Learn & Understand

The Bearing That Fails on a Bell Curve

The bearing load calculator computes an L10 rating life, a number of hours, but that figure comes wrapped in a subtle and important qualifier: it is the life 90% of identical bearings would exceed, not the moment any particular bearing will fail. This reflects a profound truth about how bearings, and many engineered components, wear out: not at a fixed, predictable instant, but statistically, scattered across a distribution. Understanding why bearing life is inherently probabilistic explains the odd-looking L10 rating and the whole discipline of designing for reliability.

Failure Without a Fixed Date

A rolling bearing does not carry an expiration stamp. Under load, its surfaces endure repeated stress with every revolution, and over time microscopic fatigue accumulates until a tiny flaw spreads and the bearing fails. Crucially, take a batch of seemingly identical bearings, run them under identical conditions, and they will not all fail together. Some give out early, some last remarkably long, and the rest fall in between. Failure is spread across a range rather than fixed at a point, because the microscopic origins of fatigue vary randomly from one bearing to the next.

Describing a Spread, Not a Point

Because failures scatter, the honest way to describe bearing life is with a statistical distribution rather than a single number. The lives of a population of bearings form a spread, with some failing soon, most failing around a central range, and some enduring far longer. Any statement about "how long a bearing lasts" must therefore really be a statement about this distribution, a probability of survival to a given time, not a guarantee. This is a fundamentally different way of thinking about durability than a fixed lifespan, and it is the reality that engineers must design around.

Why 90 Percent?

The L10 rating picks a specific, conservative point on the distribution: the time by which 10% of bearings would be expected to have failed, or equivalently, that 90% would survive. This is a deliberately cautious benchmark, chosen so that designing to it means the great majority of bearings comfortably outlast the rating. Quoting the life that most bearings exceed, rather than an average that half would fall short of, gives engineers a dependable figure to design and plan maintenance around. The "10" in L10 is literally that 10% failure fraction built into the definition.

Designing With Probability

This statistical view has real consequences the calculator makes usable. Because the L10 life is extraordinarily sensitive to load, rising steeply as load falls, small reductions in load dramatically extend the reliable life, which the calculator's load-cubed relationship captures. And because failure is probabilistic, sensible practice is to replace bearings preventively within the reliable window rather than waiting for a breakdown that could strike anywhere in the distribution. The calculator turns load, speed, and rating into an L10 figure, embodying a mature engineering mindset: that durability is not a fixed promise but a probability to be estimated, respected, and designed around.

If the bearing load comes from a rotating shaft, cross-check its sizing with the Shaft Diameter Calculator; for cyclic-load failure of other parts, the Fatigue Life Calculator.