More torsen discussion (long)

[C1J1Miller <C1J1Miller@aol.com>]

I assume that all interested parties have read Gleason's paper on the torsen
differential as presented on Jeff Goggin's web page:

http://www.mindspring.com/~audidudi/Torsen.htm

I've quoted a couple sections here.

The use of a 60:40 split center torsen in racing audis was reported; this is
possible with the torsen as the following excerpt from Gleason's paper shows:

"...The Torsen differential can also be designed so that different bias ratios
are associated with different directions of
differentiation between drive axles. That is, this design permits one drive
axle to support a larger proportion of total drive axle
torque than the other. The interaction of forces within the differential which
give rise to this design characteristic are very
complex. However, it may be generally explained that such different bias
ratios are obtained by relatively raising the coefficient
of friction which is effective between the end face of one side gear and the
differential housing. This tends to increase resistance
to the rotation of the drive axle associated with the one side gear with
respect to the other drive axle. 

This feature may be particularly advantageous in 'center box' applications
where the Torsen differential is used to interconnect
drive shafts to the front and rear drive axles with a single power source. In
this application, it is possible to set different limits to
the maximum proportions of torque which can be unequally divided between the
front and rear drive axles. Since division of
torque is automatically achieved by the Torsen differential in proportion to
available traction, a separate 'torque splitter' is not
required. In fact, such a predetermined torque split may detract from the
designed torque biasing operation. "
.............................
this section describes where the spider's bite may come into play.
.............................
"...
In turning situations, the outside wheels of a vehicle travel over more
distance than the inside wheels. Accordingly, the inside
and outside drive wheels must rotate at slightly different speeds (i.e.,
differentiate) to maintain rolling traction with the road. A
torque division between drive axles at the bias ratio is a precondition for
differentiation under all circumstances of operation.
Essentially, in order for one drive wheel to rotate faster than the other, the
drive wheel having greater resistance to rotation
slows with respect to the differential case and transfers torque to the other
wheel contributing to its faster rotation. The Torsen
differential resists transfers of torque between drive wheels in proportion to
the torque applied to the differential housing, and
this results in a larger proportion of the applied torque being delivered to
the slower rotating drive wheel. Therefore, bias ratio
should be selected to provide the maximum traction advantage that will still
allow both drive wheels to deliver significant
portions of engine torque in turns. 

However, even in turning situations, the Torsen differential enhances traction
management. Since torque is already distributed in
increased proportion to the inside drive wheel, it is exceedingly unlikely
that the outside drive wheel will ever exceed available
traction and 'spin up'. Alternatively, should the torque of the inside wheel
exceed available traction in a turn, it is equally
unlikely for this wheel to 'spin up' since such a 'spin up' would still
require a difference in traction between drive wheels which
exceeds the bias ratio. Ordinarily, when the inside wheel exceeds available
traction, differentiation ceases and torque is divided
in more even proportion between drive axles determined by the maximum torque
that can be sustained by the inside drive
wheel. Thus, in all directions of travel, the Torsen differential will resist
'spin up' of either drive wheel by instantly dividing
torque between drive axles in proportions up to the bias ratio to match
prevailing traction conditions 
. 
6  CENTER  BOX  APPLICATION 

Although the differential has been mostly described with respect to its use
between drive axles, it should be understood that
analogous performance can be expected from use of the differential as an
operative connection between drive shafts to the
front and rear axles. For example, traction management is enhanced in such
'center box' applications by assuring that more of
the traction of the front and rear drive wheels is available for use. "
.........................
To me, this reads that more power is sent to the wheels that move slower; as
Jeff reported from filming his car, oversteer can cause the rear wheels to
rotate slower than the front, causing the torsen to send more power to the
rear.  This would shift the weight of the car to the rear, hopefully stopping
the oversteer before the additional power caused wheel spinning, which would
then send power to the front...  there's the see-saw of power that has been
reported.  Seems that what we need to see (in higher-power audis, anyways) is
a biased torsen, that cannot send too much power to one axle.  Instead of the
75/25 split (?), it seems it would be easier to modulate a diff that could
only send say 60/40, or even one with a front bias as that's where all the
weight is...Doesn't the new S4 have a reduced split?

Left foot braking would seem to throw in a monkey wrench, as the torsen would
see that as torque, vs. engine braking, which is coasting (no torque) per the
paper.

Anyone care to open a dialog with Gleason on what audi uses, and how to alter
the torque split characteristics?

Lastly: anyone ever experience the torsen spider's bite in a V8 manual with
the center and rear torsen?  Does the limited slip rear lessen the potential?

Chris Miller, Windham NH, c1j1miller@aol.com