Every bearing in a manufacturer’s catalog has two numbers printed in the same row as its dimensions: C and C0. Most maintenance teams glance past them. Some reliability engineers use them to size a replacement. Very few organizations use them as they were designed — as a real engineering tool for predicting and extending bearing life.
Bearings that fail before their calculated life almost always failed for one of three reasons: they were undersized for the actual load, the load case didn’t reflect real operating conditions, or the static load capacity was never checked for the shock and startup conditions the bearing actually sees.
The Two Numbers That Define Bearing Capacity
| C — Dynamic Load Rating | C0 — Static Load Rating |
|---|---|
| Standard: ISO 281 | Standard: ISO 76 |
| Definition: Load for 1 million revolutions at 90% survival probability (L10 baseline) | Definition: Load causing 0.0001 mm permanent raceway deformation at the most heavily loaded contact |
| L10 = (C / P)p | S0 = C0 / P0 |
| Use for: Normal rotating service — motors, pumps, fans, gearboxes running at speed | Use for: Stationary or slow (<10 RPM), shock loads, startup transients, standstill vibration |
C: Three Things That Matter in the Definition
- Constant load: C assumes steady, non-varying load. Real equipment rarely applies constant load — belt drives pulse, gearboxes generate dynamic forces, process conditions vary. These variations require load adjustments before using C in a life calculation.
- One million revolutions: This is not the service life — it’s the reference point for the formula. At one million revolutions with load equal to C applied, L10 life equals exactly 1 million revolutions. Everything else is a calculation from that baseline.
- 90% survival: L10 life is a statistical prediction. 10% of the bearing population is expected to fail before reaching that life. If you need better reliability, the calculation includes reliability factors that adjust the prediction.
C is not a maximum operating load. Bearings can and do operate above their dynamic load rating at reduced calculated life. It’s not a safety limit — it’s a reference point for a life prediction formula. Running at C means running at the load where L10 life equals exactly 1 million revolutions.
C0: Three Situations Where It Governs
C0 is defined by a 0.0001 mm permanent deformation threshold at the most heavily loaded rolling element-raceway contact. That sounds trivial — in practice, it’s enough to produce audible noise, vibration, and accelerated fatigue in a bearing that would otherwise have been serviceable.
|
Stationary under load
Overhead cranes at rest, clamping fixtures, static structures with bearing supports. |
Very slow rotation (<10 rpm)
Insufficient rotation to distribute load across rolling elements over time. C0 becomes the controlling parameter. |
Shock and impact loading
Momentary peak loads during startup, emergency stops, or process upsets are evaluated against C0, not C. |
The ratio C0/C signals a bearing’s design intent. Ball bearings: C0/C of 0.5–0.8 (rotation-optimized). Roller bearings: often above 1.0, reflecting greater contact area and resistance to deformation.
The L10 Life Calculation: Turning C Into a Prediction
| L10 | Basic rating life in millions of revolutions (90% reliability) |
| C | Dynamic load rating from the bearing catalog (kN or lbf) |
| P | Equivalent dynamic bearing load — what the bearing actually sees (kN or lbf) |
| p | Life exponent — 3 for ball bearings, 10/3 (~3.33) for roller bearings |
|
L10h = (10⁶ / 60n) × L10 where n = speed in rpm |
- Doubling the load reduces L10 life by a factor of 8 (2³).
- Increasing load by 25% reduces life by roughly 2×.
- Reducing load by 20% nearly doubles predicted life.
Equipment that runs at off-design conditions — pump at higher flow, conveyor at higher tension, fan with unbalance — applies more load than the OEM assumed. None of these changes require large load increases to produce dramatic life reductions.
Worked Example: Load vs. Life
Bearing: C = 80 kN |
Speed: 1,500 rpm |
Design load: P = 50 kN |
Baseline L10: (80/50)³ = 4.096M rev ≈ 1,000 hours
| Scenario | Applied Load | Calculated L10 Life | Bearing Type |
|---|---|---|---|
| A — Design load | 50 kN | 1,000 hrs (baseline) | Ball (p = 3) |
| B — 20% overload | 60 kN | 578 hrs (−42%) | Ball (p = 3) |
| C — 40% overload | 70 kN | 357 hrs (−64%) | Ball (p = 3) |
| D — 20% overload | 60 kN | 694 hrs (−31%) | Roller (p = 10/3) |
Life values are presented for illustrative purposes under idealized conditions. Actual life depends on lubrication, contamination, installation, and application-specific factors.
The most common misuse of L10 calculations: using the catalog design load rather than the actual operating load. If your bearing life predictions consistently miss in the optimistic direction, ask: what’s the actual load on this bearing, and how does it compare to what the OEM used to size it?
The Equivalent Dynamic Load (P): Where the Calculation Gets Real
Combined Radial and Axial Loading
| Fr | Radial load |
| Fa | Axial load |
| X, Y | Load factors from catalog, depend on bearing type and Fa/Fr ratio |
If your application has significant axial loading and you’re sizing only on radial load, P is lower than reality and your life prediction is optimistic. This is a common oversight in pump bearing selection, where hydraulic thrust is sometimes estimated rather than calculated.
Variable Load Conditions
Constant load is the exception in real industrial equipment. The standard approach calculates a mean equivalent load:
| Pm | Mean equivalent load |
| P1…Pn | Discrete loads at each operating condition |
| q1…qn | Fractions of operating time at each condition |
Strain gauges, current monitoring on motors, or shaft load cells can provide load data far more accurate than design-point assumptions. In critical applications, that measurement investment pays back in better life predictions and earlier intervention before failures.
When C0 Is the Number That Matters
Static Safety Factor (S0) — Shock and Impact Loading
The static safety factor quantifies shock load risk: S0 = C0 / P0 where P0 is the peak static equivalent load. ISO 76 recommends minimum S0 values based on application severity:
| S0 Value | Application Condition | Typical Use |
|---|---|---|
| ≥ 1 | Smooth, vibration-free, no shock | Precision equipment, controlled environments |
| ≥ 1.5 | Normal industrial, moderate vibration | General rotating equipment, standard duty |
| ≥ 2–3 | Heavy shock loading | Crushers, presses, hammermills |
| ≥ 4+ | Very heavy shock, stringent noise/vibration | Mining equipment, severe impact duty |
Startup and Shutdown Load Transients
Starting torque in large motors and driven equipment generates load transients that differ from running loads. In applications with significant inertia — fans, large pumps, heavy rolls — the load during acceleration can exceed the steady-state operating load by a factor of two or more. For high-cycle start-stop applications, evaluating peak startup loads against C0 is necessary for an accurate life prediction.
Standstill Vibration — The Underappreciated Failure Mode
Equipment that remains stationary while nearby machinery transmits vibration — standby pumps, idle conveyor drives, parked cranes — experiences fretting corrosion and false brinelling at rolling-element contact points. The rolling elements oscillate slightly under vibratory load without completing full revolutions, breaking down the lubricant film and wearing the raceway surface.
False brinelling on standby equipment is frequently misdiagnosed as installation damage or poor lubrication. The giveaway is the pattern: evenly spaced marks at rolling-element intervals on the raceways, sometimes with reddish-brown fretting-corrosion debris. If your standby pumps are failing bearings that look like this, the root cause is vibration transmission — not the bearing selection or the lubrication program.
Beyond Basic L10: The Modified Life Equation
| Lnm | Modified rating life |
| a1 | Reliability factor (adjusts for survival probability above 90%) |
| aISO | Life modification factor (lubrication quality, contamination, material fatigue limit) |
| L10 | Basic rating life from the standard formula |
a1 — The Reliability Factor
The basic L10 assumes 90% reliability. For critical equipment, a1 adjusts for the required confidence level:
|
a1 = 1.00
L10 — 90% reliability (baseline)
|
a1 = 0.62
L5 — 95% reliability
|
a1 = 0.21
L1 — 99% reliability (requires ~5× capacity)
|
aISO — Lubrication and Contamination
aISO is calculated from two inputs: κ (kappa) — the viscosity ratio (actual ÷ minimum required viscosity), and eC — the contamination factor (0.1 for heavy contamination → 1.0 for clean, filtered oil).
| Viscosity Ratio (κ) | Film Condition | Effect on Bearing Life |
|---|---|---|
| κ < 1 | Inadequate Film | aISO < 1.0 — bearing consuming life faster than formula predicts |
| κ = 1 | Minimum Film | aISO approaches 1.0 in clean conditions |
| κ > 1 | Full Film | With clean lubricant, aISO can exceed 1.0 — real life exceeds L10 baseline |
Moving from a clean, well-filtered oil application (eC near 1.0) to a moderately contaminated environment (eC of 0.3–0.5) can cut aISO — and therefore bearing life — by 50 to 70% at the same κ value. This is the quantitative basis for what experienced reliability engineers already know: contamination kills bearings.
A facility that uses the modified life equation to set bearing replacement intervals instead of relying on OEM recommendations alone. They measure actual viscosity at operating temperature, track lubricant contamination levels through periodic oil samples, and calculate aISO for each critical application. When contamination increases, the modified life calculation flags bearings for earlier inspection before failures occur.
Using Load Ratings in Your Maintenance Program
|
1
|
Know Your Actual Loads
Catalog ratings are only useful if you know what load they’re being compared against. For equipment reapplied, modified, or running at off-design conditions, actual load measurement is worth the effort. Motor current monitoring, shaft strain measurement, and process data review provide far better estimates than catalog assumptions. |
|
2
|
Check Both C and C0 for Every Application
Build the habit of also checking C0 against peak loads — startup, shutdown, emergency stops, and process shock loads. For any application with significant shock or vibration at rest, S0 should be a documented specification alongside the L10 life target. |
|
3
|
Use the Modified Life Equation for Critical Assets
For critical equipment, the basic L10 calculation is insufficient. Calculate κ at operating temperature. Assess contamination levels. The difference between the basic and modified life prediction tells you how much of the theoretical bearing life your actual operating environment is consuming. |
|
4
|
Set Replacement Intervals Based on Calculated Life
If your L10h calculation gives 8,000 hours at 90% reliability, a planned replacement interval at 6,000–7,000 hours provides reasonable margin. Adjust based on consequence of failure and the cost of planned versus unplanned replacement. |
|
5
|
Use Failure Data to Validate Predictions
Track actual bearing service life against calculated L10. When bearings consistently underperform predictions, the calculation is wrong — the load assumption is off, lubrication conditions aren’t what you assumed, or there’s a systematic installation problem. The comparison is a diagnostic tool that improves over time. |
The most reliable plants don’t just track when bearings fail — they track why they failed and what life they achieved relative to the L10 prediction. A bearing that fails at 40% of its calculated life isn’t a bearing problem. It’s a problem with the load case, the lubrication, the installation, or the environment. Treating it as a bearing problem means you’ll see the same failure again.
C combined with the life formula gives you a starting point for predicting how long a bearing should last. The modified life equation gives you a realistic estimate based on actual operating conditions. C0 protects you from the failure modes — shock, standstill vibration, startup transients — that the dynamic calculation doesn’t address.
Used together, these tools let you move from reactive bearing replacement to a managed, predictable reliability program where you know how long bearings should last, why they’re falling short when they do, and what to do about it. That’s the difference between a maintenance team that responds to failures and one that prevents them.
