ABSTRACT
In the previous chapter, we studied various kinds of transducers, measuring and analysis equipment, and exciters used in the vibration and acoustics field especially related to rotating machinery. These transducers produce very weak signals in the form of charge or voltage. Through amplifiers these signals are augmented and fed to the data acquisition system to digitize these signals into a computer. In this chapter, the main focus is to study methods involved in processing these signals in various required forms in the time domain (mean, variance, probability density function, autocorrelation function, cross-correlations, etc.) and frequency domain (Fourier series, Fourier transform, digital Fourier transform, fast Fourier transform (FFT), and full spectrum). Various graphical forms of displaying these measurements are presented. Error involved in the measurement (mechanical, electrical, or digital) itself will be addressed. These include error due to the aliasing effect, error in digitizing, scalloping losses and leakage error. Tuning of sampling and signal periodicity, and effects due to windowing, will be discussed as ways to improve the data quality. Fourier series, Fourier transform, discrete Fourier transform, FFT, and full spectrum will be covered in detail, as they are very important in rotating machinery analysis. In Chapter 17, various fault conditions and their effects on the vibration signature will be examined.
