Roll centers

[scott fairchild]

Is there such a thing as too low regarding roll centers? I know there is with ride height...

[alangbaker]

You'll get a lot of debate on this point, but the answer (I believe) is: it's complicated.

I think that the one disadvantage that I can see might be that an extremely low (or extremely high for that matter) centre means that there is going to be a lot more track change with suspension movement.

Also, just as a high roll centre can lead to the car jacking up under lateral loads (think VW Beetles and—particularly—Corvairs), a low roll centre leads to the car "jacking" down.

[S Lathrop]

The only restraint on roll center location is that you do not want it to pass through ground level as a car goes through its normal suspension movements. Whether it is above ground level or below depends on the design of the suspension and the tires you are using. What happens as the roll center passes through ground level is subject to a good debate. But I feel that having it stay on one side or the other of the ground is a good practice.

What is really important is to try and understand what the designer of you particular car intended things to be and start from that point and develop the car to suite you as the driver. You need to understand all the suspension variables and how they interact with each other.

[scott fairchild]

I'm pretty sure mine is staying below ground plane (Z10C), but wanted to explore making different spacers to get it lower if there was a benefit.

[Daryl DeArman]

Is there such a thing as too low regarding roll centers? I know there is with ride height...

Depends on what you did to get it there .

I'd be more concerned with how much it moves around during roll/bump/droop and how different the RC heights are front and rear.

[alangbaker]

The only restraint on roll center location is that you do not want it to pass through ground level as a car goes through its normal suspension movements. Whether it is above ground level or below depends on the design of the suspension and the tires you are using. What happens as the roll center passes through ground level is subject to a good debate. But I feel that having it stay on one side or the other of the ground is a good practice.

And the sum of all I've read on the subject suggests very strongly to me that it is of complete indifference if the roll centre should happen to pass through the ground level. See: that's what I mean...

What is really important is to try and understand what the designer of you particular car intended things to be and start from that point and develop the car to suite you as the driver. You need to understand all the suspension variables and how they interact with each other.

Agree with that 100%

[DaveW]

IMO, and I've had this discussion with Steve L, the theoretical singularity indicating that when the roll center goes through ground level, geometric roll stiffness goes to infinity and the car cannot roll is invalid. Bill Mitchell, in his force-based suspension analysis which came after geometry-based analyses that predicted the singularity, discovered that the singularity did not really exist.

If one wants a practical data point, consider that theoretically, the front suspension on a FV has a roll center exactly at ground level due to its trailing arm setup. If that were actually the cause of a singularity that resulted in infinite front roll stiffness, all Vees would terminally understeer. And we all know that the issue with Vee handling is exactly the opposite, requiring (close to) zero-roll-stiffness rear suspension and a very stiff front anti-roll bar.

So, IMO, roll center height effect on handling is a continuum, and does not have a singularity or discontinuity at ground level.

[Jayeff]

I'm pretty sure mine is staying below ground plane (Z10C), but wanted to explore making different spacers to get it lower if there was a benefit.

So the real question here is, " will lowering it even further increase grip?"
I'm pretty sure the answer is maybe.

[Fmracerbob]

The location of the roll center is more than just it's height relative to the ground plane. It also needs to be considered with relation to lateral location relative to the center line of the chassis as the car moves in both vertical and roll motions. And the transfer of weight is effected by the relationship of the distance from the center of gravity(mass) to the roll center. It's a lever arm effect complicated by the fact the CG height(mass) varies long the length of the chassis. Just working with the roll center height at one chassis position can be misleading. The forces acting on the car are not static, they are dynamic as the car rolls and ride height changes.

Having begun analyzing geometry starting in the early 70's with one of the first two dimensional programs right up through Bill's software I've seen a lot of different configurations. In a discussion some years ago with Aurthor Mallock (creator of the Mallock race cars in England), who also had been analyzing geometry for a long time , we found we both had come to the same conclusion that just focusing on the roll center height in relation to the ground is insufficient.

We agreed that the there were a couple of conditions that almost always were present on cars that handled well, and not so with poor handling cars. As tracked through the range of ride height and roll angles as the car moves around the track, the good geometry almost always had minimal lateral movement along with minimal changes to the distance from the roll center to the center of gravity( mass). These seemed to be more important than the actual height of the roll center. Imagine a pole vaulter ,who after planting the pole in the little box , finds the pivot point moving around as they try to raise up and over the bar. Not likely to work well.

As S Lathrop said, it's not so simple. Kind of like alignment, changing either camber or castor, or even toe, can have effects on the other settings, including corner weight.

[alangbaker]

I would agree with what you said...

..except the roll "centre" is anything but the place around which the car rolls.

There is not movement of the "pivot point" because the roll centre (as typically defined) moves.

The answers are in force-based analysis.

[jharvey]

Just make sure the front and rear are close to the same.

Geometry is a rabbit hole few people truly understand. While are chasing a supposed roll center problem, you could create and anti-squat/anti-dive issue.

Unless a major tire change takes place on a car, most handling issues can be traced to some other issue, or with driver itself. These cars have been around forever.

[Jnovak]

Ford Motor Co has a laboratory that can define where the force based roll center positions are located etc.

They have had just about every type of race car in the lab. From F1 to NASCAR. Lot of learning happened in that lab!

I am very glad that i had the opportunity to work on that project!

[DanW]

Ford Motor Co has a laboratory that can define where the force based roll center positions are located etc.

They have had just about every type of race car in the lab. From F1 to NASCAR. Lot of learning happened in that lab!

I'd love to have a tour of that lab from Dave Harmison.

[Dave Harmison]

How'd I get pulled into this? I retired a couple years ago.

I'm reading but not commenting on an open forum. I'll discuss it at the track with people that want to talk but I'm not writing a thesis here.

[Jnovak]

I'd love to have a tour of that lab from Dave Harmison.

i think that Dave might be able to do it. Call him up.

[Jnovak]

How'd I get pulled into this? I retired a couple years ago.

I'm reading but not commenting on an open forum. I'll discuss it at the track with people that want to talk but I'm not writing a thesis here.

i did not know that you had retired Dave.. CONGRATS!!

sorry that I stated that you could do it. I did not bring your name up!