This document from SKF expresses bearing speed in terms of 'n dm', see screenshot below. As rotational speed of the bearing is important for grease selection, what range of rotational speeds do bicycle wheel bearings experience, for wheel designs old and new?

(Very often we refer to the bike bearings being low speed, without quantifying this. I'm looking to quantify the range to put this 'low speed' into context and demonstrate if/why bike bearings are only ever considered low speed.)

enter image description here


Searching for "ndm rotational speed" finds that D and d are outer and inner diameters of the bearing. n is not the usual symbol for speed, but neither is the final unit "millimeters per minute divided by pi" an usual unit. And here we go:

A slightly optimistic approximation for road bike speed would be 10 m/s (36 km/h, 22.5 mph) and wheel circumference 2 m. That gives us roughly 10 m/s / 2m * 60 = 300 rpm. The bearing track diameter is around 50 mm (slightly under 2 inches and yes, the sum divided by 2 is the average of inner and outer diameters).

All together this gives us 300 rpm * 50 mm = 15000 weird units. This is very well inside "very low" category for roller bearings, and you can double the speed or halve the wheel size before you reach the "low" category. For ball bearings, there is no "very low" category, and you need to exceed the approximation by factor 6.66 before you break out of "low" category. You can change all numbers involved in the calculation by quite large amounts, and the result will always stay "very low" or in extreme cases "low".

  • 2
    More precisely: Wheel circumference is approx. 0.7m * pi = 2.2m, and max. speed is around 100km/h = 28m/s (road bike, excellent road, extreme slope downhill), which gives a rotation speed of 12.7Hz = 764rpm. So, realistically, you won't get more than 38200 mm*rpm with a bike bearing. May 23 '20 at 10:12
  • 2
    You are using the words "precise" and "realistic" in a sense I have not heard before.
    – ojs
    May 23 '20 at 10:32
  • 1
    :-) Yeah, this does sound strange. My meaning is, that I derive a somewhat precise (off by no more than 20%) upper bound to anything that a standard bike (broad definition) ever has to endure, but which is still within the achievable range by going really hardcore. In other words: I've done 74.5km/h myself, tour-de-france riders do more, but only bikes with a special fairing built for scoring records can do more than 120km/h. May 23 '20 at 10:56
  • 2
    to hit 40m/s you're more likely using 406 or 306 wheels, so the circumference will be about 1.5m or 1m. But you're also probably using unlubricated ceramic bearings without seals, so there's multiple differences from an upright bicycle right there.
    – Móż
    May 23 '20 at 11:35
  • @cmaster-reinstatemonica yes, I understood that the question was about typical values, not precise upper bound. As for precision, it should be common knowledge that any of the numbers in traditional tire sizing, including 700 in 700c do not refer to physical measurements of the wheels. 100 km/h also isn't also very precise either as typical speed or upper bound.
    – ojs
    May 23 '20 at 12:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.