If you’ve spent time working with vibration spectra, you’ve probably come across this familiar scenario: a peak shows up close to a known bearing fault frequency but not quite on it. It’s tempting to write it off or over-interpret it, but the truth is, this kind of near-match is both common and worth digging into. Several real world factors can cause these slight deviations. For instance, when the load on a bearing change, it can shift the contact angles inside, which in turn slightly alters the frequencies at which faults manifest. Slippage between rolling elements and raceways especially under heavy loads or inconsistent lubrication can also lead to a mismatch between the theoretical fault frequency and what we actually see in the data.
In the example below and regardless of the low amplitudes: is 7425CPM related to the BPFO :7342CPM?
These effects aren’t just noise; they’re part of the diagnostic puzzle. In the sections that follow, we’ll break down why these differences happen, what they mean, and how to interpret them with more confidence when analyzing spectra.
1. Bearing Fault Frequencies Overview
Ball bearings produce predictable vibration frequencies when defects occur on components such as the inner race, outer race, rolling elements, or cage. These are commonly defined as:
These frequencies are derived from the bearing geometry and the rotational speed of the shaft. However, real world deviations from ideal assumptions such as slippage or axial load can cause these frequencies to shift.
2. Rolling Element Slippage: A Hidden Variable
What is Slippage?
Slippage occurs when rolling elements do not maintain pure rolling motion but slide slightly at the contact points. This is caused by insufficient traction, uneven loading, poor lubrication, or transient motion (e.g., during startup).
How Slippage Affects Fault Frequencies
- BSF is the most sensitive to slippage, decreasing by 2–5% or more in moderate to heavy slip conditions.
- BPFI and BPFO also decrease slightly, typically by 0.5–2%.
- FTF may remain stable or increase marginally.
Condition Estimated Slippage (%)
- Light radial load 0.5 – 1%
- Moderate load 1 – 3%
- High axial/misaligned load 3 – 5%
- Poor lubrication 5 – 10%
These changes, although seemingly small, are enough to cause diagnostic error if theoretical values are used without correction.
3. Axial Load and Contact Angle
Axial Load Increases Contact Angle
When a ball bearing, particularly a deep groove type, is loaded axially, the contact angle (β) increases. This geometric shift alters the path of rolling elements and thus affects fault frequencies.
Bearing Type Typical Contact Angle Range
- Deep Groove Ball 0° to 8° (max ~10° under load)
- Angular Contact 15° to 40° (by design)
Fault Frequency Drift Due to Contact Angle
At a contact angle increase from 0° to 10°: – BPFI increases by ~0.3% – BPFO decreases by ~0.5% – BSF increases by ~0.2% – FTF may slightly decrease or stay stable. This occurs because the path length and angular relationship between the rolling elements and raceways change, especially for BPFI and BSF.
4. Combined Effects: Slippage vs. Axial Load
These two effects can oppose or compound each other:
Frequency Slippage Effect Axial Load Effect Net Result
- BPFI ↓ Decreases ↑ Increases May offset each other
- BPFO ↓ Decreases ↓ Decreases Additive decrease
- BSF ↓ Strong decrease ↑ Slight increase Likely net decrease
- FTF → Stable / ↑ → Stable / ↓ Minimal change
For example, an inner race defect might show a BPFI slightly below theoretical, not because the geometry is wrong, but because slippage is countering the contact angle increase.
5. Practical Implications for Diagnostics
- Always allow for 2–3% tolerance when matching spectral peaks to calculated fault frequencies.
- Consider operational context: high axial load may suggest increased BPFI, but slippage could mask this.
- Use time waveform and phase data alongside spectra to identify ball-pass events more reliably.
- Employ a slippage-corrected calculator when working with low-speed or poorly lubricated bearings.
6. Effect of Rotational Speed on Diagnostic Clarity
Rotational speed amplifies the consequences of slippage and contact angle variation. While the percentage drift in fault frequencies due to these factors remains relatively constant, the absolute drift in Hertz (Hz) or CPM becomes more significant at higher speeds.
For example: – At 1800 CPM (30 Hz), a 2% shift in BPFI corresponds to roughly 3 Hz or 180 CPM. – At 3600 CPM (60 Hz), that same 2% shift becomes 6 Hz or 360 CPM. – At 6000+ CPM, it could exceed 10 Hz, making it increasingly difficult to confidently match spectral peaks.
This drift becomes especially problematic when fault frequencies are close to integer multiples of the shaft speed (1×, 2×, etc.). Even a small shift can make a true bearing fault peak appear blended with harmonics or misidentified as gear mesh, belt pass, or structural resonance.
Diagnostic Risk:
- A slipped BSF may align too closely with a harmonic and be missed entirely.
- A drifted BPFI may appear between 4× and 5× shaft speed and be overlooked if an analyst expects perfect alignment.
Best Practices:
- Always compare both frequency and phase information.
- Use sideband patterns and time waveform analysis to distinguish overlapping sources.
- Maintain a tolerance band of ±2–3% (or more at higher speeds) when matching peaks to theoretical fault frequencies.
Conclusion:
In bearing fault diagnostics, precision matters but so does realism. Rolling element slippage and axial loading are normal phenomena in rotating machinery. By understanding how they influence fault frequencies, analysts can improve diagnostic accuracy, reduce false positives, and better predict failure severity. Incorporating both factors into spectral interpretation leads to smarter, more reliable condition monitoring.
Thank you Betavib for sharing this informative article with us!
Filed under:
Condition Monitoring by Diana Pereda