Re: 5kTQ swaybar upgrade???

[Scott Fisher <sefisher@cisco.com>]

Pat Martin, questioning a comment about adding roll stiffness to the
front of a car and thereby reducing understeer, asks some very good
questions:

> Than please enlighten us with what makes the books wrong. 

If by "the books" you mean the Carroll Smith/Fred Puhn "classics," then
it isn't so much that the books are wrong (they're not) as that they're
written for purpose-built race cars, not production sedans. 
Purpose-built race cars typically use upper and lower A-arms of unequal
length, designed with enough built-in adjustability that it's possible
to dial in a camber curve so that the wheels stay upright under the body
roll conditions that the driver is likely to experience at racing
speeds.  Think of the front wheels of a Formula 1 car on the grid, with
the tops of the tires looking like they're several inches closer than
the bottoms of the tires.  Now watch that car negotiate a fast corner
and you'll see that the outside front tire is perpendicular to the
pavement.  

Most sedans, even good ones, aren't set up this way.  In particular, a
MacPherson-strut suspension (such as at the front of my '83 CGT --
there, MAC) will let the front tires develop positive camber
proportional to the body roll.  Under these conditions, the front tires
lean over on their sidewalls, the inside edge of the tire's contact
patch becomes unloaded, and grip goes down significantly; you feel this
in the driver's seat as understeer or "plow".  In addition to MacPherson
struts with their inherent linearity of camber gain during body roll,
many other suspension designs are built to add positive camber during
body roll (at the front) because it provides "confidence-inspiring
understeer."  

What Smith is talking about is a car that's already balanced and
possessing performance-oriented suspension geometry.  In that formula
car I was talking about, increasing the front roll stiffness has the
effect of *pulling up on the inside wheel* -- my favorite way to
visualize increased roll stiffness.   That is, as the car rolls to the
outside of a corner, high roll stiffness can actually lift the inside
wheel off the ground.  At that point, all the weight at that end of the
car is being supported by the outside tire; if that weight's lateral
load exceeds the tire's ability to stick, that end will slide off
course.  If you stiffen the front of the car to such a degree that it
lifts the inside front wheel, you will eventually see understeer because
only one tire is on the track instead of two.  Likewise, if you stiffen
the rear (and who here has ever watched a Rabbit go around an autocross
course with its rear wheel two feet in the air?) sufficiently, the
inside rear will come off the ground.

> For instance why would putting more weight on the front of the car 
> IE:increasing roll stiffnes, make a car handle better. 

Increasing front roll stiffness (which doesn't "put more weight" on the
car, as I think you can now visualize) can reduce understeer *IF* the
understeer results from body-roll induced camber gain.  Making sense? 
That is, if the tire leans outwards in the corner in an amount
proportional to the degree that the car is rolling, then by making the
car roll *less* at a given speed, the tire doesn't lean so much at that
given speed.  And in general, keeping the tire upright in a corner
produces more grip than a tire tilted over even a few degrees.

I've experienced this on several competition cars, both FWD and RWD (but
not yet AWD).  In both cases, they were cars designed with a fair amount
of positive camber gain at the front when the car rolled.  By installing
a thicker front anti-roll bar, the front tires stayed vertical at higher
speeds than they did with the thinner bars.  This resulted in more
rubber being in contact with the road surface, which reduced the
tendency of the front end to slide at comparable speeds.

Note that in such a car you'll *still* understeer off course if you
exceed the tires' tractive limits, but you'll do it at a higher speed
than before.  Whether this is a good thing or not depends on how good
you are at catching the slide before it becomes terminal. :-)

> Maybe I am crazy but I thought that the whole idea was to distribute 
> the weight as evenly as possible.

Not really -- at least not entirely.  The whole idea is to maximize
adhesion at the tire's contact patch under all conditions of dynamic
load, as a tire is in motion.  Weight distribution is part of that, but
only part, and it's the hardest part to change about a car.  

It's really a game of balance -- and not weight balance, but balancing
the effects of different modifications against the things you can't
change about the car.  It's about analyzing the car's inherent
weaknesses and doing what you can to compensate for them, to make up for
its shortcomings and enhance its strengths.

So in the case of a production car whose suspension adds positive camber
at the front under body roll, experience shows that putting on a stiffer
front anti-roll bar can *reduce* understeer, no matter what Fred Puhn
and Carroll Smith say -- but that's not because they're wrong, it's
because they're talking about purpose-built race cars.  Making the front
of a McLaren F1 car stiffer will probably increase understeer; making
the front of an Audi 5000 stiffer will probably reduce it, or at least
move the speed at which understeer finally happens a few ticks up the
speedo.

And that *is* the whole idea, isn't it?

--Scott Fisher