Bearing Defect Frequencies
Each bearing defect produces impacts at a characteristic frequency that can be calculated from the bearing geometry and shaft speed. The four fundamental defect frequencies:
BPFI (Ball Pass Frequency Inner race) = (Nb/2) × RPM/60 × (1 + (Bd/Pd) × cos α)
BSF (Ball Spin Frequency) = (Pd/2Bd) × RPM/60 × (1 − (Bd/Pd)² × cos²α)
FTF (Fundamental Train Frequency) = 0.5 × RPM/60 × (1 − (Bd/Pd) × cos α)
Nb = number of balls; Bd = ball diameter; Pd = pitch circle diameter; α = contact angle
These frequencies are available from bearing manufacturers or can be obtained from software databases. Most condition monitoring software calculates them automatically from the bearing model number.
Four Stages of Bearing Degradation
| Stage | Vibration signature | Remaining life | Action |
|---|---|---|---|
| 1 — Subsurface fatigue | Elevated gSE/HFD in 5–60 kHz range; no visible spectrum peaks | Months (weeks to months) | Increase monitoring frequency; plan replacement |
| 2 — Surface defect | BPFO or BPFI harmonics visible in envelope spectrum; possibly with ±1× RPM sidebands | Weeks to months | Schedule replacement at next planned shutdown |
| 3 — Defect growth | Multiple harmonics in envelope; defect frequency peaks in conventional velocity spectrum; increased overall vibration | Days to weeks | Expedite replacement; increase inspection frequency; consider run-to-planned-window |
| 4 — Advanced damage | High broadband noise floor; loss of distinct peaks (defect spreads across many frequencies); audible noise; elevated temperature | Hours to days | Emergency replacement; risk of catastrophic failure if not replaced |
High-Frequency Detection (HFD) and gSE
Conventional vibration analysis (10–1,000 Hz) misses early-stage bearing defects. The key is to analyse the high-frequency range (2–50 kHz) where bearing impacts excite the structural resonance of the bearing housing and sensor mounting.
The most common metrics:
- gSE (g Spike Energy): SKF proprietary metric measuring high-frequency energy — simple single-number trended against alert limits
- HFD (High-Frequency Detection): Similar broadband high-frequency energy metric
- Kurtosis: Statistical measure of impulsiveness in the time waveform — high kurtosis (>6) indicates impulsive, fault-related signals
- Envelope spectrum: Demodulate the high-frequency band around a resonance frequency; the modulating low-frequency content reveals the defect frequency
Ultrasound for Bearing Monitoring
An ultrasound detector (40 kHz) provides the earliest indication of bearing degradation — detecting microscopic impacts before they grow to a size detectable by vibration analysis. Method: route each bearing with an ultrasound detector as part of a weekly or biweekly rounds. Log the dBµV reading; trend against baseline. A rise of 8–10 dB from baseline is an early-warning trigger. A rise of 20+ dB indicates progressed defect — escalate to full spectrum analysis.
Bearing Regreasing During Monitoring
Ultrasound is particularly useful for optimising regreasing intervals. Regrease while monitoring with an ultrasound instrument — when greasing, dBµV drops immediately as the lubricant film is restored. Continue adding grease until the level stops dropping and begins to rise (indicating excess grease — stop immediately to avoid seal damage).
This technique ensures correct grease quantity — neither under-lubrication (which causes premature bearing failure) nor over-lubrication (which causes overheating and seal damage). See our motor bearing greasing guide for quantity calculations.