The Ultimate Guide to Suspension and Handling: Part 9 It’s All in the Geometry – The Roll Center

,

Believe it or not, a drift car is perhaps one of the cars most sensitive to roll center location.

 

The biggest effects that roll center location has on a car’s handling have to do with how the car responds to steering input in a corner, how much the car rolls over in a corner and how friendly the car’s balance and mechanical grip at the limit are.

 

Here is a good picture I stole showing what a roll couple is and how it affects body roll.

 

The distance between the roll center and the center of gravity is called the roll couple.  The CG location for each end of the car can be located by jacking the car up a known distance from side to side while it is on corner scales, and observing the change in corner weights. Then the data is fed into an equation to give you the coordinates of the CG. Since it’s fairly safe to say that most people don’t have a precisely flat surface and expensive corner scales, it’s usually safe to estimate theat the CG for the front suspension is around crankshaft height. In the rear, it’s usually at the floor of the trunk. The roll couple is the lever arm that centrifugal force working on the CG uses to make a car lean over in a turn about the roll center.

 

Think of a moment arm as a lever arm where centrifugal force works to lean the car over.

 

The longer the roll couple, the more leverage centrifugal force has on on the suspension through the center of gravity and the more the car will want to roll in a turn. A longer roll couple makes cars slower to respond to steering input. The resulting weight transfer from a long roll couple and high roll angles does not have that much effect on overall weight transfer but it will increase dynamic weight transfer which can make a car twitchy and harder to control at the limit of traction.

 

Although intuitively, it seems like a lot of body roll would cause a huge amount of weight transfer, it is not the case.  Body roll negatively affects vehicle response to driver input by slowing it.  The dynamic weight transfer associated with the roll can make the car twitchy at the limits of adhesion where there is little forgiveness for control overshoots.  This is one of the reasons why you want to limit roll.

 

If you draw a free body diagram, you can see that the CG shift laterally under roll is very small, even with this super exaggerated diagram with tons of roll.  Typically the weight transfer due to roll is under two percent, even in a car that rolls a lot.  You are not gaining much by limiting roll.  Lowering the CG and or increasing track width are much more effective means of reducing weight transfer.

 

19 comments

  1. If i understood correctly, you want the roll centres close to the centres of gravity, but always below them? I looked at the rear axle multilink setup on my car and it looks like the roll centre, unloaded, is somewhere just above the boot floor. Im guessing the centre of gravity is somewhere there aswell. I have lowered the car about an inch at back because it got very light and skiddish under 0.8+ G braking, now i think the roll centre was going above the centre of gravity. Anyway, problem completely gone with lower and stiffer oem springs. No body roll in the car to speak off, so will not lower the front axle as the control arms are about horizontal already.

    1. You don’t, necessarily want the roll centers close to the CG and certainly not above them because that would induce jacking. Also checking the roll center unloaded is not very useful, it should be checked at ride height.

  2. Unloaded meant without load in the car, it was most certainly at ride height. The car is an alfa romeo giulietta, with a rear setup very close or even identical to a 2013 dodge dart. All the links on the body are higher than on the wheel side and just inches below the boot floor. I ordered the car with a dynamic chassis setup which had slightly stiffer and higher springs at the back. Cars with a heavy glass roof get the same springs to correct their ride height and attitude, but mine being lighter got very dynamic indeed. The front however seems very well sorted

  3. Ok, I looked at the Dart rear suspension and the roll center probably isn’t that high. It looks to be about 8-10″ off the ground on the Dart or just under the midline of the wheel.

    1. I ll try to work out the lines in detail from upper and lower mount, and from center of tyre contact next time its on a 4 post lift. Today i went around a roundabout a few times and between 1 and 1.05G lateral the rear right tyre starts squeeling. I can keep it there indefinitely and ESC doesnt quick in but i would have expected the squeel to start at the front. The rear tyres are new and the fronts are nearly worn so that will have an effect aswell. I will investigate the rear camber and toe settings next to see if i can raise the G limit some. Dont want to mess with it much as its very tolerant in changing direction right now. was thinking to modify the front lower control arms for a bit more track width and caster/camber but will leave it for now or just add a little caster as it would be easiest. Love to see your work here though Mike. Thanks for your time!

  4. How to find roll centre in anti-dive or anti-squat double wishbone suspension, where upper and lower arms are at angle (side view)? Thanks

    1. The roll center isn’t anti-dive or anti-squat and isn’t related. I think you need to read the article because I am not sure you understand what it is.

  5. Hi mike, could you suggest a way (if any) of raising the rear roll centre on a multi link rear suspension, the likes of that found on the mk5 golf?

    1. There a lot of ways to do it but I suggest not trying to mess with this unless you really know what you are doing and even why you want to do it.

  6. Amazing info and very concise and understandable, many many thanks. But, uh….typo? Second to last paragraph says: “If the space between the two lines is less in the front of the car, with an upward sloping Mike axis, the car will tend to understeer …….. If the distance between the lines is greater at the front and less in the rear, the car will understeer…” Should not the last example say “oversteer” and not say “understeer” again?

  7. Roll moment depends on both distance og cg to roll center and distance of roll center to ground .If you lower roll center the distance of roll center feom cg increases whereas distance from ground increases.So will the roll moment increase or decrease?

  8. I lowered a vehicle with McPherson strut by 35 mm does it mean that the roll centre is closer to the C.G now as it isn’t an extreme lowering?

  9. Thank you for the reply mike !
    So if I reduced body roll of a vehicle using stiffer lowering springs, did the body roll decrease because of decrease in load or weight transfer in vehicle? Or is it just because of reduced roll centre to C.G distance?

    Thanks in advance

    1. Mostly because of the stiffer springs. It’s hard to say about the roll couple height without actually measuring whats going on.

Leave a Reply

Your email address will not be published. Required fields are marked *

*
*