Vibration Monitoring Guide
Vibration monitoring detects developing mechanical problems before they cause failure. Learn what to measure, how to interpret results, and which faults each frequency reveals.
Why Monitor Vibration?
Every rotating machine vibrates. The vibration pattern (amplitude, frequency, direction) contains information about the machine's mechanical condition. By monitoring vibration, you can detect bearing wear, imbalance, misalignment, looseness, gear damage, and many other problems months before they cause failure.
Vibration monitoring is the most powerful predictive maintenance technique for rotating equipment. It allows you to plan maintenance during scheduled downtime rather than reacting to unexpected failures.
What to Measure
- Overall vibration level: The total vibration amplitude across all frequencies. Simple to measure. Good for trending. Rising overall level indicates a developing problem but doesn't identify the cause.
- Frequency spectrum (FFT): Breaks the vibration down into its component frequencies. Each mechanical fault produces characteristic frequencies. This identifies the specific problem.
- Phase: The timing relationship between vibration at different points. Used to distinguish between imbalance (in-phase) and misalignment (often 180 degrees out of phase).
- Time waveform: The raw vibration signal over time. Useful for diagnosing impacts (bearing defects, gear tooth damage) and low-frequency phenomena.
Common Fault Frequencies
| Fault | Characteristic Frequency | Notes |
|---|---|---|
| Imbalance | 1x running speed | Most common fault. Vibration at shaft rotation frequency |
| Misalignment | 1x and 2x running speed | Angular or parallel misalignment between shafts |
| Looseness | 1x, 2x, 3x+ running speed | Many harmonics. Mechanical looseness in mounting or bearings |
| Bearing defect | High frequency (BPFO, BPFI, BSF, FTF) | Specific frequencies calculated from bearing geometry |
| Gear damage | Gear mesh frequency and sidebands | Number of teeth x shaft speed |
| Belt defect | 1x and 2x belt frequency | Belt rotation frequency |
| Electrical | 2x line frequency (100 Hz at 50 Hz) | Motor electromagnetic faults: eccentric rotor, broken bars |
| Oil whirl | 0.43 to 0.48x running speed | Journal bearing instability. Sub-synchronous |
The bearing defect frequencies
Rolling element bearings have four characteristic frequencies: BPFO (Ball Pass Frequency Outer race), BPFI (Ball Pass Frequency Inner race), BSF (Ball Spin Frequency), and FTF (Fundamental Train Frequency). These are calculated from the bearing geometry (ball diameter, pitch diameter, number of balls, contact angle). The bearing manufacturer provides these frequencies or they can be calculated. Vibration at these frequencies indicates bearing damage.
Monitoring Strategy
- Install vibration sensors (accelerometers) on bearing housings. For most machines, one radial and one axial measurement per bearing.
- Take regular measurements. Monthly for critical machines, quarterly for important machines. PlantLogica's continuous monitoring system can measure automatically.
- Trend the overall vibration level. Rising trends indicate developing problems.
- When the overall level exceeds the alarm threshold, take a frequency spectrum (FFT) to identify the specific fault.
- Compare the spectrum to the machine's characteristic frequencies (running speed, bearing defect frequencies, gear mesh frequencies) to identify the fault.
- Monitor the trend of the specific fault frequency. Rising amplitude means the fault is worsening. Schedule repair before failure.
ISO 10816 vibration limits
ISO 10816 provides vibration severity limits for different machine types and sizes. These are guidelines, not absolute limits. The most useful approach is to trend your machine's vibration against its own baseline. A machine that has always vibrated at 2 mm/s and is now at 4 mm/s is more concerning than a machine that has always been at 4 mm/s. Know your machine's normal vibration level.
Tip: For the most critical rotating equipment, install continuous vibration monitoring connected to PlantLogica. The system learns each machine's baseline, detects trends automatically, and alerts you when vibration deviates from normal. This catches problems weeks or months before failure, allowing planned maintenance during scheduled downtime.