Snow Driving

David Eaton dave.eaton at
Fri Dec 13 23:11:31 EST 2002

>From: QSHIPQ at
>Date: Thu, 12 Dec 2002 08:10:42 EST
>Subject: RE: Snow Driving
>In a message dated 12/11/02 2:55:40 PM Central Standard Time,
>Dave.Eaton at writes:
>>nope.  you can't change the engine-side torque to the diff other than by
>>gearing or speed. "trg" to use your [incorrect] definition (the centre
>>torsen doesn't actually have a ring gear, but the carrier is connected to

>>the gearbox output).  have a look at a torsen, identify the "ring gear"
>>[carrier], look at what it is connected to.  you are mistaking the
>>allocation of torque across the torsen with the torque input to it.
>The INPUT Dave, is referred to as Trg, so as to not confuse it with Teng.
>The summ of T1 and T2 = Trg.  You are confused a bit because the SAE
>literature references Trg as applied to a rwd vehicle.  In the case of an awd

>audi mit torsen, Trg is before the f/r differential final drives, but after

>the gearing.  So, my statement appears correct, that Trg is reduced,  cuz by

>definition T1(front) +T2 (rear)= Trg, NOT Teng.  As you exceed traction by

>spinning a front wheel, T1 is reduced, so then by definition is Trg.  If I

>don't have that right, please do help me understand.

no, trg cannot change.  as you have stated, trg=t1+t2 (the 2 output shafts).
 all that changes is the mix between t1 & t2 - this is, by definition, the torsen
doing its thing.  if the torque can't be supported by one input shaft, it is
sent by the spur gears to the other side.  trg stays the same.

>>The torsen couldn't care if you had a LSD front, LSD rear, EDL front EDL rear

>>it reacts to only 1 EFFECT.  Attempted speed differentiation.  As the poster

>>listed, he was speaking of a straight line, in which case, the torsen will

>>work the same with or without locking the rear diff, as would a locker.
>>>completely incorrect.  the torsen will react to the loss of traction from
>>>single tyre.  with the normal open front/rear diffs, this has a fundamental

>>>effect on the "input torque" to the torsen centre.  a locked or edl rear

>>>diff will fundamentally alter this behaviour.
>I don't see how Dave.  Since EDL is not capable of crossing the center axle,

>there is no difference in HOW the device works.  You can lock the rear in a

>straight line (as was the example), and the effect on the torsen won't
>change.  As weight shift rear occurs, the fronts will slip first in that
>straight line.  As such locking the rear won't affect the conclusion of the

>characteristics in a straight line.  EDL *might* have an effect, but that
>requires wheel spin first (since EDL is based on wheel speed, not attempted

>wheel slip differences), at which point the torsen has already shifted to max

>rear.  OR, you could propose a start on snow (not what the lister presented,

>but...), in which case there still is no difference, because the fronts will

>spin first, then the rears, but AT max TBR of 78r/22f.  That won't change
>with or without the rear diff locked.  T2 might be higher with rear diff
>locked, but the behavior of the torsen is the same, the torque allocation is

>still max rear.

perhaps a personal example of this would help to explain.  in my old house i
had a steep drive with an up-hill, off-camber driveway - the lhs was a lot higher
than the rhs.  on entry (from the left) the front of the car would follow the
camber, but once it came to the rear, the car nose was rising, but weight was
still at the front.  the car would therefore lift the rhs rear tyre.  if speeds
were slow enough the car would be stuck because the torsen (while it was distributing
torque) would not be supporting enough torque to power the front (which had
the traction) up the hill.

the only way to progress was to lock the rear diff.

once this was done, the torsen "saw" the traction available at the rear, and
supported torque distribution (both front and rear) enough to progress up the
hill.  this had nothing to do with the steering angle on the car (there wasn't
any anyway), simply the traction available to allow the torsen to differentiate
enough torque.

the torsen is very much aware of the traction available at either end of the
car, and therefore (given the presence of open diffs at either end). it is therefore
correct to say that the torsen is aware of the traction of the *least tractive*
wheel at either end of the car.  the only way to vary this behaviour in the
early (generation 2) cars is using the rear diff lock.  modern torsen cars have
edl working across the front & rear diffs, and esp working across the centre
diff, so have much less need for driver involvement...

in the scenario above, with an unlocked centre diff, the car would hang in exactly
the same fashion as with the torsen.  locking the centre would allow 100% of
the torque to the front of the car (none is able to be supported by the rear).
in either scenario driver intervention would be needed.

'95 rs2
'90 ur-q

More information about the quattro mailing list