Table of Contents
Advertisement
11
Fast Fourier Transform (FFT)
11.1 Background
In simple terms, FFT is a mathematical algorithm that converts a data presen-
tation (graph, for example), shown as a function of time, into a data presenta-
tion shown as a function of frequency and vice versa.
This is useful in the representation of vibration measurements. For example,
Velocity graphs, in the FLIR mobile and Windows apps, are derived from FFT
analysis of Acceleration measurements.
11.2 Time-Domain and Frequency-Domain Methods
Time and frequency domain vibration measurement methods can both be em-
ployed to study vibration measurements.
The time domain method provides a glimpse into the source of vibration but
not ideal for analysing multiple frequency vibration signals.
The frequency domain method is more effective; especially when evaluating
amplitude and phase characteristics of vibration signals. The frequency do-
main method is effective finding bearing defects and identifying shock pulses
and friction activity.
11.3 Determining Machine Defects by Vibration Characteristics
Each defect causes vibrations having unique characteristics. You can deter-
mine, more easily, the root cause of a problem if you understand these char-
acteristics, and that's where FFT comes in handy.
You can determine a lot about an asset, even when it is working normally, by
looking at FFT data graphs. For example, in normal operation, the shaft rota-
tion frequency (fundamental) will be represented on the graph on the left fol-
lowed by a series of harmonics whose amplitude will be approximately a third
the amplitude of the fundamental.
Additional signals are present in a normal system representing the vibrations
of the structure that the machine is mounted on.
For faulty systems, the following changes in FFT graphs may appear.
• Shaft balance issues cause large radial variations and marked increase in
amplitude of the fundamental frequency.
• Misalignments are shown as an increase in frequency amplitude at the 2nd
harmonic.
#NAS100051; r. AC/76460/77189; en-GB
48
Hide quick links:
Advertisement
Table of Contents
