Rotating/reciprocating machinery
produces vibration signatures depending upon the mechanism involved.
Faults may occur at motor, rolling element bearings, gearboxes,
belts, fans and other electrical/mechanical components. It is
strongly necessary to detect these problems at an early stage and to
avoid serious damage and catastrophic failure. The purpose of
analysis is to identify the fault frequencies so that root cause can
be addressed and corrective action can be taken.
Rotating machinery faults usually
associate with strong harmonics and sidebands. Therefore, the fault
frequencies can be distinguished from the other frequency contents
by identifying the harmonics or sideband components. Envelope
analysis is a useful tool to extract the sidebands caused by
amplitude modulation, while cepstrum analysis is to separate
harmonic families.
In this tech note, the
fundamentals of envelope and cepstrum analyses are briefly
introduced with examples of rolling element, broken rotor bar, and
gearbox faults. The physics associated with the faults are also
discussed. The readers do not need to worry about the theories
behind the algorithms.
For rolling element bearings, when
the rolling elements strike a local fault on the inner or outer
race, or a fault on a rolling element strikes the inner or outer
race, an impact is produced. The bearing frequencies can be
categorized as BPFO (ball passing frequency outer race), BPFI (ball
passing frequency inner race), BFF (ball fault frequency), and FTF
(fundamental train frequency). Please see the last page of this tech
note for the calculation of these fault frequencies. Like rolling
element bearings, a faulted gear tooth also generates impact once
per revolution when meshing with the other gear.
The impacts generated by gearbox
and rolling element faults superimpose upon the vibration signal,
resulting in amplitude modulation as shown in Fig. 1, and thus cause
sidebands in the spectrum around the frequency bins associated with
the vibration signal. The sidebands mingle with the frequency
components of the vibration signal so that it is hard to distinguish
them in the spectrum. Impacts in time domain generate many harmonics
extending to very high frequency in frequency domain. Often some of
these harmonics excite resonance in structure, bearings or sensors.
Exact location of the resonance is usually not known and cannot be
determine easily. However, the resonance amplifies the modulating
and carrier signals. Envelope analysis when applied in this region
is a useful tool for amplitude demodulation. It should be noted that
it is not easy to relate the amplitude of the signal to the fault
severity. Envelope analysis is a useful tool for amplitude
demodulation. The envelope analysis function in VibraQuest is based
on an improved Hilbert transform method. This improved demodulation
method attenuates the influences from high frequency contents and
makes the envelope frequency easier to identify.
Figure 1 illustrates a simulated amplitude
modulated sinusoidal signal. The vibration signal is called the
carrier. The red curve indicates the envelope which is directly
caused by the impacts mentioned above. The envelope analysis is to
extract the frequency of the envelope so that the faults caused by
rolling element bearing or gearbox can be identified.
RESOURCES 
