If two surfaces rub together with a normal force between them, friction is generated. You could add lubrication between those two surfaces and reduce that friction.

Bearings essentially consist of two surfaces rubbing together and, as before, under load, friction will occur. As before, you could lubricate the bearings and reduce that friction.

So absent a better understanding, I can't see how bearings add much of anything to efficiency. This must be wrong or all of industry would be spending money frivolously on unnecessary parts.

So my question is, in what way is my argument wrong?

  • $\begingroup$ In terms of mechanical efficiency you may have a point. But in terms of economic efficiency bearings change the wear characteristics of the system, prolonging the useful functioning period of a part of a device & thus prolonging the period between replacement of either the part of the device. Also, a badly worn part can negatively affect other parts which reduces their useful operating period which can increase maintenance costs or result in parts or the device being replaced sooner than would otherwise be necessary. Which can mean more money being spent. $\endgroup$
    – Fred
    Apr 17, 2019 at 3:45
  • 1
    $\begingroup$ The balls in a bearing act like wheels, allowing both the inner and outer races to counter-rotate with no sliding action required. It's exactly like the difference between dragging a tire across the ground and rolling it. $\endgroup$
    – Drew
    Jul 10, 2020 at 8:06

2 Answers 2


Your initial assumption is a bit flawed. You can not just draw the conclusion that all friction is the same. It is like saying all birds i know of can fly thus all birds can fly. Which is obviously untrue. Moreover, continuing this similie, you also imply that there is only one kind of flight.

In reality friction is a highly complicated phenomena that includes a lot of the things we don't deem worth thinking about in our day to day life. In fact if you have ever seen the formula:

$$ F_f= \mu F_n $$

You have seen a physicist saying i have no idea of how it works but seen a simple property. Everything in your question is hidden in the symbol $\mu$ and the formula is approximate. Also by the way the formula is actually $F_f <= \mu F_n$ which is even less useful to answer the question asked.

See, the friction coefficient is highly dependent on not just the material pair, but also the quality of surface finish, surfaces capability to carry the load, how the materials react on the interface, temperature, lubrication and geometry. But it gets better, rolling is way different it has much lower friction so a roller bearing has just the fraction of the friction a conventional sliding contact pair. Even better you can have no solid contact at all. Typically this is achieved by lubrication, but there are even bearings that do this naturally like airbearings or magnetic bearings that don't even have touch anything.

It is not always even about the efficiency increase. Rather about controlling the wear, or some other characteristic. See by having specialized pieces you can change them if damaged, you can spread the load so they last longer, and you can get these cheaper than manufacturing them yourself. The bearing can even have a function in aligning the shaft properly and thereby reduce vibration.


Bearing have rolling elements specifically designed to insulate the two moving surfaces from coming into contact with each other and let the move freely (almost).

There are many types of bearings, ball bearings, cylinder bearing, cone bearing, etc. Each designed to best serve its purpose with least amount of stress, friction, wear and tear. Some resist normal stresses, some lateral and some angular stresses. Wikipedia bearings link

But all of them share one important feature and that is intermediary rolling elements rolling in races or tracks to rotate between the two surfaces and change the direct rubbing into ideally friction less rolling. here is Wikipedia diagram, by I, Solaris2006, CC BY-SA 3.0,

https://commons.wikimedia.org/w/index.php?curid=2496983 diagram, .

  • $\begingroup$ But the balls are rubbing against the races. So to me, it seems as if they add nothing. $\endgroup$ Apr 17, 2019 at 1:04
  • 2
    $\begingroup$ They are not rubbing, they are rolling. If the outer ring of the bearing is fixed, the ring of balls "rotates" at half the speed of the inner ring. (The individual balls rotate about their own axes very much faster, of course). $\endgroup$
    – alephzero
    Apr 17, 2019 at 1:08
  • $\begingroup$ Hmmm so apparently sliding and rolling friction are different and have different coefficients of friction: physics.stackexchange.com/questions/149409/… $\endgroup$ Apr 17, 2019 at 1:09
  • $\begingroup$ Correct! Of course for a simple journal bearing what you said in your OP is basically correct - except that the bearing is a still convenient way to keep the two rubbing surfaces lubricated, without the oil getting everywhere else and making a mess. $\endgroup$
    – alephzero
    Apr 17, 2019 at 1:12
  • 1
    $\begingroup$ @JordanMcBain, there is a huge the difference between rolling and rubbing. Two rigid bodies rolling against each other have no friction, zero. But rubbing against the races and tracks because the forces are nearly orthogonal, they produce very little friction. As we get to close up picture of the ball bearing stresses, it's not rigid, there is a minuscule friction or loss of energy due to the fact that they create a tiny furrow in their path, (and also they get flat a bit too) which fills in immediately after and heats up and absorbs tiny amount of energy, not even close to rubbing friction. $\endgroup$
    – kamran
    Apr 17, 2019 at 2:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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