Talk about torque on the rear freehub.

[roubaixtuesday]

Guy does have a point that torque will break a freehub, so a big rider might well do so. However power is torque x rpm.

So the big guy won't turn over the pedals as fast or as efficiently as the lighter guy who turns the faster.
A diesel engine might produce
250n/m at 3000 rpm

But a petrol

125n/m at 6000rpm
but they are still producing the same HP.

However the diesel will need a bigger drive shaft to account for the torque.

@presta nailed it upthread.

Torque at the hub is independent of sprocket size when climbing a hill at constant velocity.

The torque is determined by the gradient of the hill, the weight of bike & rider, and the size of the wheel. Increasing the sprocket size just reduces the tension on the chain and the torque on the chainset.

Torque = mgrsin(arctanG)

m: mass
g: 9.81 m/s/s
r: wheel radius
G: gradient

Counterintuitive but true. Physics for the win.

 

[rogerzilla]

Modern *hub* geared bikes usually have a minimum recommended chainring/sprocket ratio. 22 x 22 will destroy a Shimano Nexus hub if you ride it up a steep hill. The torque at the hub shell may be independent of the gearing, but the torque at the sprocket is not.

 

[albion]

Modern *hub* geared bikes usually have a minimum recommended chainring/sprocket ratio. 22 x 22 will destroy a Shimano Nexus hub if you ride it up a steep hill. The torque at the hub shell may be independent of the gearing, but the torque at the sprocket is not.

Ebike motors also add torque thus many of the new Nexus hubs are designed stronger for that reason.

 

[bonzobanana]

Ebike motors also add torque thus many of the new Nexus hubs are designed stronger for that reason.

From what I've read Nexus hubs vary hugely in their capacity to handle torque. The simple 3 speed model being by far the strongest and the Nexus 7 being one of the worst as it is actually a more complicated 9 speed hub with 2 gears mapped out. Hub motors massively reduce the load on hub gears (front hub motor) so they can extend the life of Nexus hubs and even provide a long life for hub mechanisms with a large number of gears. However when it comes to mid-drive it gets more complicated. Often companies like Bosch will provide low power mid-drive motors for use with hub gears, it may only provide 40Nm of torque in order to reduce the Nexus wear rate but still means a far higher incidence of hub gear failure and generally a much shorter life for the hub gears. A typical rider provides 20-50Nm of torque so 40Nm is still a huge increase. That's 3x the power going through the Nexus compared to the weakest riders. Ideally mid-drive and hub gears are best avoided especially if you want to use a high power mid-drive motor perhaps a model designed for e-mountain bikes.

Some mid-drive motors are at 100Nm peak torque. They may only maintain that for short periods and it might mean a discharge rate from the battery around 800-900W but still a huge amount of power going through a freehub when you add it to the riders own power. I mean if you had a 32T chainring at the front and 48T at the rear with a 100Nm motor plus a cyclist that can produce 40Nm of power that is 140Nm x 1.5 which is 210Nm of torque going through a freehub. A huge amount of torque. I don't know what freehubs they typically use on e-mountain bikes with high power mid-drive motors but they must be good. To be honest I've read a lot of forums regarding e-mountain bikes and it doesn't come up a huge amount about freehub failure when you compare it to all the postings regarding motor failures and chain snaps etc. Freehubs do seem to survive it seems. So there must be a way of making very strong freehubs suitable for mid-drive motors. Such freehubs would delivery incredible long life surely for standard bicycles.

 

[albion]

It was redesigned with SG-C3001 range made to better suit to ebikes.

However I imagine the 5 speed ebike one is still stronger.

 

[bonzobanana]

It was redesigned with SG-C3001 range made to better suit to ebikes.

However I imagine the 5 speed ebike one is still stronger.

My understanding of the Nexus 7 is it doesn't have a 1:1 gear where the hub gear mechanism just locks to the hub itself which gives maximum efficiency and long term reliability if you favour this gear. The Nexus 8 has this but the Nexus 7 is a more complicated 9 speed mechanism that locks out 2 gears one of which is the 1:1 direct drive gear. If you are careful with the Nexus 8 you can set the gearing up so the 1:1 gear is the most commonly used gear its either gear 4 or 5 depending on the shifter used. The point is with the Nexus 7 you are forced to use one of the more mechanically complicated and less efficient gears at all times. These gears are also a little more vulnerable to the the extra torque of mid-drive motors.

However a quick look at the Nexus 5 seems to show it has a 1:1 direct drive gear as its lowest ratio climbing gear so when you are putting through the power of the mid-drive motor to climb hills you are in the mechanically most robust gear so it seems like the 5 speed Nexus is pretty well optimised for e-bikes from a mechanical viewpoint. I didn't know this previously so seems like a great option for an ebike especially if you are going for a mid-drive motor rather than hub motor.

Personally I would avoid the Nexus 7 and a mid-drive motor but then I'm a heavy rider.

 

[albion]

Yes, and why I mentioned that it is especially suited to an ebike where some loss of efficiency is irrelevant.
Also, as mentioned previously it is near ideal for front wheel drives.

 

[wafter]

Seems @GuyBoden's getting a lot of needlessly snarky comments on this thread..

There are (probably many more than) two ways of looking at this. The first echoes @presta's breakdown, where we want to ascend a hill at a set speed. Assuming bike speed is constant, then force at the tyre's contact patch will also be constant. The product of this force and its radius will give the torque necessary at the axle of the rear wheel. Dividing this torque by the effective radius between the rear sprocket being used gives us the force exerted at the sprocket by the chain, while dividing this torque by the radius between the centre of the axle and the engagement points of the pawls in the freehub gives the total force they have to react.

I'm not going to do a worked example, however with the above caveat that the speed is constant, so is the force at the contact patch, torque at the rear wheel axle and force acting on pawls in the hub. As per @presta's post the larger the rear sprocket, the less force has to be exerted on it by the chain to provide the fixed level of torque we desire.


However what about a more real-world-likely example, whereby our rider gets to the bottom of a substantial hill, drops into the lowest gear, gets out of the saddle and puts in a maximum effort..? Assuming the front chainring is the same size throughout and the force at the pedals courtesy of the rider is constant, so will be the force on the chain. This force acting on a larger sprocket at the back will create a larger instantaneous torque about the centre of the axle, which in turn will translate into larger forces at the pawls in the hub and at the tyre's contact patch.

On less-severe hills this lower gearing may just result in the rider travelling at the same speed as they would have with a larger sprocket; for the same power output but at a higher cadence for a lower, more comfortable force at the pedals. However if we imagine that the hills just get steeper and steeper, eventually we reach a point where the gearing is the limiting factor and the hill can only be climbed with the largest sprocket because the rider cannon provide the necessary force at the pedals to drive anything smaller.

Hence, in this example the larger sprocket does give rise to a situation where it's exerting more force at the hub pawls. This will be the case in any situation that presents enough retarding force against the rider that allows them to input max power to the drivetrain in the lowest gear without causing large accelerations (such as ascending a very steep hill, accelerating a very heavy bike & rider from a standstill or pushing into a massive headwind) will give higher force transmission at the hub's pawl as Guy suggests in his first post.

Finally it's worth remembering that the power a rider can product (which is the product of torque and rotational speed) will vary with cadence. While a rider might hit peak power at a crank speed of say 80-100rev/min, hitting a steep hill - even with low gearing - will likely drop the realistic cadence right down and out of the ideal range for peak power. Hence, gearing lower pushes us back towards this ideal range and allows greater net power input than would have been possible with a smaller rear sprocket and lower cadence it dictates.

 

[the snail]

Surely the maximum torque you can produce is your body weight on a pedal, x crank length x sprocket size / chainring size. So you will produce more torque with a bigger sprocket size.

 

[wafter]

Surely the maximum torque you can produce is your body weight on a pedal, x crank length x sprocket size / chainring size. So you will produce more torque with a bigger sprocket size.

In practice probably a bit more force to the pedal if you're pulling yourself down towards the pedals with the bar... but fundamentally your calcs are right.

 

[rogerzilla]

Surely the maximum torque you can produce is your body weight on a pedal, x crank length x sprocket size / chainring size. So you will produce more torque with a bigger sprocket size.

You will, but you only need enough torque to overcome resistance (wind, gravity, mechanical drag). Peak torque requirement at the wheel only arises when climbing a steep hill. Sprinters apply a huge torque at the pedals, really pulling up on the bars, but will be using a very high gear, so the torque at the rear wheel is modest.

 

[the snail]

You will, but you only need enough torque to overcome resistance (wind, gravity, mechanical drag). Peak torque requirement at the wheel only arises when climbing a steep hill. Sprinters apply a huge torque at the pedals, really pulling up on the bars, but will be using a very high gear, so the torque at the rear wheel is modest.

If you're worried about breaking your hub, then peak torque is a factor? To say that torque at the back wheel is the same for the same climb/load is kind stating the obvious.

 

[Ajax Bay]

slightly concerned that having these modern large rear sprockets (46t+) on a roadbike freehub is creating too much torque for some freehub internal ratchet and pawl designs

So the assertion/concern is for the freehub ratchet/pawls. Those operate (more or less in all freehubs) at about 15mm from the axle centre. The driving force on the wheel has to pass via the freehub and its pawls/ratchet system.
You'll be pleased to learn that you can reduce your "slight concern" to 'negligible'.
A couple of your freehubs failed you, Guy. Don't blame the torque. @fossyant "I'm going with shoot hubs"

Torque at the hub is independent of sprocket size when climbing a hill at constant velocity.

Counterintuitive but true. Physics for the win.

Correct.
When a rider is riding at max power (combo of force on pedal, crank length and cadence) then their effort is exerting a force (Ff) on the bike/rider system accelerating/countering gravity and wind resistance (discount rolling resistance and drivetrain loss).
That (NB max) force (Ff) converts to a torque (622 rimmed wheel with tyre/ground contact area at a radius of 340mm) of Ff x 0.34 (in Nm). This torque is what the spokes have to transmit from the hub.
At the freehub ratchet/pawls, which is the interface that matters in the context of this thread ("too much torque"), at about 15mm from the axle, the force (what breaks stuff in extremis) (NB max) is Ff x0.34/0.015 (in N).
The number of teeth (and associated radius of the sprocket) is irrelevant (to the stresses in the freehub) except inasmuch as sensible gear length choice allows a rider's maximum power to be realised: see also Mathieu van der Poel (but not Chris Hoy: track fixed).

Seems @GuyBoden's getting a lot of needlessly snarky comments on this thread..

My corrections to his needless concern which they then backed up with flawed analysis are not needless. My first comment was straightforward and my second suggesting Pro Team consultancy opportunities was in response to his keeping digging despite others showing him the way (physics).
Here's an Conversation article for you @GuyBoden

 

[albion]

Too much torquing.

 

[Yellow Saddle]

I'm slightly concerned that having these modern large rear sprockets (46t+) on a roadbike freehub is creating too much torque for some freehub internal ratchet and pawl designs.

I've changed over to using Hope freehubs, that are based on mtb design and seem a stronger design.

So, lets talk about torque concerning freehubs with large rear sprockets. What do you think?

Now that the group has pointed out that sprocket size is irrelevant ( in spite of advertising claims), what about the Hope design do you consider MTB-ish? Why is it a MTB design?