KevinK2

This was part 3 in a series, hitting on roll centers and lowering. Author suggested lowering the CG would not reduce weight transfer (cornering) if roll centers dropped much more than CG. Wrong. Even came up with a misleading CG axis that he calls his own. Can't believe this was published .. Dave C is usually correct, but this guy Mark X was way off. Truth is, for lowering a reasonably stiff chassis:

1) total weight transfer will be proprtional to the single point CG_1 height for the total car weight.

2) body roll increases proportional to distance from single point CG_2 (for the sprung weight only) distance to the roll axis. Stiffer springs tend to offset increased roll due to lowering.

Guy was trying to use front and rear cg's in his analysis, which was his major blunder.

jgrewe

iTrader: (1)

Wouldn't the fact that the roll center usually lowers more than the CG affect weight transfer? I have not seen the article but the car will require more roll stiffness,either from springs or bars, to corner with the same degree of lean if the RC(s) drop(s) more than the CG. I wonder if he was thinking about Fr/Rr roll centers, as far as I know ther is only one CG. What's on the cover I'll see if I can find the mag.

DamonB

I'll have to stop by a newstand and read that article.

It's sad that it was published if what you say is true. Now more people will read it and not realize it's fundamentally flawed

DamonB

No. Roll center is merely the axis that the chassis rolls around as a result of weight transfer. The reason weight transfer is present is due to the fact that the reactions input into the chassis originate from the tire contact patches at ground level, and the CG of the vehicle is some distance above ground level. If you could somehow keep the tire contact patches and the CG in the same plane there would be no weight transfer.

Roll center doesn't matter. We could weld solid bars in place of our springs and the car would no longer roll, but the exact same weight transfer would still be present because we didn't change CG location, track width or wheelbase.

Gene

iTrader: (1)

The ghost of Carroll Smith will haunt them for that.

KevinK2

Correct, if RC drops more than CG, more roll stiffness needed to keep same lean (see my point 2). But, that has nothing to do with total weight transfer.

There are really 3 mass centers to be considered:
1) total mass (predicts total weight transfer)
2) just the sprung mass (predicts weight transfer due to body roll, and due to related shear loads at the roll centers ... not carried through springs)
3) unsprung mass (wheel tire axle brakes ..)

Although many "guides" just use the total mass center, my guru, Fred Puhn, describes all 3 in "how to make your car handle".

It's the August issue.

jgrewe

iTrader: (1)

The weight transfer part is where I'm having trouble bringing up data from the old, small hard drive between my ears. I've got to find my old books and reread. Slowly but surely this stuff is coming back after not thinking about it for almost 10 years.

peejay

The problem/myth/idea that needs to be beaten into the ground with fire axes, is that people seem to think that body roll = weight transfer.

No.

Lowering the car lowers the CG and reduces weight transfer.

Lowering the car *IN MOST CASES* lowers the roll center faster than the CG drops, making the roll couple (distance between roll center and CG) longer. All else being equal, this means you get more body roll.

I can think of two instances where this is not the case: swingarms where the suspension pivot is parallel to the axle (think front suspension on an old Beetle or 356), because the roll center is the ground no matter what. And, solid axles. Depending on the type of lateral support used, the roll center can stay with the chassis, stay with the axle, or something in between. 1st-gens are awful for lowering because the rear roll center stays with the axle, so when you lower the car, you shorten the rear roll couple. Dramatically. That's why lowering springs don't lower much in the back, tend to be *softer* than stock in the back, and you still want to throw away the rear swaybar...

jgrewe

iTrader: (1)

So the only thing you can do to affect weight transfer is to get the CG as low as you can for any given track width? All the while trying to keep the suspension geometry at angles that keep your roll centers and camber changes etc. etc. doing what you want, where you want, and when. I think I'll skip finding the article after all, I learn(sometimes re-learn) more here.

KevinK2

Based on Puhn theory, there is no short rear roll couple to worry about. Assuming a stiff chassis, there is the roll couple at the sprung cg only. The problem with the 1st gen, as described, is the high rear roll center, that takes the rear sprung body weight and creates a "shear" induced moment to the ground, for too much weight tranfer, even with most of the roll resistance biased to the front. In theory, a Z-bar help provide take verticle loads, and provide no roll stiffness.

I had similar issues autoxing and tracking a triumph with rear transverse leaf spring and true swing axles. Final versions of the car had a center pivot connection between ne "swing spring" and frame for about zero rear roll stiffness. I made adjustable 9/16" rear sway bar, easily flexed by hand, to dial in exact roll stiffness needed in rear.

JEC-31

And the ghost of Colin Chapman will just laugh...

DamonB

Here's a great article from Mark Ortiz that discusses roll centers and weight transfer:

http://www.auto-ware.com/ubbthreads/...n=0&page=0#602

JEC-31

Excellent reading - thanks for the link Damon