Vibration-based Condition Moni

ISBN-13: 9781119477556 / Angielski / Twarda / 2021 / 448 str.

Robert Bond Randall

Spis treści

Chapter 1 Introduction and Background1.1 Introduction1.2 Maintenance strategies1.3 Condition monitoring methods1.3.1 Vibration analysis1.3.2 Oil analysis1.3.3 Performance analysis1.3.4 Thermography1.4 Types and benefits of vibration analysis1.4.1 Benefits compared with other methods1.4.2 Permanent vs intermittent monitoring1.5 Vibration transducers1.5.1 Absolute vs relative vibration measurement1.5.2 Proximity probes1.5.3 Velocity transducers1.5.4 Accelerometers1.5.5 Dual vibration probes1.5.6 Laser vibrometers1.6 Torsional vibration transducers1.6.1 Shaft encoders1.6.2 Torsional laser vibrometers1.7 Condition monitoring - the basic problemReferencesChapter 2 Vibration Signals from Rotating and Reciprocating Machines2.1 Signal classification2.1.1 Stationary deterministic signals2.1.2 Stationary random signals2.1.3 Cyclostationary signals2.1.4 Cyclo-non-stationary signals2.2 Signals generated by rotating machines2.2.1 Low shaft orders and subharmonics2.2.2 Vibrations from gears2.2.3 Rolling element bearings2.2.4 Bladed machines2.2.5 Electrical machines2.3 Signals generated by reciprocating machines2.3.1 Time-frequency diagrams2.3.2 Torsional vibrationsReferencesChapter 3 Basic signal processing techniques3.1 Statistical measures3.1.1 Probability and probability density3.1.2 Moments and cumulants3.2 Fourier analysis3.2.1 Fourier series3.2.2 Fourier integral transform3.2.3 Sampled time signals3.2.4 The discrete Fourier transform (DFT)3.2.5 The fast Fourier transform (FFT)3.2.6 Convolution and the convolution theorem3.2.7 Zoom FFT3.2.8 Practical FFT analysis and scaling3.3 Hilbert transform and demodulation3.3.1 Hilbert transform3.3.2 Demodulation3.4 Digital filtering3.4.1 Realisation of digital filters3.4.2 Comparison of digital filtering with FFT processing3.5 Time/frequency analysis3.5.1 The short time Fourier transform (STFT)3.5.2 The Wigner-Ville distribution3.5.3 Wavelet analysis3.5.4 Empirical mode decomposition3.6 Cyclostationary analysis and spectral correlation3.6.1 Spectral correlation3.6.2 Spectral correlation and envelope spectrum3.6.3 Wigner-Ville spectrum3.6.4 Cyclo-non-stationary analysisReferencesChapter 4 Fault Detection4.1 Introduction4.2 Rotating machines4.2.1 Vibration criteria4.2.2 Use of frequency spectra4.2.3 CPB spectrum comparison4.3 Reciprocating machines4.3.1 Vibration criteria for reciprocating machines4.3.2 Time/frequency diagrams4.3.3 Torsional vibrationReferencesChapter 5 Some special signal processing techniques5.1 Order tracking5.1.1 Comparison of methods5.1.2 Computed order tracking(COT)5.1.3 Phase demodulation based COT5.1.4 COT over a wide speed range5.2 Determination of instantaneous machine speed5.2.1 Derivative of instantaneous phase5.2.2 Teager Kaiser and other energy operators5.2.3 Comparison of time and frequency domain approaches5.2.4 Other methods5.3 Deterministic/random signal separation5.3.1 Time synchronous averaging5.3.2 Linear prediction5.3.3 Adaptive noise cancellation5.3.4 Self adaptive noise cancellation5.3.5 Discrete/random separation (DRS)5.4 Minimum entropy deconvolution5.5 Spectral kurtosis and the kurtogram5.5.1 Spectral kurtosis - definition and calculation5.5.2 Use of SK as a filter5.5.3 The kurtogramReferencesChapter 6 Cepstrum analysis applied to machine diagnostics6.1 Cepstrum terminology and definitions6.1.1 Brief history of the cepstrum and terminology6.1.2 Cepstrum types and definitions6.2 Applications of the real cepstrum6.2.1 Practical considerations with the cepstrum6.2.2 Detecting and quantifying harmonic/sideband families6.2.3 Separation of forcing and transfer functions6.3 Modifying time signals using the real cepstrum6.3.1 Removing harmonic/sideband families6.3.2 Enhancing/removing modal properties6.3.3 Cepstrum pre-whiteningReferencesChapter 7 Diagnostic Techniques for particular applications7.1 Harmonic and sideband cursors7.1.1 Basic principles7.1.2 Examples of cursor application7.1.3 Combination with order tracking7.2 Gear diagnostics7.2.1 Techniques based on the TSA7.2.2 Transmission error as a diagnostic tool7.2.3 Cepstrum analysis for gear diagnostics7.2.4 Separation of spalls and cracks7.2.5 Diagnostics of gears with varying speed and load7.3 Rolling element bearing diagnostics7.3.1 Signal models for bearing faults7.3.2 A semi-automated bearing diagnostic procedure7.3.3 Alternative diagnostic methods for special conditions7.3.4 Diagnostics of bearings with varying speed and load7.4 Reciprocating machine and IC engine diagnostics7.4.1 Time/frequency methods7.4.2 Cylinder pressure identification7.4.3 Mechanical fault identificationReferencesChapter 8 Fault simulation8.1 Background and justification8.2 Simulation of faults in gears8.2.1 Lumped parameter models of parallel gears8.2.2 Separation of spalls and cracks8.2.3 Lumped parameter models of planetary gears8.2.4 Interaction of faults with ring and sun gears8.3 Simulation of faults in bearings8.3.1 Local faults in LPM gearbox model8.3.2 Extended faults in LPM gearbox model8.3.3 Reduced FE casing model combined with LPM gear model8.4 Simulation of faults in engines8.4.1 Misfire8.4.2 Piston slap8.4.3 Bearing knockReferencesChapter 9 Fault trending and prognostics9.1 Introduction9.2 Trend analysis9.2.1 Trending of simple parameters9.2.2 Trending of "impulsiveness"9.2.3 Trending of spall size in bearings9.3 Advanced prognostics9.3.1 Physics-based models9.3.2 Data-driven models9.3.3 Hybrid models9.3.4 Simulation-based prognostics9.4 Future developments9.4.1 Advanced modelling9.4.2 Advances in data analyticsReferences

O autorze

Robert Bond Randall, is Emeritus Professor in the Mechanical and Manufacturing Engineering Department at the University of New South Wales in Australia. His research focus is on vibration analysis and signal processing applied to machine condition monitoring. He is the Chief Investigator for three Australian Research Council research grants since 2016 alone.