Hub torque

[KM2]

Will a 105 Shimano 9 speed road hub be able to withstand the torque generated by a 42 rear cog?

[Cyclothesist]

Yes. Assuming human powered.

[KM2]

Thanks.

[TheBomber]

This is not a question I’d ever considered. Assuming there is a limit to what the freehub can withstand (when human powered), wouldn’t the limit be determined by the overall gear size (eg ‘20 inches’j rather than a sprocket size?

[Cyclothesist]

This is not a question I’d ever considered. Assuming there is a limit to what the freehub can withstand (when human powered), wouldn’t the limit be determined by the overall gear size (eg ‘20 inches’j rather than a sprocket size?

It is the turning force (torque) transmitted through the chain to the hub via the sprocket that matters. An average rider even cycling up a 30% gradient isn't likely to trouble the hub because they can't generate enough power to deliver a damaging torque. If you're a 'Pogačar' doing a maximum effort pulling a loaded trailer something might give - likely the back wheel will spin or the chain will break before the hub.

[Chris Jeggo]

This is not a question I’d ever considered. Assuming there is a limit to what the freehub can withstand (when human powered), wouldn’t the limit be determined by the overall gear size (eg ‘20 inches’j rather than a sprocket size?

Yes, agreed.

[cyclop]

This is not a question I’d ever considered. Assuming there is a limit to what the freehub can withstand (when human powered), wouldn’t the limit be determined by the overall gear size (eg ‘20 inches’j rather than a sprocket size?

Just so.A 42 rear,42 front would give a 1.0 ratio(about 28inches)...exactly the same as a 30 rear,30 front for instance.A 3or 4 speed Sturmey Archer hub gear has strict limits on chainring size as I found to my cost many years ago when a too small chainring destroyed the hub.

[Cyclothesist]

I guess that tells you that the SA hub in question isn't designed for coping with steep gradients. I assume there's a limit to how strong they can make those tiny gears and pawls

[TheBomber]

Thanks all. Maybe I should have considered it when putting a Deore freehub on the tandem, but we never had an issue so that probably confirms we were not a strong pair.

[Carlton green]

This is not a question I’d ever considered. Assuming there is a limit to what the freehub can withstand (when human powered), wouldn’t the limit be determined by the overall gear size (eg ‘20 inches’j rather than a sprocket size?

The chain tension puts a load on the bearings, small chainrings up the chain tension. The gear size is a good first guide but what really matters is the opposing force or load. Riding up 1 in 3’s in a 20” gear might damage something, but riding on the level in a 20” gear never will.

[Brucey]

I guess that tells you that the SA hub in question isn't designed for coping with steep gradients....

quite so. Also you should try to distinguish those failures which are instant and inevitable, given a surfiet of torque, and those which are really a form of fatigue failure, where it seems like the final overload was responsible, but actually the part had been cracked for some time, and would have failed sooner or later anyway. It is easy enough to mistake one for the other.

IMHO the torque can only repeatedly be applied to the hub when climbing a steep hill or with a heavy load. Otherwise the effect of just a single pedal stroke (assuming the brakes are not set to drag) is to accelerate the bike so much that a gearchange is required. It is somewhat debatable whether this single pedal stroke ever reaches full force or not.

[Brucey]

BITD a lot of freewheel makers used to fit their products with two pawls, fitted 180 degrees apart on the body centre. These fell into two main categories. 1) with an odd number (n) of teeth built into the outer part of the freewheel body, and 2) where 'n' is even. In the latter case, the two pawls are meant to engage simultaneously, but hardly ever do, especially in a new freewheel. 1) includes such freewheels as Normandy, and some versions of Sun Tour's Pro Compe model. Such freewheels had a lower torque rating, but about double the number of clicks. Most shimano freehubs (and their copies) have two pawls set 120 degrees apart within an 18 toothed ring. This gives just 18 clicks per rev. and a good torque rating if the parts were machined properly and both pawls are engaged. Unfortunately it seems impossible to guarantee the latter in any new 2-pawl device, such are the tolerances required. Many a race has been won by a canny rider using an old 'known good' freewheel body equipped with new sprockets.

The result of having 180 degree pawl separation and worn freewheel bearings is often that the freewheel becomes noisy in the higher gears, usually with a terrible clonking sound with two clonks per wheel revolution. By contrast if the pawls are separated by 120 degrees and they are both engaged, a biasing force is generated having the effect of quelling any rattles from loose bearings. Should only a single pawl engage, the freewheel body will start to move from side to side, if there is sufficient free play in the freewheel bearings. The free play gives every opportunity for the second pawl to engage, whereupon they both share the work.

For this to always happen, the movement via the free play needs to be enough to allow for any likely discrepancy in pawl length as well as any lack of concentricity between the freewheel bearings and the teeth of the ratchet, else single-pawl engagement can persist.

[andrew_s]

You could argue that low gears are better,

The same rider riding up the same hill at the same speed will give the same hub torque regardless of gearing (torque x wheel revs = power)

However, twiddling a low gear allows smoother pedalling than getting out of the saddle and stomping in a higer gear, so although average hub torque is the same, the peak torque is higher in the higher gear than in the low gear, and it's peak torque that's going to break something.

I dare say that gearing limits are a more politically acceptable way of restricting the use of hub gears on steep hills than rider weight limits would be.