Vibration Analysis in Predictive Maintenance
What is Vibration Analysis?
Predictive maintenance platforms utilize condition monitoring setup to collect and analyze the data of rotating machines in order to detect faults and failures in them. The most important parameter is the vibration occurring in these machines which are observed and then utilized to classify the early defects arising in them.
This methodology of detecting faults and anomalies in the rotating or reciprocating machines by continuously monitoring its vibration levels is more called Vibration Analysis.
When machinery occurs some faults….
Analysis & Solving the problem…
What are the faults that can be detected with Vibration Monitoring?
All types of faults that occur in rotating or reciprocating machines mentioned below can be easily determined by the Vibration Analysis methodology.
- Bearing Faults
- Gear Defects
- Belt wear and tear
- Pump Cavitation
How Vibration Analysis is done?
The sensors installed, which are mostly accelerometers are used to gather data of various vibration levels from the rotating machines.
An accelerometer sensor functions by detecting the mechanical motion of an object/surface and converting the same into electrical signals corresponding to the magnitude of these vibrations. These vibrations data gathered are then viewed on a software platform and are analyzed by a user to predict faults in the machines.
The data collected is either viewed in the form of Amplitude VS Time or Amplitude VS Frequency.
Amplitude VS Time Graph
Amplitude VS Time graph is read with the help of an ISO Standard chart called ISO-10816 Vibration Severity Chart. The chart describes various satisfactory and unsatisfactory values of vibrations corresponding to different types of rotating assets. The rotating assets are categorized into types such as CLASS 1, CLASS 2, CLASS 3, and CLASS 4 and each type has its definition based upon the power and size of the machine.
Faults that are considered to produce low-frequency unusual vibrations such as - Unbalance, Misalignment and Mechanical Looseness are detected by using the ISO 10816 Standard Chart. The frequency range that is applied to this chart is (10-1000) Hz.
This type of measurement provides a user with information about the overall health condition of the machine.
Amplitude VS Frequency Chart
The composite vibrations emanating from a point of a machine can be dispersed into a spectrum of vibration frequencies to observe each and every frequency separately. This can be done with the help of a mathematical model called FFT or Fast Fourier Transform. The graph generated is also called SPECTRUM GRAPH. To identify defects from this graph, the notion that each type of fault displays its own signature or characteristic is taken into consideration.