RE: a yaw'ning chasm....

["Dave Eaton" <Dave.Eaton@clear.net.nz>]

no worries jeff, it's a common misconception.  same groundhog day feeling
here as well (coupled with a certain disbelief that the fundamental
principles of differentials have been misunderstood from the start - no
wonder the torsen thread went no-where).  so i'm not getting into a
meaningless debate, just stating the facts.  these can be verified by
reading any number of papers or books on differentials.  i'm happy to
provide references if required.

the fact is that the normal open differential is a *torque* splitting
device.  within fractions of a percent it will reliably divert 50% torque to
each output shaft.  it will do this *regardless* of the output shaft's
respective speed.  (in fact the sum of the rotational velocities of each
shaft will always be equal).  i'll skip a description of what torque
actually is here, it's a given.  in other words the differential cares
nothing about output shaft *speed*, only that 50% of the torque goes each
way.  hence in an awd setup, with all open differentials, you get a reliable
25% or torque at each axle.  regardless of conditions.  however, across each
differential, you allow output shaft speed differences (the rotational
velocities for each output shaft will always be equal though).  this ability
to allow *speed* difference but not *torque* difference is the fundamental
property of the differential and an advantage when cornering obviously.

the even torque split is obviously a disadvantage when one wheel (or shaft)
has less traction than the other, since the driving torque to each wheel
(shaft) will be a function of the tractive force on the tyre with *less*
traction...

the only way you can change this fundamental property (disadvantage) of
differentials is to lock them in some manner.  this means that you try to
limit the slip in low traction conditions while maintaining the differential
speed capability (which is required to prevent windup etc).  when you use a
simple dog clutch (audi generation 1 system), you lock their relative
*speed*, but now allow *torque* to move around the drivetrain.  by
definition.

michael's case is correct.  visualise the front inside wheel lifted with an
open centre.  we have equal torque through the system, a spinning wheel
(it's tractive force) is the torque "determinant", and you're going
no-where.  with a locked centre, you now have the ability to move torque
through the system, and 100% of torque now goes to the rear, 0% to the
front.  the result is drive.  with the audi generation 1 system it also
brings significant understeer because in normal conditions more torque is
being sent up front as a result of the weight distribution mostly.

one of the non-recognised difficulties of the locked centre diff is that the
chassis designers have to increase the weight of the awd system because the
shafts have to cope with a potential 100% torque.  limiting the amount of
torque shifted through the system means that they can get away with less...

i probably haven't explained this very well, if people like i can scan some
more more learned descriptions from more qualified people than i - and then
we can move away from the realm of opinion ;-)

hth,
dave
'95 rs2
'90 ur-q
'61 mb fintail

-----Original Message-----
From: JustaxPHX@aol.com [mailto:JustaxPHX@aol.com]
Sent: Monday, 24 May 1999 03:19
To: Dave.Eaton@clear.net.nz; quattro@audifans.com
Subject: Re: a yaw'ning chasm....


> he generation 1 locker *must* and does shift torque when it is *locked*.
>  (0%-100%-0% torque shift front to rear is certainly possible).
>
>  however, when all diffs are *open* *no* torque can be shifted.  all
torque
>  outputs are equal, and 25% by definition (for awd).
>
>  you need to think about it some more.  once you understand this
fundamental
>  principle, this will help your understanding of this whole issue..

Why do I suddenly think I'm starring in my private version of the movie
"Groundhog Day," where the same events keep happening over-and-over again
despite my best efforts to change them?

Needless to say, I respectfully disagree with Dave's comments above.  And if
anybody figures out how a "locked" diff can transfer torque, please be sure
to drop me a note before I go offline for the summer at the end of the
month...

JG