Vibration monitoring is one of the most established and reliable methods of condition monitoring.
Vibration Monitoring & Vibration Analysis
Let’s start the conversation…
In rotating machinery like motors, fans, pumps, and compressors, vibration behaviour often changes when faults develop. These changes can signal potential failure and reduced reliability.
By regularly measuring and analysing vibration, you can detect issues early, allowing for predictive maintenance to address or mitigate faults before they lead to costly breakdowns and proactive maintenance to address design issues causing poor reliability or operational problems.
For more details, see our guide on the benefits of predictive maintenance and how it helps improve reliability while cutting costs.
What is Vibration Analysis?
Vibration analysis is the process of extracting valuable insights from vibration signals to identify latent faults in machinery. It involves two key steps:
- Fault Detection: Has the vibration pattern changed enough to indicate a potential issue?
- Fault Diagnosis: Can we pinpoint the specific fault by, e.g. analysing the vibration frequency in relation to the machine’s components?
Vibration can be measured in various ways, such as displacement, velocity, and acceleration. It can also be expressed in root-mean-squared (RMS) values, peak values, or peak-to-peak values, depending on the fault type and machinery involved. For example:
- RMS vibration velocity (10 to 1000 Hz) is commonly used to detect issues like unbalance, misalignment, and looseness in machines running between 600 to 12,000 rpm.
- Peak acceleration (up to 12 kHz) is more effective for identifying faults in rolling bearings.
- For large machines with journal bearings, we measure shaft motion relative to the bearing housings, typically using peak-to-peak displacement units.
Time Waveform and Spectrum Analysis
In addition to these measurement techniques, we can analyse the raw time signal, known as the time waveform, to identify repetitive impacts within the machinery.
A smooth-running machine should operate without impacts, so any detected impacts can be linked to specific faults based on the machine’s speed and geometry. The impact’s amplitude also helps assess the severity of the problem.
More commonly, the time waveform is broken down into discrete frequencies, resulting in a plot of amplitude versus frequency known as a spectrum. This spectrum analysis is a cornerstone of vibration diagnostics. By correlating vibration frequencies to shaft speeds and component geometries (such as bearings, gears, belts, and impellers), we can accurately diagnose the root cause of the vibration changes.
Advanced Signal Processing Techniques
Beyond basic spectrum analysis, additional signal processing techniques can enhance fault detection and severity assessments. These techniques help improve the accuracy and confidence of diagnosis. They include:
- Signal Demodulation (Envelope Analysis)
- Phase Analysis
- Time Synchronous Averaging
- Bump Testing
- Modal Analysis and more.
These advanced techniques allow us to pull out specific features from vibration signals, providing deeper insights into potential issues and improving our ability to diagnose and address faults early.
For businesses that rely on high-performing machinery, predictive maintenance is a game-changer. Contact Us today to discover how our solutions can help you reduce downtime and extend the life of your equipment.