Shaking, continuous movement about a point, periodic or random oscillations, quivering or trembling motion are terms that can be used to best define the word vibration. Vibration is a mechanical phenomenon and understanding the basics and fundamentals of vibration analysis are very important in forming a solid background to analyse problems on rotating machinery. Vibration analysis is also applied for the purpose of condition based monitoring. Through it, many types of faults can be detected and effectively managed especially early enough in the life of a machine. Vibration analysis helps us to detect failures early hence enabling timely planning and scheduling of repair work or maintenance activities.
Vibration analysis has to do with the vibration spectrum or time waveform and a lot of failure modes can be detected by its implementation. The vibration spectrum provides important frequency information that can pinpoint the faulty component. Vibration analysis is often referred to as predictive maintenance. Bearing defects, imbalance, misalignment, bent shaft, mechanical looseness, gear faults and electrical faults are examples of faults that can be detected through vibration analysis.
Vibration analysis is a broad term. To truly identify a failure mode, there are several considerations that must be addressed. These include but are not limited to: The data type, frequency ranges, resolution alarm types, alarm amplitudes and rate of failure.
Steps to Implementing Vibration Analysis
To implement a vibration analysis program here is a summary of five steps one could follow:
Align the data with the failure modes:
The data types used in vibration analysis are targeted at detecting individual fault areas. The data type selected depends on the specific fault targeted. Understanding the data type is important in selecting the correct tool to apply to the failure mode.
Velocity- will detect imbalance, misalignment, looseness and late stage bearing faults and is focused on general purpose data.
Acceleration-detects gear faults, rotor bars and other higher frequency faults.
Displacement-for low RPM equipment, it measures actual shaft movement with eddy current probes.
Establish alarm criteria for the data:
Only by truly understanding the failure mode can the appropriate alarm be developed. Overall alarms will indicate approaching failures but will not help in detecting specific failure modes. The type and level of alarms are based on specific failure modes.
Comply to and manage your collection frequencies:
The frequency range in which the fault is found in will drive the selection of data types. Failure modes may be found in sub harmonics, low range, mid range or high range. This refers to the frequency of the fault or how often within one revelation does the fault happen. One must comply to the set frequency ranges. Don’t base your program on a standard time-driven collection frequency (e.g. “we collect data every three months”) without consideration of failure modes.
At this point, the failure mode is detected. One must set up data collection points to address identified failure modes.
Provide decision support
When failures are detected, provide the decision support to minimise the impact on your business goals.
figure 1: Steps to Vibration Analysis
Though the job of vibration analysis is usually performed by a certified Vibration analysis it is important that reliability leaders and professionals have a fair knowledge of it. A well run vibration analysis program will surely control the number of asset installations and predict a multitude of failures well in advance.
Reference/Further Reading: www.reliabilityweb.com