The basic dynamic load rating C is defined by ISO 281:2007 as the constant radial load a group of identical bearings can endure for one million revolutions before 10% fail by rolling-contact fatigue [S4][S1]. The basic static load rating C0, defined under ISO 76, is the load that produces a permanent indentation of 0.0001 times the rolling-element diameter at the most heavily loaded contact, which is the criterion that controls a bearing that is stationary, oscillating, or running below roughly 10 rpm [S4].
On a typical deep-groove ball bearing the static-to-dynamic ratio C0/C sits in the 0.5 to 0.8 range, while roller bearings commonly exceed 1.0 because the line contact carries more load before the same brinelling limit is hit [S7]. That ratio is the first sanity check a bearing engineer applies when reading a catalogue page, and the ball bearing reference page covers the geometry that drives it.
Where C applies and where C0 takes over
Dynamic load rating C is the input to the fatigue equation L10 = (C/P)^3 x 10^6 revolutions for ball bearings, where P is the equivalent dynamic load combining radial and axial components [S1][S2]. The same equation is written for roller bearings as L10 = (C/P)^(10/3) x 10^6, which is why comparing C across ball and roller bearings of the same bore without converting exponents is misleading [S4].
C0 controls three regimes. First, a stationary bearing supporting a sustained load, such as a parked wind-turbine main shaft or a stalled conveyor idler. Second, an oscillating bearing that never completes a full rotation, for example a slewing ring that swings through a partial arc. Third, any rotation slow enough that brinelling rather than subsurface fatigue is the dominant damage mode, conventionally below 10 rpm for ball bearings [S4][S3]. Outside these three regimes, the fatigue-limited C is the relevant figure and C0 is only checked as a sanity bound. A different ball bearing geometry, such as an angular-contact or duplex pair, shifts both ratings but not the C-versus-C0 decision tree; the ball bearing entry walks through the contact-angle variants that change both numbers.
Calculating equivalent load P and static safety factor s0
The equivalent dynamic load for a combined radial and axial case is P = X Fr + Y Fa, where X is the radial factor, Y is the thrust factor, and Fa is the axial component; for miniature and instrument bearings with outside diameter at or below 0.625 inch, NHBB publishes X = 0.56, Y = 2.10, and e = 0.16 as sufficiently accurate [S2]. The rotation factor V is 1.0 for inner-ring rotation and 1.2 for outer-ring rotation, which is the small but real reason a bearing that rotates around its outer race fatigues faster than the same bearing loaded the other way [S2].
For the static check, the safety factor s0 = C0/P0 must clear a minimum that depends on bearing type and shock severity. Standard guidance is s0 ≥ 1 for ball bearings under normal load, s0 ≥ 1.5 for roller bearings under normal load, and s0 ≥ 3 for roller bearings exposed to shock loads [S4]. An equivalent static-capacity formula often used in distributor material is C0 = (P0/F0) x C, where F0 folds in bearing type, load direction, and lubrication conditions and is taken from the manufacturer's catalogue [S5]. A useful cross-check on a candidate selection is to ask whether a disc coupling on the same driveline has been similarly de-rated; the service factor logic for disc couplings follows the same pattern of multiplying catalogue torque by a shock-and-duty multiplier.
Asymmetry of the L10 exponent under load swings
The L10 exponent operates on a load ratio, not a delta, so a 25% load change is not symmetric around the design point. Cutting load to 0.75 P extends ball-bearing life to (1/0.75)^3 ≈ 2.37x rated and roller-bearing life to (1/0.75)^(10/3) ≈ 2.61x rated [S4]. Increasing load to 1.25 P drops ball life to 0.8^3 ≈ 0.51x and roller life to 0.8^(10/3) ≈ 0.48x of rated [S4]. The asymmetry is small per cycle but compounds quickly in fatigue-limited equipment, which is why heavy-industry sizing rules of thumb lean conservative rather than nominal.
A practical derating target on rotating machinery is a load ratio C/P of 3 or more, which on the ISO 281 formula gives 27x rated L10 for ball bearings and 38x for roller bearings [S1]. Where the calculated C/P falls under 3 the catalogue should be revisited before the bearing is released for purchase, since a 2:1 ratio is only 8x rated life for ball bearings and 10x for roller bearings [S1].
Decision matrix: which rating drives the selection
For a clean rotating application with steady load and no shock, use C with the L10 = (C/P)^3 formula and verify C0 as a bound only; the typical C0/C ratio of 0.5 to 0.8 on a deep-groove ball bearing means the static check usually passes automatically once fatigue life is acceptable [S7][S4].
For a stationary or oscillating bearing, including a parked wind-turbine rotor, a slewing ring, or a slow-pivoting crane pedestal, the controlling number is C0, and the selection criterion is s0 = C0/P0 against the thresholds of 1 for ball bearings, 1.5 for roller bearings, and 3 for roller bearings under shock [S4].
For mixed regimes, where the bearing both rotates and holds a sustained external load (pump shafts under deadweight, gearbox idler gears), size for fatigue using C, then recheck s0 against the heavier of the rotating load and the sustained static load; the larger of the two is what the catalogue must satisfy [S3][S4].
For shock or impact environments, such as a rolling mill back-up roll or a crusher eccentric, the static path dominates because peak brinelling force is far above the rotating equivalent load; raise the static threshold to s0 ≥ 3 for roller bearings, and consider a cylindrical-roller alternative because the line contact raises the C0/C ratio above 1.0 [S4][S7]. The same conservatism pattern shows up in adjacent rotating-equipment specs; for example, the API 598 seat test for butterfly valves uses a similar idea of upgrading the duty multiplier on the static side rather than the dynamic side.
What C0/C ratio of 0.5 to 0.8 actually tells you
For deep-groove ball bearings, a C0/C band of 0.5 to 0.8 means the brinelling limit is reached at a lower absolute load than the million-revolution fatigue load, which is the correct trade-off for a part that is expected to rotate most of its service life [S7]. A ratio below 0.5 would indicate a fatigue-optimised bearing that cannot hold much static load; a ratio above 1.0 would indicate a static-optimised bearing that is over-built for fatigue and is likely heavier and more expensive than the duty needs [S7].
For linear rolling bearings the same logic applies with a different deformation limit; the C0 rating is the load that produces a permanent raceway indentation of roughly 0.0001 times the ball diameter at the most heavily loaded contact point, which is the criterion used in linear-rail catalogue entries [S8]. Once that criterion is exceeded at standstill, brinelling is permanent and the bearing will never run true again, regardless of how much fatigue margin the C rating shows on paper [S8].
Common spec errors and the standards they violate
Using C to size a stationary or oscillating bearing is the most common mistake and typically leads to undersizing, because C0 on the same part is roughly 0.5 to 0.8 of C and the bearing will brinell before the calculated fatigue life is reached [S4][S7]. The reverse error, using C0 to size a continuously rotating bearing, is less harmful in safety terms but wastes material, since the fatigue-limited L10 of an oversized part climbs with the cube of the load ratio for ball bearings [S4][S1].
A subtler error is dropping the ISO 281 reliability modifier a2 when target reliability is above 90%; for 99% reliability (L1 life) the L10 result must be multiplied by a modifier that NHBB publishes for both 440C and 52100 alloys, and the same rule applies in miniature and instrument-bearing catalogues [S2]. A related error is ignoring the 1.2 rotation factor for outer-ring rotation, which understates the equivalent load P and overstates calculated life by up to 20% [S2].
Track the next node on the C-versus-C0 discussion: ISO 281 and ISO 76 revision cycles, and the ABMA #9 / #12 standards cited by NHBB for C values that include race-to-ball conformity, which is the conformity treatment that makes the published C values match real test data rather than pure-geometry theory [S2]. Watch for the ABMA/ISO harmonisation work on bearing-life modifiers, which historically has been the vehicle for adding lubrication-contamination and fatigue-load-factor adjustments to the bare (C/P)^p equation.
The underlying component specifications are covered under dynamic compactor, and dynamic balancing machine.