Re: Tire wear, all-seasons, and stuff

[Scott Fisher <sefisher@cisco.com>]

Josiah says:

> Common AutoX wisdom says overinflating your tires decreases 
> the contact patch, so overinflating your rear tires would cause 
> oversteer, not understeer.

...in response to Michael Williams, who wrote:

> Todd Phenneger decided to speak these words:
>
> > For the track I would up this to about 45 in back to induce more
> >Oversteer.
>
> thats interesting, technically, this would increase understeer, not
> oversteer...

The answer?  You're both right, it's a floor wax AND a dessert topping!

Seriously: 

Raising or lowering pressure can either increase or decrease grip,
depending on whether you're already over or under the optimum pressure
for a tire. And therefore simply saying you're adding or reducing
pressure at the rear isn't everything you need to know to determine
whether that action will result in more understeer or more oversteer.

In particular, this is a subtlety that isn't covered explicitly in any
of the usual books; it's usually taught in the paddock at an autocross
or track session, and then only if the person doing the instructing has
experienced this subtlety first-hand.

Here's how it works.  In most good references on suspension setup for
various kinds of competition (I'm thinking at least of Fred Puhn's
well-known introduction "How to Make Your Car Handle" and in one or
possibly both of Carroll Smith's "Prepare to Win" and "Tune to Win"),
you can find a diagram that shows the relationship of adhesion to
pressure in a typical pneumatic tire.  It's a curve that's usually shown
with grip along the vertical axis and air pressure along the
horizontal.  Higher up = higher grip, farther right = higher pressure. 
On a simple, first reading of this curve, it would seem that lowering
pressure would reduce the grip  at that end of the car, hence (as
Michael says) the standard advice is to raise the pressure at the rear
to increase understeer.  That isn't always the case.

The key point: the curve drops on both sides, but it drops a little more
slowly as you move to the right (that is, higher pressure).  There is a
point somewhere in the middle where grip is highest; if you keep adding
air above that point, you will start to lose grip (as Josiah says),
because the contact patch becomes smaller.

But the curve falls off more sharply as you go *lower* in pressure than
it does if you go *higher*.  Meaning (and I'm not going to try to do an
ASCII diagram of this) that if, say, 36 PSI tests out to be your optimum
grip, then you will have less grip at 32 PSI than you will at 40 PSI. 

Clear?  There are two lessons hidden in here:

1 - if you need to err in your pressure settings (would that be then
your err pressure? SORRY), err on the high side, because being off by
the same number of PSI will cost you less in absolute grip.  Besides,
it's faster to drop a couple PSI between autox runs than it is to raise
pressure (especially if, as I used to, you use an air pump that runs off
the cigar lighter instead of a high-pressure air bottle).

2 - if you want to reduce the grip at one end or the other for some
reason, you're safer (see lesson 1) to *raise* air pressure than you are
to *lower* it.

In addition, being over the optimum pressure rating for a given tire
makes that tire's drivability a lot more stable and predictable, at
least in my experience.  

If you're having trouble visualizing the chart and its effect on your
car's handling, here's a story that may make more sense.  At the end of
my first full season of autocross competition back in 1989, I attended
an autocross school and tuning session put on by the SCCA San Francisco
Region.  I'd been running a 1984 VW GTI (in ES, Josiah, not DSP, so
tires were some of the only things I *could* play with), and had read
the same books Michael quotes, so I ran lots of pressure up front and
really low pressures in the rear -- like 40 psi in front and 32 in the
rear.  Eventually I found that 36 psi (for the tires I was using, Yoko
A008s) produced the best rollover characteristics, so I ran 36 psi in
the fronts and 32 to 30 in the rears.

My instructor explained the asymmetrical curve thing to me and had me
run the car at 36 psi in front and *40* psi in the rear.  This still
gave me good turn-in characteristics but made the car more stable and
predictable as well, because the rear tires were no longer flexing so
much -- they just built up their slip angles more quickly.

So in short:

1 - there's an optimum pressure for each tire on each car, and it's
obtainable experimentally with nothing more complex than chalk and a big
circle of asphalt

2 - tires that are lower OR higher than that pressure will have less
grip than tires AT the right pressure

3 - tires that are HIGHER than the optimum pressure lose LESS grip per
PSI than tires that are LOWER than the optimum pressure

This, unfortunately, makes a hash of the rule that Michael (and others,
including me till ten years or so ago) learned about raising/lowering
tire pressures at one or the other end of the car to fine-tune the
handling.  

But it works.

--Scott

P.S.  Note that for real race tires -- meaning tires with the words NOT
FOR HIGHWAY USE cast into the sidewalls and a smooth layer of black
gummy stuff where street tires would have a tread pattern -- pressures
are typically in the range of 18-20 PSI, not 32-40 PSI.  Josiah, are you
using Hoosier Autocrossers, or is something else the hot ticket this
year? -- Scott