Additional torsen stability factor - was Alignment

["Smith, Kirby A" <kirby.a.smith@lmco.com>]

quk@isham-research.demon.co.uk (Phil Payne)states, in a pearl of clarity,
but not necessarily universal validity:

	In message <000d01bf16c9$c4843720$59916420@fremontpeak>
bbell@surview.com writes:

	> Seems everyone agrees the torsen can shift torque rapidly from one
axle to
	> another. Everyone also seems to agree that front drivers tend to
understeer
	> and rear drivers are more prone to oversteer. Then *why* do we
have such a
	> hard time getting agreement that the shifting torque (F/R)will
transfer
	> behavior from Understeer to oversteer? This does not seem to be a
great leap
	> of logic to my small mind.

	I agree.  Except that (at least on both of my Torsen cars) it flat
	doesn't happen.  Bear the situation in mind - a car very close to
the
	limit and under power.  Classic understeer is simply the front axle
	starting to slip - the Torsen then shifts power backwards and the
	result is that the rear slips by the same amount - classic
understeer
	then becomes a graceful and predictable four-wheel drift.

	In this situation, you can simply centre the wheel and steer with
the
	throttle.  If the car is aligned properly.

And also:

	In message <v03130304b42baa973376@[192.168.0.1]> Phil Rose writes:

	> BTW, my "off-road" experience happened about a week or two after
having a
	> 4-wheel alignment done by a *presumably* reputable, ol'-timer type
	> alignment specialist. However the guy managed to erase the
alignment data
	> before I could request a copy. I watched him do the alignment, but
will
	> always wonder: just _what_ were the alignment specs he used?

	And that's the most interesting new piece of information I've seen
for
	a year.  It's also no surprise.

	It seems to be taking some people a long time to abandon their dogma
	(and I can do without the anonymous private hate mail, whoever it
is)
	but I think this is conclusive - spider bite is caused by
	out-of-specification alignment.

In summary, Phil believes its all due to alignment.  In keeping with Gary
Lewis' request for new factors, I submit the following, with all due
respect, and in full anticipation of becoming a flack target.

I suggest that while alignment can change a car's intrinsic handling (U or
O) and thereby make the vehicle more or less controllable, a U-O-U
oscillation, such as reported by Scott and Jeff, requires two lags through
the servomechanism made up of vehicle and driver.  Where might these lags
come from?  One is from chassis dynamics, i.e., the delay in vehicle
attitude and slip response due to a steering input.  The second, I assert,
is from the driver, i.e., the delay in steering response due to a "seat of
the pants" input.  If these times are commensurate in a given
servomechanism, and the "gain" is high enough, the system _will_ oscillate
and likely go unstable.

Notwithstanding how well any of the "protagonists" understand, or agree upon
theory, we have to take at face value that competent drivers have achieved
different results.  This suggests to me that they, together with their cars,
either have different pairs of reaction times or different "gains" or both.

I believe this question is unresolvable until a controlled experiment is
defined and performed.

Kirby A. Smith   New Hampshire USA
1988 90q Titanium gray, 182 kmi
1988 90q Stone gray, 187 kmi 
1995 S6 Pearl effect, 82 kmi