旋转机械安全运行有着重大意义。实践证明,在旋转机械碰摩过程中总是会产生丰富的声发射信号,因此,声发射技术为碰摩故障的识别和诊断提供了又一途径。与基于振动信号诊断方法相比,基于声发射的诊断方法对碰摩故障的诊断更具有优越性。本文以声发射技术为手段,对基于声发射技术的碰摩故障识别和诊断方法进行理论和实验研究。主要工作包括以下三个方面:1. 碰摩声发射的参数分析。采用了均方根、信号幅度峰值、上升时间和振铃计数等典型声发射参数对不同工况下碰摩声发射的特征及传播特性进行了分析,提出了鉴别不同种类声发射信号的方法;2. 碰摩声发射的波形分析。主要采用小波分析方法对碰摩声发射的时频特性及传播特性进行分析,着重研究了小波尺度谱和再分配尺度谱在碰摩声发射时频特性分析中应用。其中,讨论了小波基函数的选择问题和小波消噪方法,提出了基于信号熵值的小波基选取原则。另外,对多尺度下碰摩声发射的能量分配特性进行了研究,提出了相应的特征参量;3. 讨论了基于互相关分析技术的声发射源定位问题。将互相关分析技术分别与小波变换、小波模极大值以及带通滤波器结合,并针对碰摩声发射源的定位问题,对相应的三种定位方法进行了详细的对比研究,给出了三种声发射源定位方法的适应条件。
【Abstract】 Rubbing between the rotor and the stator is the frequent fault in the rotatingmachine, which has great harmfulness to the safe and stable running of the rotatingmachine. Thus, it is very significant to study the diagnosis method of the rubbingfault of the rotating machine. It has been proved that rubbing between the rotor andthe stator can produce abundant acoustic emission (AE) signal. Therefore, AEtechnique can be used to study the diagnosis method of the rubbing fault. And AEbased diagnosis method of the rubbing has obvious superiority compared with thevibration based one.This dissertation uses AE technique to perform theoretical and experimentalstudy of AE based diagnosis method of the rubbing fault. The main contents includethe following three aspects: 1. Parameter analysis of the rubbing AE. Classical AEparameters, such as rms, peak amplitude, rising time and AE counts, i.e., are used toanalyze the feature and the propagation characteristic of the rubbing AE. And thecorresponding identification methods for different kinds of AE are proposed. 2.Waveform analysis of the rubbing AE. The wavelet analysis, especially waveletscalogram and reassignment scalogram, is utilized to analyze the time-frequencycharacteristic and propagation characteristic of the rubbing AE. In this study, otherproblems, such as the choice of wavelet family and wavelet transform basedde-noising method, are discussed. In addition, energy percentage of the multi-scalesis calculated to obtain corresponding features of the different AE signals. 3.Cross-correlation based AE source locating method is discussed. Combingcross-correlation method with wavelet transform, wavelet modulus maxima andband-pass filter respectively to perform locating the rubbing AE source throughexperiment. The results are analyzed and compared. And corresponding applicationcondition of these three methods are concluded.
- 【网络出版投稿人】 清华大学 【网络出版年期】2006年 08期
- 【分类号】TH17
- 【被引频次】31
- 【下载频次】1212