I will say the jury is still out on how this system works. With the Phil Wood axle, the Q and chainwheel position has unique adjustability. On the other hand, with a regular cotter the crank walked in on the axle until the inner chainwheel rubbed. It looks really neat, but my Darwinian engineering sense says there was a reason TA jumped ship, and merely being peculiar or machining difficulty wouldn't scare TA away. On the other hand Spence Wolfe apparently sold plenty of these cranks in his day from what I could gather. Stay tuned. By the way, this chicken is holding on to a conventional Cyclotourist triple and BB until I figure this out.
As to torque, each of the four surfaces, the two flats and the two on the "truk", "frob" or "jeu de grain" picks up some load. One flat and one of the frob surfaces are directly loaded and the other two relieve pre-load. The combined stress distribution could be be the subject of quite a dissertation.
Joe
At 10:59 PM 12/11/00 -0500, karl.frantz@juno.com wrote:
>
>
>On Mon, 11 Dec 2000 22:18:21 -0500 Sheldon Brown
><CaptBike@sheldonbrown.com> writes:
>> Karl Frantz wrote:
>>
>> >I have my doubts about aluminum holding up with cotters
>>
>> Bingo! These cranks didn't hold up well in practice. The cotter
>> hole would get deformed and stretched by the pressure of the cotter.
>
>That's about what I figured would happen. Odd that SR would do that; they
>usually seem to make OK stuff. I guess the key word is "usually" :-)
>
>> >(for that matter, how did that TA design fare in service?).
>>
>> The T.A. system doesn't use the cotter-like frob to resist torsional
>> loads; that's handled by the shape of the axle and the matching hole
>> in the crank. All the T.A. doodad does is hold the crank in position
>> laterally so it won't slide on or off of the axle. There's very
>> little stress in this direction, so there's no need to wail on the
>> "jeu de grain" with an inertial impact alignment instrument.
>
>What I wondered about was whether the "jeu de grain" placed any preload
>of the spindle into the cornered pocket of the crank to prevent the joint
>moving. This is the main issue with the tapered-square spindle - it has
>to have enough preload to overcome the pedal forces. I guess the TA
>design, if it's not a press fit, doesn't work the same way. Hmm... as
>long as the forces from torque don't exceed the yield strength of the
>aluminum arm, you're OK - and there's a fair bit of area over which it is
>spread, plus the shape means the force doesn't necessarily all bear on a
>corner. It might wear through fretting though.
>
>I can guess it was a bit more expensive to make, but shouldn't be too
>bad. I suppose it got dropped because everybody else used the
>square-taper?
>
>Karl Frantz
>Sterling, Massachusetts
>--------------------------------------------------
>Karl.Frantz@juno.com