"John D." wrote in message
...
On Mon, 12 Apr 2010 10:35:15 -0400, "Ed Huntress"
wrote:
"John D." wrote in message
. ..
On Sun, 11 Apr 2010 10:23:22 -0400, "Ed Huntress"
wrote:
A great deal of previous posts deleted
As you describe it
cornering force alone overcomes every other force and lifts the entire
back of the vehicle enough for the camber, which would have probably
been at least a degree or so negative as a result of body roll to
suddenly go, from your pictures at least 10 degrees positive.
To go through the whole thing, analyzing all of the vectors, is fairly
complicated. My chassis books are not handy and I'm not going to do it
now.
I'm sure you can do it yourself, if you've analyzed chassis before and
think
through all of the loads.
But one fact here may help: When the cornering force is first applied, the
downward force measured at the differential is zero, because the car
springs
are supporting the weight of the car. So when that upward vector is
applied -- the resultant of the couple I described before -- it's lifting,
initially, against zero resistance. In other words, it takes very little
force to lift the rear end of the car a little bit. And doing some numbers
in my head, I'll guess that the lifting force, at maximum cornering of
around 0.6 g or so (the most you'll get out of those tires), is well over
100 pounds. Then the jacking reinforces itself because the bottom of the
outside tire is moving in toward the centerline of the car, and the
vectors
result in more upward force for a given cornering force.
No Ed, the differential is attached to the frame so weight
(disregarding chassis flexibility) would be the same at the two spring
supports and at the deff..
So what's holding up the differential? What's supporting its weight? It's
the springs. Follow the forces, using vectors. Jeez. g
Unfortunately all my references (long gone after 50 years) were in
reference to single seat cars, or full bodied cars designed as rule
beaters by being the minimum width at the cockpit for two seats. but
still essentially single seat chassis which generally moved around a
roll center.
John, ALL cars roll around their roll center.
I suspect that what is happening with the "road cars" is that the body
is actually tending to rotate around the outer spring, in essence
attempting to roll over.
It can't. It rolls around its roll center.
This would lift the diff. high above any
normal position and of course create wildly positive camber.
No. It's not roll that causes the rear to lift. Roll CAN'T lift a car. It's
rotational force. It compresses the outside spring and decompresses the
inside one in equal amounts -- a rotational force that results in equal and
opposite (up and down) forces on opposite sides of the car.
The lift at the pivot point is the resultant of a force vector, the vertical
force resulting from the couple we talked about. That's the only place it
can come from.
It doesn't seem likely that simple cornering forces applied to the
narrow tires that y'all were using, which would likely be, at best,
modified street compounds, could generate sufficient torque to raise
the entire rear of the car without some assisting force.
If they get enough grip and you're cornering hard enough, the car jacks.
Then, if you're lucky, the tires slide and you just spin out, rather than
rolling over.
By the way, here are a couple of URL's showing non-jacking swing axle
cars.
Pre WW II
http://jalopnik.com/5310050/jenson-b...cedes+benz-w25
Great car.
Post WW II
http://vodpod.com/watch/2937590-merc...96-jochen-mass
Another great car.
Also keep in mind that the body roll, while it's acting downward on the
outer spring, is dead neutral on the car's centerline -- and thus on the
diff. It is not supplying any downward force to oppose that lifting force
vector. The car is rolling *around* the roll center: one side depresses
while the other side lifts.
It gets complicated with further considerations of body roll and its
effect
on camber, and with considering the offset between the car centerline and
the actual pivot point on each half axle (this is shown in the
illustration
linked to above). I'll leave those things for you to work on.
On the off chance that we discuss this again, let's clarify one more
thing,
so we don't get tangled in terminology. We've been loose in the use of the
term "oversteer." Oversteer refers only to the effect of slipping, which
is
the result of tire-tread distortion in cornering. The tire is pulling in a
direction offset somewhat, angularly, from the rolling direction of the
tire. It does NOT refer to the effect of a car's rear end coming around
because of *sliding*. Most of what we're discussing here is the result of
sliding, and is not oversteer. Once a car starts to slide, other dynamics
take over.
It's important to keep that clear because the transition from slipping to
sliding is where a lot of the nastiness occurs. A rear-engined car will
not
necessarily oversteer (later Porsche 911s do not oversteer). But a
front-engined car *can* oversteer. An out-of-the-box 289 Cobra, with full
independent rear, oversteers. So does a Bugeye Sprite, with its solid rear
axle. But a Corvair with a stiff front stabilizer bar understeers. Put on
a
rear stabilizer bar, and it oversteers.
Suspension geometry and dynamics determine whether a car oversteers or
understeers. Front-to-rear weight bias *can*, and often does, determine
what
a car does in that regard once sliding begins. But with swing axles,
traction on the rear tires diminishes so rapidly as the car jacks that it
can snap you from understeer to violent tail-end sliding with a snap of
your
fingers. That occurs whether the car is rear-engined, like a VW bug, or
front-engined, like the Triumph Spitfire in the photos above.
I agree that if/when the rear "jacks" awful things happen however, as
seen in the above scenes, properly designed race cars don't jack. At
least the two examples from Mercedes didn't.
They didn't jack because they were race cars that were designed to
compensate for it. They used various suspension tricks. The W25 had
significant negative camber. The W196 had the Mercedes low-pivot swing axle,
which lowered the vector angle of the lifting forces.
If you're going to talk swing axles, you have to study the Mercedes
low-pivot design. The pivot point was below the axle centerline. It required
an extra gear sector in the diff -- typical MB (over)engineering.
Remember when we discussed that negative camber overcomes jacking? It was in
the illustration for which I provided a link. That's one of the ways you
prevent it. Remember that the subject was the Chevrolet Corvair? g It did
not have negative camber (except in the John Fitch modification, which I
have discussed here at length). It also did not have Mercedes-Benz low-pivot
swing axles.
Are we clear on this now?
Not that I have driven one but from what I read the Mercedes swing
axle cars were easy to drive. You mention Fangio (admittedly a poor
example of car drivibility as apparently he could drive anything)
clipping the same spot in the corner each lap.
Again, that was a low-pivot swing axle. Its geometry was much better.
The usual swing axle horror stories are about Auto Union with it's
earlier swing axles and monster engine, but I suspect that engine
power and torque was as much at fault as the axles. There were stories
of them spinning the off wheel during acceleration from engine torque.
I believe they added a ZF posi-traction to stop the wheel spin.
They were slick engineering, all right. Ferdinand Porsche designed them, at
the same time he was designing the Volkswagen.
And, even after they went to the De Dion rear axles the cars never
earned a reputation for being easy to drive.
They just didn't have enough experience yet with high-performance
mid-engined cars. They're tricky. John Cooper was the first one to get it
right, around 1955. DeDion tubes are what you do when you haven't figured
out unequal-length, non-parallel double wishbones yet.
If I haven't confused you yet, you may see that we're actually dealing
with
two kinds of handling transitions. One is from understeer to oversteer,
or,
conceivably (I can't think of an example), vice-versa. Some advanced
rear-engined performance cars transition from understeer to oversteer.
Actually, a Bugeye Sprite does that too, but the transition occurs at
fairly
low speeds. It transitions into oversteer just driving it smartly around
town. (It's the result of the rear-spring configuration; later Sprites do
not do that.)
Without doing more then sketching the rear of a car in a "jacking"
position to try and see what forces are acting on it, I suspect that
the killer is the body not rotating around any central roll axis but
rather around an axis through the outboard spring which combined with
the geometry of the swing axle causes the wheel to have really
excessive positive camber.
It can't rotate around the outboard spring. It rotates around its roll
center.
The other is from slipping to sliding, which can cause any car --
understeering or oversteering, front-engined or rear -- to transition from
any kind of handling to a tail-end slide. Or, in some instances (Allard
J2;
some early Lotus race cars; modern front-wheel drive cars) to a front-end
slide. That has nothing to do with applying power and causing a power
slide,
which is another factor that Bill mentioned.
It's complicated, but what's important here is that swing-axle cars can be
deadly because they're prone to snap transitions from any one condition to
a
severe rear-end slide. Thankfully, no one builds them anymore.
Over steer is a complex action with practically everything effecting
it. Weight bias, suspension design, power, driver actions. But fair
enough to eliminate skidding as normally one doesn't design for
deliberate skids (except for the circle track guys who do some strange
things).
You can simplify it a lot by remembering that it's the result of tire
contact-patch distortion, not sliding. The difference between the expected
rolling direction of the tire and its actual direction, which results from
the distortion, is called the slip angle. The greater the lateral force on
the tire, the greater the slip angle.
When the slip angle of the rear tires is greater than that of the front
ones, you have oversteer. And vice-versa.
But the
Internet is so slow here in the Marina that any research will have to
wait until I get back home to a faster connection.
I had damned Nadar for all these years when he was right......but what
the hell, I'm not going to start lauding him with phrase, I plead the
rights of RCM to continue my own cockamamie view point, evidence to
the contrary be damned :-)
Cheers,
John D. Slocomb
(jdslocombatgmail)
You aren't alone in damning Ralph Nader. I did, too, for years, until I
got
the hang of what he was doing.
It is interesting how many people get by doing what Nadar did/does -
talk with an authoritative tone of voice, throw out some sort of
statistics to "prove" the point and viola, an expert. Never mind the
Internet, just get around the "yachting" crowd and listen to them...
I have. g
The worst are consultants.
--
Ed Huntress