roll centre

[silente]

Hi all!

I am new of this interesting forum but i start immediately asking something about roll centre.

What is in your opinion the effects of changing roll centre hieght?

I know that, if the distance between roll centre and cg grows up, the same happens for the elastic weight transfer. So, for the same value of lateral g you should have more roll. But the total weight transfer doesn't change. what is changing is only the part of this weight transfer which pass through spring, bars etc...

So what exactly happens to the balance of the car changing the rc height? For example, if i make the front rc height bigger...

[graham bahr]

not quite sure what your asking, but maybe thats because i've got a headache,

but the roll centre as i understand it is the point midway of the front and rear axles which the car rotates or rolls around, you picture the front lower outer ball joint and draw an imadginary line through it and through the inner pivot on the chassis, the point at which it would coincide with the same line drawn from the otherside of the car is the roll centre, the higher the line the more the cars natural resistance to body roll, lowering cars usually lowers the roll centre and decreases the anti roll bars effiecnecy, the reason cars dont actually roll further when lowered is because most lowered springs are stiffened.

generally we cant do much about the roll centre, it is where it is, and we just tune the chassis around it, although by moving suspension pick up points we can adjust it to a degree, caution is required, altering the roll centre by a substanical amount at only one end of the car, can lead to some very strange handling

[silente]

ok...

i know the way to determine the position of the rc and i know "matematically" what happens if you move it up or down. What i asked is the effects on car's balance of moving the roll centre...

[graham bahr]

what do you mean by balence? are you talking about basic handling traits like over/understeer? ability to change direction, body roll etc?

[silente]

Originally Posted by graham bahr

what do you mean by balence? are you talking about basic handling traits like over/understeer? ability to change direction, body roll etc?

yes!

What i like to know is how and why car behaviour is affected by changing the rc height. I know that when you change the distance between rc and cg the elastic weight transfer change, but i don't understand what happens to car balance because the total weight transfer should not change! What is changing is only the way this wheight trasfer is using to pass load from one tire to another...

[graham bahr]

ok, moving the roll centre on its own, shouldn't change the basic balence of the car, unless you considerably alter it at one end of the car only,

if you move the roll centre without making any other changes to spring or roll bar weights it doesn't normally have any significant effects as its only effecting the roll stiffness, however depending whether you raised or lowered it it will either magnify or reduce existing tendancies, like how well the car turns in, or roll induced oversteer.

i suppose the reason it has an effect is that it changes the speed of the weight transfer and will have an effect on the total amount of weight transfer, just in the same way you can tune and to a large extent control the speed of (but not the total amount) by adjuating the damping

having said that its rare to change RC on its own, most lowering of cars also lowers the RC but you dont notice because its usually done in conjuction with stiffer springs ARB's etc.

for the most part unless your designing from scratch the RC is what it is and what your stick with, after you've lowered the car to get the COG down and more favourable (for race use) suspension arm angles, unless your able to relocate piviots higher up in the car, although usually you cant move them anywhere near enough make siginificant changes to the RC.

[retro_msport]

On the race escorts we lower the rear RC to get it somewhere near the front RC as the front is near the ground on most Escorts and being live axled is at the center of the diff (23" dia wheel, rear RC of 11.5"), with the Watts link i do this drops down to around 5"-7" which means the car corners flatter without the front inside wheel waggling in the air on power, and the rear inside wheel going light on turn in.

Its nothing new V8 OZ tourers use it and the "mumford link" is designed to get the RC on a live axle as low as possible.

[silente]

i work on formula cars like formula 3 and formula renault...

we change sometimes the roll centre height and we have small effects on the balance above all in corner entry.

But standing on theory, the weight transfer must not be affected by this change, only the way the load is transferred from one side to another..and for sure also the way the suspensions work, because of the change in the amount of roll for a certain lateral g.

Other ideas? Somebody has done other experiments on this change?

[graham bahr]

Originally Posted by retro_msport

the car corners flatter without the front inside wheel waggling in the air on power,

and there was i thinking all the best RWD saloons do that, AKA mk1 cortinas and bmw 2002's and now my E30 now its working properly,

that said i have a feeling it will be doing rather more of that than i want as the front struts seem to of come back from leda a bit on the short side, i've heard of single seaters having zero droop but never a saloon, might have to make up some spacers to lower the top mounts so i can experiment

[Goran Malmberg]

Roll centre is a well discussed subject and surrounded by a lot opinions.
Accordingly I hav an opinion of my own here. The weight transfer is (w*g*cgh)/Tw no matter the height of the Rc. If the cgh is 400 mm and the Rc is 200 mm the car could be said to have 50% antiroll. 50 % of the moment over the 400 mm arm is going throught the springs of the car and 50% is geometrically going directley to the ground.

This antiroll force is very brute to the tire grip and destoys grip, as it makes the car partley "sidway unsprung". If rising the Rc at one end of the car this will have a similar effect as mounting a stiffer rolbar at that end. The difference is that the part of the Wt that goes over antiroll is totally ded weight transfer. (Eexept for the tires). It is better to use an heavier roll bar that also make use of the shock absorbers to dampen the moments of roll, making the weight transfer more controllable.

So, I suggest using close to ground Rc at both ends of the car, and A-arm geometry tha keeps the Rc at the same level during wheel travel.
If Rc alters even the over-underster of the car will alter during wheeltravel in a corner. What we have to lok for is where the forceline intercept with the middle vertical line under Cgh of the car, or should we say the angle of the forceline should be keep constant during wheel travel. Where the left and right forceline intercept dosen matter at all, the car is having its real "hinger" Rc in the midlle of the car anyway. Also, it is the outer loaded side forceline that carry the moost of the weight anyway.

Goran Malmberg

[silente]

i think that you have read mark ortiz pages!

i use to separate elastic and geometric weight transfer, like you do.

The elastic one is dependant from rc height and it influence the part of wt which pass thorugh the springs, bars etc.

Have u made experiments to say that with an higher rc (so with a lower distance between rc and cg) the axle we are working on loose grip?

Because one time i had the same experience than you: i changed rc height (i used an higher one) on the front and the front lost grip in corner entry. But it's also true that the total wt doesn't change!

You mean that the difference is dued to the fact that with an higher rc the wt transfer is less controllable? maybe because the driver feels less the wt transfer when it is happening...

i also supposed that the reaction you have changing rc height depends from your suspensions angles (like camber for example) because what is changing is the tendency of the car to roll...

anyway..thx for all your replies!

[retro_msport]

Originally Posted by graham bahr

and there was i thinking all the best RWD saloons do that, AKA mk1 cortinas and bmw 2002's and now my E30 now its working properly,

No all the best escorts corner flat .....

http://www.rsownersclub.co.uk/video/nwxp.wmv

[Goran Malmberg]

Originally Posted by silente

i think that you have read mark ortiz pages!

I have read moost things and Mark is one favorite. I also have had personal contact with M. Anything in the geometry has an influecnce on Rc and I have made my own physical experiments with geometry and Rc. This makes you find out that I have a slightley different aproach to the problem than M. Anyway, I must spend the day with my family, but I will be back answering your Q soon.
Goran Malmberg

[TEAM78]

read up on roll moment distribution
That will explain how RC's affect the balance of a car

[silente]

Originally Posted by TEAM78

read up on roll moment distribution

where?

[Lukin]

We use front and rear RC adjustments. As has been said here and in that article, raising the RC effectively stiffens that end of the car. Since we run a front engine RWD V8, entry US on turn in/throttle drop is a big problem.

Changing the rear roll centre changes the roll axis, and changes the roll distribution % towards the rear without effecting the overall roll moment. As more load transfer is geometric, it increases the speed of load transfer on the rear axle as well as reducing the roll angle.

The combinations of these two things (increased rate of load transfer and lower roll angle) make a RRC change the first point of call for a car that has US and feels lazy and unresponsive.

On the front, lowering the FRC is used from time to time to cure US too, the only problem is it slightly reduces the responsiveness of the front end, not that it can't be cured with a front damper adjustment.

I thing Goran's comments are good for some types on car/tyre combinations, but not always. Our tyres can cop a pretty high RC (as a function of COG), especially on new tyres. Also, don't forget the influence of the track.

Take two tracks; a fast, smooth circuit with long quick flowing corners (Phillip Island, Catalunya) compared to a bumpy street circuit with 90 degrees corners (Long Beach, Surfers Paradise). Say for example you know what mechanical balance the car likes and you achieve this by a combination of springs, RC and bars.

At the smooth fast circuit I'd say lower RC, stiffer springs and bars as the longer corners might make jacking an issue, and the smooth circuit won't punish stiff springs and hurt in one wheel bump.

At the street circuit, I'd say high RC, soft springs and soft bars. The nature of the corners reduces the chance of jacking causing problems, and the higher RC allows softer springs and bars without hurting the overall balance. High RC on the rear can sometimes hurt exit, but with the bumps there will be a good gain from running softer springs (reducing contact patch load varation) and softer bars (one wheel bump on a bumpy circuit).

Merry Xmas all!

[silente]

Originally Posted by Lukin

Say for example you know what mechanical balance the car likes and you achieve this by a combination of springs, RC and bars.

normally i look to the mechanichal balance without analizing the roll moment distribution, which is connected with rc. normally i use to look to the roll stiffness % (front/(front+rear)) but i don't look to the roll moment distribution..

In which way you look at it and in which way you use it to define the mechanical balance of the car?


merry christmas to everybody!

[Lukin]

Originally Posted by silente

normally i look to the mechanichal balance without analizing the roll moment distribution, which is connected with rc. normally i use to look to the roll stiffness % (front/(front+rear)) but i don't look to the roll moment distribution..

In which way you look at it and in which way you use it to define the mechanical balance of the car?

I define the mechanical balance as the ratio of front:rear vertical loads for any given lateral acceleration (ie. steady state). Without accurate vertical load v traction, it's almost impossible to use the ratio of horizontal loads produced.

Very Similar to this: http://tentenths.com/forum/showthread.php?t=77238

[silente]

just one question...

why the total weight transfer in that sheet is influenced by rc height?

i thought that moving rc what is changing is not the total wt, but only the value of the elastic and the geometric one...

[Goran Malmberg]

Hi again!
I will start of by saying that there is allways exeptions for everything. We all have different cars, tires and other conditions that must be part of the equation. I am trying to discuss a little more independent from “outside” factors as there is factors enough involved anyway.

Lets say we have a wheel rate 27 N/mm for the spring and 13 N/mm for the bar. The bar is 22 mm in diameter. We change the bar for a 19 mm and the result is 9 N/mm wheel rate. We may all estimate for ourself what the difference in front grip and feel would bee from such change, but I assume that we can all agree that there WILL be a difference noticeable.
We had a total wheelrate of 27+13=40 /mm. And we lowered that to 40-4=36 N/mm or 10%.

We all like a BALANCED car. This made me come up with something I named “zero” car. I don’t have any drawings here to picture what I mean, so I try by using words.

The Zero car has parallel A-arms for ground level roll centre. No SAI angle and long A-arms that allow the spindle to be located in the wheel for zero scrub. Parallell A-arm to chassis mounting for zero antidive and antisquat. It ALSO has the same track width as wheelbase, and an equal weight distribution.

When the zero car is subjected to forces, acceleration, braking and lateral G it will get THE SAME NUMBER OF DEFLECTION IN ALL DIRECTIONS.
We may now make it look more like an ordinary car and and shorten the track width. As the moment of roll is w*g*cgh/wheelbase OR track width the sidway deflection-roll will rise from less track width. To get back to pitch-roll balance we now need to add ROLLBARS of the same amount as the reduction in track width.

What I am saying here is that as soon as we leave the “zero car” set up, a compensation setting for the action, has to be made. In the case of a narrower track we must add roll bars. If the SAI is getting large, caster has to be addet to prevent from positive camber in turns. If scrub is added, to in must be used to compensate for forces created around SA. And so it goes on.

We may add an inclined roll axis, which means a higher rollcentre at one end of the car than the other. We then got a certain amount of antiroll, and that portion of antiroll has its own antirolldistribution front to rear acordinly to the angel of the rollaxis. Then we got a sprung weight transfer AND an unsprung weight transfer to take into account to determine over understeer balance.

We may also add some antidive and antisquat, and is should become obvious that we have involved ourself in some pretty complex equation to get back to the “zero car” balance.

Alright, despite the “ANTI” factors built in to the car, we still got the zero spindle-upright configuration. Lets say we like some more conventional spindle layout that include a 7 dgr SAI angle and a 100 mm scrub radius. A camber compensative upper A-arm angle for better cornering camber-grip. This action may alter the wheelrate by 50%, during wheeltravel. If the 10% change created by a rollbar is clearly detected, imagine what will happen with the mentioned cambercompensative-big scrub layout during roll and pitch suspension movements.

So an A-arm layout that keeps control over roll centre height location during spring travel is imperative to car balance. If to many “factors” are built in to the design we may end up with a car that alter its behaviour in different situations.

Goran Malmberg

[Lukin]

Originally Posted by silente

just one question...

why the total weight transfer in that sheet is influenced by rc height?

i thought that moving rc what is changing is not the total wt, but only the value of the elastic and the geometric one...

If you look at the Magic Number sheet, cells A4:E6 you will see the contributions of each type of WT. Those cells are the 'summary' for non suspended, suspended geometric and suspended elastic both front and rear. The sum of each axle is in F5 and F6.

Changing the RC values changes the numbers for NS, SE and SG, but not the total, which is correct and what you said.

The weird thing about that chart is the roll angle. For any given lateral acceleration and car setup, I thought it should be calculated based on springs, bars and RC height. If you have the RC on the COG there is no roll angle. When you go up on the RC height you lower the roll angle etc.

Also the colours are horrible and its about as user friendly as a Volkswagen.

Goran, I am not saying I disagree with anything you say, but you seem to be starting more from a design point, not a track point. Due to tech regs in our series, we have very little scope for adjustment in the things you talk about. I think most people on this forum are in the same boat. Gotta get the basic tuning parameters down and sorted before we get the gas axe and milling machine warmed up.

[silente]

i usually analize the problem in another way...

I use to look to a mechanical balance based on the roll stiffness produced by spring and bars and, obviously, track, for front and rear. This because normally, if u have not the correct drawings of the car is really difficult to find exactly the position of the cg and of rc.

But what you use is not roll stiffness distribution, but the antiroll moment distribution. So you use also the rc height and cg height.In which way are you able to find the correct position of rc and cg?

What i still don't understand completly is how many importance has the position of the rc on the balance of the car. Using the antiroll moment distribution you calculate how much the car resist to roll on every axle. For sure in this way you can insert in your calculation also the ride height, via the cg position and you can also distinguish the various "pieces" of the weight transfer.

With my method it is impossible to determine the influence of the ride height on the balance of the car. And also the influence of the rc position..

But what i am normally interested in is not how much the car resist to roll but how much total weight transfer i have on every axle..What i am loosing??


In your opinion which differences are there between the two methods?

[ubrben]

The tyre generates force as a function of the load it carries. To accurately look at vehicle balance you need a way of calculating the normal load on each tyre under a given lateral acceleration - the roll angle should be viewed as a byproduct of this calculation provided you are not too dependant on platform control for aero reasons. The spreadsheet Lukin has posted is the easiest way of doing this. More complex methods would be multibody simulation for example.

Just looking at roll stiffness because you can't get accurate CG and RC information is fine. But doesn't change the fact that you're effectively blind to a large part of the vehicle balance picture.

How to measure CG height? - Lukin's spreadsheet has a page telling you how to do it.

How to measure RC heights? - Set the car on a flat patch put graph paper under the car, use a plumline and a height gauge to calculate all the pickup points and then draw a diagram to at least get the RC height in 2D.

Use the spreadsheet that Lukin has posted because it covers the minimum you need. It would be pointless to simplify it further IMO.

Ben

[Goran Malmberg]

Originally Posted by Lukin

Goran, I am not saying I disagree with anything you say, but you seem to be starting more from a design point, not a track point. Due to tech regs in our series, we have very little scope for adjustment in the things you talk about. I think most people on this forum are in the same boat. Gotta get the basic tuning parameters down and sorted before we get the gas axe and milling machine warmed up.

Lukin.
You are right, very good of you to point that out. I am talking form a blank sheet of paper, just for basic info. From this point we could analys
plus and minus when a car is built in a certain way to suit within a special type of racing.

I have done a lot of work using physical models, and it is this "basic" information I ty to share to other racers.

I am almost allways discussing form this point of view, if not asked to come up with an opinion for a certain solution, so dont take offence from my comments.

Goran Malmberg

[jonners]

am nothing but a simpleton and a lot of the very learned stuff posted on this goes way above my head but why has nobody mentioned roll couple - or perhaps I've missed it????

Surely the original enquiry is a roll couple query - or perhaps I need to go back to the books???

I don't want to ruffle any feathers here - and this applies regardless of whether this is a roll couple issue and/or if I've missed something that has already been said - but some of you engineering types don't half go on about what you want to tell everybody without thinking about relevance

I blame the way engineering is taught - you know what exam questions are getting at so just reproduce absolutely everything you've learnt.....

...I've already got my coat....