Sale ends todayGet 30% off any course (excluding packages)

Ends in --- --- ---

What should my roll center movement look like?

Suspension Tuning & Optimization

Forum Posts



Tech Articles

Discussion and questions related to the course Suspension Tuning & Optimization

= Resolved threads


Ultimately I just don’t know what “good” looks like. I measured my suspension and used VSUP to model it (car is a 2011 STI McP up front and multi link in the rear). Granted my measurements are imprefect and VSUP has its limitations, but beyond having the front roll center “close to the ground” I don’t really know what is good. As modeled, the front roll center quickly moves below the ground which I have been led to believe is not a good attribute. I’d like to understand this better so I can make educated decisions about modifying my suspension geometry in the future.

-Should I care about dynamic roll center location?

-If so, what should I look for in regard to McP roll center movement and multi link roll center movement?

I know the answer is always “it depends” but any insight would be greatly helpful. Thank you. I’ve attached pictures of what I have modeled in VSUP so far.

Attached Files

What you want, is to see the roll center move up and down with ride height changes. This allows you to use ride height/rake to change the balance of the car (by modifying the amount of instantaneous weight transfer through the roll center).

What you try to minimize is the migration of the roll center left/right as the car rolls. I've seen geometries that result in the roll center at the contact patch (I think that happens with heavy preload / drop limited suspension with soft roll resistance).

Thanks for the reply! So, looking at what I have in VSUP, my front roll center migrates a lot, but not the rear. The main difference that seems to drive the difference is the instant center location. The closer the instant center is to the center of the car, the less the roll center migrates. However, the closer the instant center is, the smaller the front virtual swing arm becomes, which would generate great track width and camber variation for a given deflection. Any idea where the right amount of trade-off is? Also, I'm doing this in two dimensions, so the simulation is inherently wrong. Also, I'm just some guy trying to learn, so please let me know how wrong I am.

It's a pretty complex set of compromises that you need to make with suspension geometry and there isn't really a 'right' compromise so we come back to 'depends' which I understand isn't overly helpful so let me see what I can do to add some clarity.

First of all, without a significant redesign of the suspension pickup points you're going to be somewhat limited on what you can do to manipulate the roll centre height and roll centre migration. Let's deal with these terms separately. The static roll centre height and the centre of gravity height for each axle will dictate how much the chassis will roll for a given cornering force. Typically when we lower a car the roll centre is lowered MORE than the c of g meaning the car will want to roll more which isn't desirable. Fortunately for popular cars there are roll centre correction kits which more or less gets the roll centre height back to where the factory intended. However that's not to say the factory got it right as their design criteria almost certainly aren't to provide optimal handling on a race track.

A key element when you're considering roll centre height is the roll centre axis (which is an imaginary line drawn between the front and rear roll centres when viewed from the side). By raising or lowering the roll centre at one end of the car only, you alter this axis which will affect the handling balance of the car from understeer to neutral or over steer. Ultimately though it's a case of making a change and assessing the effect of that change - I can't tell you what you should do.

Now on to roll centre migration. While we often talk about roll centre like it's a static point, it will move around as the car rolls and pitches. You'd probably need to use something like Optimum Kinematics to plot this accurately. In the perfect world we'd have little or no roll centre migration but that's not very realistic so we're always going to end up with some. The less we have, the more predictable and consistent the car is going to feel.

This leads me back to the start though - If you're dealing with a production car then the roll centre migration is going to be defined by the factory geometry and within reason you're stuck with that. You can (and if you've lowered the car, should) consider roll centre correction kits for roll centre height but again you may want to do some careful testing to see how this affects the handling balance.

We usually reply within 12hrs (often sooner)

Need Help?

Need help choosing a course?

Experiencing website difficulties?

Or need to contact us for any other reason?