ABSTRACT
Vibration analysis is an effective tool to identify and predict failure using Signal
Processing based on condition monitoring. Accordingly, machine faults can be
detected and diagnosed using different post processing of vibration signature. The
present study is to prove that Machine Health and availability achievement becoming
easier by using different vibration analysis techniques. Three different types of
nonstationary signals are addressed in this paper to evaluate the technical conditions
of Paper Machine Forming Roll. Forming Roll consists of frequency drive motor,
gearbox and a symmetric rotor supported on two similar bearings. The present study
is focusing on Forming Roll Bearings Condition Monitoring. A cost wise comparison
of maintenance strategies is made to confirm that Condition based Maintenance is
the most cost-effective strategy. In the past portable data-collectors were used to
measure vibration levels, but Condition based Monitoring had taken place to avoid
any sudden fault developing before next time schedule. Applying different vibration
analysis techniques such as Spectrum, Waveform and Impact Demodulation will
support fault detection, diagnosis and decision-making. The present study proved
that Impact demodulation is recommended to specify faulty bearing source location
and severity, while time waveform or spectrum analysis techniques can detect fault
progress but couldn't distinguish severity due to bearings similarity. Based on Impact
Demodulation technique decision is made to exchange only the higher severity faulty
bearing and postponed the other one for next planned shutdown. Consequently, the
machine availability and health can be achieved by reducing vibration impact around
100 % through continuous condition monitoring.