ISO/TS 16281 — Bearing Internal Load Distribution, Contact Angle & Stiffness Calculator

ISO 281 reduces the applied loads to one equivalent load P and computes life from the ratio C/P. That is the right method for selecting a bearing from a catalogue, and it cannot see inside the bearing. ISO/TS 16281 solves the internal equilibrium instead, so it produces the load on each rolling element, the contact angle each one runs at, the size of the load zone, and the bearing's stiffness. Because life goes as the inverse cube of contact load, the worst-loaded element governs — and only the internal method can tell you what it carries.

No. MechanixCalc implements the standard's internal load-distribution method and reports the per-element load, the contact angles, the load zone and the bearing stiffness. The modified reference rating life L_nmref — which integrates the contact stress along the raceways and applies the standard's life-modification treatment — is not implemented. The life figures on the bearing calculator are ISO 281 L10 and Lnmh. Treat the internal load distribution as design insight, not as an ISO/TS 16281 compliance calculation.

Misalignment tilts one ring relative to the other, which shrinks the load zone and concentrates the load on fewer rolling elements. On a 20-ball angular-contact geometry under combined radial and axial load, one milliradian — about 3.4 arc-minutes, an ordinary mounting error — takes the load zone from eleven loaded elements to nine and raises the worst contact by roughly 14 %. Because rolling-contact fatigue life goes as Q⁻³, that is about a third of the life. That figure is for a shaft stiff enough to IMPOSE the misalignment, which is the usual case and the calculator's default; a bearing free to tilt instead finds its own alignment, and there the same milliradian can lower the worst contact rather than raise it. Which boundary condition applies is a real modelling choice, and the calculator names the one it used on every result. An equivalent-load calculation shows no change under either.

The number of rolling elements, the ball diameter, the pitch diameter, the free contact angle, the raceway osculation (groove radius divided by ball diameter) and the mounted diametral clearance. None of those appear on the catalogue page a bearing's load ratings come from, so MechanixCalc can estimate them from the bore and outside diameter to get you started. That estimate is shown, not hidden, and it is editable: a real ball count depends on the series, the cage and the filling method, and two bearings with identical boundary dimensions routinely differ. Replace the estimate with the bearing's own geometry before relying on the result.

Because nothing else can supply it. A shaft-deflection model, a rotordynamic model and a thermal-growth model all need the bearing's radial, axial and tilting stiffness as an input, and a bearing catalogue publishes load ratings rather than stiffness. The stiffness is the tangent of the load–displacement relation at the operating point, so it falls straight out of the internal equilibrium — and it changes with load, clearance and preload, which is exactly why a single tabulated figure would not do.

The angle in the designation is the FREE contact angle, measured with no load on the bearing. Under thrust the rings shift axially and the contact point climbs the groove, so the operating angle opens — on a 15° row under 4 kN of pure thrust it reaches about 21.4°. Under combined load it also varies from element to element around the race. Holding the angle at its free value understates the axial capacity and the axial stiffness.

No. MechanixCalc's implementation covers ball bearings, where each contact is a Hertzian point contact with a 3/2 load-deflection exponent. A roller's line contact needs the roller to be divided into laminae along its length, with the load distribution solved along that length as well as around the race — that is a different calculation and it is not implemented. The ISO 281 life calculation on the bearing page covers roller bearings as usual.

Centrifugal loading of the rolling elements is not modelled. It throws the balls outward, so it ADDS to the outer-raceway contact — roughly 10 % of the worst contact at 20 000 rpm on a 90 mm pitch diameter, and more above that. That omission is not conservative for high-speed duty, which is why every result says so. The rings are also treated as rigid: for a stiff housing that is conservative, because a compliant ring would spread the load over more elements, but for a thin-section ring in a soft or out-of-round housing, ovalisation concentrates the load instead. Gyroscopic and cage effects and any thermal change in clearance are outside the model.