Bearing life calculations are the foundation of mechanical design and maintenance planning, but they are also the most misunderstood numbers in the reliability engineer's toolkit. The L10 life printed in catalogs and calculated by software represents a statistical prediction under perfect conditions. Real bearings in real plants live in a world of contamination, misalignment, inadequate lubrication, and variable loads.
Understanding what L10 means, what it does not mean, and how to bridge the gap between calculated and actual life is essential for anyone who selects bearings, schedules maintenance, or investigates failures.
What L10 Life Actually Means
L10 is the number of revolutions (or hours at a given speed) at which 10 percent of a population of identical bearings will have developed fatigue spalling under identical load and speed conditions. Flip it around: 90 percent of bearings will last at least this long.
The key word is "fatigue." L10 predicts subsurface fatigue failure, the natural end-of-life mechanism for a properly installed and lubricated bearing. A tiny crack initiates below the raceway surface, propagates to the surface, and a flake of steel pops out. That is spalling. It is inevitable given enough load cycles, and L10 tells you when to expect it.
But here is the disconnect: less than 10 percent of bearing replacements in industrial plants are due to fatigue spalling. The vast majority fail from contamination, inadequate lubrication, misalignment, or improper mounting. These failure modes are not captured by L10 at all. A bearing with a calculated L10 of 100,000 hours may fail in 2,000 hours if the seals are damaged and grinding dust enters the housing.
L10 = (C / P)p × 106 revolutions
C = basic dynamic load rating (from catalog)
P = equivalent dynamic load
p = life exponent (3 for ball, 10/3 for roller)
L10h = L10 / (60 × n) hours
n = speed in RPM
Basic L10 Bearing Life Calculator
Calculate L10 bearing life in hours and years from dynamic load rating C and equivalent load P. Includes reliability-adjusted L10a for 95% and 99% confidence levels.
The C/P Ratio: The Number That Drives Everything
Bearing life is driven by the ratio of the dynamic load rating C to the applied equivalent load P. Because the life exponent is 3 (for ball bearings), doubling the C/P ratio increases life by a factor of eight. Halving it cuts life to one-eighth.
This extreme sensitivity means that small errors in load estimation have large impacts on predicted life. If you underestimate P by 20 percent, the predicted life is 2x too optimistic. If you overestimate P by 20 percent, you are 50 percent too pessimistic and may select an unnecessarily large bearing.
Common load estimation errors include: forgetting belt tension on a driven shaft (adds 2 to 3 times the torque load as a radial force), ignoring gear mesh thrust on helical gears, assuming constant load when the duty cycle varies, and not accounting for the weight of the shaft and coupling assembly.
C/P < 4: Very heavy load, short life, check carefully
C/P 4 to 8: Normal industrial application
C/P 8 to 15: Light to moderate load, long life
C/P > 15: Very light load, consider a smaller bearing (may have minimum load issues)
Equivalent Dynamic Load Calculator
Calculate equivalent dynamic bearing load P from radial and axial forces. Supports deep groove, angular contact, cylindrical roller, self-aligning, and spherical roller bearings with correct X and Y factors.
L10a: When 90% Is Not Good Enough
L10 gives 90 percent survival probability. For many industrial applications, that is adequate. But for safety-critical equipment, remote installations, or machines where unplanned downtime costs tens of thousands of dollars per hour, 90 percent is not enough.
The reliability adjustment factor a1 multiplies L10 to give L10a life at higher confidence levels. For 95 percent reliability, a1 = 0.62. For 99 percent reliability, a1 = 0.21. That means the 99 percent confidence life is only one-fifth of L10. A bearing with a 50,000-hour L10 life has an L10a life at 99 percent of only 10,500 hours.
For maintenance planning, L10a at 95 percent is a practical design target for equipment that runs continuously. It means that only 5 percent of bearings will fail before this interval, giving you reasonable confidence that a planned replacement will catch most failures before they happen.
90% survival (L10): a1 = 1.00
95% survival: a1 = 0.62
96% survival: a1 = 0.53
97% survival: a1 = 0.44
98% survival: a1 = 0.33
99% survival: a1 = 0.21
Why Actual Life Rarely Matches Calculated Life
In well-maintained applications with clean environments and proper lubrication, actual bearing life often exceeds L10 by a factor of 3 to 5. The bearings die of old age before fatigue gets them. In poorly maintained applications, actual life can be one-tenth of L10 or less.
The extended rating life standard (ISO 281:2007) adds an aISO factor that accounts for lubrication conditions and contamination. A bearing in clean oil with proper viscosity can have an aISO of 5 to 50, dramatically extending the calculated life. A bearing in contaminated grease with marginal viscosity can have an aISO below 1, reducing calculated life below the basic L10.
The practical lesson: L10 is a comparison tool, not a prediction tool. Use it to compare bearing options against each other. Do not use it to predict when a specific bearing will fail. For maintenance scheduling, combine calculated life with condition monitoring (vibration, temperature, ultrasound) to replace bearings based on actual condition rather than calendar time alone.