Overview

A total analysis of high-speed rotating equipment requires a complex blend of performance and vibration data. The trend toward total analysis is growing with the problems of an energy shortage and the need for maximum plant utilization. Performance analysis is essential in the efficient utilization of turbo-machinery and, when coupled with vibration analysis, is an unbeatable tool as a total diagnostic system.

The real-time analyzer plays a very important role in presenting vibratory data in a manner that can lend itself to a trending data system. This important role of the spectrum analyzer will be explored in detail in this chapter. Also, the role of the spectrum analyzer will increase with a better understanding of statistical techniques in vibration analysis.

Basically, spectrum analysis transforms a displacement/time chart into an amplitude/frequency chart known as a spectrum. This analysis consists of decomposing a time-varying signal into its component pure tones. Pure tones are sinusoidal wave forms of constant frequency and amplitude. This decomposition is done digitally upon a signal by a minicomputer using the Fourier transformation or by filtering the signal.

Signals generated by high-speed machinery are very complex in nature and are generated by several forces with a net effect that masks the pure tones. The random portion of the signal, which is blended with the pure tones, is called noise. The ratio of the total amplitude (area under spectrum) to that of the noise is called the signal-to-noise (S/N) ratio. Sometimes this ratio is expressed in decibels, or db,...