Home > Bike Types > Road Bike

IB16: Wheels Manufacturing reinvents the rear hub, pushes axle bearings to the limit

39 Comments
Support us! Bikerumor may earn a small commission from affiliate links in this article. Learn More

wheels manufacturing bicycle hubs designed by jeremy parfitt

See the bearing behind the end cap in the pic above? That’s not the freehub bearing, that’s the axle bearing, which is independent of the freehub. The design is unlike any other, and it’s a project from Alchemy Wheel Work’s Jeremy Parfitt that’s being made by Wheels Manufacturing.

Truth be told, these hubs aren’t exactly new. We saw them six years ago at Interbike, and not much has changed in the interim…except that now you can actually order them. Key updates from the original are a threaded, enclosed bearing cup for the driveside that no longer puts the bearing in direct contact with the freehub body, eliminating pedaling stresses from harshing the ride, and vice versa…

wheels manufacturing bicycle hubs designed by jeremy parfitt

What makes the Orc UL rear hub special is that the driveside axle bearing sits on the very outside edge of the axle, very close to the dropout. In most (every) hub, the driveside axle bearing sides inboard of the pawl/ratchet mechanism, meaning about 35-40% of your hub is not fully supported on the axle. That means extremely hard efforts or unusually stressful bending (think full suspension bikes) can “bend” the hub, causing less-than-ideal engagement and uneven stresses on the system that cause friction and drag. The challenge in putting the bearing on the far outside is, of course, that the freehub body’s insides are much smaller and are usually filled with two bearings of its own.

In the image above, the Driveside bearing sits inside it’s own shell (C), which threads directly into the hub shell (B). The axle (D) runs all the way through, with both bearings resting directly on it. An end cap (E) threads onto the other side of the axle. To assemble it, the freehub’s bearing is pressed onto the silver part of the red shell (B), then the freehub body (A) is slid over it, then the axle bearing (C) is threaded into place to lock it all together.

wheels manufacturing bicycle hubs designed by jeremy parfitt

wheels manufacturing bicycle hubs designed by jeremy parfitt

The freehub body rolls on a single bearing, and it’s completely isolated from the axle.

wheels manufacturing bicycle hubs designed by jeremy parfitt

You’ll need a large allen wrench to unthread the driveside axle bearing’s cup.

wheels manufacturing bicycle hubs designed by jeremy parfitt

The non-driveside bearing is pressed directly into the shell and sits on the very edge.

wheels manufacturing bicycle hubs designed by jeremy parfitt

wheels manufacturing bicycle hubs designed by jeremy parfitt

The front is designed the same way, with bearings pressed into the very outside edges of the shell. Another benefit to their overall design is that the flange spacing has been pushed out as far as possible for better bracing angles (stronger wheels). Claimed weights are a mere 200g for the rear, 66g for the front.

wheels manufacturing bicycle hubs designed by jeremy parfitt

So, what’s changed from the original spotted years ago? This, from Jeremy at Alchemy:

It is a revised design. The small number of failures of the first generation ORC-UL hubs was not due to an inherent problem with the design. It was due to the placement of a bearing shoulder at the base of the post. This created a stress riser which caused the problem. The revised design has eliminated the stress riser and increased the diameter and wall thickness of the post.

To confirm that the problem had been solved, I sent a number of wheels with the revised hub and the first generation hub as a control to ACT labs in California. They ran what they called the European standard test. All of the hubs (including the control) passed the first test. I then asked them to increase the load significantly above the standard (this amounted to a rider weight of 330 pounds). This test caused the control hub to fail, but not the revised hub. They kept running the test until rim failure began. So, the new hub survived a load lest that was significantly harsher than the Euro standard.

Wheels Mfg will make the body, shell, axle, end cap, will use Enduro bearings and source the pawls and ratchet ring elsewhere. They’ll eventually do thru axle and disc brake versions, too, but for now it’s standard 100mm front/130mm rear spacing and SRAM/Shimano 10/11 freehub bodies only. Retail is $440 for rear, $180 front. Choose from silver, black or red and 20, 24, 28 and 32 hole drillings on both, plus 16 and 18 hole options up front.

WheelsMFG.com

Subscribe
Notify of
guest

This site uses Akismet to reduce spam. Learn how your comment data is processed.

39 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Alex
Alex
8 years ago

Sounds like a good design, but for that price you’d expect nicer machining on the hubs themselves. Something the way White Industries hubs look.

STS
STS
8 years ago

“The freehub body rolls on a single bearing, and it’s completely isolated from the axle.”

Certainly not or rather not alone on a single bearing since this won’t work. They also added a bushing right under the pawls. Mavic also does that and customers with their hubs have a lot of problems because they don’t relube that bushing from time to time. And if you don’t relube the bushing will wear out the hub body where it’s supposed to run on a lube film.

Greg
Greg
8 years ago
Reply to  STS

This is the only real problem I have with this hub. It does look, however, that with no chain torque, that the medium sized central bearing could keep the bushing from making significant contact with the hub shell. Only under torque would the bushing be pressed into the shell. Maybe.

Cassmaster
Cassmaster
8 years ago
Reply to  STS

I’ll 2nd that. The number of times I’ve had customers come in complaining of chain suck or drag on mavic wheels is astounding. Many hardly even know that service of the freehub bushing is necessary.

asdf
asdf
8 years ago
Reply to  Cassmaster

Ah, THAT’S why people always had issues with Mavic. I worked in a shop for a few months and never knew.

Morten Knudsen
Morten Knudsen
8 years ago

Flange spacing is the most interesting part of the design

aeroking
aeroking
8 years ago

If you look at early Velomax hubs they used the same thread in drive side bearings holder. Although that design used a Shimano Cassette body it’s still the same execution.

Cleo
Cleo
8 years ago

“That means extremely hard efforts or unusually stressful bending (think full suspension bikes) can “bend” the hub, causing less-than-ideal engagement and uneven stresses on the system that cause friction and drag.” – Great opportunity for a quantification. What is less than ideal?

Ryan
Ryan
8 years ago

Not totally convinced about the engineering when it comes to having just one freehub bearing in what looks to be a central location. I would expect that to cause the cassette and freehub to tweak under torque when in the innermost and outermost gears. Wheels Mfg has a pretty good reputation though, so if they’re willing to put their name on it maybe it’s good. I’d still like to see some aesthetic touches for that kind of price. These hubs look like they were designed in the USSR around 1963. Maybe bad vodka makes them look cooler.

anonymous
anonymous
8 years ago

I thought the entire point of a freehub was to support the dropout end of the freehub with a bearing. I don’t see the point of this. And no, not almost every hub. Shimano cup and cone hubs (and cheap copies) and Mavic hubs have the wheel bearings on the dropout side of the pawls, and Shimano and Mavic make up a huge part of market share.

Frank
Frank
8 years ago

A clever design. Further about the freehub bearing, Extralite also uses a similar system with one bearing and one plastic bushing. Frequent disassembly for lubrication is required.

Smokestack
Smokestack
8 years ago

A very modernized version of a Helicomatic hub. Good thinking.

bc
bc
8 years ago

Easton’s ECHO hub move the driveside hubshell bearing quite far out, but use a huge bearing (basically an angular contact headset bearing) as support for the freehub, instead of a bushing.
http://brimages.bikeboardmedia.netdna-cdn.com/wp-content/uploads/2013/07/Echo-Hub.jpg

myke2241
myke2241
8 years ago

Reinvent…. who is writing these headlines! I see the same old engagement system used for years! = old. With a single bearing and a heavy or strong rider this system is going to bind.

Plus since this is a central mass your not going to feel the weight savings much while riding.

Frank
Frank
8 years ago
Reply to  myke2241

It’s not as bad as that. Please note that under load there is no relative movement between freehub and hub, and that the plastic bushing is supported on a surface belonging to the hub. There is movement between hub and freehub only when coasting, when there is no load.

dypeterc
8 years ago

Kappius still have the best rear hub design yet for outboard bearings. Too bad the design is so proprietary

asdf
asdf
8 years ago
Reply to  dypeterc

Yep, it’s too bad we couldn’t standardize on their design instead of XD. Oh well, it’s the bike industry so we’re more or less obligated to choose sh*tty standards (and change them).

Dylan
Dylan
8 years ago

I bent an axle once riding my road bike (then my only bike) off-road; it was a Shimano ‘Exage’ (i.e. cheap) rear hub from 1987. That would have been in 1991, shortly before I bought my first MTB. I haven’t had a similar problem in the last 25 years.
As for “extremely hard efforts or unusually stressful bending (think full suspension bikes) can “bend” the hub, causing less-than-ideal engagement and uneven stresses on the system that cause friction and drag”, these hubs are not even available in MTB spacing…and if they were, peak load would certainly be higher in a hardtail than a full suspension bike.

TheKaiser
8 years ago
Reply to  Dylan

I agree with most of your points, although the last one is a bit more nuanced than you present. Some loads are undoubtedly higher on a hardtail, however on a hardtail the rear end of the bike is effectively 1 piece. On a full suspension, there is the potential for more flex in the rear end, which means each dropout may move independently, putting more torsional load on the rear axle, and compromising hub mechanism alignment.

Dylan
Dylan
8 years ago
Reply to  TheKaiser

I hadn’t considered the point of potential flex in the rear triangle causing off-axis loading of the dropout faces. There could be some truth in this with suspension designs with pivots in the chainstay or seat stay (other designs like Maestro/DW/VPP however also effectively have a one piece rear triangle). It would have to be a very poor, flexy frame however that allowed any significant relative displacement of the dropouts. In most respects however, any flex in the frame is elastic deformation that is going to be protective of the wheel axle.

paolo
paolo
8 years ago

These are very similar to Spada’s Hubs “Cristallo” , that are on the market since many years. http://www.spadabike.com

gatouille
gatouille
8 years ago

1- Freewheel with middle bearing + bushing under pawl isn’t a durable solution.
2- The link between extended threaded part with bearing (to widen support on axle) and hub shell seems to me too light. This area can bend and be misaligned, in this case the geometry will be bad.
Idea was good but the result seems to me not so good. Some other brand have better solution to answer to this problem.

Mike B.
Mike B.
8 years ago

Quick show of hands, how many posters here are engineers and have seen the actual force load calculations for the design?

gatouille
gatouille
8 years ago
Reply to  Mike B.

Brands don’t communicate on load calculations ; it would be better.

For the design, watch on the left side : you have a large diameter cylinder between flange and bearing, it’s good. The flange could me more outside because there is no brake disc.
On the right side, you have a smaller diameter and a bit longer cylinder between flange and bearing + cut by a threaded assembly, it’s bad.
It’s only geometry, no more.

Bushing on aluminium part (hub shell) is a not durable. It’s better to have stainless steel for example or bearing with inside & outside steel ball/needle track.

JasonK
JasonK
8 years ago
Reply to  Mike B.

Oooh! Oooh! Me! I am! And I have!

Seriously. I attended the same college as Jeremy Parfitt from Alchemy. I then went on to get a master’s degree in mechanical engineering. A few years ago, as I passed through Santa Fe, I stopped by Jeremy’s shop and got a walk-through on his hub design. I was impressed by his thought process.

I’m not sure what you mean by “force load calculations,” but if you’re referring to the influence that flange location and diameter has on spoke tension, then yes, Jeremy did those calculations. They’re actually really simple trigonometry, and I know for a fact that Jeremy has the math background to get those right. (Euclid’s Elements FTW!)

If “force load calculations” refers to something else, what is it? I’d be glad to answer your question if I can.

I’d have bought a pair of Jeremy’s hubs if I had the need and the cash at the time (though they’re not priced unreasonably given the cost to produce them and the cost of competing options).

When Wheels Mfg. was making these hubs for the Alchemy brand, there was no bushing, IIRC. The bushing may or may not be a problem. Mavic’s bushing pretty much is a problem, but that doesn’t mean they implemented their bushing in a reasonable way. In other words, bushings aren’t automatically bad, especially if they’re just acting as backup to a ball bearing. But I haven’t seen the details of the bushing design, so I won’t speculate further.

I’d love to see a through-axle disc-compatible version of this hub. But this hub has had an enormously long gestation period, so I’m not holding my breath.

Ted
Ted
8 years ago

Article says the spoke flanges are pushed out as far as possible but I see plenty more room.
Would have just been better to say pushed out as far as they wanted or reasonable.

Thomas McDaniel
Thomas McDaniel
7 years ago
Reply to  Ted

Actually the front couldn’t be any further from a frame “standards” perspective. And the rear drive side couldn’t be any further either. So the non-drive could be anywhere in space you want it, but that would defeat a logical D/ND balance of any sort. All said, as far as possible.

Sean
Sean
8 years ago

Shimano hubs have had the axle loads at the dropouts since they started making cassette hubs. I thought that was the whole reason for the cassette design in the first place.

Rideifbikes
Rideifbikes
8 years ago

Seems a lot like the Boyd eternity design. Externally they even look very similar.

MaraudingWalrus
8 years ago
Reply to  Rideifbikes

I was thinking they looked like those, too.

Christian Samuelsen
8 years ago

Could an actual engineer explain to me the force loads on _the_ freehub bearing with a 100 kg clydesdale who rides a 75 km commute, 5 days a week up three hills each being about 1.2 km in length with an average gradient of 12% each and an average speed of 28km/hr each way and the resultant effects on the longevity of the bearing?

I ask because I am going to be that guy next summer and would like to know if I should buy _another_ bearing press and an armload of bearings.

Kernel Flickitov
Kernel Flickitov
8 years ago

Considering $600 light weight road hub sets for that much commuting is… silly. You’d fit right in Portland.

Mark Rothschild
8 years ago

“Fear the Vodoo Doughnut”

the other Andy
the other Andy
8 years ago

I have to ask: Why buy these when you can buy DT 240s for significantly less?

Tom
Tom
8 years ago

seems like few of the manufacturers of these outside bearing hubs manage to make the axle/bearing retention system last – there’s a big stress riser where the DS bearing cup threads onto the hub shell. Obviously, Alchemy had problems with this, I suspect Boyd also did with Eternity hubs.

GR
GR
8 years ago

Have thousands of TROUBLE-FREE miles on a set of wheels built around the Alchemy hubs that predate this Wheels Manufacturing version.

Congratulations to Jeremy Parfitt! He’s been overcome many challenges to bring this Made In America hub design to the market.

Cody
Cody
8 years ago

Also nobody see the $440 price tag on the rear hub…?

Kernel Flickitov
Kernel Flickitov
8 years ago
Reply to  Cody

So what.

King R45 rear $450
Extralite Cyber rear $489
DT 180 $555

troykatt
8 years ago

I have 12,000+ miles on my set and they are still the best set of hubs I’ve ever owned.

Subscribe Now

Sign up to receive BikeRumor content direct to your inbox.