Torsen Tech

[QSHIPQ@aol.com]

Reading with interest Jeff Goggins extensive technical posts on the Torsen
personality.  Having explored the center Torsen phenomenon several times
myself, and had some interesting off-line discussions with Jeff and Graydon
regarding both at the track and off-road Torsen behavior, I want to add some
specifics to his excellent explanations.

Jeff touched on a couple of points that really hit what the torsen is
"thinking" while one carves into a turn.  Basically, the unit is sensing a
slower rear wheel rotation on initial turn in, so more power is transmitted to
the rear axle, per Jeff's post.  Let's explore that more fully to understand
the handling of a Torsen at the limit thru a turn.

Turning Exercise:
Car:  1991 91 200tq, Stage II modded computer
Turn:  90 right at the end of a long straight.  Dry track.
Enter the braking zone.  Loading the cars' nose heavy front end, chassis
dynamics will give intial understeer as you steer for the apex.  So, Entry
Strategy (as per most fwd):  Late brake, oversteer before the apex, and drift
out with power.  Not quite.  

As you oversteer to correct the understeer inherent to the chassis, you add
power to initiate drift.  You now have the front wheels turning faster than
the rears (like the speed skater on the inside vs outside of a turn), and the
torsen "thinks" - Wheel slip in front, add torque to the rear.  Well ok, now
you have the wrong line for the understeer you initiated into the turn.  Now
the chassis is exibiting oversteer, your line is for understeer.  What
happens?  The back end will go out fast.  Your correction?  Off throttle,
steer for the apex again.  Torsen:  Back to front wheel understeer condition,
oversteer again, add throttle, back end out.  My definition of hunting torsen
awd thru a turn.  Happens on all the torsens, the smaller cars have a more
control advantage, because the COG doesn't change as much as the larger and
effectively softer sprung cars for a given turn, but it's still there.  Those
with less power have less of a problem, since WOT oversteer is hard in the
dry, but those pups on a slick surface will have a bigger problem, since a
short wheelbase car will do things quicker than the longer ones, including
spin.

Next let's look at what happens beyond that oversteer, where you have the nose
90 degrees+ to the apex of the turn, the "expert" line.  Ok, so you've entered
the turn, added throttle, back end comes out.  One of two things will happen.
A)  in low traction, you keep your foot planted, and the torsen eventually
will "find" that both axles are spinning at the same speed, and you have a
50/50 split.   OR B) You lost your balls and lift throttle.  <A> is a better
proposition, but takes some serious practice to get correctly.  I saw a couple
of posts (Jouko) that indicate that in an exercise this can be easily
modulated.  I argue for a given turn or exercise maybe so.  For a full track
or conditions which change, maybe not so.  

The difference between a race driver and one who "exercises" is repetition,
exactly doing the same thing every lap of a race.  Few do, specifically, the
number that don't in the street world exceeds exponentially the number that
do.  On the track, Dave L experienced the "changing" condition issue, torsens
at the limit can repeat, but as steamboat got more slick, the handling of the
tosen did too.  Why?  Because the torsen will reach the equal torque split
(50/50) at a different chassis to apex angle as conditions change.  Where you
found 50/50 at 90 degrees to the apex the first time, will be less as the
traction lowers.  Lift throttle doesn't help either.

Lift throttle will give you oversteer as you are at the 90 degree mark.  Why?
Well, when you were at WOT going into the turn, the car squats to the rear,
lifting the front, reducing the weight on the front tires.  When you lift at
or near the 90 degree mark, the weight now goes back to the fronts on a front
heavy car, and so does the torsen hunt, now you have a lift throttle oversteer
condition (common fwd), and the back will just come right around.  A spinner. 

So what is the answer?  I'm not convinced there is one.  My experience has
been to stay on the throttle and try to "find" that 50/50 split, whatever
chassis to apex angle that happens to be at.  On dry pavement, it is at 90
degrees, on slick, less.  I can share that whatever you choose to do, <A> or
will take time, and a lot of it, btdt.  Maybe more than one can consider
reasonable, or more than you have available to you on the street.  The finess
required to master the torsen is beyond the best of drivers.  Maybe a good
reason why you don't find torsens center diffs in race cars.  Fixed splits are
a proven better way at the limit.  50/50 is better, most, even audis own
formidable race cars of track and rally, are even more torque biased to the
rear.

Given exercises at 11/10ths can make you better at understanding the "hunt",
but I argue, not necessarily mastering it.  So that old "primitive" Gen I
locking diffs, might just be a better setup for those looking for true
performance driving.  For those interested in the 7/10ths world in which we
drive day to day, the torsen is probably a good thing.  It certainly is for
those (like my former wagon owner) who have never or don't use the knob
supplied in the Gen I lockers.  

I watched the exercises Freddie was giving several at the RA QClub event, and
really didn't see the correlation to the track with the pointers.  Why?
Because he wasn't teaching a track technique, just a constant control exercise
(being familiar or more familiar with an exercise, doesn't necessarily
translate to a faster LAP in a torsen car).  Remember, the turn has to be
completed.  As you finish the 90 right, and you drifted perfectly, you now
have to get the car to the exit of the turn.  So, you keep the power on and
steer for the exit, torsen starts to hunt again.  Why?  Because, as you hit
the apex the torsen went 50/50, as you start to exit the turn, the torsen will
now go back to a more understeering fwd character.  This is the part that
makes me think that the torsen and audi chassis are setup for fwd character
instead of rwd, given the customary street car engineering preference to
understeer than oversteer.  

Sound complicated?  It is.  And audi added to this torsen persona, the
seemless and remote feeling of security as you play below the dark side
threshold (7/10ths), read:  Very little feel as to the consistent What and Why
of it's character. I have yet to see a torsen car that has been mastered on a
track.  In an exercise?  Sure.  But the world in which we all drive is filled
with constant varying exercises, and the torsen adds a sense of security for a
lot of drivers up to a point.  Beyond the threshold, it can be ugly and mean.
You play there, have lots of room, fun can turn costly quickly.

Gen I.  Pretty straight forward, locking the center, makes for a very
predictable understeering car.  I posted the Gen I locking detail for
Steamboat in the archives.  I also find few take advantage of all the Gen I
has to offer there.  Adding and deleting locking differentials into a speed
event takes a lot of concentration.  However, adding and deleting that rear
locker to a speed event has rewards beyond the customary set and forget.  For
either Gen I or Torsen cars.

My .02, looking forward to more discussion of the Lucifer called Torsen.

Scott Justusson
QSHIPQ@aol.com
'87 5ktqwRS2
'86 5ktqw
'84 Urq