torsen summary (at last!)

[David Eaton <david.eaton@computerland.co.nz>]

ok, for all the tired readers out there, here is my attempt at a summary
of the behaviour of the centre diffs in our beloved quattro's.  i am
attempting to simplify the pages of prose to a few scenarios whcih
should help illustrate the traits of boris the spider.  please feel free
to add light (not heat), but "kiss" please!

i will consider 3 scenarios which should cover most eventualities:

1) normal cornering will good traction.  the 99% case.
2) hard cornering with chassis roll/weight shift which unweights the
inside front wheel.  the 0.8% case.
3) very hard cornering where the car significantly oversteers, rears are
going sidewards, and opposite lock is dialed-in bigtime.  the 0.2% case.

i will consider 3 differential systems:
a) open centre differential (generation 1 quattro)
b) locked centre differential (generation 1 quattro)
c) torsen centre differential (generation 2,3,4 quattro)

for the sake of argument, i will assume 200lb/ft torque available at the
centre differential (a 20v ur-quattro above 2,000rpm, or a wr above
3,000rpm).

definitions
=======
t(min) is defined as the minimum torque sent to a [virtual] axle by the
centre diff, t[max] the maximum torque, and t[shift] the torque shift
between the two driveshafts (ie. virtual axles).

assumptions
=========
1) torsen is as with quattro a 70:30  device, or a 2.3:1 bias ratio.
2) orin is correct inthat the locked centre diff can provide 100% torque
to either driveshaft.

ok, here we go.

scenario 1
========

1a) normal cornering, open centre diff.

nominal 50:50 f:r distribution (ie. t(min) 100lb/ft, t(max) 100lb/ft)
t(min) 100lb/ft, t(max) 100lb/ft, t(shift) 0 - give or take.
some differentiation between driveshafts for the corner.
no problems. quattro is working

1b) normal cornering, locked centre diff.

nominal 50:50 f:r distribution (ie. t(min) 100lb/ft, t(max) 100lb/ft)
t(min) 100lb/ft, t(max) 100lb/ft, t(shift) 0 - give or take.
some differentiation between driveshafts for the corner.
no problems. quattro is working

1c) normal cornering, torsen centre diff.

nominal 50:50 f:r distribution (ie. t(min) 100lb/ft, t(max) 100lb/ft)
t(min) 100lb/ft, t(max) 100lb/ft, t(shift) 0 - give or take.
some differentiation between driveshafts for the corner.
no problems. quattro is working.

ok, so thats the easy stuff....

scenario 2
========

2a) inside wheel lift under hard cornering, open centre diff.

nominal 50:50 f:r distribution (ie. t(min) 100lb/ft, t(max) 100lb/ft)
t(min) 0lb/ft (rear axle), t(max) 200lb/ft (front axle), t(shift)
+100lb/ft to inside front wheel (by defintion)
all torque (100%) to spinning wheel, no motive power (all the torque is
spinning the wheel). quattro is not working.
when traction is restored, t(shift) to the rear axle is 100lb/ft to
restore 50:50 ratio.
major bite/balance problems.

2b) inside wheel lift under hard cornering, locked centre diff. (using
orin's point here)

nominal 50:50 f:r distribution (ie. t(min) 100lb/ft, t(max) 100lb/ft)
t(min) 0lb/ft (front axle), t(max) 200lb/ft (rear axle), t(shift)
+100lb/ft to rear axle.
all torque (100%) to the rear driveshaft, and rear wheels. (orin's
point).
big-time oversteering possibility, but motive power is still present.
quattro is working.
when traction is restored, t(shift) to the front axle is 100lb/ft to
restore 50:50 ratio.
significant bite problems.

2c) inside wheel lift under hard cornering, torsen centre diff.

nominal 50:50 f:r distribution (ie. t(min) 100lb/ft, t(max) 100lb/ft)
t(min) 30lb/ft (front axle), t(max) 170lb/ft (rear axle), t(shift)
70lb/ft to rear axle.
most (70%) of the torque goes to the rear driveshaft and the rear
wheels.  motive power is present.  quattro is working.
when traction is restored, t(shift) to the front axle is 70lb/ft to
restore 50:50 ratio.
some bite problems.

conclusion of scenario 2:
t(shift) is less with the torsen than with the other 2 scenarios.  the
open centre diff is downright scary, the locked centre is ok, but with a
greater "bite" than the torsen due to the greater t(shift).

scenario 3
========

3a) major oversteer/spin under extreme cornering, open centre diff.

nominal 50:50 f:r distribution (ie. t(min) 100lb/ft, t(max) 100lb/ft)
t(min) 0lb/ft (rear axle), t(max) 200lb/ft (front axle), t(shift)
+100lb/ft to front axle (due to rears not moving).
all torque to front axle, large understeer moment  (traction could
easily be lost to inside front wheel, see scenario 2a above). quattro is
working for this instant, until the (2a) effect kicks in (spin inside
front), and motive power is lost, and car continues to oversteer (after
a big 'moment')....
when traction is restored, t(shift) to the rear axle is 100lb/ft to
restore 50:50 ratio.
major bite/balance problems.

3b) major oversteer/spin under extreme cornering, locked centre diff.

nominal 50:50 f:r distribution (ie. t(min) 100lb/ft, t(max) 100lb/ft)
t(min) 0lb/ft (rear axle), t(max) 200lb/ft (front axle), t(shift)
+100lb/ft to front axle.
all torque to the front driveshaft, and front wheels, large
understeering moment. (orin's point again).  however, even if traction
is lost (as is likely - to the inside front), t(shift) to the rear will
be 100lb/ft, as described in scenario (2b), and more motive power will
be transferred to the rear.  if oversteer continues, than the scenario
could well repeat.  hello boris.
big-time understeering possibility, followed by more neutral stance.
quattro is working.
when traction is restored, t(shift) to the front axle is 100lb/ft to
restore 50:50 ratio.
some bite problems.

3c) major oversteer/spin under extreme cornering, torsen centre diff.

nominal 50:50 f:r distribution (ie. t(min) 100lb/ft, t(max) 100lb/ft)
t(min) 30lb/ft (front axle), t(max) 170lb/ft (rear axle), t(shift)
70lb/ft to rear axle.
70% of the torque goes to the rear driveshaft and the rear wheels.
motive power is present.  front less likely to lose traction than in
scenario (3a) and (3b), rears have already overcome their traction (the
car is going sidewards).  t(min) to the front axle will provide motive
power until grip is restored to the rear which is receiving t(max) of
170lb/ft. quattro is working.
when traction is restored, t(shift) to the front axle is 70lb/ft to
restore 50:50 ratio.
some bite problems.

conclusion of scenario 3:
t(shift) is less with the torsen than with the other 2 scenarios, but
the power is sent to the wrong end (the situation is rescued due to the
30% torque at the front).  the open centre diff is again, downright
scary, but the locked centre is better than the torsen as power is sent
to the front.

ok, comments please?
dave
'95 rs2
'90 ur-q