High-strength freewheels or freehubs

boudin

100 mW
Joined
May 4, 2022
Messages
44
Any hub can probably handle a high power mid-drive once or twice, but what hubs are you all using for these multi-kilowatt motors to dump 100s of Nm of torque into a wheelset, especially once gearing is considered?

The 3-pawl "Quanta" Ali special hub that came on my commercial bike decided that two uphill snow rides were enough and sheared 2 teeth and 2 out of 3 pawls. While I expect a repair on warranty, I doubt that the vendor is going to replace them with anything but the original equipment, which means it's just a failure waiting to happen again. I'm starting to research what might be a useful replacement

I'm somewhat locked into the current 190/197x12mm axle form-factor, which appears to limit my options significantly. Does anyone make an all-steel hub at that scale, or at least an aluminum one that allows replacement of the wear parts?
 
It sounds like you may be looking for a "freehub", not a hub that can use a freewheel (which is what your title wording sounds like you want).

I don't know what's available in that size that would be any better than what you have. :(

Is there a reason you can't use a "standard" hub that is threaded for a freewheel instead? If you did, then at the least you could take it off and replace it in the field if it did break. Would be annoying to have to carry around the wrench, tool, and freewheel (with it's sprocket stack), but at least you could do a field replacement, which you can't do with a freehub (unless you carry a whole spare wheel, or do a complete relacing in the field, or carry the tools/parts to replace the freehub inside the hub body).

FWIW, if you had to, you can *make* a "standard" threaded hub of the right size yourself, if you can't find one, using a steel hub body and cutting the central tube, then slipping a larger (or smaller) diameter tube of the right diameter onto that tube, then welding the set together at the right flange width for your wheel. Depending on the bearings it has you might need to replace them with ones that can handle the 12mm axle.


Then it's mostly up to finding a freewheel with the right sprocket stack for you (or that can be restacked with the ones you want), that is also strong enough to handle the loading internally (or easily and cheaply replaced when it does fail).
 
My terminology is plausibly incorrect, but I find the components I disassembled and on component sites like modernbike.com fairly entangled and apparently made a confusing portmanteau.

In my situation, both ratchet teeth and pawls have broken, so they both need replaced. I could just replace the freehub body (or perhaps the pawls if I can find replacements), but that leaves broken ratchet teeth. The ratchet teeth appear to be integral to the hub - either milled or press-fit. As such, it seems I would have to replace the hub to repair the teeth.

I'm not finding just "standard" hub bodies for sale (no freehub, but it's plausible I'm looking for the wrong thing. I'm only seeing hub and freehub combos, or standalone (seemingly manufacturer-specific) freehub bodies.

Mainly, I'm seeking manufacturer and product line recommendations, ones designed to (or at least shown capable of) handling the hundreds of N-m of torque that ebikes can produce. For example: I've seen opinions that the DT-Swiss 440 is one of the strongest, but they don't appear to be making those any more. Is the 350 "enough"?
 
The DT Swiss 350 Hybrid hub is one example of a hub that's designed for supplementary motor power. I'm not aware of it being offered in a super wide format.

It doesn't make sense to feed motorcycle level torque into a pedal drive system. You are finding out why. That said, Quando is a reputable but very price-point oriented manufacturer whose version of cheap and effective is definitely cheap, possibly effective. Fake Quando isn't reputable and probably not effective even for pedal only applications.
 
machanic said:
You might try the gas motored bicycle suppliers they have heavy duty hubs for builders pushing 10+ hp.

Do any of those bikes drive through the gears when they get to those power levels?
 
When you have the right amount of motor motor, the desire for gears concludes.
 
Ask ES member LightningRods which freehub he considers to be the strongest.

If that selection does not hold up for your application, then you might consider foregoing gears for the motor as LFP has suggested.
 
Chalo said:
machanic said:
You might try the gas motored bicycle suppliers they have heavy duty hubs for builders pushing 10+ hp.

Do any of those bikes drive through the gears when they get to those power levels?

Yes! although not always, some drive through a left side one speed chain sprocket and some use a jackshaft setup to transmit power to the right side deraillier.
 
boudin said:
I'm not finding just "standard" hub bodies for sale (no freehub, but it's plausible I'm looking for the wrong thing. I'm only seeing hub and freehub combos, or standalone (seemingly manufacturer-specific) freehub bodies.
By "standard" hubs, I wasn't referring to freehub hubs at all, but to the "old" standard threaded hubs used for thread-on freewheels (which come with the sprocket stack already on them, though most of them you can unstack and restack it differently or change to other sprocket sizes).

This page
https://www.sheldonbrown.com/freewheels.html
explains the difference visually as well:
freewheel-vs-k7[1].jpg


Freehubs will typically include the freehub mechanism itself, because from what I have read by others with such problems, they *are* manufacturer-specific.

I've only used a handful of freehub-cassette type bikes, or separate wheels for my own projects, almost all of the freewheeling applications I have had used thread-on freewheels, which gives me a lot more options for solving problems and making repairs.


However...I don't know of any standard bicycle parts system that is designed specificially to take the kind of motor power level you are after. (which is one reason some people using these types of drives specifically use thread-on freewheels, as that makes the repair process simpler and cheaper--you don't have to build a new wheel everytime the freewheeling mechanism is trashed).


I do heavy cargo hauling, which needs significant power to quickly accelerate in traffic (don't know the nm numbers, but the watts is around 4000+ during the few seconds to get up to 20mph; higher the more cargo there is). I presently use hubmotors, but want to move them out of the wheels to make better stronger wheels and enable me to change the gearing ratio between motor and wheel to whatever the single ratio is that works best for that motor, while still using a large diameter wheel for better ride quality. I don't know that I can get a reliable freewheel for this purpose, so I will probably not use one.

But I have another drivetrain design that would use a freewheel thru a shiftable-gears drivetrain (specific to each motors; the pedals would have a separate one to the front wheel for redundancy), and from what I can find for available parts vs reports of using them like this, I would just need to carry parts and tools to change the freewheel in the event of failure. :/
 
boudin said:
My terminology is plausibly incorrect, but I find the components I disassembled and on component sites like modernbike.com fairly entangled and apparently made a confusing portmanteau.

In my situation, both ratchet teeth and pawls have broken, so they both need replaced. I could just replace the freehub body (or perhaps the pawls if I can find replacements), but that leaves broken ratchet teeth. The ratchet teeth appear to be integral to the hub - either milled or press-fit. As such, it seems I would have to replace the hub to repair the teeth.

I'm not finding just "standard" hub bodies for sale (no freehub, but it's plausible I'm looking for the wrong thing. I'm only seeing hub and freehub combos, or standalone (seemingly manufacturer-specific) freehub bodies.

Mainly, I'm seeking manufacturer and product line recommendations, ones designed to (or at least shown capable of) handling the hundreds of N-m of torque that ebikes can produce. For example: I've seen opinions that the DT-Swiss 440 is one of the strongest, but they don't appear to be making those any more. Is the 350 "enough"?

There's cassettes, and there are freewheels. Both are similar looking (a gear cluster), but they mount to the hub totally differently. If talking big mid drive power, 160nm Bafang Ultra or BBSHD for example, you are very likely looking for (or should be) a cassette drive (NOT a threaded on freewheel). The biggest difference is in how they are supported on the axle. Any freewheel I've ever seen is supported by 1 bearing, and it's not even centered, being closest to the larger gears in the cluster. The cassette style hub will have a pair of bearings supporting it in a manner that makes it much stronger.

Then there's the ratchets you already know about.

The DT Swiss 350 hubs are generally considered a quality hub that will handle big power.

When purchasing, make double sure it's (or which ever one you're considering) been drilled for the right number of spokes! There are 32 hole, and 36 hole (often described as 36h). Without getting into which is better, it's going to need to match the number of holes in your rim....
 
AHicks said:
There's cassettes, and there are freewheels. Both are similar looking (a gear cluster), but they mount to the hub totally differently. If talking big mid drive power, 160nm Bafang Ultra or BBSHD for example, you are very likely looking for (or should be) a cassette drive (NOT a threaded on freewheel).

I don't agree. There are two very big advantages to threaded freewheels. First, the ratchet sits at a larger radius than the threaded portion of the hub, so forces are much lower than they are in a typical freehub. Second, every time you toss out a worn freewheel, you get a brand new fresh ratchet, which you don't when you swap cassettes.

The main drawbacks of freewheels have to do with the hub axle, which is an addressable problem. You don't have to tolerate a 10mm threaded axle when there's 19mm of space you can use for that purpose.

There's also a contextual problem with freewheels. They're used mostly on cheap bikes anymore, so most of them are cheap and awful. But with like quality, they are certain to be more robust than freehubs just because they have larger diameter and better mechanical advantage.

For my part, I've had great luck with ancient Suntour AG 14-38t 5-speed freewheels. The only time they seem to break or have any problems at all is when you go to remove them and the shabby 2-prong remover interface won't cut it.

Any freewheel I've ever seen is supported by 1 bearing, and it's not even centered, being closest to the larger gears in the cluster.

My favorite freewheels definitely have two rows of bearings. I wanna say that's also true for other multi-speed freewheels I've opened, though I usually only do that when there's no other option.

EDIT:
Yes, also true for others:
[youtube]T_vRbBRPr3c[/youtube]

Freewheels are only designed to carry load when they're locked up and not rotating. It's assumed that they won't be under chain tension when they're freewheeling, which is true for pedal bikes and the rear freewheels of mid drives. When they only spin unloaded, it doesn't matter.
 
12-C said:
https://onyxrp.com/tech/

Even at the stroke-inducing price of $500 and up to 6 weeks' lead time, that's awfully compelling. Silent, four sealed ceramic bearings, a 7075 housing and dual sprag clutches working against stainless steel friction surfaces. Moreover, they will sell you every single part individually? I suppose they had better for the price of a whole bike in many places.
 
12-C said:
I don't think it would ever need parts 😁

You hope! But the fact they use ceramic bearings means they're more about hype than engineering. Still, folks seem to like them. Might be the sunk cost fallacy at work.

For what it's worth, I've never seen a roller or sprag clutch rated for as much torque as a freewheel ratchet of equal size can tolerate with no problems at all. You have to use a bigger clutch to do the same job as a ratchet.
 
Chalo said:
You hope! But the fact they use ceramic bearings means they're more about hype than engineering. Still, folks seem to like them. Might be the sunk cost fallacy at work.
Plausible. Could also be that they decided on a price point first and are using the ceramic bearings to help a potential customer justify the price in their head. More features, regardless of their actual utility, make us biologically happy. See the Swiss Army knife if you've any doubt
 
Am I doing my math wrong here?

Say you have a 100kg human standing unsupported on the end of a 170mm crank arm at the 3:00 position, otherwise unsupported. That is roughly 166N-m of torque? 100kg * 9.8m/s/s * 0.17 == 166.77, which mirrors what you get if you do the same calculation for ft-lb.

If so, adding another 90-160N-m from a mid-drive motor puts torque values approaching 250-300N-m. Add in gearing advantages, and you could see 500N-m at the hub?
 
Chalo said:
For what it's worth, I've never seen a roller or sprag clutch rated for as much torque as a freewheel ratchet of equal size can tolerate with no problems at all. You have to use a bigger clutch to do the same job as a ratchet.

Onyx is pretty coy about what kind of torque they'll tolerate, so that underscores this observation. There are ancient posts here implying that they handle more than e-bikes produce, but when I just asked they state that 500N-m is definitely off the table. The only further commentary they offered was that "standard" mid-drive motors from Shimano, Bosch, et. al. are within their envelope.

Given that those motors appear to produce 80-90N-m of torque, my WAG is that they're in the 250-300N-m range to allow for human variation and safety margin.
 
You might be over thinking it? Any high end one will do that is focused on DH riding or the like; not a weight weenies hub.

I’m about 180lbs with BBS HD @ ~47amps on a downhill fame, so perhaps a little less power than mentioned about but not by a lot. 46t front 11-36t rear.

In the U.K. we have Hope Hubs and I have three all purchased second hand. I have rebuilt / regreased the ring and pawls a few times and they look like new. I also have Onyx Classic because it has more/bigger sprag than their Vesper model.

The Onyx is amazing, it’s not easy to justify the cost and I would not try but it improved the feel of the bike a lot due to the instant engagement. As you throttle off and on there is no driveline ‘shunt’ while waiting for engagement and it is lovely :)
 
There no way that anyone is putting 250nm through a bicycle... not even relevant.

100kg on .175m crank produces 171nm , assuming 1.7 gearing that's 292nm , assuming 26in wheels it's about 96mn at the wheel...

Maximizing drive torque for acceleration would be at the limit of wheelie-ing which depends on geometry, rider body mass, traction... and if we're going to play these scenarios we need to remember that is on bicycle components...

My 2c
 
12-C said:
There no way that anyone is putting 250nm through a bicycle... not even relevant.

I have a bike with 225mm cranks, and I weigh close to 170 kg. That's about 375 Nm at the crank when I stand on the front pedal. The small chainring has 42 teeth, while the large rear sprocket has 32 teeth. So that reduces the torque at the wheel to about 285 Nm in the same circumstances.

If that bike had motor assist, I'd probably have used shorter cranks and taller gears, so less torque from pedaling. But if the motor were a crank drive, it could still be easy to exceed 250 Nm at the rear wheel sometimes.

Wheel diameter isn't a factor in calculating torque.
 
Chalo said:
There are two very big advantages to threaded freewheels. First, the ratchet sits at a larger radius than the threaded portion of the hub, so forces are much lower than they are in a typical freehub. Second, every time you toss out a worn freewheel, you get a brand new fresh ratchet, which you don't when you swap cassettes.
One thing I found very confusing about threaded freewheels was the thread specification. If one refers to the Sheldon Brown page it lists 5 known thread sizes. Frewheel Threading.png
https://www.sheldonbrown.com/freewheels.html
Leaving aside the rather rare French version and the smaller Metric BMX thread, the first three all use the same thread pitch with a minor variation in the ISO standard 1.375 inch diameter.

However when one starts searching for threaded freewheels (most being manufactured by our good friends in SEA) that thread specification some how translates to "34 mm inner diameter". I guess that is the "Chinglish" standard.

The other problem being the typical bicycle industry "we have no bloody standardization" wide assortment of proprietary tools required for their removal.
 
LewTwo said:
Leaving aside the rather rare French version and the smaller Metric BMX thread, the first three all use the same thread pitch with a minor variation in the ISO standard 1.375 inch diameter (which is coincidentally the same as the ISO BB standard thread size).

Italian threads have a 55 degree form rather than 60 degrees. Functionally speaking, they're interchangeable, but a little more likely to damage threads when you mix standards.

Anything other than 1.370" x 24 tpi British threads is obsolete and something you'll only find on old equipment.
 
12-C said:
There no way that anyone is putting 250nm through a bicycle... not even relevant.

100kg on .175m crank produces 171nm , assuming 1.7 gearing that's 292nm , assuming 26in wheels it's about 96mn at the wheel...

Maximizing drive torque for acceleration would be at the limit of wheelie-ing which depends on geometry, rider body mass, traction... and if we're going to play these scenarios we need to remember that is on bicycle components...

My 2c

To be fair, I'm not maximizing for acceleration, I'm just trying to make sure that my bike is both rideable without assist (if I'm in the very plausible scenario where the motor or battery fails miles from civilization) and will survive the maximum reasonable input from both motor and human.

You assert that the 292nm of torque is 96nm at the wheel. What i'm interested in is not the moment at the wheel's edge, but what I'm doing to the hub itself. I'm not an engineer and may have a misunderstanding of physics, but does the relationship between the diameter of the wheel and the sprocket diameter actually affect the forces that the hub experiences?
 
Back
Top