re: U/S Torsen

[Eliot Lim <eliot@u.washington.edu>]

On Mon, 8 Jan 1996, Dave Eaton wrote:

>in my ur-quattro, i found that to avoid this situation when driving
>fast, i'd always lock the diff.  as a result the car lost some feel
>for the road, and understeered more etc. etc.

i always thought that locking the center diff led to oversteer.
???

> 
> i've found the ur-quattro with the torsen understeered a lot less than the ur-quattro with the open centre diff.
> 

because the torque split in a turn was rear biased.  the torsen would
sense that the front axle was "slipping" wrt to the rear and thus
transfer power to the rear. this is because the front wheels travel a
greater distance than the rears in a turn.  the center torsen would
treat it as "slip".

> as i understand it, the big advantage of the torsen centre diff as
>opposed to the viscous variety is:

> 1) the spped of the reaction to low-traction situations
> 2) the ease of implementation of abs.

more specifically:

the torsen (and computer controlled clutches) have a linear relationship
between its locking action and the amount of slip.

the VC has a far more exponential characteristic.  it makes the
chassis trickier to design.  for the original full time 4wd 911C4, it
was rejected for this reason.

the VC is rotational sensing, while the other two are torque sensing.
actually it is just a fancy term to describe the fact that VC locks
both in acceleration and braking, while torsen and CCC lock only
during power application.  immense effect on ABS implementation
because ABS requires independently rotating wheels to make its
determinations.

the cars that feature VC full time 4wd and ABS have to slacken the
couplings so that ABS can work properly.  it is a cheap and ugly
hack. i believe that the latest porsche c4 and turbo disconnect
drive to the front altogether during braking.


eliot