QR tension measured. How much load for a quick-release?! Wheel pull-out force predicted....

[@hisdadjames4876]

A truly fascinating solution to a problem Ive never had!

[@PeakTorque]

You'll sleep easy tonight

[@MichielvanderMeulen]

this calls for a patent

[@CrabgrassFarmer]

Hats off to you sir. Amazing content ... Now I'm off to tighten my QR skewers!

[@earthstick]

I read the open QRs with a brass 'seat' under the cam are better than the ones with a plastic seat due to less friction.

[@anneandchasmay3311]

Interesting topic well covered. As an old school rider having grown up with Campagnolo Nuovo Record steel closed cam skewers (and Record later versions) , I have always had TOTAL FAITH in their ability and design. Also I have used 1990’s Dura Ace as well and they are as good imo. I went though a period of trying many types of “lightweight open cam designs” and based on previous experience concluded they were a bad design, and a waste of time for the weight saved. Interesting to see how critical the geometry of the closed cam design in some (including your Giant skewer) is. Just goes to show how right Campagnolo got it back in the 1930’s! 👍

[@F3n0l1g4]

It would be interesting to see the numbers with a drop of oil on the cams before clamping. Like for the video as always!

[@PeakTorque]

good idea

[@kailashblades]

Absolutely! Would love to see track axle nuts too. Difference in force applied with a touch of grease between the captive washer and the nut compared to bare

[@fabit.8734]

I have just put some lube on the cams of my QR s and it s a massive difference!

[@PeakTorque]

You should be careful though, as theres a risk you put too much axial load on the m5 threads. Above 500kg and you're close to yielding a standard M5 thread.

[@Surestick88]

I've noticed that an almost guaranteed fix for a slipping MTB seatpost is to remove the collar and QR, take them apart, clean and lube them then re-install. Re-grease the seatpost while you're at it. The increased clamping force stops the post from slipping with the QR on the collar properly adjusted when, before, the seatpost was slipping even with the QR adjusted so tight it was almost impossible to close by hand.

[@PaulLangmead]

This video was really helpful. I have a fully chromed steel frame from the 90's (F.Moser Leader AX), which of course has the older style dropouts - not horizontal, but slotted - and was having loads of problems with slip in the back wheel with a variety of different. After watching this I got a set of internal-cam QR skewers and have not had a problem since. Thanks!

[@williamtai89080]

Really enjoying your videos! I am an engineer myself and I also work on my own bikes. One thing I always wonder about is the torque specifications printed on different locations of the bike. I wonder how much factor of safety they've put into those recommended torque values. Everyone always says to use a torque wrench when working on carbon parts, but the amount of tension in the bolt, which is translated to a clamping force in most locations on the bike (stem to handlebar, seatpost clamp, stem to steerer tube and etc...) is highly dependant on the coefficient of friction between the contacting surfaces. If you can imagine a very clean vs a rough contacting surface, tightening those two bolts up to the same torque will give very different tensions in the bolt and hence clamping force. Since you already have this strain gauge set up, it would be nice to see a video on this topic! Thanks again for these very technical videos!

[@bantolphbantir9605]

Some of the torque specifications are specifically set to hold parts together under normal loads, but to give way and release under loads that are too high. This is especially important with levers mounted to the handlebars, and the stem mounted to the fork in order to dissipate energy into rotational motion around the bar, instead of the part bending or breaking on impact. Magura is the only manufacturer I've seen to actually this one of the reasons for their torque specs.

[@michaelharvey2537]

Amazingly accurate setup! I always use a bit of tri-flo on an exposed cam qr, you'll feel a big difference in force required without adjusting the nut.

[@geraldmaybebaby1585]

The amount of miles I've done with really slack QRs and the wheels never fell out...:-)

[@adaycj]

Yea, since the brakes go wonky and the tire often hits the fork or stays it makes you wonder how oblivious you need to be to have one actually come apart.

[@universalsprout]

"I'm pinching pretty hard, I can get like 3kg"...which begs the question... What is @peaktorque 's peak torque?

[@lazylizard6705]

It's only limited by how long of a lever he can get :p

[@lfoster7601]

Would have loved to see a Campag one included as that was the original design.

[@paulmcknight4137]

The old Campy chromed steel QRs are bullet proof. Feel a little resistance when clamping on with the lever, and forget it. The trick is hard materials, chrome and steel. Weight weenies like aluminum but it's too soft. Steel on steel tightens very positively by feel. No need for torque wrenches. Never had to replace one of these Campy quick release skewers. They're so nice when flipping off the front wheel. No tools necessary. Thru axles came about when a few riders had front wheels drop out when disc braking that weren't snugged down, a serious liability issue. A solution looking for a problem, kind of like 13 speed cassettes, or these shortened seat stays becoming the standard of coolness. The latter tighten up the rear triangle and shave off grams, but so what? Do they absorb road buzz better than the old larger triangles, or transmit them straight into the middle of the seat tube?

[@bintangcorner890]

Always love your content mate ! Keep educating !

[@denisbassom172]

And now...Thru axles tested? Brilliant video. Thanks.

[@nicdhunter7652]

My mind is blown. Thank you! I was most interested to see just how much force was required to pull out a qr. Very impressive.

[@CezarTrifoi]

Fantastic vid, using a load tester removes ALL the ambiguity. Again.... GREAT work!

[@ottosandrini8480]

Great content! I appreciate the effort! I'd love to see some thru axles on that jig though. It does make sense that they would have less clamping force, but a visual representation is always nice

[@PeakTorque]

Yeah it would but unfortunately i cant fit an m12 axle through the jig. Maybe il try and make some sort of adapter.

[@waynosfotos]

@@PeakTorque yes that would be a little bit of work as the load box needs to be inbetween two plates that the through axle needs to screw into. Plus you need a filler piece to fit nicely between the load box and the one plate. I do think the stiffness thing would be better represented by making long plates clamped by either the QR or the TA and apply force at the end the plate showing the force required to deflect it the same distance. This is why they believe it should be stiffer. The reality is both just clamp the axle. But most can't see this as they believe a screwed item must be better than a cam clamp.

[@adaycj]

Yes, this please.

[@bimmerbent]

Why, just believe the Science. A photo of a Surly Thru Axle is all one needs to understand (a solid Al axle w/ 2 Stainless barbed end caps clamped with 2 M5 Screws)!! They're too many variables and through skewer designs, I have one fork that will not hold any hub that is not a solid axle, Many thru axle skewers are fundamentally flawed and design due to their marketing less weight.

[@andorsn]

Would have been interesting to have a DT Swiss RWS 5mm quick release in the comparison. They look like thru-axles and DT claim they offer higher clamping forces.

[@debelifratar]

How did they sort out the issue of grinding off dropout material while tightning?

[@andorsn]

@@debelifratar the serrated bit is a washer. I assume it's not supposed to move while tightening

[@hdwoernd]

I got some of them and don‘t like them.

[@TomZangle]

Agreed, I'd love to see this test. I have these and they feel like they clamp much tighter than any cam QR I've tried.

[@Tethysmeer]

I have them with titanium "axles". They always loosen and I have to retighten them.

[@tropfen]

amazing video! would love to see more technical talks about different parts of the bike.

[@georgehagstrom4022]

Several years late comment but this video is awesome. I'm curious if the ratcheting wing-nut style skewers like the DT Swiss RWS or the Corki can easily reach similar clamping forces to the Dura Ace. Also would love to know more how much of a practical difference those tensions make, what is the practical difference when you are riding your bike between the high tension achieved on the Dura Ace versus the lower ones on the others? Thanks again for busting cycling marketing myths. This changed how I think about through axles and quick releases.

[@big_icky]

So what I'm getting from this is that you'd be able to make me a power meter

[@PeakTorque]

Correct

[@swites]

I've got a bad dropout design on the rear of a 2008 carbon roadbike I now use as my commuter. One dropout is carbon, one side is alloy, so the wear is one sided on carbon side so wheel doesn't sit straight anymore. To make axle straight I have to set the wheel unevenly in dropout on one side. I use a shimano QR which is the only one strong and secure enough to keep it from slipping and therefore tyre rubbing on chainstays. I can sprint (900W for me as not very strong), go over lumps and bumps and the wheel stays put. Not ideal but a cheap and easy solution that works. Done this for years now. All other QR's I have tried from other wheelset brands slip. I only use shimano QR's on all my wheels(shimano and others) as for me they are by far the best, and most secure. Fit and forget :)

[@Primoz.r]

Do you have an option of testing any 12 mm quick release through axles on that rig? I think you should have a Maxle on your Hightower?

[@avoycendeether8869]

@Peak Torque I have a titanium bike with titanium dropouts on the rear. The rear wheel is always moving around in the dropouts & once came out during a climb in a 24hr TT. This is helping me understand that circumstance heaps. I wonder where Ti dropouts, Al QR nuts would fall on your table! :-D

[@bimmerbent]

Ti is hard material, you need Shimano skewer with steel caps (older skewer), or if you like modern Shimano skewer, try replacing end cap with older steel one.

[@MP48]

Great vid and the numbers are as expected. Would have been ideal if you had a DT Swiss RWS skewer as well. They really rate those but I'm not convinced they're any better than an open cam

[@Ninja_Revenge]

They are MILES ahead of any open cam! Only internal cam DA could compete.

[@earthstick]

I find the test of whether the QR is too tight is if you can undo it without skinning your knuckles.

[@thorn6809]

Afaik the torque force of the qr is highest, when the lever reaches an 90 degree angle and after that the torque force drops slightly when the lever moves to the end position.. It is designed to work like that.

[@palicar]

love the inner tube display stand

[@BenitoAndito]

So it seems like the QR clamping force is a function of three variables. First is perceived indentation/pain on your palm, which depends on the surface area of the lever touching your palm. Second is the friction within the lever cam that makes you press harder for the same clamping force. Third is the leverage of the lever, which depends on lever length and also any mechanical advantage that the cam might give you. It seems like the force doesn't depend on steel vs titanium axle because the stretch under load won't be very different at these forces. Is that a good summary of the most important factors?

[@TheGotoGeek]

The issue with wheels pulling out happens because the braking forces of disc brakes can loosen the QR, particularly on the front. I’ve seen this happen on skewers that I tightened myself, so I know it’s a real problem. The skewers were open cam, so I replaced them with XT skewers, which solved the problem.

[@warpedweirdo]

That would mean the disc brakes are warping one leg of the fork, inducing a change in the dropout's angle about the wheel's axis of rotation, while the other dropout's angle remains unchanged. This would produce a twisting force against one side of the QR skewer assembly not matched on the other. The nut on the QR skewer could rotate about the threads, loosening. Releasing the brake would cause the fork leg to return to normal position, which would also cause the nut to try to rotate back into its former position, re-tightening. Failure of the nut to return to its original position requires slippage between the nut and its dropout, or between the cam-end and its dropout, right? No slippage = no ratcheting mechanism, right? So... Why was there slippage? Did the original QR not clamp with enough tension? How flexy were the forks? The same problem could theoretically occur with thru-axles, couldn't it?

[@bazzbazz8839]

I am not so convinced the braking forces, loosen the QR. I think it is generally because the caliper is mounted behind the fork and almost in line with the axle. this means when you brake the resulting forces try to lever the wheel down out of the drop out. this is generally not a problem with the rear wheel as the rear caliper is likely at the front of the wheel which means the forces push the wheel into the drop out. You can try it by resting the wheel in the drop out without tightening the qr and push the bike forwards (dont get on) and brake. This is lso not a problem with rim brakes because the brake acts at the top of the wheel so the braking forces from the caliper result in a force trying to push the wheel backwards into the drop out where it cannot go.

[@peglor]

@@warpedweirdo It does occur with through axles based on my experience with one of the early Rockshox Maxles. It wouldn't budge even slightly on group spins, but was noticeably loose after doing the same route solo at about twice the average speed - very noticeable as I used to take the wheel off before putting it in the car, so it would be freshly fitted before each spin and removed after each spin. Current models seem to be better though I cycle to the trails these days, so the front wheels aren't taken off regularly. A bolt through axle will give you a lot more warning before the wheel tears out and jams in the fork throwing you into the ground as it has to unscrew completely, so the front end will start rattling as soon as the parts aren't snugged up, or in the case of the Fox/Shimano bolt through axle design, unscew from the non-disk side completely, at which point it'll be pretty obvious something is wrong, but using the brake will still not make it worse as the axle is out in from this side.

[@philipegoulet448]

It would literally never happens with nutted axles! I really wish we kept those :P

[@peglor]

Great content - confirms what I've long suspected having tried other brands and been shocked by the terrible feel, that Shimano QR's are the only ones that are properly engineered. I was one of the commenters who didn't think you'd get 500 kgf tension on a QR, and while I'm mostly correct, it looks like Shimano have shown it's not as far off as I expected. As a qualifier to the calculations you've shown, the pull out force from a disk is multiplied up by the ratio of the disk diameter and the wheel diameter, so braking at 0.5 g, where weight transfer is entirely to the front wheel (Back wheel is no longer carrying weight but not lifting and is doing no braking either) on a 29" (0.73 m) wheel with a 160 mm rotor (0.22 times the tyre diameter), a 100 kg bike and rider combination will see a braking force of 490 N (100 kg x 9.81 m.s^2 x 0.5), which will be a pullout force of 2258 N or 230 kgf. The rider + bike weight of 981 N (100 kgf) is assumed to be applied entirely onto the front axle in this calculation, pushing the wheel into the dropouts. The disk pullout force is only applied to one side of the hub however, so you need to half the pullout force calculated in the video to get a more realistic comparison (And subtract half the rider + bike weight being applied to the axle, so you end up with 180 kgf pulling the hub out at the disk side (230 - 50 = 180 kgf) compared to 120 kgf friction force for the steel hub and carbon dropout option or 180 kgf for the lubed Al frame, hub and QR - so there are cases where this will not be enough). It does mean that the design is mostly safe until you factor in that not everyone tightens QR's properly, not all QR's are well made (As shown here) or maintained with clean and lubricated cam surfaces and, particularly on forks where the legs are flexy enough to do a lot of relative twisting, QRs will unscrew with hard riding, particularly off road as seen by numerous people, including me, where all this blew up in the early 2000's. This loosening in use issue is why so many forks have dropouts with 'lawyer lips' (the popular internet name for the raised surround on the dropout that means you need to unscrew the QR before the wheel can fall out of the dropout). Forces from rear disks don't pull the wheel out of the frame (Except for that one Klein frame in the '90s where the dropouts pointed backwards, which was a pain to change wheels on, but that frame didn't have disk mounts in the first place). The moral of the story really is that Shimano QR's are the only ones still made with some engineering oversight - though I'd like to see a Hope QR tested too - they seem to know what they're doing. The other interesting thing here is that for a given braking force, a bigger disk rotor will produce less force to pull the wheel out of the dropout.

[@markkirschling9340]

You have to be aware that as you are loading up the quick release clamping force, you begin compression loading the axle assembly. This will increase your wheel bearing preload, and can result in additional bearing preload that is undesirable. Increasing bearing wear and rotating resistance. I always check my hub bearing preload after clamping in the bike dropouts. Campy and Dura Ace are my favorite QR’s. You can clamp consistent......not over or under clamping.

[@letsgocamping88]

If you look at the drawing the load runs along the axle not the bearing end caps. So shouldn’t be any preload

[@markkirschling9340]

This will vary based the hub design. The easiest way to verify is quite simple. Mount your wheel in your bike frame drop outs with very light clamping force on your quick release. Spin your wheel and observe how it fee spins. Next clamp your quick release with the normal or full clamping force. Spin your wheel again and observe the free spin. If is spins the same as the light clamping load your good to go. If it slows down significantly you may what to adjust your bearing preload.

[@willo7979]

Per specifications, axle assembly’s locknut is to be tightened around 30Nm, that’s double the specified fork locknut’s torque, hence reduce the possibility.

[@m.talley1660]

I have watched both the leadup and this video and not read the comments so don't know I am repeating a query. Is the surface area of the axle nut contacting the dropout worth factoring? Just prior to the mass-change to thru-axles there were oversized contact axle-end designs with increased surface areas added to forks. I recall marketing that pitched this as a near-equal to through axles in the torsional specs helping provide more suspension fork steering precision. Made by Fox forks if I remember. There was the "Gary Fisher Control Column Front Hub (FCC)" for road/cross as well. Both date from 2010-11. Great content - thought provoking. Cheers

[@robbeelsas]

Those Prime skewers are just rebranded Novatec skewers by the way

[@branes2000]

That's a really interesting and worthwhile video - I've always used Shimano QRs vs the various lightweight ones as they've always felt more secure. Nice to see this proven. I do feel however that it answers a different question to that asked in your other video, which was focussed on 'stiffness'. I would still claim that the thru-axle is 'stiffer' by just about any measure. The clamping force is obviously largely irrelevant in the thru axle case - but it seems clear to me at least that the average thru axle will stretch far less than a steel or titanium (esp.) QR rod, thus stopping the hub from moving (twisting? not sure of the proper engineering term, I was an applied mathematician) in the dropout under lateral loads. Of course, happy to be proved wrong if you can devise an equally comprehensive experiment to dis/prove this either way.

[@hgh96]

They have nothing to so with the stiffness they just clap hub between fork legs

[@theadventurebiker]

Fantastic and very interesting and informative demonstration!!! Cheers 👍

[@kamalfarouque3982]

Had external cam skewers on chrome dropouts. Big wheel slippage on climbs . Could never get enough clamping force. Installed dura ace closed cam skewers and never had a problem again. I’ll never install an open cam skewer on any bike I ride !

[@TheMusicman1654]

This video series has been really awesome! Thanks for the work you do. Do you think in the future you could talk about the merits of the “wave” wheels that are growing in popularity? I.e. the Princeton Carbon Works or Zipp 454 NSW.

[@dario3716]

Thank you for the video. I was asking myself these questions. It would be interesting to analyze the forces acting on the two connection points between the wheel and the frame when cornering.

[@grahamm2015]

Of the top of my head 10% error between prediction and result. Interesting set of results and as ever well explained. Is the takeaway we need a sleeve extender for a qr 😊

[@tednruth453]

I'd be interested to know how well Tulio Campagnolo's first qr levers compare. What about track axles and wheel nuts too!?

[@guidodejong1633]

ISO 4210 - bicycle safety, specifies a wheel pull out force of 2300N (230 kg). This should be relatively easily achievable with a quick release that has some serrated contact surfaces.

[@All4Grogg]

That Tullio gentleman was a smart guy.

[@AndrewWalter]

I have the same lightweight exposed cam QR. Not sure where it came from but I put it on once and immediately thought I'm not riding with that thing on there. I keep going back to a really old Mavic QR from a set of crossmax wheels. They feel like they had some sort of progressive cam but maybe they are like the giant. Really reliable, and easy to use but might not be getting the highest clamping.

[@kvaks3000]

Confirms what Sheldon Brown claimed about QR releases. He recommended the design that is used on the Dura-Ace one.

[@willo7979]

that's my go-to place as well whenever i have doubts.

[@MrGarycoww]

Would be interested in seeing the clamp force on hope qr .

[@hectorhector737]

nice video, how about explaining the difference between 25.4 and 31.8 handlebar

[@PeakTorque]

One is bigger than the other

[@cccpkingu]

@@PeakTorque In fact the 31.8 is the bigger one.

[@assaultedpeanut9]

@hector hector if you don't know why there are two sizes, it's because 25.4mm was an old-school size (25.4mm = 1 inch) but modern bikes have a larger size, I'm guessing for strength (ultimate and fatigue) and maybe stiffness

[@hectorhector737]

@@assaultedpeanut9 thanks sir, but what i like to know if a certain force were applied at both handlebars, at what point failure will come.

[@PeakTorque]

@@hectorhector737 stiffness of the tube goes up with d^4, so it could be said the 31.8 will be a lot lot stiffer, and potentially thinner wall.

[@MozOnBikes]

Very interesting video, thanks for doing the demonstration. It’s made me think of some issues I had with front quick release back in the early/mid 2000s on quite long travel front suspension forks before things moved to thru axles. Back then on some cheaper Marzocchi (Z3 long travel and some dj3) the threads of the skewer would end up deformed and stretched out. Like every few months I’d notice it was hard to undo, inevitable find the thread shagged. Your videos of late has made me think that there must have been significant radial deflection between each lower fork leg only resisted by a little brace with bolted to each side of the sliders. When I finally upgraded to a 20mm non qr front axle on a later model Z1 the stiffness was noticeable, especially on steep shoots into catch berms. I can only assume that the splaying/rotational forces from jumping, drops and generally being a wannabe free ride bro seen at the dropouts were able to exceed the elastic limit of the skewer. Thanks again for the vid.

[@PeakTorque]

Classic forks those! I had dj3s and a pair of Shivers. They were the dream. Don't make em like that anymore

[@peglor]

It was very common to see the steel locknuts on Shimano hubs slowly eating into the suspension for dropouts if they were ridden hard as there is plenty of flex between them. Newer designs have bigger contact areas between hubs, dropouts and axles and stiffer fork lowers to reduce this significantly.

[@jrcp87]

I lost the metal protecting shim on the inside of my rear drop out on my Shiv...I realized it too late and those serrations ate up some of the carbon in the drop out...now I am on the process of modifying a washer to epoxy it and place and fix the mess.

[@Rindertvdt]

@Peak Torque - very interesting insight in the forces at play. A question though; What’s your opinion then on rear pannier racks that are supported by the QR - e.g. Tubus Disco or some of their other racks with a conversion kit. Would that present a radical change in the forces as you presented here since there would be a dynamic load pressing down on the QR axle itself, or would the clamping load be able to negate that? Cheers!

[@bobqzzi]

With disc brakes, a poorly designed drop out and quick release will result in pull-out. There were quite a few cases in mountain bikes with suspension forks, discs and quick releases. On road bikes it's not an issue, but the direction of torque out of the slot generated by the disc brake, combined with uneven bump forces was sufficient to create a hazard. First they changed the orientation of the drop out so braking force no longer vectored out of the slot. Then they went through axle

[@Adam-xo7cv]

I've always used the generalized "if the lever leaves a mark on your hand, it's tight enough." Hasn't failed me yet.

[@torqueytorque8861]

Fascinating video--thanks for sharing! However, isn't there a clamping threshold above which the bearing pre-load is exceeded, resulting in excessive drag and possibly crushed bearings?

[@PeakTorque]

Not really a risk in modern hub design anymore. The bearing preload is passive to the clamping loads in most designs now either with: 1. Toleranced spacer tubes between the bearings' inner races, 2. A dedicated axle with shoulders to lock rhe inner races Both options prevent over squashing of the inner races towards eachother. There are some cup and cone angular contact bearings (like Shimano), that have nut adjusted preload on a thread which can be effected by clamping, but you can account for that by winding off a bit of preload before clamping.

[@torqueytorque8861]

@@PeakTorque Thanks for the reply! I only know from my relatively new DT Swiss 240s hubs which have the shouldered axle you mention, and it SEEMS (by watching the wheel spin down) that the bearings do experience more drag as clamping force goes beyond a certain point which I haven't the ability to quantify. Perhaps others will chime in with their experiences.

[@PeakTorque]

@@torqueytorque8861 the distance between the shoulders is finely toleranced so when clamped up, the internal bearing clearances are closed up and a preset preload is applied, after which more clamping *shouldn't* add any more. With preload there will be slightly more drag, but the wheel will be stiffer (due to removed bearing clearances) and the bearings will last longer! The caveat is, if the shoulder distance is out of tol, you can squash the bearings too hard!

[@torqueytorque8861]

@@PeakTorque Thanks again! And I echo the others who'd love to see you use your tension measurement device to chart other clamping mechanisms including DT Swiss RWS (recommended 15nm torque), as well as 15mm and 12mm thru axle bolts. Keep up the great work.

[@assaultedpeanut9]

@@torqueytorque8861 This sounds like another video topic to cover ;)

[@Maxinatorization]

Archer Sully mentioned this as well, but can you do a first principles look at QR's and disc brakes? (I'm 99.99% sure this is why thru axles exist) I'd be keen to see your take on the leverage that caliper location provides on popping skewers out of dropouts, the rear in particular. Excellent video all the same though! Good to see someone applying engineering principles properly.

[@willo7979]

Yes agree with you Max. when disc brakes was first introduced on mtb, the brake caliper were pretty low, almost to the same height as dropout compared to where they are on fork today. This was where problem starts. Newer fork has the caliper positioned higher and closer to fork for many reasons that make sense: more stout, sturdy, avoid long adapter, suits various disc diameter. Most critically it makes rotational interaction between caliper and hub axle away from dropout direction, and a more secured construction is then adopted from motorcross where disc brakes was carried over: rod through axle, aka Thru Axle. this is evidenced in newer line up from many renowned brands, where disc brake models are now coupled with Thru Axle, rim brakes models are not.

[@luftens]

Did you check what bolt torque you were able to do up each of the QRs to using the lever? It would be interesting to know

[@matteofornero4326]

Very interesting, as always. What do you think about the speed-release axle by Mavic? Basically the dropout where you insert the axle is a normal, open dropout. Does it imply a higher or lower clamping force with respect to a standard thru axle? (supposing the same torque is used)

[@wbhandy]

The lack of friction between carbon and aluminum is precisely the problem carbon assembly paste is meant to solve, I wonder if putting some on the carbon dropouts might be helpful.

[@Incaensio]

I've always used Shimano closed cam skewers. For sure worth the cost and weight.

[@wazzup105]

Extra points for propping the Arduino onto an inner tube :-) Also, normally there's small springs in there... Does that make any difference one way or the other? (or should I just leave them out?)

[@peglor]

The springs are to open the gap between the QR and the hub evenly on both sides of the bike. It makes the wheel easier to fit. The force the springs create is negligible compared to the clamping force, so the only reason to get rid of them is to save weight or if they get bent out of shape. Or if you're running very thin steel dropouts and your hub has a relatively long axle to dropout interface, where the QR might be just crushing the spring onto the hub axle rather than squeezing the dropouts onto the hub. This is very rare in modern bikes, but I have a 2001 DMR frame which this is an issue on with certain QR skewers.

[@markfisher7962]

Also extra points for the eclectic collection of utility spacers used. Perhaps half my socket set is used more as spacers than as wrenches.

[@heksogen4788]

Wow i've never thought they clamp with such force. I mean now i understand why fork ends bend. It's a frigging 300kg of force. How much shear force a 200mm disc brake exert on these QR clamps, and will 300kg of clamping force be enough to not dislodge wheel from QR mount?

[@bazzbazz8839]

Educational video. You mention the clamping load. As I understand it most bicycle manufacturers mount the caliper to the rear of the front fork the braking force on the disc results in a force which is acting downwards trying to pull the wheel out of the dropouts (on the front wheel). How large is this force? I would suggest that if we assume a deceleration rate of 1g it would be possible to work out this resulting force which is likely the one that would pull the front wheel out if it wasn't for the addition of things like (lawyers lips) which a lot bicycles have.

[@peglor]

I calculate it in another comment on here. Search for my username on the webpage.

[@jeanluc397]

Awesome video explanation as always. I was deciding between thru bolts til now so going old school QR for sure now to decide which one. BTW, when do wheels ever pull out in reality? I've ridden 40 yrs and never had a wheel pull out while riding that I know of ??

[@simonchaddock3694]

My rear dropout says 9nm to tighten and they seem fine and the front aswell i wondered wether the force would be sufficient to tighten the 6mm bolt but alls been fine so far

[@MAEBikr]

Nice to see that you have dampened the LCD from the bench with the tube :)

[@henrikmanantancarstensen8849]

What about the Sram 15mm Maxle QR? best of both worlds i think? Also the torque spec for my allenkey rear axle is a lot higher than the 4Nm you used for the 12mm thru axle. Its specced at 165 ib/in (18,6 Nm). But nice to see the difference in clamping force + the importance of adjusting the QR

[@jimmatheson9125]

What about allen key tightened QR skewers (steel vs ti)

[@Max__apex]

That dent is calibrated 😂

[@WeRideSouth]

lol came here to say this.

[@jeidun]

isn't the main difference between the qr and thru axle is that the thru axle can align the wheel better? for disc brakes to avoid disc brake rub

[@Cerv3ra]

That's not what the press says. They say better rigidity. (Lies)

[@FightFilms]

If you mean that the TA eliminates human error in installing the wheel, then yes. IDK why it would align any better than a properly installed QR. Any ideas? I don't experience this problem. Just put the skewer/wheel in the same place every time.

[@willo7979]

well strictly speaking, a large diameter Thru Axle tightened onto forkend will make this Thru Axle rigid and sturdy on the fork. Will this makes the wheel aligns better? Well this question has to be more specific. Are you asking if Thru Axle will align the wheel better to disc brake? Yes and no, because to align better with disc brake requires geometry of forkend to be more rigid, therefore it depends. Or are you asking if Thru Axle will make the wheel aligns better by itself as if be laterally more rigid? Maybe some but not significant as main lateral rigidity is still from the rotating assembly starting from bearings, thru hub, spokes, and rim.

[@carlosflanders518]

That conical washer was in the opposite direction for the final giant skewer test. I think the washer flange will have affected result?

[@tommechelmans9471]

Great content, only one noob question remains from my side.... How much clamping pressure do I really need on my bike? In other words what are the average vertical forces, that during a bike ride on the road, my wheel axes get exposed too?

[@andrewpintar1620]

curious how this compares to a nutted axle type (eg horizontal dropouts for a SS)....to the spreadsheet i guess....

[@mortlow6688]

Could you test the tension of the dt swiss rws qr for me?

[@r.davies2702]

Interesting. Me - immediately gets the GT85 out and cleans all the grease off the dropouts.

[@johnmurray9746]

Great video! One question, though: it appears that your hub pullout forces are based on 4 friction faces (unless I'm misunderstanding), but wouldn't it just be the 2 hub faces? The only other faces would be between the QR and the outside of the fork, but you've already shown that there is no radial load carried by the QR skewer since it makes no contact with the hub assembly.

[@peglor]

Once the hub starts to move it'll apply load to the QR, so to remove the wheel outright, the QR to dropout friction will be in play.

[@johnmurray9746]

@@peglor agreed, if you're calculating maximum pull-out force. But my understanding is that he's really trying to calculate the maximum load that can be carried by friction during normal operation without any relative motion between the hub and fork, right?

[@robertwrightphoto]

possibly the Giant cam (similar to the Bontrager cam) with that weird cam profile is designed to limit the amount of load you can apply to avoid crushing carbon dropouts?

[@PeakTorque]

That's being very kind! Possibly. Benefit of the doubt? This type of pointy cam profile helps it stay closed too.

[@robertwrightphoto]

@@PeakTorque getting over the hump prevents it from opening. right.How many times have you ridden past other cyclists with the skewers done up backwards:)

[@mrtnlstr]

Related, I always wonder how much clamping force is too much on the carbon fork by QR, especially after seeing the carbon blades visibly compressed by the clamping force.

[@craigbesa5672]

Interesting stuff. Re the lightweight, the one that I have is designed to close the cam with light effort then do a half turn to completely tighten. Once done, you can't pull up the cam. Would be interesting if that same principle applies to the one you have there and see the force then applied. Obviously to remove, half a turn to loosen then flick the cam open. Thanks for the content.

[@peglor]

Who makes this? Sounds like the worst of both worlds, given the cam is effectively redundant except as another point of failure, while the threads are doing all the work. Does the end of the skewer slide against the dropout of the frame as you turn it to tighten?

[@robertstrobel7338]

Was hoping for an Allen key style skewer greased and torqued to 7nm. Not drop out greased, but rather greased under the head of the bolt and threads.

[@Dolmar-Rick]

I noticed you put the spacer on the giant skewer on wrong way round so the shim bit was outwards to the qr nut. So looked like the nut was skewiff and not clamped flat...would that affect the clamping pressure?🤔

[@BurnDuration]

She definitely does not want to have a “low pull out force.” Definitely a safety issue.

[@driventomadness117]

Low pull out force could cause decoupling and incorrect re-insertion.

[@BurnDuration]

driventomadness117, also the repercussions of insufficient pullout force will manifest approximately ~9 months after the first load cycle.

[@budbarlang3729]

remember to practice safe riding

[@joelhume2360]

Can you run this test with some 'thru-axles' and share your findings?

[@detmer87]

Great work!

[@appa609]

You'll never actually get u = 1 for Al-Al contact. That number is for a metal-metal contact but any Al surface that's been in air for a while is Al2O3, so your actual contact friction is like 0.5

[@jowjor]

Maybe an other column with a factor to take in account some out of spec parallelism may lower these numbers by a lot?

[@SandyW4196]

Between the ISP and the carbon dropouts, the TCR Adv SL seems like a pretty difficult bike to live with

[@PeakTorque]

Alu dropouts on mine as its discs. Alu dropouts on my adv pro rim too. So all good 👍

[@thelmaviaduct]

Are there any advantages to using a DT Swiss QRS 10mm thru axle vs QR in a QR frame? Are there any safety issues using the above method, with a 1mm alloy wall tube spacer in a 12mm hub and 10mm DT Thru QRS going through it all?

[@adaycj]

Nice video, and thank you. But your original supposition was that the thru axle wasn't better than the QR. Are the thru axle loading videos coming? I was never worried that a QR couldn't provide enough clamping force, the billions of miles each year on them proves they work. You did prove that the QR load is extremely variable. You didn't even test for variables like contamination (not that it is needed), just different components and you are approaching a 80% difference in tension/compression. Without an answer too "how much tension is enough" all we have proved is that the interface is likely to have wildly varying tension with things as simple as user application of a lever, or brand of QR. I have had a QR wheel shift in the dropouts while riding, and it was tightened. It is easy to blame "user error" but what does that mean? I left it floppy loose? I didn't dent my hand when I tightened it? Or I failed to service it? There is no way for a user to quantify those actions ... well ok ... floppy loose is observable. So my question is still this. Does a thru axle hold the tension more consistently over repeated applications? Does the ability to use a torque wrench mostly eliminate the huge swing in tension? BTW you did prove that your bolt chart wasn't in line with reality. The question wasn't "can I approach ideal tension with ideal hardware", it was is the tension equal. It wasn't. Nice work, I'm enjoying this video series.

[@FiveMinuteVelo]

As an engineer I find this very interesting. However, as a cyclist I'm not sure what use information is. I don't see the need to put a large axial force through the axles, particularly the front one, as it's mechanical keying of the axle in the dropout that reacts the forces through the chain and wheels. I tend to use wimpey QR skewers and do them up quite lightly. In 40 years of cycling I've never had a problem. I did however see somebody have a problem once when they were doing their QR up so tightly that they broke it - lucky for them I cobbled something together using a spare spoke and some washers. . Terry

[@PeakTorque]

Hi Terry. That worries me a bit as an engineer you don't want a high axial load! The worry is if the clamping load isn't tight enough the wheels slips and frets in the dropout faces. On modern lightweight bikes these faces are pretty delicate, especially carbon, so you really dont want any stick-slip. Yes, QR dropouts are slot with a stop which can take a shear load of the hub, but in my view priority is a high friction face contact, shear load as a backup. Cheers

[@willo7979]

@@PeakTorque agree with PT, especially when someone pulls handlebar side by side in standing. probably next vid from PT.

[@FiveMinuteVelo]

@@PeakTorque Perhaps I should have been clearer, I don't leave the QRs loose, I just don't tighten them so much they leave a dent in my hand (lol). I'm not an erratic rider and I don't do big sprints or try to get 'air time' so I'm probably not going to test those interfaces that much. Also I am from the time of cup and cone bearings where doing up the QR too tightly could have a detrimental effect on the bearing and races (not a problem with cartridge bearing I admit). I suppose I'm not a fan of tighter is better - tight enough is tight enough - which is why, as you know engineers specify torque values for critical applications. I have however just had the thought (sorry I'm a bit slow) that QR axles are different to through axles (The TA axle is a draw-bar?). The QR axles positively locate in the dropout slots where as the TA true axles butt against the dropout sides and friction does the resisting. This does assume that the TA draw-bar works like a QR skewer. I have not examined a TA draw-bar to check its fit in the frame and hub so I don't know what part it can or does play as a structural shear/bending member (not just a tension member like a QR skewer). Terry

[@HulluJanne]

I'm interested in the situation of clamping a hub with cup & cone bearing with a QR and noticing the bearings getting tighter. What are the parts that are stretching and/or compressing? I've fixed loads of retro MTB's and I've come to the conclusion that to get the right preload on the cup & cone bearings, one should adjust them by trial and error, testing the preload with the QR's as tight as common sense tells me. Seems like one of the most common place of failure of bikes in general are the hub bearings in cup & cone style hubs.

[@peglor]

One technique for adjusting cup and come bearings is to take the spring out of the nut side of the QR and use it to clamp the drive side of the hub to a plate with a QR sized hole. Then the locknuts can be adjusted with the QR at working tension, so not trial and error. Newer hubs with bigger axle diameters are less prone to this, but ti's still there. Cartridge bearing hubs can suffer from this with no way of adjusting it out of the axle end applies load to the inner bearing race. On some of them, the axles can slide through the bearing as the QR squeezes it shorter, so QR load doesn't create any side load on the bearing, but this is an unusual design as it's harder to make.

[@richard0crewe]

It'd be good if you could do a video on building that load-cell. It's the sort of thing people go mad for on Hackaday.

[@allthingsTechrelated]

I was trying to remove my front wheel but it wouldn’t come out until the QR was very loose.

[@Ninja_Revenge]

I use only 2 models of 5mm QR: - DT Swiss RWS steel - Dura-Ace 7900/9000/9100/9200 That’s it. I would shed weight anywhere on the bike, except of cockpit and axels.

[@willo7979]

To the general public like us, PT's series of videos really keep us thinking and discussing: Is Thru Axle really better than QR. thinking along what's evolving in cycling industry, i believe track cycling could be a good starting point. look, if Thru Axle is so good in improving rigidity/stiffness or whatever the marketing guy would like us to believe, this theory has disproved itself in track cycling, because it is not being used in track cycling. therefore Thru Axle must be something to do with something else: 1. disc brakes, 2. pure marketing, and 3. (again) safety (sorry PT, this i insist haha) even better, track bike uses the old school direct bolting of axle onto dropouts.

[@adaycj]

But other than crashes, a track bike is exposed to the least amount of dynamic forces. No pot holes, no bunny hops, no jumps, no rocks ... so why would a track bike be the gauge for thru axles? Look at DH, enduro, AM, trail, XC, and even BMX. All have trended towards thru axles. Maybe only the track types can see through the marketing hype?

[@willo7979]

@@adaycj thanks, but i thought i list down 3 points though. Safety is there accountable for all you've listed: DH, enduro, AM, trail, BMX. Never heard about historically they induce rod through forkends with closed loops to 'increase stiffness'. it's for something in common: disc brakes, suspension forks (mostly), and safety, it's plain simple, safety, you do see that they moved away from QR in those use case and environment. Peak Torque has a vid here discussing how it's being marketed to road bike market with the claim 'Thru Axle increases stiffness': kzbin.info/www/bejne/gX7HpIarjayhn7c and thanks for reminding, marketing hype too imho. by the way, will let track cyclist to comment if they agree 'track bike is exposed to the least amount of dynamic forces'.

[@willo7979]

@@adaycj and you’d pointed out the distinctive obvious; track cycling do not go through potholes, bumps, terrains etc, but still track bike needs high stiffness and high rigidity. Hence the absence of Thru Axle on track bike, on hindsight, is telling us Thru Axle has nothing to do with stiffness, but Thru Axle is there for conditions such as shocks, bumps, accidents and asymmetrical left/right force and impact loading, that’s physically coupled on bikes with disc and shocks. So, considering where Thru Axle is being promoted and marketed (mtb, disc, front shocks), to what they say (eg. Improve stiffness), I’d say seeing is believing; and Thru Axle is meant for mtb, shocks, disc brakes etc., where safety (and legal) is priority. These people do not simply increase stiffness of their bike on offer just because they wanted to, but it is to push sales of new bikes and to solve problem with their newer disc models, yes because there's a problem with their older disc models. For example, if you like rim brakes, and you still wants stiffer frame/fork that comes with Thru Axle. sorry you won't get one. Don’t believe in their bulls.

[@spiloFTW]

I read somewhere that a QR can bent the axle and fuck the bearings if you clamp it too hard.

[@jonathanfeldman5459]

Is the number of friction faces for wheel pull-out not 4? Two on the hub, two on the skewer? One isn't pulling out without the other and the normal reaction is basically the same.

[@PeakTorque]

Yes, 4. That's what I used in the calc unless I'm wrong. Cheers

[@nroose]

Why are you using different spacers for the different QRs? I guess it shouldn't affect the results, but maybe it is? Seems like it would be better to just use the same ones.

[@thomasvmanning]

All of my quick release nuts have a serrated steel insert as the interface too the frame.

[@BenitoAndito]

What makes the reading nonlinear above 1000 kg? Is it the metal plate, a property of the strain gaugue or just a limitation of the Arduino?

[@PeakTorque]

The plate seems to go from uniform elastic deformation to some weird buckling effects that depend greatly on the end constraints. Its still more or less linear but not as accurate.

[@Onigure]

I was searching for "drouputs" meaning in Spanish (my native language) and google corrected me every time.

[@anotheryoutuber_]

i run a fixed gear mtb with a shimano xt quick release to hold my rear wheel with very little to no movement this is an extreme load and can def vouch for QRs ability to hold.

[@petrcivela9413]

What kind of washers would you use between carbon dropout and alu QR?

[@PeakTorque]

Flat ones for sure, not serrated. Flat grit blasted. Look at Look TT stem video for an example