QR v Thru Axles

[ctznsmith]

I think this is one of those areas where tradition seems to have trumped functionality. If we were to start from scratch and design an axle/frame interface and we assume that a (very) quick way of removing the wheel wasn't a requirement, then we wouldn't end up with what we know as a QR system. I'm not denying that the QR works fine but thru-axles and the benefits they bring (bigger bearings, bigger contact surface and stiffer axle) seem like a sensible step in the right direction with no real downsides (other than backwards compatibility).

How do you get bigger bearings with Through axles? Surely if your hub shell diameter remains the same (which it generally does) and you increase the hole running through the middle your bearings get smaller?

e.g.
M8000 has 3/16 bearings.


Vs

M8010 has 5/32 bearings. (despite apparently changing the hub shell?)

[Bearbonesnorm]

How do you get bigger bearings with Through axles? Surely if your hub shell diameter remains the same (which it generally does) and you increase the hole running through the middle your bearings get smaller?

That was the problem with Isis cranks wasn't it? Bigger axle = smaller bearings = crap.

[Alpinum]

I think this is one of those areas where tradition seems to have trumped functionality. If we were to start from scratch and design an axle/frame interface and we assume that a (very) quick way of removing the wheel wasn't a requirement, then we wouldn't end up with what we know as a QR system. I'm not denying that the QR works fine but thru-axles and the benefits they bring (bigger bearings, bigger contact surface and stiffer axle) seem like a sensible step in the right direction with no real downsides (other than backwards compatibility).

How do you get bigger bearings with Through axles? Surely if your hub shell diameter remains the same (which it generally does) and you increase the hole running through the middle your bearings get smaller?

e.g.
M8000 has 3/16 bearings.


Vs

M8010 has 5/32 bearings. (despite apparently changing the hub shell?)

For many hubs; as long as they're set up for QR, the QR axle runs in a (diameter reducing) sleeve. Once you go TA, the sleave is removed. This way the manufacturers can make one hub body and use one bearing for different axles.

Same goes for 135/142/148 mm hub widths.
Not always the flange is wider, but just a mere spacer used.

[Alpinum]

No perceptible difference in performance imo. If you're a heavy or hard rider you may get a tad less disc rub / 'shing', that's about it. TA certainly a benefit on Sus forks but I read your post as talking about rigids? Functionally it's marginal ime.

And bent axles. Coming back from my last trip I had to saw the axle apart to be able to pull it through my SON hub.
I really should have invested in a new TA SON hub, this would've allowed me to design a laterally stiffer fork. This was the only weak spot of my expedition bike.

Oversized fork, yet the design of the QR was a major restriction in making a fork like the frame (12 mm axle); strong enough to feel responsive even with 4 days worth of water and 30 days worth of provisions, including gear to ascend some of the highest volcanoes on the planet.


It was quite sketchy in places, having fun on descents with a fully loaded bike and a weak fork-hub interface. I'm glad it didn't break, but was afraid it would.
Steering was soft and braking lead to a wandering front wheel.
This was with a very heavily loaded bike, but I remember bending and braking QR on our old DH/FR Bikes back when TA hadn't yet found their way onto mountainbikes.

I'm done with QR.

Something has to flex and if it can't then it will find the next weakest link, usually to terminal effect once all other options are exhausted.

No, there's just more force transfered to the rider. I notice this when I ride my stiffest bike. I deliberatly wanted it to be that way and love the precision, but it is indeed a lot of work to ride it appropiately. It's not exactly comfortable when ridden steadily, but I can go really fast (where I find most fun) safely with it.


And anyone remember the thread about the bolts for todays racks being too weak?
Well, they held fantastically. There was absolutely no problem with them. Different to the QR axle, I had absolutely no problems removing them

[Squirrelking]

And bent axles. Coming back from my last trip I had to saw the axle apart to be able to pull it through my SON hub.
I really should have invested in a new TA SON hub, this would've allowed me to design a laterally stiffer fork. This was the only weak spot of my expedition bike.

Oversized fork, yet the design of the QR was a major restriction in making a fork like the frame (12 mm axle); strong enough to feel responsive even with 4 days worth of water and 30 days worth of provisions, including gear to ascend some of the highest volcanoes on the planet.

Steering was soft and braking lead to a wandering front wheel.
This was with a very heavily loaded bike

I'm done with QR.

That's quite an extreme case though and not really representative of the vast majority of situations the vast majority of riders would find themselves in.

No, there's just more force transfered to the rider. I notice this when I ride my stiffest bike. I deliberatly wanted it to be that way and love the precision, but it is indeed a lot of work to ride it appropiately. It's not exactly comfortable when ridden steadily, but I can go really fast (where I find most fun) safely with it.

I think we're at cross purposes here. When I was talking about stiffness I meant the "flex" in the fork or allowing the fork to twist (moreso dual crown forks) under extreme loading conditions. Allowing the energy to dissipate throughout the fork rather than in one place is just better and results in less stress raisers.

but I remember bending and braking QR on our old DH/FR Bikes back when TA hadn't yet found their way onto mountainbikes.

Can't comment on that really but by the time I was showing an interest around the '99 mark 20mm was already a thing. The 90's were a steep learning curve for everyone as manufacturers couldn't just wing it like they used to and the engineering expertise caught up with reality.

[Alpinum]

That's quite an extreme case though

I know. That's why I recalled riding in my teenage years. More on that bellow.

The 90's were a steep learning curve for everyone as manufacturers couldn't just wing it like they used to and the engineering expertise caught up with reality.

Exactly. DH forks back then (up to about '95/'96), exept for hugely expensive specialities like the Foes, all came with QR. This was a time before RS Boxxers, stanchions still less than 30 mm, MCUs for spring etc. Marzocchi Z1, one of the best performing DH/FR forks at the time, RS Judy DH, 3G Urtho - all QR and it wasn't a good thing. Often we had to saw the axles apart to get them out. More so at the rear. In my case this was in teenage years.

I'm sure today I put far more stress on the material than 24-25 years ago. I go much faster, hit objects harder and yet, even my heaviest trailbike weighs less than 13 kg, its frame incl. the rear shock is more than a pound lighter than eg a Surly Krampus frame (same size) yet it's much stiffer thanks to smart material and engineering that has finally cought up with what folks do on their bikes. Of you don't want a stiff bike, chose a different model/brand - it's so easy today.

Why (go back in time and) use a wobbly system that even in everyday use has it's downsides?

Seemingly even on a road bike...

I'm really glad even my everyday/off road touring bike has 12 mm TA.

When I was talking about stiffness I meant the "flex" in the fork or allowing the fork to twist (moreso dual crown forks) under extreme loading conditions.

I get you. It's a fine line. I remember chatting to the DVO guys about their upside down fork. I was already preferring the precision and feedback of a Fox 40 on my DH bike, but they had an interesting point - with the added arch, they could easily adjust (especially) the torsional stiffness and thought the pre 2012 Fox 40s to be too stiff. Interestingly, Fox eventually made them flex more, agreeing to a development in racing where the feedback was all about balancing stiffness with compliance.

We win as we can freely chose what we want to ride. How much compliance or how stiff.

Not too long ago I got some thin Ti-seatposts for a 3 weeks trip on bumpy tracks. Rather than going with what the frame would've required, I went with a smaller diameter and a adapter sleeve, giving us a more comfy ride.
On other bikes I chose burly carbon rims to make the overall feel of the bikes more precise. It's up to us and with a bit of knowledge we can get exactly what we like.

Allowing the energy to dissipate throughout the fork rather than in one place is just better and results in less stress raisers.

Yes, completely agree. A QR interface will struggle to do that, so lots of the stress is located in one spot only. I've never seen a fork fail thought. It was always the axle.

[Squirrelking]

Hmm...

As I said, I'm not the hardest on my kit (I'm 10.5 stone soaking wet) but I wouldn't say my experience is unique. I used to race the SDA's on a hardtail with XT hubs all round (the Stratos MX6 up front dumped once its oil and the frame it was attached to buckled before the wheel even noticed). In fact I've never owned a thru axle rear. Fast forward to today and I've thrashed the utter life out of the Revelations on my Trailstar and the Hope XC up front is showing no signs of distress. I even raced the Naughty Northumbrian with them last year. Probably got a lot to do with the fact that they have actual spring properties and damping as opposed to rubber elastomers on 1.9" tyres. I sure as hell wouldn't want to try the same on the Jett's I started out with! The engineer in me suspects suspension tech (or a lack thereof) is a prime culprit for axle failure.

Don't get me wrong, I'm well aware that nothing here is anything more than anecdote and far from it for me to tell anyone they are wrong. Just the other side of the coin.

[jameso]

And bent axles.

Big risk with that extended axle type though. A lot of bending load on the unsupported section between the hub shell and the fork end. That's why I didn't buy the extended axle version of the SON for my Jones, not convinced it'd be a good idea with a TA even. Esp under the load you had there! : ) Trip looks good... SON have a 150mm full-width shell hub now?

[Alpinum]

And bent axles.

Big risk with that extended axle type though. A lot of bending load on the unsupported section between the hub shell and the fork end. That's why I didn't buy the extended axle version of the SON for my Jones, not convinced it'd be a good idea with a TA even. Esp under the load you had there! : ) Trip looks good... SON have a 150mm full-width shell hub now?

Yeah, it's not confidence inspiring and by far not just seemed but also turned out to be the weakest link of that build. Yet it was the unsupported area at the fork's dropouts itself. This has in my experience always been the area where the axle would bend using different hubs, QRs, forks…

If you make the dropout thinner to gain more support on the axle, you cause a different problem (fork dropouts too thin).

It's a dog chasing his own tail. You can only get out of it by going with a TA. I can remember some frames in the late '90 to have had some strange sort of proprietary screw axles, but they were reserved for the continous prototype like frames.
Wheelies would've solved the problem too.

Yes, SON have a 150 x 15 mm TA hub.

There's a little story to why I took some risks with sticking with the 135 mm QR:
I didn't like the looks of it either and it had just been announced in the very winter, when I built up my first fatbike - and I was really happy to see it available. Chosen frame and fork had front and rear QR and it was a very lightweight (and wobbly/soft/comfy/etc.) ti frame, so there was a theme going on anyways ;-)

Due to the functionality (like new) of my by now I think 7 years old 135 mm QR SON I had a fork made to go with it, since the fork being less than half the price of the hub, let alone re-building the wheel with the new hub.

Once I test rode the new bike with a load close to what I had on it during the actual trip end of last year, I was afraid I'd kill myself, it was really, really weak up front. I know myself, if there's a nice descent I like to fly. This combination was a bit scary. It was a strange and tense fun on the descents...

Riding the bike "hard" at home (little test ride, to bed in the pads and check gearing etc.) was able to get the QR loose. I worked into roughening the contact areas which may have helped a bit. During the trip I regularly checked the QR and rarely had to retighten it. When removing the wheel to fix the damaged tyre (twice - beyond the maxalami stage) I didn't realise the axle was bent, or it wasn't bent at that point.
Only when wanting to pull the axle out of the hub completely to get the bike ready for the flight I realised the damage.
Was a Hope QR.

Never again. I was asking for trouble, but slipped away without.
I tried to save money and risked getting hurt, with a crappy riding sensation.
Bonus: due to the amount of give, the front was quite comfy on hits.

To make it short:

The whole QR design is too weak for what can be done (doesn't mean it should be done - but go figure, some are curious) by bike. It always has been and still is.

[Squirrelking]

Was a Hope QR.

Without wanting to sound flippant that could well be your problem right there, for the loosening issue at least. External cams are terrible in comparison to even the cheap Shimano skewers and Hope skewers have never had a great rep. Mine only secures once the cam is past the 180degree mark, never had that issue with Shimano.

You're right enough about where the axle bends, classic bending moment. I'd have thought the solution would have been to have more axle in the dropout to spread the force over a larger area.

[Alpinum]

Without wanting to sound flippant that could well be your problem right there, for the loosening issue at least. External cams are terrible in comparison to even the cheap Shimano skewers and Hope skewers have never had a great rep. Mine only secures once the cam is past the 180degree mark, never had that issue with Shimano.

Completely agree. I might not have closed it properly at the beginning.
Gosh... it's not even idiot proof!



Where possible, I ran DT 10 mm skewers in QR frames. More idiot proof.

Well... we're all still alive, can't have been that bad after all

[jameso]

the unsupported area at the fork's dropouts itself

Don't know if the SON has a stub 9mm section of axle end cap or the axle sitting in the fork end, was it the axle itself that bent rather than the whole extended section? Al or steel axle? Still, high loads need more material. OP was talking about rigid gravel / ATB stuff and I think QRs work OK there.

And not wanting to pile on... : ) but Hope QRs won't help, they do have a rep for being weak clampers - that is a sign of why TAs are a better design though, less need for clamp force along the axle.

[Fat tyre kicker]

I use Salsa skewers and have no problems