DeTomaso Mailing List: May 2001, Message #13

In a message dated 4/30/01 15:31:29, jhansen@Covad.COM writes:

<< I have heard from, one
person that it may be too much swaybar since it isn't a wide wheekbase model
like a GT4 or 5.  My car has already had the spring spacers removed and has
17" Hall Ultra wheels with 335/35 on the back and 245/40 on the front.  It
is a fast and fun road car that may see some track work in the future and
could use a little more in the handling department and I was going to get
one of the GTS 7/8" rear bars that everyone seems to like but since I have
these new 1" chrome ones I was wondering how they would work. >>

>>>Based solely on empirical data gained at a PCNC skidpad event back in '92, 
I would suggest that 1" bars front and back may be a bit too stiff.  Below is 
the text of an article Jack and I wrote following that event, in which we 
tested the same car back-to-back with different swaybar configurations.  Much 
to our surprise, we saw a massive gain when moving from the stock rear bar to 
the GTS rear bar, and then incremental losses with each subsequent "upgrade" 
to ever-larger bars.  

The car was a stock-bodied Pantera with 15 inch wheels fitted with the 
then-obligatory Pirelli P-7 tires.  You can speculate regarding the effects 
of modern 17" wheels/tires all around, but I'd be inclined to guess that if 
anything, the increased rear-to-front tire width would probably result in 
increased understeer.  Thus it may be possible that the optimal setup would 
be a larger rear bar than front bar, i.e. 1-inch rear and stock 7/8 front.  
But we never thought to test that configuration.

Finally, when the article was written, there were no off-the-shelf chassis 
stiffening kits available.  Having heard glowing reports from various folks 
who are using the two competing packages (Hall Pantera/Precision ProFormance 
or Pantera Parts Connection), I would further speculate that a car fitted 
with chassis stiffeners may benefit from ever-larger swaybars.  What's really 
needed is another thorough test.  Hmm....

Here's the article:

   The Shade-Tree Mechanic   
Swaybars: What’s Right for the Pantera?
By Jack DeRyke and Mike Drew
What is a sway-bar, really?  A swaybar is a piece of steel barstock that 
connects the suspension together on opposite sides of a car that has 
independent suspension.  The bar can be on either the front or rear 
suspension.  Fastened between the left and right side suspension pieces, the 
bar acts as a twisted torsion-bar when one of the wheels hits a bump or the 
vehicle leans in a corner.  The spring-force of one bar end twisting up tends 
to force up the opposite side wheel, which is being extended by the body 
rolling away from it.  In this way, the suspensions for the two sides of the 
car are “evened out” as far as the forces acting on them, but more 
importantly, the “light side” wheel is forced up while the car body is 
forcing it down. 

This adds cornering force to the (relatively) lightly-loaded inside wheel and 
removes some of that same force from the heavily-loaded outside wheel.  It 
also tends to keep the wheels and tires more straight-up-and-down so the 
tire’s traction-footprint is maximized. Too, the reaction force from the bar 
to the body of the car will affect the “lean angle” of the body, which 
affects the driver’s perception of how things are going down under him, as 
well as preventing the tires from tipping over onto their low traction 
sidewall areas.  A drastically-tilted body and driver’s seat, together with 
reduced tire traction, is not a confidence-builder for speeding up your 
cornering!

A high-performance car like the Pantera has anti-sway bars at both ends of 
the car.  There is a relationship between the two bars that affects the 
cornering, called “roll distribution”.  Increasing the front roll stiffness 
loads the outside front tire and unloads the inside rear tire.  This tends to 
cause understeer by increasing the front-tire “slip angle”.  This is done by 
increasing the stiffness of the front anti-sway bar.  Conversely, increasing 
the stiffness of the rear bar will unload the front tire, increase the load 
on the rear tire and the car will “oversteer”.  Understeer is when the 
steering wheel has no effect on the direction of the car, which goes off the 
outside of a turn.  Oversteer occurs when the rear attempts to pass the front 
in a turn, and the car goes off backwards.  It has been described thusly: 
“Understeer is when you see what you’re gonna hit, while with oversteer you 
don’t.” 

Mid-engine cars are often designed to be neutral, neither over- nor 
under-steering.  Unfortunately for those with racing aspirations, when Ford 
became involved with the Pantera, they redesigned the front suspension for 
understeer.  Understeer is what nearly all street cars have straight from the 
factory, because it’s self-correcting: if you find yourself too fast in a 
corner, slow down a bit and the steering works okay again.  But what they did 
not design out is another characteristic of mid-engine cars: ultra-quick 
steering response.  Because the main mass of the car is very near the center 
of balance, things happen extremely fast in a mid-engined auto when one end 
or the other begins sliding.  My experience in autocrossing Panteras leads me 
to say that, if the front or rear tires break loose in a turn, it’s very 
difficult to catch the car without what a one writer called “a lot of untidy 
elbow-flailing from the driver”.  Both the Porsche 928 and 944 were designed 
from a “clean sheet of paper”, but this “twitchiness” led the Germans to 
design them both with the main masses at each end rather than concentrated in 
the middle, just to slow down their handling for the benefit of the average 
driver. 

A complicating factor in this balancing act between front and rear bar 
stiffness, front and rear tire sizes and steering response time, is the 
limited-slip differential.  The tighter your “posi” unit is, the more 
tendency there is for oversteer, because the rear axle acts like it has no 
differential, or like what dirt-track guys called a “Lincoln-locker”, from 
their practice of using a (Lincoln) arc-welder to weld the spider gears 
together for racing.  This is particularly troublesome because the degree of 
“lock” from a posi-unit changes with gear lube temperature, confusing your 
diagnosis of handling woes.  A quick check of the amount of traction one gets 
from a limited-slip: GM’s test is to wet down the concrete under one tire, 
then lay a two-by-four in front of the other wheel.  If the rear wheel can 
barely climb the two-by-four, a brand-new limited-slip is set just right!  
But, by this time in our 20-plus-year-old Panteras, we probably don’t have to 
worry about the posi being too tight!

OK, so what’s the effect of changing swaybar sizes a bit?  A stock Pantera 
has a .750" bar in the rear and a .845" one in the front; the lever arm on 
the front bar is shorter, so it is stiffer-acting than the longer rear as 
well.  When 10" rear wheels are mounted, the factory recommended increasing 
the rear bar to .875".  That 1/8" increase raised the bar stiffness by 
roughly 50%.  A 1" rear bar would be 3.15X stiffer. When Rich Agiorni and I 
autocrossed his ’71 back in the ’80s, we ran a 1-1/8" rear bar and decreased 
the front to .675"!  For autocross/solo-II’s fairly low-speed turns, this 
worked great.  The hollow bars we used were 5% less stiff than solid bars 
would be, but 40% lighter!  I’m not sure what the car would’ve done with this 
chassis setup at real high speeds, either.  I suspect it might’ve been a real 
handful, though, because it was so super-quick-handling in turns!  
Interestingly, a noted midwestern Pantera owner has terrific success 
autocrossing his car with a stock rear bar and no front bar at all! 

Larry Stock had a new motor built after losing the mostly-stock original at 
Las Vegas ’92 to a spun rod bearing.  We think this occurred due to Larry 
using Goodyear race slicks with a stock oiling system.  The new motor is 
substantially stronger than before, so he really wanted to maximize the 
handling of his now not-so-stock Pantera.  A 100-ft diameter skidpad was laid 
out in a lot across from his business.  The surface was an average 
asphalt-bonded stone, about five years old.  A G-analyst from Valentine 
Research was solidly mounted in his car per directions, on the console back 
in the knick-knack tray.  This position seemed to be very close to the center 
of balance of the car.  The good folks at Pantera Performance Center in 
Denver, Colorado, provided a variety of swaybars for us to test and compare.

The car is a ’72 Pantera, with stock suspension components except for Carello 
shocks, running 15x8 front and 15x10 rear aftermarket wheels with Pirelli 
P7’s in the standard 225/50-15 and 285-50/15 sizes.  The wheels had stock 
offset and required no flares or spacers to fit.  The shocks were adjustable, 
but were not optimized for the various swaybar combinations used.  To obtain 
the best handling characteristics, the ride was lowered, and the front end 
was set for 1/8" toe-out and -1.5° camber.  

In Test 1, the car used stock front and rear swaybars; 0.845" front, .750 
rear.  It was able to maintain 0.79 g’s on the described skidpad.

Test 2:  the rear bar was exchanged for a .875" unit.  This is the so-called 
GTS or Group 3 swaybar setup, recommended when running larger-than-stock rear 
wheels/tires.  The car was now able to maintain 0.93 g’s with this single 
change.  Since the bar change took over half an hour, the tires had obviously 
cooled back down to ambient before starting Test 2.

Test 3:  figuring that if some is good, more is better, the front bar was 
exchanged for a 1.0" bar, while the rear stayed at .875".  Now, the car could 
only hold 0.91 g’s — a slight step backwards from Test 2.

Test 4:  the rear bar was increased to 1.0" along with the 1.0" front bar 
from Test 3.  The car could only hold 0.89 g’s while in this configuration, a 
further step backwards.  This is especially confusing, since it is the 
configuration used by some for road racing, and was recommended by the folks 
at Pantera Performance as the “killer setup”.  However, they have used this 
combination mostly on cars with 15x10 front and 15x13 rear wheels, i.e. Group 
4 or GT-5 cars, which not only have much more rubber on the ground, but 
possess a much wider track, as well.  

For Test 5,  the car was returned to the test 2 configuration and Larry’s 
favorite gumball race tires were mounted.  In studying the printouts from the 
G-analyst for this Test, it can be seen that the car now turns left at 1.10 
g’s and turns right at 1.25 g’s.  Left turns are smooth and predictable while 
the right turn segment was ragged and obviously right on (or over) the edge.  
With cornering power like this, it is easy to see why Larry had oiling 
problems on the racetrack.  The oil in the stock pan rode right up the side, 
away from the pump pickup.  After a few episodes of the pump sucking air, 
there were no bearings left!

Most people can turn to the left faster than to the right, simply because in 
a hard left turn, the whole car is pivoting around the driver’s seat.  During 
right turns, the driver’s position is also pivoting with the car, so his 
point of reference is constantly changing during the turn.  This introduces 
another variable or two which destabilizes the driver’s reference points.  
The result is, you tend to slow down (unless you are in a ‘banzai’ mode, but 
this is difficult to maintain for long, and often results in a trip to the 
tules).

So why didn’t the progressivly stiffer bars give progressivly better 
performance?  The answer probably lies not with the swaybar or suspension, 
but with the chassis itself.  Noted engineer Kevin Cameron recently wrote a 
piece on the relationships between frame and suspension stiffness in 
motorcycles, and used automobiles to explain his theories.  He described the 
fact that any suspension is three springs in series — first is the tire, 
deflecting to absorb the smallest bumps; next is the suspension spring and 
associated components; behind that is the flexibility of the chassis itself, 
being deflected by the forces transmitted through the suspension.

Stock car racers learned years ago that there is no point in putting stiffer 
suspension on a chassis too weak to support it.  As a bump pushes the wheel 
up, the spring, shock absorber and sway bar resist the motion, passing on the 
force to that corner of the vehicle, which bends upward as well.  Once the 
bump has passed, the shock absorber prevents the wheel from snapping back, 
rebounding off the pavement again, and continuing to oscillate.  But the 
chassis has no shock absorber to damp its motion so it continues to vibrate 
up and down.  This continuing motion can be just as disturbing to tire grip 
as running without a shock absorber.  The correct response would be to 
stiffen the chassis (notoriously weak in the Pantera, particularly in the 
rear, although extrordinarily strong by street-car standards), to force more 
of the bending to occur in the suspension, less in the chassis.  However, 
this would require major re-engineering of the chassis, and the costs of the 
design and fabrication would be prohibitive to all but the most die-hard 
Pantera crazies.

It seems that most Pantera owners would be well-served by changing their rear 
bar to a .875" unit and leaving the front bar alone, particularly if the 
tires have been upgraded to at least 1980’s-spec sizes.  Besides increasing 
the overall cornering abilities of the car, the driving characteristics will 
likely become much more neutral, without the plowing of the front end that 
Pantera owners have endured for decades.  However, such a setup will create 
the need for circumspection when driving, for once the car reaches it’s 
(higher) limit, it will probably be much less forgiving!