Bearing life L10 is the number of revolutions that 90% of bearings from a given production batch will reach or exceed before the first sign of rolling contact fatigue appears on the raceway or rolling elements, calculated according to ISO 281:2007. This is the foundational parameter every design engineer must understand when selecting bearings for any application — from electric motors and gearboxes to industrial conveyors. If you need a refresher on bearing fundamentals, read the overview article before continuing.
This article walks through ISO 281 life calculation step by step, from the basic L10 formula to the modern adjusted L10a method — with two complete worked examples you can apply directly to real designs.
Basic Life L10
Basic rating life L10 is defined in ISO 281:2007 as: the number of revolutions that 90% of bearings in an identical group will reach or exceed before the first evidence of fatigue spalling on the raceway or rolling elements. The number 10 in L10 indicates a 10% failure probability — meaning 90% survival probability.
The fundamental formula:
L10 = (C / P)^p x 10^6 revolutions
Where:
- C — Basic dynamic load rating, in kN. This is the load at which 90% of bearings achieve 10^6 revolutions. The C value is found in manufacturer catalogs — for example, SKF 6308 has C = 31.9 kN, ZVL 6308 has C = 31.5 kN.
- P — Equivalent dynamic bearing load, in kN. A calculated load that converts the actual combination of radial load Fr and axial load Fa into an equivalent purely radial load.
- p — Life exponent. p = 3 for ball bearings (Hertzian point contact). p = 10/3 ≈ 3.33 for roller bearings — including cylindrical roller, tapered roller, spherical roller, and needle roller bearings (Hertzian line contact).
The difference between p = 3 and p = 10/3 reflects different fatigue mechanisms: point contact creates higher peak Hertzian stress over a smaller area compared to line contact, resulting in different nonlinear load-life relationships.
Converting to Operating Hours
In practice, engineers need life in hours. The conversion formula:
L10h = L10 / (60 x n)
Where:
- L10h — Life in operating hours
- n — Rotational speed in rpm
- 60 — Conversion factor from minutes to hours
Quick example: If L10 = 500 x 10^6 revolutions and n = 1,450 rpm:
L10h = 500,000,000 / (60 x 1,450) = 5,747 hours
Recommended Minimum Life by Application
Each application has different minimum life requirements. Recommended L10h values per SKF and ISO guidelines:
- Large electric motors (continuous): 40,000 - 60,000 hours
- Small electric motors (continuous): 20,000 - 30,000 hours
- Industrial gearboxes: 20,000 - 40,000 hours
- Conveyors: 30,000 - 50,000 hours
- Industrial fans: 40,000 - 60,000 hours
- Centrifugal pumps: 20,000 - 40,000 hours
- Crushers: 10,000 - 25,000 hours
If the calculated L10h falls below the recommended value, you need to select a larger bearing (higher C) or reduce the load.
Equivalent Dynamic Load P
The equivalent dynamic bearing load P is the hypothetical purely radial load that, if applied to the bearing, would produce the same life as the actual combination of radial load Fr and axial load Fa. This is the most critical step — and the most common source of error — in bearing life calculation.
The general formula:
P = X x Fr + Y x Fa
Where:
- Fr — Radial load, kN
- Fa — Axial load, kN
- X — Radial load factor, dimensionless
- Y — Axial load factor, dimensionless
Determining X and Y for Deep Groove Ball Bearings
For deep groove ball bearings (6000, 6200, 6300 series), the straightforward rule:
- If Fa/Fr ≤ e: then P = Fr (axial load is negligible). X = 1, Y = 0.
- If Fa/Fr > e: then P = 0.56 x Fr + Y x Fa. X = 0.56, Y is looked up from a table based on the ratio Fa/C0.
The e-factor (threshold force ratio) depends on Fa/C0, where C0 is the basic static load rating. Typical e values for deep groove ball bearings range from 0.19 to 0.44.
Example: SKF 6308 has C0 = 19.0 kN. If Fa = 1.5 kN, then Fa/C0 = 0.079, from the table e ≈ 0.27, Y ≈ 1.6. If Fr = 4.2 kN, then Fa/Fr = 0.357 > e = 0.27, so P = 0.56 x 4.2 + 1.6 x 1.5 = 2.35 + 2.40 = 4.75 kN.
Determining X and Y for Tapered Roller and Angular Contact Bearings
For tapered roller bearings and angular contact ball bearings, additional considerations apply:
- The Y factor depends directly on the contact angle α. Larger α → smaller Y (better axial load capacity).
- Tapered roller bearings always generate an internal axial force when carrying radial load. Internal axial force = 0.47 x Fr / Y.
- When mounted in pairs (back-to-back O or face-to-face X arrangement), axial force balance between the two bearings must be calculated.
For 40° angular contact bearings, X is typically 0.67 and Y = 0.68. For standard tapered roller bearings (α ≈ 15°), X = 0.4 and Y ≈ 1.5-1.7 depending on the series.
Static Equivalent Load
In addition to P for life calculation, the static equivalent load must be checked:
P0 = X0 x Fr + Y0 x Fa
The requirement is P0 < C0/s0, where s0 is the static safety factor. s0 ≥ 1.0 for normal applications, s0 ≥ 1.5-2.0 for shock loads or low-noise requirements.
Adjusted Rating Life L10a
The adjusted rating life L10a (also called L10m in SKF notation) incorporates real-world factors that the basic L10 formula ignores: reliability levels above 90%, lubrication quality, contamination level, and material fatigue limit. ISO 281:2007 defines:
L10a = a1 x a_ISO x L10
Where:
- a1 — Reliability adjustment factor
- a_ISO — Life modification factor per ISO, combining the effects of lubrication, contamination, and fatigue limit
Reliability Factor a1
When reliability greater than 90% is required, a1 reduces the calculated life:
| Reliability (%) | Designation | a1 |
|---|---|---|
| 90 | L10 | 1.00 |
| 95 | L5 | 0.62 |
| 96 | L4 | 0.53 |
| 97 | L3 | 0.44 |
| 98 | L2 | 0.33 |
| 99 | L1 | 0.21 |
Interpretation: If L10h = 20,000 hours and 95% reliability (L5) is required, then L5h = 0.62 x 20,000 = 12,400 hours. Railway and aerospace applications typically require L3 or L2.
Critical note: Many engineers misunderstand L10 as meaning "the bearing will last at least L10h hours." This is incorrect. L10 means 10% of bearings will fail before that point. The actual median life (L50) is typically about 5 times L10.
The a_ISO Factor
The a_ISO factor (life modification factor aISO) is the core of the modern ISO 281:2007 life calculation method, integrating three factors: lubricant film quality (κ), contamination level (ηc), and material fatigue limit (Cu). The value of a_ISO ranges from 0.1 (very poor conditions) to over 50 (ideal conditions).
Viscosity Ratio κ (Kappa)
κ is the ratio between the actual kinematic viscosity of the lubricant at operating temperature and the minimum viscosity required to form a full EHL (elastohydrodynamic lubrication) film:
κ = ν / ν1
Where:
- ν — Actual kinematic viscosity of the lubricant at operating temperature (mm²/s)
- ν1 — Minimum required viscosity (mm²/s), dependent on mean bearing diameter dm and rotational speed n
Determining ν1: Look up the chart in SKF or Schaeffler catalogs. Inputs: dm = (d + D)/2 (mm) and n (rpm). Example: bearing 6308 has d = 40 mm, D = 90 mm → dm = 65 mm. At 1,450 rpm, ν1 ≈ 10 mm²/s.
If using NLGI 2 lithium grease with base oil ν = 100 mm²/s at 40°C, at an operating temperature of 70°C, ν drops to approximately 25-30 mm²/s. Therefore κ = 25/10 = 2.5 — an acceptable lubrication condition.
κ classification:
- κ ≥ 4: Full EHL film, thick lubricant layer, a_ISO can be very high
- κ = 1-4: Mixed lubrication, acceptable if contamination is low
- κ < 1: Boundary lubrication, metal-to-metal contact, severe life reduction
- κ < 0.4: Critical, lubricant type or operating temperature must be revised
Contamination Factor ηc
ηc reflects the cleanliness of the lubricant and the operating environment:
| Contamination Level | Description | ηc (ball bearing dm < 100) | ηc (roller bearing dm < 100) |
|---|---|---|---|
| Very clean | Clean room, oil filtered < 3 μm | 0.8 - 0.6 | 0.9 - 0.8 |
| Clean | High-quality grease, good seals | 0.6 - 0.5 | 0.8 - 0.6 |
| Normal | Standard grease, standard seals | 0.5 - 0.3 | 0.6 - 0.4 |
| Contaminated | Dusty environment, poor seals | 0.3 - 0.1 | 0.4 - 0.2 |
| Heavily contaminated | Abrasive particles, no seals | 0.1 - 0 | 0.2 - 0 |
Note: ηc values in this table are reference values from ISO 281:2007, Table 14. Exact values depend on dm ratio, load, and specific bearing type.
Fatigue Load Limit Cu
Cu (fatigue load limit) is the load below which, theoretically, the bearing has infinite life if lubrication is perfect and no contamination is present. Cu values are found in catalogs:
- SKF 6308: Cu = 0.655 kN
- SKF 22220 EK: Cu = 24.5 kN
The ratio ηc x Cu / P is the primary input for looking up a_ISO from catalog diagrams. A higher ratio yields a higher a_ISO.
How to Look Up a_ISO
The 4-step process:
- Calculate dm and look up ν1 from the chart (or use an approximation formula)
- Determine actual ν at operating temperature → calculate κ = ν/ν1
- Determine ηc based on contamination conditions
- Calculate ηc x Cu/P, combine with κ, and look up a_ISO from the diagram in SKF/Schaeffler/ZVL catalogs
The a_ISO diagram consists of a family of curves: the horizontal axis is ηc x Cu/P, the vertical axis is a_ISO, and each curve corresponds to a κ value. When κ ≥ 4, a_ISO can reach 10-50. When κ < 0.4, a_ISO is typically below 0.5 regardless of ηc x Cu/P.
Worked Example 1: Electric Motor with 6308 Bearing
Input Data
- Application: 15 kW electric motor, 4-pole, belt drive on end A
- Bearing at end A (load side): 6308-C3 (deep groove ball bearing)
- Speed: n = 1,450 rpm
- Radial load: Fr = 4.2 kN (belt tension + rotor weight)
- Axial load: Fa = 0 kN (no axial load)
- Operating temperature: 70°C
- Lubrication: NLGI 2 lithium grease, ν_base oil = 100 mm²/s at 40°C
SKF 6308 Catalog Data
- C = 31.9 kN (dynamic load rating)
- C0 = 19.0 kN (static load rating)
- Cu = 0.655 kN (fatigue load limit)
- d = 40 mm, D = 90 mm
Step 1: Calculate Equivalent Dynamic Load P
Since Fa = 0, the load is purely radial:
P = Fr = 4.2 kN
Check static safety factor: s0 = C0/P = 19.0/4.2 = 4.5 > 1.0 — safe.
Step 2: Calculate Basic Rating Life L10
Ball bearing, so p = 3:
L10 = (C/P)^3 x 10^6 = (31.9/4.2)^3 x 10^6
C/P = 7.595
L10 = 7.595^3 x 10^6 = 438.2 x 10^6 revolutions
Step 3: Convert to Hours
L10h = 438,200,000 / (60 x 1,450)
L10h = 438,200,000 / 87,000
L10h = 5,037 hours
Step 4: Calculate Adjusted Life L10a
Determine κ:
dm = (40 + 90) / 2 = 65 mm. At n = 1,450 rpm and dm = 65 mm, from the chart: ν1 ≈ 10 mm²/s.
Grease with ν_base oil = 100 mm²/s at 40°C. At 70°C, using viscosity index VI ≈ 95, ν ≈ 28 mm²/s.
κ = 28 / 10 = 2.8
Determine ηc:
Indoor motor, high-quality grease, good 2RS seals or effective labyrinth seal: ηc = 0.5 ("clean" level).
Calculate ηc x Cu/P:
ηc x Cu/P = 0.5 x 0.655 / 4.2 = 0.078
Look up a_ISO:
With κ = 2.8 and ηc x Cu/P = 0.078, from the SKF diagram for ball bearings: a_ISO ≈ 1.5
Result:
L10a = a1 x a_ISO x L10h = 1.0 x 1.5 x 5,037 = 7,556 hours
Step 5: Evaluation
| Parameter | Value | Comment |
|---|---|---|
| L10h (basic) | 5,037 hours | Below 20,000h recommendation for motors |
| L10a (adjusted) | 7,556 hours | Still below recommendation |
| Static safety factor s0 | 4.5 | Good |
| C/P ratio | 7.6 | Reasonable |
Assessment: The calculated life of 5,037-7,556 hours falls well short of the 20,000-hour recommendation for continuous-duty electric motors. Root cause: the belt drive load of 4.2 kN is substantial for a 6308 bearing. Solutions:
- Upsize the bearing to 6310 (C = 48.5 kN) → L10h = (48.5/4.2)^3 x 10^6 / 87,000 = 17,568 hours
- Reduce belt load by adjusting the transmission ratio and belt tension
- Switch to a coupling if the speed ratio permits — eliminates belt load entirely
In practice, many industrial motors with belt drives using 6308 bearings still run 3-5 years (approximately 13,000-22,000 continuous hours) because: (1) actual loads are typically lower than calculated loads, (2) high-quality grease pushes a_ISO to 3-5, and (3) L10 represents the 90th percentile, not the average life.
Worked Example 2: Conveyor Pulley with 22220 EK Bearing
Input Data
- Application: Drive pulley for a coal conveyor, continuous operation
- Bearing: 22220 EK (self-aligning spherical roller bearing — see the spherical roller bearing article)
- Speed: n = 300 rpm
- Radial load: Fr = 45 kN (pulley weight + belt tension)
- Axial load: Fa = 8 kN (belt misalignment + 5° incline)
- Operating temperature: 55°C
- Lubrication: Circulating oil ISO VG 220, ν = 220 mm²/s at 40°C
SKF 22220 EK Catalog Data
- C = 285 kN (dynamic load rating)
- C0 = 270 kN (static load rating)
- Cu = 24.5 kN (fatigue load limit)
- d = 100 mm, D = 180 mm
- e = 0.32 (for Fa/C0 = 8/270 = 0.030)
- Y1 = 2.1, Y2 = 3.1 (from catalog)
Step 1: Calculate Equivalent Dynamic Load P
Check: Fa/Fr = 8/45 = 0.178
Compare with e = 0.32: since Fa/Fr = 0.178 < e = 0.32:
For spherical roller bearings per ISO 281:
- When Fa/Fr ≤ e: P = Fr + Y1 x Fa
- When Fa/Fr > e: P = 0.67 x Fr + Y2 x Fa
Since Fa/Fr = 0.178 < e = 0.32:
P = 45 + 2.1 x 8 = 45 + 16.8 = 61.8 kN
Important note: For spherical roller bearings, axial load is always included in P even when Fa/Fr < e (unlike deep groove ball bearings). This is a critical point that many engineers overlook.
Check static safety factor: P0 = Fr + Y0 x Fa (Y0 ≈ 2.1). P0 = 45 + 2.1 x 8 = 61.8 kN. s0 = C0/P0 = 270/61.8 = 4.4 — safe.
Step 2: Calculate Basic Rating Life L10
Spherical roller bearing (roller type), so p = 10/3:
L10 = (C/P)^(10/3) x 10^6 = (285/61.8)^(10/3) x 10^6
C/P = 4.611
L10 = 4.611^3.333 x 10^6 = 130.2 x 10^6 revolutions
Step 3: Convert to Hours
L10h = 130,200,000 / (60 x 300) = 130,200,000 / 18,000
L10h = 7,233 hours
Step 4: Calculate Adjusted Life L10a
Determine κ:
dm = (100 + 180) / 2 = 140 mm. At n = 300 rpm and dm = 140 mm, from the chart: ν1 ≈ 18 mm²/s.
ISO VG 220 oil: ν = 220 mm²/s at 40°C. At 55°C, ν ≈ 85 mm²/s (from viscosity-temperature chart).
κ = 85 / 18 = 4.7
κ ≥ 4 — full EHL film, excellent lubrication condition.
Determine ηc:
Coal conveyor — dusty environment, but circulating oil with 25 μm filter and labyrinth seals: ηc = 0.3 ("normal" to "contaminated" due to coal dust).
Calculate ηc x Cu/P:
ηc x Cu/P = 0.3 x 24.5 / 61.8 = 0.119
Look up a_ISO:
With κ = 4.7 and ηc x Cu/P = 0.119, from the SKF diagram for roller bearings: a_ISO ≈ 3.5
Result:
L10a = a1 x a_ISO x L10h = 1.0 x 3.5 x 7,233 = 25,316 hours
Step 5: Evaluation
| Parameter | Value | Comment |
|---|---|---|
| L10h (basic) | 7,233 hours | Below 30,000h recommendation for conveyors |
| L10a (adjusted) | 25,316 hours | Approaching recommendation thanks to good lubrication |
| κ | 4.7 | Full EHL film |
| a_ISO | 3.5 | Good, driven by high κ |
| s0 | 4.4 | Safe |
Assessment: L10a = 25,316 hours approaches the 30,000-hour recommendation for conveyors. To improve further:
- Upgrade the oil filter from 25 μm to 10 μm → ηc increases to 0.4-0.5 → a_ISO rises to 5-8 → L10a reaches 36,000-58,000 hours
- Add contact seals outside the labyrinth seals to reduce coal dust ingress
- Keep oil temperature below 55°C with an oil cooler → maintain κ ≥ 4
This example illustrates a key principle: for heavy-loaded roller bearings, lubrication and contamination control have a greater impact on life than upsizing the bearing.
At a paper mill in Binh Duong province, an engineer calculated life for a 22317 EK bearing on a dryer roll shaft — Fr = 28 kN, Fa = 3 kN, n = 800 rpm, lithium complex grease at 85°C. Result: L10h = 12,400 hours — below the 20,000-hour requirement. Rather than upsizing the bearing (costly housing modification), the engineer switched to polyurea grease (higher ν at 85°C) and upgraded seals from felt to labyrinth. The κ ratio rose from 1.8 to 3.5, ηc from 0.3 to 0.5 — a_ISO jumped from 1.2 to 4.8, pushing L10a to 59,500 hours.
Bearing Life Calculation Software
Manual calculation is appropriate for quick checks and training, but complex designs require dedicated software. Here are the four most widely used tools.
SKF Bearing Select
- Cost: Free, web-based (no installation required)
- Features: Bearing selection by application, L10 and L10a calculation, lubrication analysis, PDF report export. Complete SKF catalog included.
- Strengths: Intuitive interface, fast results, good English support. Automatic a_ISO calculation from κ and ηc inputs.
- Limitations: SKF bearings only. No multi-bearing shaft system analysis.
- Best for: Maintenance engineers, quick design checks, comparing alternatives.
Schaeffler BEARINX
- Cost: Free online version (BEARINX-online Easy). Full version requires a license.
- Features: Complete shaft system analysis, multiple bearings on a single shaft, time-varying dynamic loads, FEM housing analysis.
- Strengths: Most powerful tool for gearbox and shaft system design. Calculates load distribution among individual rollers.
- Limitations: Steep learning curve, FAG/INA catalog only.
- Best for: OEM design engineers, R&D departments.
NSK Bearing Doctor
- Cost: Free, iOS/Android app
- Features: Bearing damage diagnosis from photos, L10 calculation, quick NSK catalog lookup.
- Strengths: Mobile convenience, AI-based image diagnosis useful for field maintenance.
- Limitations: Simpler calculation features compared to SKF Bearing Select, NSK catalog only.
- Best for: Maintenance engineers, field troubleshooting.
Timken Syber
- Cost: Free after account registration
- Features: Shaft system analysis, adjusted life calculation, tapered roller bearing clearance optimization. Especially strong for tapered roller bearings (Timken's core expertise).
- Strengths: Most accurate tapered roller bearing calculations, preload and endplay analysis.
- Limitations: Primarily Timken catalog, interface is dated.
- Best for: Tapered roller bearing applications (automotive hubs, machine tool spindles).
Note: All of the above software tools use ISO 281:2007 as their calculation foundation. Results will be equivalent given the same inputs — the differences lie in each manufacturer's proprietary C and Cu values and their specific a_ISO determination method (SKF uses its own method called a_SKF, which sometimes yields different results from the generic ISO diagrams). ZVL also follows ISO 281, so you can use any of these tools with C and Cu values from the ZVL catalog.
Common Mistakes in Bearing Life Calculation
Based on 9 years of technical consulting, these are the 8 most common errors engineers make:
1. Ignoring Axial Load
Many engineers use P = Fr while forgetting axial loads from: belt tension (axial component), bevel/helical gear forces, vertical shaft weight, thermal shaft expansion. For deep groove ball bearings, if Fa/Fr > e but axial load is ignored, the actual P can be 30-60% higher than Fr alone.
2. Using the Wrong Life Exponent p
p = 3 applies only to ball bearings. Cylindrical roller, tapered roller, spherical roller, and needle roller bearings all use p = 10/3. Using p = 3 for roller bearings overestimates L10 by approximately 15-25%.
3. Treating L10 as Guaranteed Life
L10 = 20,000 hours does not mean the bearing will last at least 20,000 hours. It means 10% of bearings in the batch will fail before 20,000 hours. If you install 100 identical motors, approximately 10 will need bearing replacement before 20,000 hours. The shortest actual life in a group (L1) is only 21% of L10.
4. Ignoring Contamination Effects
When calculating basic L10 (without a_ISO), the result assumes perfect lubrication and zero contamination. In reality, cement plants, coal mines, and food processing environments have low ηc (0.1-0.3), driving a_ISO below 0.5. Actual life may be only 30-50% of the calculated L10.
5. Using the Wrong Viscosity
Looking up viscosity at 40°C instead of at the actual operating temperature. ISO VG 100 oil has ν = 100 mm²/s at 40°C but only 12-15 mm²/s at 80°C. Using ν = 100 to calculate κ produces an error factor of 5-8x.
6. Not Accounting for Dynamic Load Additions
Catalog loads are typically static. Real-world additions include: vibration from rotor imbalance, process shock loads (crushing, pressing, stamping), inertia loads during start/stop, resonance effects. SKF recommends a dynamic load factor fd = 1.2-1.8 for moderate-impact applications.
7. Ignoring Installation Effects
A 0.5° shaft misalignment on a deep groove ball bearing can reduce life by 50%. Excessive interference fit reduces internal clearance → increases contact stress → reduces life. ISO 281 does not account for installation quality — this is why theoretical L10 is always more optimistic than field reality.
8. Calculating for a Single Operating Point
Many machines operate at multiple load conditions. A water pump motor runs 60% of the time at 100% load, 30% at 70% load, and 10% at 50% load. The equivalent life must be calculated using Miner's rule:
1/L_total = q1/L1 + q2/L2 + q3/L3 + ...
Where qi is the time fraction at load condition i, and Li is the L10 at that condition.
Actual vs Theoretical Life Comparison
Actual average bearing life is typically 2-8 times the calculated L10 under good conditions, but can fall below L10 when operating conditions are poor. Understanding this relationship helps you correctly interpret calculation results and make sound design decisions.
Why Field Life Often Differs from Theory
Factors that make actual life HIGHER than L10:
- L10 is the 10th percentile, meaning 90% of bearings last longer. Median life L50 ≈ 5 x L10.
- Design loads are usually worst-case; actual operating loads are 20-40% lower.
- Modern bearing materials (clean steel, advanced heat treatment) exceed ISO assumptions — SKF has introduced the a_SKF factor to reflect this.
Factors that make actual life LOWER than L10:
- Installation errors: misalignment, excessive interference fit, wrong clearance, improper tooling
- Poor lubrication: wrong grease type, insufficient quantity, over-greasing, missed relubrication intervals
- Contamination: dust, water, metallic wear particles
- Electrical discharge machining (EDM) from variable frequency drives (VFDs) causing shaft currents
- Resonance vibration, unexpected shock loads
Field Case 1: Cement Plant
A cement plant in central Vietnam installed 22320 E bearings on an induced draft (ID) fan, n = 990 rpm, Fr = 55 kN. Calculated L10h = 18,500 hours. In practice, bearings failed after 6,000-9,000 hours — just 32-49% of L10.
After investigation, the root cause was clinker dust ingress through worn labyrinth seals → actual ηc ≈ 0.15 instead of the design assumption of 0.4. Solution: added V-ring contact seals outside the labyrinth seals + increased regreasing frequency from quarterly to monthly. Result: bearing life increased to 14,000-22,000 hours.
Field Case 2: Packaging Line
A food factory in Binh Duong province used 6205-2RS bearings on a packaging conveyor, n = 1,200 rpm, Fr = 1.1 kN. Calculated L10h = 62,000 hours. But bearings failed after 8,000-12,000 hours.
Root cause: Daily high-pressure water washdown of the conveyor. The 2RS seals resist dust but cannot withstand direct high-pressure water jets → water ingress → raceway corrosion → premature fatigue. Solution: switched to 6205-2RS1/W64 (FKM seals with better water resistance) + installed splash guards to deflect wash water away from bearing housings. Result: bearing life exceeded 30,000 hours.
Field Case 3: VFD-Driven Motor
A textile factory in Dong Nai province installed a 22 kW motor controlled by a variable frequency drive (VFD) operating at 200-1,450 rpm. The 6310-C3 bearing had a calculated L10h = 28,000 hours at maximum load. But the non-drive end bearing failed after 4,000-6,000 hours with characteristic "washboard" (fluting) damage patterns.
Root cause: Shaft current (EDM) from the VFD's high-frequency switching creating common-mode voltage on the shaft, discharging through the bearing oil film → electrical erosion. This failure mode is completely outside the scope of ISO 281 life calculation. Solution: installed an electrically insulated bearing (SKF INSOCOAT or hybrid ceramic) at the non-drive end + shaft grounding brush at the drive end. Result: bearings achieved the design life of 28,000+ hours.
Lessons Learned
L10 and L10a are design tools, not precise predictions of when a bearing will fail. To bring actual life close to theoretical predictions:
- Control contamination — Seals, filters, environment. This delivers the highest ROI of any single measure.
- Lubricate correctly — Right type, right quantity, right interval. Maintain κ ≥ 1, ideally κ ≥ 2.
- Install properly — Specialized tooling, clearance verification, shaft alignment. Installation errors account for 16% of premature failures.
- Monitor condition — Vibration, temperature, grease analysis. Early detection enables planned replacement instead of unexpected breakdown.
- Calculate correctly — Use L10a instead of just L10. Do not ignore Fa, do not use the wrong p, do not neglect contamination.
When all five practices are followed, actual field life typically reaches L50 (median life) — approximately 5 times the calculated L10. That is the target every maintenance engineer should aim for.