Certainly with modern tires on a road car camber gain is fairly small potatoes, in the linear range, and roll centre height is important.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies http://eng-tips.com/market.cfm?

Let me try a simplification here. Roll center height determines how much a vehicle will roll in a turn. Camber gain keeps the wheels cambered (tilted) correctly relative to the road surface as the vehicle goes through this rolling motion. A high roll center reduces the tendency for the chassis to roll and thus lessens the need for camber gain. There are other issues related to high vs. low roll centers and control arm linkage design to lessen camber loss on chassis roll, and these can introduce new factors affecting handling. But one thing at a time. And remember, the amount of roll has no effect on weight transfer from inside to outside wheels on a curve. Only the track width and height of center of gravity affect this weight transfer.

I always liked the modular design of the early Corvair front suspension and wondered why they weren't more universally adapted like VW front ends (now there's an example of a zero camber gain suspension). I fear the Corvair front suspension may have been designed for heavy understeer to compensate for the natural oversteer tendancy of the rear end (like GM did with their muscle-cars so they go off the road nose-first instead of tail-first).

I got my drivers license in a '61 Corvair. And they were as NASTY as Nader said they were. I spun it several times. Even my conservative sister went off a freeway ramp in one of these.

Bob

Sorry that's wrong in detail. RCH directly controls weight transfer on a given axle, that's why it is important.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies http://eng-tips.com/market.cfm?

I think the center of gravity height controls weight transfer, not the roll center (moment center) height.

The floor is open...................

Bob

I'm one of those on the other list...
Since the Corvair (in reasonably streetable condition) has the outer front going into positive camber w/r the road surface, we were considering approaches to get a bit more camber gain, to keep it flatter on the road.
The main approach in consideration was a aftermarket upper ball joint, to effective lengthen the spindle and alter the geometry a bit.
We'll ass/u/me they're dealing with the rear suspension in some fashion...

Re stock handling characteristics - I've never driven a stock Corvair hard enough to spin it.
I did spin my modified Corvair, back when it had the 327 midships. Crappy old tires on the rear probably didn't help.
I did spin my early Beetle, too. And that one had good fresh radials on it.
The results in both cases were the same - I came to a stop at 90 degrees to the intended direction of travel.

cheers
Jay

Jay Maechtlen
http://www.laserpubs.com/techcomm

A swing axle 1963 Spyder Corvair without a camber compensator on bias ply tires comes around real quick when driven by a knucklehead 19 year old snapping the steering wheel on trailing throttle in a fashion that a beat up MGA never would. I, I mean he applied lots of steering lock thru the slow stock steering, did not spin out, but it sure felt like we were way beyond 45 degrees to traffic. Adding a JC Whitney camber compensator made it much more controllable.

Over the next 10 years or so 3 different late model Corvairs never bit the knucklehead like that, although perhaps the k'heads driving skill improved a bit too.

yeah, the stock steering was dreadfully slow, too. When the rear goes, you have to be really quick to catch up.
The OP did mention 2nd gen Corvair, which (I presume) means the 65-69 series.
Anyway, the rear traction/behavior at limit is another topic - we were looking at front adhesion, and if improved camber control was worthwhile.

So, based on the opinions I've heard here, the answer is "maybe not worth it"?
Obviously, the owner could just use really stiff ARBs and call it good.
regards
Jay

Jay Maechtlen
http://www.laserpubs.com/techcomm

Quote (Tmoose)

He also offered the opinion that "handling is affected more by roll centers than camber gain.

IMHO, the «ideal» suspension should have a roll center close to the ground to minimize the tire scrub. Furthermore, it should be designed such that the roll center location varies as little as possible with suspension motion; Especially small displacements.

But this is true when you have a high roll stiffness (which is determined by the springs and the roll center), when suspension motion is very limited under lateral acceleration. Since there is little suspension motion, there is no need to care for camber gain as much.

If you a have a small roll stiffness, then there will be a lot of suspension motion and it is almost inevitable that the roll center will be all over the place. In such a case, it might be more important to look at the camber gain to maximize traction.

So - as usual in the case of suspension design - the answer would be: «It depends».

Thanks - seems like a street-driven car might benefit more, because you'd want a softer on the street than on the track.
It should be real entertaining to see how mine works out - it has longer spindles in front, and I'm giving the rear quite a bit of camber gain. I will probably run it without ARBs at first, or maybe just a small one up front.
cheers...
Jay

Jay Maechtlen
http://www.laserpubs.com/techcomm

http://www.mediafire.com/view/t2m2qi33cclou32/weig...

Formula for weight transfer. p is the roll centre height at a given axle. textbook stuff.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies http://eng-tips.com/market.cfm?

Greg
I'm looking forward to studying those formulas. Where do they define the terms?
Thanks.

Bob

Greg,

I stand corrected. There is a weight transfer attributable to body roll/roll center height. In the building design world we call this P-delta effect. If you have a very high center of gravity, body roll will allow this center of gravity to shift laterally, like a leaning building, and add to the weight transfer. Interestingly, this is a static weight shift induced by the dynamics of centrifugal force.

In most cars today this weight shift is low enough to be ignored in most literature.

Bob

On a quick check on a typical large RWD sedan ignoring the RCH contribution reduces the weight transfer on the front wheels by 10%, which is no biggy by itself, but that shifts the understeer gradient by 0.3 deg/g, which would be subjectively quite significant

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies http://eng-tips.com/market.cfm?

... in overall effects even though the linear range understeer gradient is, by itself, not easy to assess subjectively on a road - very easy on a skidpan of course.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies http://eng-tips.com/market.cfm?

Quote (BUGGAR)

I stand corrected. There is a weight transfer attributable to body roll/roll center height. In the building design world we call this P-delta effect. If you have a very high center of gravity, body roll will allow this center of gravity to shift laterally, like a leaning building, and add to the weight transfer. Interestingly, this is a static weight shift induced by the dynamics of centrifugal force.

I just want to to point the poor choice of words to make sure everything is clear: Weight transfer is attributable to centrifugal force only; nothing else. The suspension design has no effect on the weight transfer magnitude, only its distribution between front and rear axles.

If you have no suspension (like a go-kart), the chassis acts like a rigid beam and weight transfer distribution is basically a static problem with one load and two supports. If the supports are springs, the load is centered and we assume the beam cannot lean, then the stronger spring will take more of the load than the soft one. It is easy to visualize when we assume one spring has an infinite stiffness and the other has zero stiffness, it is easy to see that the load will be entirely supported by the strong spring.

But a suspension is combination of springs and rigid links, which both can react to weight transfer. The roll center measure the importance of one compare to the other. If the roll center is at ground level, the weight transfer distribution is entirely directed by the springs; If it is at CG level, it is entirely directed by the chassis and suspension links (the suspension links are «locked» with the chassis from the point of view of the centrifugal force). Anything in between is a mix of both.

But since most suspension designs aim at having the roll center near the ground, it can be safe to assume that weight transfer distribution is only affected by springs. As shown in the next figure, the exception are cars with a beam axle, which are extremely difficult to design with a roll center near the ground (hence the comment of Greg talking about a «typical large RWD sedan»)

Good discussion. See if the attached makes sense.

Bob

• http://files.engineering.com/getfile.aspx?folder=6420a58b-2a93-4642-b7d1-93

Carrol Smith in "Tune to Win" (page 37 and 38 in my old tattered copy) says lateral load transfer is primarily a result of CG height and track width, but then goes on to say "Actually, the lateral load transfer picture is a bit more complicated than I have yet indicated. It is generated in four separate ways:
(I)By the side forces generated by the tires as they resist centrifugal force. These forces are reacted on the sprung
mass through the roll centers.
(2) By physical compression of the outboard springs due to roll and by deflection of the anti-roll bars.
(3) By the jacking tendency inherent in any independent suspension system.
(4) Lateral displacement of the c.g. due to roll has a minor effect which we will ignore.

Folks who bought legit copies, but who may have left them at home or lost a few sections due to the horse glue binding self-destructing can find a searchable working copy on the Internet

Quote (BUGGAR)

See if the attached makes sense.

I see your point. The CG goes towards the outside wheel, increasing the weight transfer. I stand corrected as well.

Good day all,
most intrigued by the above but more so perhaps by the fact that no one has mentioned the interelated issue of the "jacking effect" on roll centres, especially on swing axle [corvair] layouts except in TMoose,s last post.{Carol Smith} I have always understood that roll centre migration in any form should be kept to absolute minimum, but is it really only confined to dedicated race cars as it in its turn influences [aggravates the jacking effect] and the issue goes round and round. Is it not, as it seems always a question of compromise in design.
All the best Golfpin

Any suspension with the roll center above the ground will experience a jacking force; That's the whole point of resisting the rolling motion.

But if the roll center is above the CG, the jacking force becomes larger than the weight transfer and thus the body actually lifts off. That is when it's an issue as you mentioned.

Swing axles are particularly bad at this because the problem is not the roll center height at rest but their short lever arm. As the wheel moves down, the roll center goes up rapidly and excessively. It's not good on any vehicle that experiences the slightest lateral acceleration, race car or not.

I think maybe that Spitfire's drive is off the gas too. That is when I got into trouble.
As do early Porsche owners, and even BMW owners with semi trailing arm suspensions.

The story in Denis Jenkinson's Race driver book has a section on Sir Stirling Moss driving various Mercedes swing axle or maybe low pivot swing axle race cars. He was able to make them initially with tons of steering lock, but later applied lots of throttle when the tail started to come out. Supposedly some German engineer said, yah we make it that way on purpose, but only very knowledgeable drivers know how to use it. Stirling was apparently 1 of 2 factory drivers that figured it out.

"Hark, the Herald axles swing!"

All independent suspensions with instant centers above ground level will experience some jacking effect. Trouble arises not only if the instant center is above the CG, but also if the spring rates are too soft. As the body lifts up, the jacking effect goes up, that increases the "jacking" effect, and if the "jacking" effect is increasing faster than the spring forces are falling away, up and over she goes!

Some swing axle suspensions were fitted with something called a "camber compensator" which was an extra spring tied between both sides with almost the opposite of an antiroll bar. The idea was that if both sides of the suspension were in droop, it added some spring force to resist further extension into droop.

Swing axle suspensions of the normal design have the roll-center above axle centerline; the M-B low-pivot system was a bit different. That is a very high roll center by modern standards.

Beam axle suspensions can get away with having a high roll center because the force that would otherwise lead to jacking is (correction: should be, if the suspension is properly designed) contained within the axle, it (mostly) doesn't pass through the bodyshell and thus affect the ride height.

"Side scrubbing" is another side effect of a high roll center, and that doesn't matter if it's beam axle or IRS. Go over a one-wheel bump, and the tire contact patch kicks sideways. You feel it in the vehicle as a sideways kick. People used to being in modern cars will feel odd ride motions if they are in a 4x4 vehicle with beam axles front and rear, and this is part of it.

Quote (Buggar)

Good discussion. See if the attached makes sense.

Bob

The sketch does make sense, but it needs to be shown with a more realistic amount of roll. Drawn closer to 4.5° than 45° would provide a better sense of the magnitude of actual CG migration being relatively small.


Norm

Norm,
You've never followed an overloaded mini-pickup in TJ?

Actually I was trying to figure this out late at night and used the principle of exaggeration to make it clear.

Bob