Rohloff Speedhub

Chachi said:

I just put together a new bike using a Rohloff Speedhub 14-speed internally
geared rear hub. One basic question comes to mind now that I've ridden the
bike a few times. Why does a bike with a Rohloff Speedhub need such a
formidable torque arm while a convential rear hub/derailleur equipped bike
does not?

Our Nexus 8 speed hub has a similar arm, but that also has a roller brake!

The Rohloff one is overbuilt - look at a Thorn with the purpose-made
Rohloff dropouts and you can see that such a long arm on other bikes is
unnecessary.

Chachi said:

I just put together a new bike using a Rohloff Speedhub 14-speed internally
geared rear hub. One basic question comes to mind now that I've ridden the
bike a few times. Why does a bike with a Rohloff Speedhub need such a
formidable torque arm while a convential rear hub/derailleur equipped bike
does not?

Conventional hub merely rotates on the axle. Splines transfer the torque from the cogs to the hub bypassing the axle. The only torsional load is from friction.

Gear hub generally needs to use the axle as a torsional anchor upon which the gears turn.

Tom Ace said:

To step torque up (or down), some part (sun, or planet carrier,
or annulus) of a planetary gear stage needs to be kept from
rotating; a QR or axle nut isn't enough for that purpose,
hence the torque arm.

So why can all the other gearhubs make do with a simple keyed washer to
prevent the axle from rotating?

Jasper

Jasper Janssen said:

So why can all the other gearhubs make do with a simple
keyed washer to prevent the axle from rotating?

With its wide range, the Speedhub needs more torque
support than other internal gear hubs do.

In first gear, the Speedhub drives the rear wheel with 3.6 times
the input torque at the sprocket. Thus the torque arm supplies
a torque of 3.6 - 1, or 2.6 times the input torque.

Compare that to a Shimano Nexus 8, whose first gear drives
the wheel with 1.9 times the input torque. The axle needs to
be held with a torque of 1.9 - 1 = 0.9 times the input torque.

Even so, the length of the Rohloff arm may well be overkill.

Tom Ace

Chachi said:

Are the direction and magnitudes of the forces so significant that the
torque might move the axle in the dropout?

Which direction might the axle move, forward or back?

Neither. Torque (wants to) move the axle around, not forward or back. It's
chain tension that wants to move it forward, but that shouldn't be a real
problem as long as you tighten the nuts sufficiently.

Jasper

quote='Jasper Janssen'On 4 Nov 2005 21:52:35 -0800 said:

To step torque up (or down), some part (sun, or planet carrier,
or annulus) of a planetary gear stage needs to be kept from
rotating; a QR or axle nut isn't enough for that purpose,
hence the torque arm.

So why can all the other gearhubs make do with a simple keyed washer to
prevent the axle from rotating?

Jasper[/QUOTE]

In the case of my Sachs 7 speed, that keyed washer prooved to be seriously inadequate.
Dan

quote='Tom Ace'Jasper Janssen said:

So why can all the other gearhubs make do with a simple
keyed washer to prevent the axle from rotating?

With its wide range, the Speedhub needs more torque
support than other internal gear hubs do.

In first gear, the Speedhub drives the rear wheel with 3.6 times
the input torque at the sprocket. Thus the torque arm supplies
a torque of 3.6 - 1, or 2.6 times the input torque.

Compare that to a Shimano Nexus 8, whose first gear drives
the wheel with 1.9 times the input torque. The axle needs to
be held with a torque of 1.9 - 1 = 0.9 times the input torque.

Even so, the length of the Rohloff arm may well be overkill.

Tom Ace[/QUOTE]

The torque transmitted to the axle shaft is inversely proportional with respect to the input torque as it is to the hub shell. While input torque is multiplied to the hub shell most in the lowest gear, it is multiplied most to the axle shaft in the highest gear.
Dan

Per Chachi:

Quoted message said:

Is it wise to use the Rohloff OEM2 axle plate in lieu of the torque arm on
this frame?

http://www.rohloffusa.com/frame.htm + "Products" + "Speedhub 500/14" +
"SpeedBone" (or) "Axle plates/torque arms"
--
PeteCresswell

Dan Burkhart said:

The torque transmitted to the axle shaft is inversely proportional
with respect to the input torque as it is to the hub shell. While input
torque is multiplied to the hub shell most in the lowest gear, it is
multiplied most to the axle shaft in the highest gear.

I don't follow what you're saying.

I gave the examples I did (lowest gear) because they
required the greatest torque support at the axle.

In its highest gear, the Nexus 8 steps torque down to
0.6x the input torque. The axle needs to be held fixed
with a torque of (0.6-1) or -0.4 times the input torque.

Tom Ace

Chachi said:

I can understand that answer, thank you. I'm also hoping that someone with
specific knowledge about the Speedhub will answer the following...

The bike I just built up is a Surly Karate Monkey with horizontal dropouts
and International Standard disk brake mounts. I think the currently
installed torque arm is fugly.

Is it wise to use the Rohloff OEM2 axle plate in lieu of the torque arm on
this frame? I'm not so much concerned about the disk brake mount since the
bike is built like a tank and meets the Rohloff stated requirements. I'm
more concerned about the horizontal dropouts.

You're right to be concerned about the horizontal fork ends, because
depending on where you set the axle, the distance from the axle to the
disc brake mount will vary. The OEM2 mount is intended for use with
vertical dropouts, where there's a fixed distance betwixt the axle and
the lower disc mount.

This might work, but would probably limit your usable range of axle
positions in the fork ends.

Quoted message said:

Are the direction and magnitudes of the forces so significant that the
torque might move the axle in the dropout?

Not likely.

Quoted message said:

Which direction might the axle move, forward or back?

The axle's torque varies in direction, going one way in high gear and
the other way in low gear. I don't see how the axle could actually move
back, though due to the chain tension.

The risk is that the axle could _rotate_ causing the axle nuts to loosen
up, if the OEM2 axle plate doesn't have a snug grip.

Quoted message said:

The hub I have is a "TS" or Touring Special with solid axle and nuts (no
quick release).

Dan Burkhart said:

The torque transmitted to the axle shaft is inversely proportional
with respect to the input torque as it is to the hub shell. While
input torque is multiplied to the hub shell most in the lowest gear,
it is multiplied most to the axle shaft in the highest gear.

That's true of hubs such as the SRAM that have direct drive in the
middle of the range.

It's not true of the Rohloff, though. With the Rohloff, direct drive is
11th out of 14 gears.

Top gear is 1.467:1, while bottom gear is .279:1, a much larger gear ratio.

See: http://sheldonbrown.com/gears/internal

Sheldon "Rohloff" Brown
+---------------------------------------------------+
| If you oppose making marriage legally available |
| to all adults, you are promoting promiscuity. |
+---------------------------------------------------+
Harris Cyclery, West Newton, Massachusetts
Phone 617-244-9772 FAX 617-244-1041
http://harriscyclery.com
Hard-to-find parts shipped Worldwide
http://captainbike.com http://sheldonbrown.com

Sheldon Brown quoted Dan Burkhart and added:

Quoted message said:

Quoted message said:

The torque transmitted to the axle shaft is inversely proportional
with respect to the input torque as it is to the hub shell. While
input torque is multiplied to the hub shell most in the lowest gear,
it is multiplied most to the axle shaft in the highest gear.

That's true of hubs such as the SRAM that have direct drive in the
middle of the range.

I disagree. In this thread, I showed the math (giving just a few
significant figures, to keep the numbers simple) for the Nexus 8.
The situation is similar with the SRAM 7 speed hubs: the axle
needs the most torque support in the lowest gear, not the highest.

Dan and/or Sheldon: please show the math to support your claim.

Tom Ace