Re: Torsen Threads <*all*>

[Scott Fisher <sefisher@cisco.com>]

I think I see some light.  Scott J. quotes from 885140:

885140 "Vehicle Behavior When 
Acclerating in a Turn: (.9cf, max load, 80q) "...  However, the front
wheels 
follow a wider radius that the rear wheels on the circular course, so
that 
.2% forced slip occurs between the two axles, which reduces the wheel
slip 
under traction at the front wheels. and increases the slip at the rear 
wheels."

Wheel slip and slip *angle* are two different phenomena, related in many
ways, but quite distinct.  And it's clear that 885140 is talking about
wheel slip.

Wheel slip is a purely rotational phenomenon, when the torque applied to
a tire's contact patch exceeds its ability to stay planted.  Let's say a
tire has a circumference of 1 meter (little tire).  If the car travels
100 meters and the tire has rotated 100 times, there is zero wheel
slip.  If, however, the car travels 100 meters and the tire has rotated
105 times, there is 5% wheel slip.  Wheel slip is a predominantly linear
effect, the result of the interaction between the tire's contact patch
and a linear force vector parallel with the tire's direction of travel.

Slip angle is the sum of that linear force vector and the car's
"cornering force," defined as the amount of energy absorbed by the tires
in order to keep the car on the road in the direction you've pointed
it.  Simply put, slip *angle* is the difference between the centerline
of the tire's rotation and the direction the car is actually going. 
Torque is certainly one contributing factor to a tire's slip angle, as
is the amount of vertical load on that tire, the side load on the tire
(that is, the car's sideways inertia as it rounds the corner), the
tire's camber, and as has been pointed out elsewhere, the tire's static
toe setting which is effectively a built-in slip angle.  

So in short: wheel slip is a linear phenomenon due to the difference
between available wheel torque and available traction.  Slip angle is a
lateral phenomenon due to the demands being placed on a tire by
cornering loads, among these demands being vertical load, lateral load,
camber, toe, torque, and available traction.  

885140 goes on to say:

"However, if the wheels of one axle reach the limit of adhesion, 
the drive torque is diverted to the other axle (within the range of the 
torque split limits), independent of the forced slip rsulting from the 
vehicles circular path."

This, again, is not talking about slip angles, it's talking about wheel
slip, and it's the same factor that acted as the "parking brake" on our
locked-rear RWD autocross car.  "Forced slip resulting from the
vehicle's circular path" is exactly the effect we observed, when trying
to push the car onto the trailer, as the inside rear wheel hopping and
skittering on the pavement.  And it had next to nothing to do with slip
angles, given that we were two guys trying to manhandle a little car
onto a trailer at maybe 1/10 of a mile per hour.

So -- is there somewhere else in 885140 or in the other literature which
outlines how a Torsen detects and responds to slip *angles*?  The
paragraph cited here says nothing about slip angles, but a great deal
about how Torsens respond to wheel slip, which are as Phil's compatriots
might put it, as different as chalk and cheese.  And while I've
excerpted from an excerpt, I don't think I've cut out anything that
pertains to the question.  I can see how a slip angle will affect the
available linear traction of a tire -- there's only so much grip the
contact patch can afford, and if it's exceeded you slide in one
direction or another depending on the greatest force vector -- but I'm
not entirely clear on how that effect would act as an input to the
center Torsen.

Thanks to Scott J. for posting the excerpt, it's been a great help.  I'm
still interested in pursuing the relationship between wheel slip and
slip angle, though at present it seems as though slip angles are
partially the *result* of the Torsen's torque-shifting activity (as they
are with traditional LSDs, locked diffs, and even open diffs, each in
their own way), and not one of its inputs.  

--Scott Fisher