How to Calculate Bearing L10 Rating Life and Modified Life (ISO 281)

ISO 281 defines the L10 life as the number of revolutions (or hours at a given speed) that 90 % of a group of identical bearings will complete or exceed before the first evidence of rolling-contact fatigue pitting. The basic formula relates the dynamic load rating C from the bearing catalogue to the equivalent dynamic load P your application applies, raised to the life exponent p. For ball bearings p = 3; for roller bearings (cylindrical, tapered, spherical) p = 10/3.

The equivalent dynamic load P is not simply the radial load on the shaft. When an axial load is also present, ISO 281 combines radial and axial components through bearing-type-specific X and Y load factors (for deep-groove ball bearings, the X/Y pairs depend on the ratio Fa/C0, interpolated from Table 11-1; for angular-contact and roller bearings, catalogue-specific factors apply). For a purely radial load with no axial component, P equals the radial load Fr directly.

The basic L10 life assumes adequate lubrication and a clean environment. ISO 281:2007 introduces the life modification factor aISO, which corrects for actual lubrication conditions through the viscosity ratio κ = ν/ν₁, where ν is the actual kinematic viscosity of the oil at operating temperature and ν₁ is the reference kinematic viscosity the bearing needs at its speed and mean diameter dm to form a full elastohydrodynamic (EHD) film. The reference viscosity ν₁ is defined analytically in ISO 281:2007 Annex B: for n ≥ 1000 rpm, ν₁ = 4500 · n^−0.5 · dm^−0.5; for n < 1000 rpm, ν₁ = 45000 · n^−0.83 · dm^−0.5 (all in mm²/s).

When κ ≥ 1 the EHD film is complete and aISO reaches its maximum for the given contamination level. When κ < 1 the film is incomplete (mixed lubrication), aISO drops sharply, and predicted life falls — often dramatically. κ < 0.5 represents severe boundary lubrication, which the ISO 281 model defines as a floor condition (aISO = 0.1). The contamination parameter ηc · Cu/P — where ηc is the contamination factor (0.8 for clean conditions to 0.1 for heavy contamination) and Cu is the bearing's fatigue load limit from the catalogue — further scales aISO; higher contamination reduces life even at adequate viscosity. The modified life also applies the reliability factor a₁ from ISO 281:2007 Table 2 (a₁ = 1.0 at 90 %, 0.64 at 95 %, 0.25 at 99 %) to target reliability levels other than L10.

Fatigue life addresses the long-term rolling-contact behaviour, but a bearing can also be permanently damaged by a single overload that causes plastic deformation at the Hertzian contact. ISO 76 guards against this with the static safety factor s0 = C0 / P0, where C0 is the bearing's static load rating from the catalogue and P0 is the equivalent static load combining radial and axial components (P0 = max(0.6 · Fr + 0.5 · Fa, Fr) for radial bearings).

A value of s0 ≥ 2 is the general design target for applications subject to vibration or shock loading; s0 < 1 is a failure condition indicating that the selected bearing is undersized for the peak static load and permanent deformation is likely. The static check is independent of life — a bearing can have adequate L10 life but fail on s0 if it sees occasional overloads.

ISO 281 gives the calculation method, but it does not prescribe a target life — that comes from the application. Common benchmarks from bearing manufacturer design guides and machinery standards: electric motors 20 000–30 000 h; industrial gearboxes 30 000–50 000 h; compressors and pumps 40 000 h; machine tool spindles 20 000 h; continuous-process plant up to 100 000 h. These targets are at 90 % reliability (L10h); if the application demands higher reliability, apply the a₁ factor to back-calculate the required C/P ratio from the ISO 281 equation.

The most direct way to improve calculated life is to reduce the equivalent load P (redesign to lower loads, redistribute across more bearings, or change mounting), increase the dynamic capacity C (select a larger or different bearing series), or improve the lubrication to raise κ and therefore aISO. The MechanixCalc sensitivity chart quantifies each of these levers with a ±20 % variation on load, speed and dynamic capacity.