rec.autos.simulators

Ken

Let's take a super simplified lab example. A large block of *** sitting
on the concrete lab floor. Let's say it will take 200 lbs. of force to
overcome it's cf and start it sliding across the floor.

What happens when you push it with 50 lbs. of force?
( I'm using 50 lbs. because it is .25 of  200, and the .2 g you use as an
example is .25 of the cornering force a decent road car can generate.)

Well, it pushes back with 50 lbs. of force and other than that just sits
there.
Push it with 100 lbs. of force and it pushes back with 100 lbs. of force and
continues to just sit there.

Push it with 201 lbs. of force and it will accelerate under 1 lb. of force
by sliding across the lab floor.

A simplfied tire does the same thing. If you only load it to .25 of it's
force generating capacity it has a reserve capacity. Load it to .2 g at a
real force of 200 lbs. and it is still capable of generating 600 lbs.
additional force. Accelerate the car to .4 g and the tire responds by simply
generating the additional force neccessary to balance the load and the car
remains on the same arc. You can continue this right up to the limit, which
is *defined* as the point where the arc of the car must widen to balance the
force in accordance with the rules of circular motion. By definition if you
are below the limit the car will follow the same arc whether you increase or
decrease speed as long as the car stays below the limit. This is essentially
the same way banked turns work, and why at a particular speed a car can hold
it's wheels straight and follow a curved path with 0 side loading, using
gravity to supply a centripital force rather than the tires. Correct? What
about this don't you understand?

You can go out test this on your own street car, can't you? .2 g is well
below a speed likely to gain the interest of the authorities on turns you
can find around any  city.

You understand all this? So perhaps you can see my quandry at your question.
I'm using a very simplified model here that dosn't take the fact that the
capacity of a tire to generate force is variable with the slip angle. At .2
g this is irrelevant to the driver because of the huge reserve capacity of
the tire. It's only the small window around the limit that things get
interesting, where the reserve capacity of the tire is small and may
actually disappear. Are you being confused about the idea of slip angle and
what it does and what it is? Am I being confused about what you are and are
not confused about?

The next issue  is your specific question about the reaction of GPL.

GPL does not use a simple single tire in the lab model. It uses a model of 4
tires attached to a semi-rigid body. It resolves the vector sum of those
forces. How can the behaviour of such a model be used to deduce a
hypothetical simplified single tire model? Isn't that rather like trying to
deduce an Elephant's DNA sequence by looking at the elephant, rather than
the other way around? I can set up a GPL car to exhibit just about any
behaviour you want. The static loadings are user variable. Roll resistence
is user variable. Tire pressures are user variable. Damping rates are user
variable. All on each corner of the car seperately. Different chassis have
different inate properties such as roll center, weight bias, suspension
geometry, etc. Even aerodynamic properties are taken into account.

And the arc taken by a car with 4 tires attached to it will be the vector
sum of the forces on all four tires right? Each with a different load and
perhaps even a different cf. Slip angle may even be different on all four
tires depending on suspension geometry. The degree to which the arc changes
will be the dependant on the ratio of the total slip angles of the front
tires to the total slip angles on the rear, right? A car with larger slip
angles at the rear will oversteer and one with larger slip angles at the
front will understeer, correct? And the slip angles are user settable to a
large degree, yes?

I fail to see what you are attempting to deduce from the behaviour of GPL,
or how you feel you can deduce anything at all by it.

I can leave you with the same answer I gave before. In GPL if you are
cornering at .2 g and increase your speed to .4 g the arc will change by not
changing at all, which is exactly what it should do, and exactly what your
street car will do.

Can you pick up a copy of Skip Barber's Going Faster? It's available off the
shelf at at any Borders or Barnes and Nobles, as is Van Valkenburgh's Race
Car Engineering and Mechanics. I highly reccomend you suppliment the
Milliken's work with these to give you a drivers perspective on the hard
science.

As for the car in your model not wanting to return to a straight line, have
you modeled all the self-centering torques? A tire is a spring. When it
moves at a slip angle it loads in torsion, and thus generates an equal and
opposite torque. Caster and camber also need to be modeled correctly for
their self-centering torques.

I really have no idea whether I'm being helpful here, or just condescending.
I MEAN to be helpful. Honest.

Yours,

                Kevin F. Gavitt

A couple of random comments....

.....

Fine for a simple tire, but not quite true for real ones!  In a tire the
"block of ***" in the print is constantly changing, the tread *** and
the carcass structure all deform, etc.  As I said before, nothing is simple
about tires!

Nice book, except that the plots of tire data are wrong -- after the
peak, they show a drop off in force...and this doesn't happen.  After
the peak, the sideforce stays pretty constant, out to 90 deg slip angle.
  The fact that the force "stops increasing" with additional slip angle is
enough to cause the driver to "lose it".

???  Are you talking about tire self-aligning torque (or pneumatic trail)?

-- Doug

                Milliken Research Associates Inc.

FWIW I use the UK branch of Amazon & have never had a problem with them. I
ordered your book from them on 30th Dec & it was shipped on 4th Jan - not
bad considering that due to the millennium celebrations this was the next
working day. Maybe amazon.co.uk has a different approach to their parent
company?

I had previously tried to order it via Barnes & Noble and it took them
about 2 months before finally admitting that they weren't going to supply
it....

Cheers,
Richard

PS: Thanks for writing such a fascinating book!

   Doug, I assume you are correct, since you wrote the book, but I have also
read in other sources that the grip level will decrease slightly (in some, such
as Nascar tires) after a certain slip angle range has been exceeded.  I'm
looking right now at a chart in "Circle Track" Annual 1995 that shows a Nascar
tire's lateral acceleration increasing with slip angle until about the 5 degree
point (1.2g), then dipping slightly and rising again around the 9 degree point,
then falling sharply after that.  
   Perhaps this is because the direction of the force as measured in this
magazine was perpendicular to the car's chassis, and not perpendicular to the
direction the tire was facing.  That would be a huge difference!  
   So could we conclude that:

   Yes, I included self-aligning torque (a greatly simplified, fudged pneumatic
trail type set of equations that approximated the effect), suspecting it would
cure the problem, but the forces that the torques generated on the center of
gravity of the car more or less cancelled each other out.  It had very little,
if any, effect on straightening the car.  
 ** I missed your other post with quote I showed above, Kevin, could you please
repost it?**

   Doesn't this go against the concept of dynamic vs. static friction?  Or is
the tire always sliding a little somewhere in the contact patch (if you drag
out the magnifying glass) so the abrupt change between dynamic/static is lost
gradually, resulting in a smooth curve?  

  Of course, as you said, this would depend on the tire.  But don't high
performance racing tires operate this way?  Peak at a high grip, then drop off
afterward?  Street tires probably don't do this, but racing tires?

Todd Wasson

">

Tire men are the wizards of archane knowledge as far as I'm concerned.

To me trying to understand, let alone describe, the actual dynamic
properties of a tire are like trying to work a three body problem on an old
crank arm adding machine.

Doug Milliken wrote ...

I'm biting. What are you talking about here?

Tony

Sorry for all this off-topic chatter.

Thanks for the update on Amazon UK -- seems like they must have
put in a stock to ship it so quickly.

-- Doug

-- Doug