With traditional Quick Releases, the fork or frame attaches firmly through pressure (applied by the QR) to the axle. The axle becomes effectively part of the frame or fork.

With the (more modern) through-axles, the axle is simply inserted through the two pressfit ball bearings.

through-axle image credit

The axle is not at all fixed to the smaller-diameter (inner) part of the ball bearings. What stops the axle from rotating against the inner part of the ball bearings?

Is it simply that the friction at that joint is so much higher than the friction inside the ball bearings? If the axle can rotate, even a little through a layer of grease/oil, the two surfaces would gradually wear out, increasing the chance of rotation at that joint.

  • I'm not convinced it would be a problem if the axle did rotate against the bearing. As long as it only happens a little bit, no significant heat would be generated, thus the grease would stay thick and ensure a reliable film between the surfaces, which means wear is essentially zero. I think it would only become a problem with worn bearings, because then the whole thing would start to act as a simple bushing, and then the grease is indeed not enough for protection. Jul 2 at 22:58

3 Answers 3


MaplePanda's answer covers the vast majority of thru-axle hubs. Usually there's an endcap between the bearing inner race and the thru-axle.

However, there are thru-axle hubs where the thru-axle directly contacts the bearings. It's most common to see on 110x20mm hubs, which at one time was the only common thru-axle size, and then largely became DH/FR-specific once 15mm came out and immediately took over on trail and XC bikes.

Here is a picture of a Specialized 110x20 hub like this: enter image description here

The answer to your question is that such hubs typically use a tubular spacer, which can be seen on the right in this image. It bridges the gap between the bearing inner races. Then there's an endcap on the outside of each bearing that compresses down on the other side of the inner race when the thru-axle is tightened. The result is that the bearing inner races are pinched by the thru-axle, and it is this pinching force that keeps them reliably stationary and prevents the contact between the thru-axle and the inner race from becoming a dynamic surface. The tubular spacer is doing the same thing as the shoulder on the main axle of more common thru-axle hub designs.

It's possible to imagine designs where friction between the inner race and the thru-axle they're in direct contact with is enough to prevent the inner race from rotating on the axle, but a hub manufacturer would likely never do that because they don't have control over the tolerances of the thru-axle, so it would be easy for it not to work correctly. A side note is that the main reason you'd even be considering such a design is if you were making a hub where the bearings lack an axle shoulder or tubular spacer. Some bike hubs have dabbled with this kind of concept, such as QR and nutted axle hubs that use angular contact bearings on a standard axle with a locknut type design to adjust preload, but in a thru-axle design it would be difficult or impossible to do in a way where preload from the thru-axle is properly regulated.

  • I see. So after we tighten the axle and during use, the axle maintains a tension along its length, pushing the system together, and the opposing compression passes through a narrow multi-part outer axle, as thin as the thickness of the inner races. This means that the threads the axle attaches to are much less likely to be worn, since the axle never reaches the end of the nut built into the frame/fork.
    – Sam
    Jul 2 at 17:25
  • But if the threads do get worn, are we then stuck with the end of life of the frame or the fork? Also, I vaguely recall the existence of a third, hybrid, system that uses sealed ball bearings but that is tightened by a QR. How that third system works now makes sense. Indeed, the big advantage with through-axles is not so much that we no longer have to mess with greasy cones and bearings, ...
    – Sam
    Jul 2 at 17:25
  • ...but that a QR getting loose will eventually be catastrophic (delayed by the “lawyer lips”), but a through-axle getting undone is a mild event. The inner races of the bearings might spin a little around the axle, but we still have a lot of time to retighten, and as long as it's not completely undone, the rider will be just fine.
    – Sam
    Jul 2 at 17:25
  • 1
    It's possible for damaged thru axle threads to be the end of a frame or fork. On the frame end there aren't very out there where the threads are part of the frame as opposed to the hanger, i.e. they're replaceable. Generally speaking though they're all big, and overtigthening or crossthreading them would be difficult. Jul 2 at 18:39
  • 1
    As for cup and cone and sealed, there are both types of bearings for both types of hubs. Now that a lot of bikes are coming with thru-axle frames and forks, there are a lot of cup and cone thru axle hubs being put out into the world and there's nothing unusual about them. And as always, all Shimano hubs are cup and cone, thru axle or not. Jul 2 at 18:41

With the (more modern) through-axles, the axle is simply inserted through the two pressfit ball bearings.

Oops! You misunderstood something there because the image shown has a crucial omission. The pictured axle is missing its end caps. When these end caps are tightened down, they press on the inner races of the bearings (usually against a shoulder on the axle), thereby locking the axle to the bearings just like in a QR system. These end caps usually have a raised shoulder which ensures they only clamp onto the bearings' inner races.

enter image description here

These are the end caps from a DT 240s hub. You can clearly see the raised shoulder on the left one.

Don't forget that a through axle provides a clamping force when tightened, just like a QR skewer does. Sometimes the end caps and axle are threaded and can function as described above on their own. Other hub designs (such as the DT 240s) have press-fit or snap-fit end caps, which simply use the TA clamping force to hold onto the bearing inner races.


The thru axle shares a similarity to a QR sqewer in that it effectively squeezes the drop outs to the hubs lock nuts and/or end caps. Rather than threading into a separate nut on the outside of the drop-out, the fork or rear of the bike has to accept the threads of thru-axle.

The removable part in a thru axle hub, like a QR skewer, slips thru the center of the actual axle.


Note that in this diagram, the center tube--the thing many call the thru-axle--is not shown. The axle in the picture (part of the thru axle assembly)is held in place with a locknut on the nondrive side. One must loosen that to remove the axle out that side. The right side has an end cap that can be removed by hand, as can the freehub body. Note the position of the sealed cartridge bearings (#5 in diagram). These are designed with 2 races, an inner and outer race that sandwich the ball bearings. The seals close the gap between the inner and outer races. The inner ring or race is held captive by the pressure exerted by the inner thru axle transferred upon the face of the inner race. The outer race is an interference fit (press fit) into the hub, effectively becoming one with the hub shell. Now the outer axle (pictured) goes thru both bearings as a slip fit--the fit is tight in the hole of the inner race but obviously you can remove it be hand. When the dropouts are squeezed by the inner thru axle, they are forced against end caps of the outer thru axle. This creates a situation where neither the inner nor the outer axle rotates as the wheel turns, but rather the hub, connected inside by press fit to the outer race of the bearings and outside by spoke tension to the rim, rotates with the rim and outer bearing race around the thru axle assembly (the outer axle, inner bearing race and the squeezing inner thru axle) is stationary.

The necessarily tight tolerances of this system doesn't allow much room for misalignment (ideally the bearing faces are perfectly parallel) or play (slight movement off of the perpendicular plane of rotation). Misalignment causes premature wear to the cartridge bearing. Even in a good quality system, after some time one can see where the inner race of the bearings make a mark on the outer thru axle as the miniscule amount of necessary play between the bearing races cause the metal of the axle to become shinier where the inner race frets the metal of the outer thru axle.

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.