Better to specify what type of vehicle, what lateral capability of vehicle, how much roll is expected and axle configuration (independent, solid, etc.), weight distribution and range of tire properties. for example, if roll per lateral g is low and tire camber stiffness is low, worry about some other aspects of the suspensions. 3 other combinations of these terms. If it rolls a LOT and tire camber stiffness is very high, then you might start to consider getting more serious about it, IMHO. I suppose your understeer budget ought to be considered first.

Hello Cibachrome,
For example, we have double wishbone suspension RWD high performance road car such as BMW m3. Or 911(maybe not double wishbone). They have fair amount of roll compare to full race car, but still not as much suspension tralve if we compare to things like off-road buggy. For those car, do we need to worry about the RC changing between front and rear?
Maybe it is not good example, but I am trying to say if there is a on road track day which has amount of roll and maybe because package, it has short double wishbone arms. Will Rc migration difference between front and rear make driver feel harder to control and we need to beware of?

Thanks

Well this seems to be the BMW recipe for roll camber specs for M, X, C class vehicles.

• https://files.engineering.com/getfile.aspx?folder=ce3972f9-5f8f-4f47-be0c-e

Their ingredients for roll centers..

• https://files.engineering.com/getfile.aspx?folder=afd9ba44-6c32-4843-b5a8-9

Roll Stiffness ...

• https://files.engineering.com/getfile.aspx?folder=8e6b8e8b-8d35-4aa3-8066-f

And one more key ingredient: the tire's size dependent reaction to the roll control issue. Try to figure out why this is meaningful. BTW: I've never been a fan of the roll center migration religion because cars like these roll 4 - 5 deg/g at most. 1 - 1/5 of this might be the tires alone. If your analysis is 2D and kinematic only, you ought to get out of playing in a sandbox and see what 3D workings with real suspension compliance elements does to your storybook analysis. If this was so important, then simple simulations with fixed roll center lateral locations would do a terrible job of predicting load transfer and understeer out to the limit. Instead, they do VERY well if the car structure is any good. Welcome to the real world!

• https://files.engineering.com/getfile.aspx?folder=f02498ad-2128-4fd9-96df-f

Hello Cibachrome,
Thanks for these informtion. I was having Logging In issue at this website for while.

From graphs I can see OEM cars like to have larger static camber at back and lower camber gain at back. I think the reason is to make sure a little under steering under emergency for safety.
But for race car I think we need front and rear camber gain to be as close as possible? so if there are RC changing, they will change at same rate? Be honestly, I have never had any experience of how RC migration will affect the handling..

What I have seen at race track. Formula car has almost no camber gain at all, upper arm is shorter at wheel side, I think only because they want to keep scrub radius lower.Also, some Touring cars has short arms and almost no camber gain. But,some of track cars like Radical has much larger camber gain.(maybe not Raical, But kind of car)
Those cars roll so little during turning, they will not have much help from camber gain. But why some of cars have big camber gain, some of have almost no camber gain. As you said RC migration is not big problem. Otherwise everyone will run small camber gain for track car.

What's your example of a car that has "big camber gain"?

Chevrolet Corvair? Fiat 126? VW Beetle before 1967?

Ford Mustang S197?

There are reasons why I gave those specific suggestions ... can you see why I suggested them?

You presume that any sort of camber by roll or camber by side-force has some contribution to make. If a turnip has no blood, why would you bother squeezing it ? Teams know what their tire camber stiffnesses are, including the effect of slip angle and vertical load on it. If they have any camber stiffness to work with, they monitor predicted Mx to know when the sidewall will buckle (snap-thru) or chatter.

Hello BrianPetersen,
I am not sure, because they are old?😃

Recently I saw a front suspension picture of “Factory five 818” track car. That upper arm is so much shorter than lower one.

Hi Cibachrome:
Its making sense👌

"Looks" can sometimes be deceiving.

The first Ford Mustangs had front suspension that looked like this -
http://image.mustangandfords.com/f/31934550+w650+h...

A relatively slight relocation of the upper control arm inner pivot ( Arning/Shelby Control Arm Drop) did some very good things for "handling."
https://www.mustangirs.com/pdf_articles/mustang_mo...

Quote (cibachrome)

Teams know what their tire camber stiffnesses are, including the effect of slip angle and vertical load on it. If they have any camber stiffness to work with, they monitor predicted Mx to know when the sidewall will buckle (snap-thru) or chatter.

I'm assuming that 'chatter' produces a juddery feeling. What would that indicate, and would it be associated with an independently sprung drive "axle"?


Norm

"Chatter" is not associated with drive axle designation. It is more like a "shimmy" within the tire as lateral sections of the tread in contact with the road excite an nth mode standing wave in the carcass. Certain combinations of pressure, camber, slip and Fz create hot (as in melted) spots which generate a stick-slip phenomenon and this forcing function excites a normal mode. It is/can be very destructive to the tire as well as the suspension elements depending on the mode number. More evident with pressure below minimum recommended.

All that being said, the chatter phenom is more widely noticed in motorcycle racing bikes under full power in a turn with heavy camber angles. But the 'feature' of this tire resonance is also seen in stock car front tires (that I'm familiar with). The tread is reckoning with the peak of the mu-slip curve. One side is over and one under. There is an Mz reaction from the inner and outer tread band longitudinal forces which 'shimmy's the wheel. There may be some videos around showing chatter on a tire Force & Moment machine. When running at speed, the vibration will be severe and make you look around to see if the building is going to self destruct ! I've even seen it on a race tire being tested at 3.5 kph. It doesn't sound like frying bacon anymore (like it should).

Some folks talk about 'power hop' in the context of 'chatter', too, and this is a similar resonance but different degree of freedom from the one I think we are talking about here.

Thanks. I suspect I've gotten into a little 'chatter' in the Subaru (Legacy 2.5GT) in a corner taken unusually, ummmm, enthusiastically. OE tire and wheel sizes. Can't remember ever having experienced it in any other car, and the S197 (significantly wider wheels and tires - but the same mfr & model tires as the Subaru) sees road course track time.


Norm

From a tire usage standpoint ('what the ducklings call Optimization'), a simple but capable cornering simulation can show you that, as is often the case, the pair or tires on the steered axle is not coordinating their Fy generation. One of those tires will be very unhappy and complain about it loudly. If you have the nonlinear tire data and some Ackermann function values, you can investigate to find that the outer (higher vertical load) tire is actually fighting the inner one. So, to get more front grip you need to set up a better Left - Right steer angle pair. It can be very puzzling as the 'best' function is tire construction specific (The old "Do I need positive or negative Ackermann ? question). I think I can show you an example of what a ridiculous toe advance that might require. But, as you give the front more grip, you are dramatically lowering understeer even to the point of closed loop instability at a high lateral g-level (i.e. the driver can't save it no matter what). So, you need to lower the rear cornering compliance a bit too, to save the day. Also the large toe-in or toe-out angle differences at that steer angle can make the vehicle nearly impossible to drive straight. (Ackermann is hard to package, so they change static toe instead). Obviously tire wear and aero and braking and wheelhouse clearance concerns come along with the deal, too. just like in-laws. Hence the popularity of 'parallel steer'.

Here's a cheezy example from a premier mid-sized Euro Sled simulation. A streamlined 4 term Pacekja4Lite tire model for the car's tires. You will need to know or iterate how much load transfer is apportioned to the front meats. In the curves shown on the plot, the actual amount is one of those dashed lines. In this case, the tires want some toe-in as you approach Max-Lat (at say 5 degrees of slip angle to give the front end some jump. But the car is already pretty balanced so there is not much to be gained. However, by adding the 1 degree or so of toe-in, we've cut about 0.6 deg/g of understeer out of the balance act, which is no small amount for this car. My numbers aren't exact because I've only considered Fy and not any Mz or Mx or camber influences, but you get the picture and the right tools easily hand you the in-law compromises.

Put a slick on or a VERY large size and width tire and that plot may mysteriously flip upside down. Cars with 2 different tire/wheel recipes can really muddy the water because the rear may get into the act if it gets lonely.

• https://files.engineering.com/getfile.aspx?folder=1da35ff3-e401-48f7-8bfe-e

I’m not Sure what “chatter” is. My Alfa has special front suspension which moves upright position forward and back while you turn the wheel. At parking speed, front tyres jump at full lock. Is it chatter?

Is it possible to tune Ackerman or Toe without data? Some people measure front tyre temperature and compare both sides after skid pad. But Weight transfer based on speed as well as Ackerman, skid pad test shouldn’t be useful

Sorry those thing above are not related to the topic.
Anyway, I saw discussion on SAE forums about Rc motion yesterday. Someone mentioned that If RC position is changed during squat or lift or roll. Roll stiffness may change at same time. If front and rear change at different rate, we may having problem of alternating understeer and oversteer?
Rc point shift around is fine, but we have to watch the other side of the car.

Dose this make any sense?

Problems, issues and questions related to combined cornering, braking and ride motions are easily 'handled' by studying them under controlled circumstances using computer simulations making use of mass and inertia data, K&C relationships and nonlinear tire data. Features of cars such as yours are usually designed (as is synthesized) to have specified reactions to race track loading conditions. So, most SPECULATORS worry themselves into Catch-22 nightmares. In fact, synthesis avoids all your worries because of the applied constraints, the magnitude of the parametric changes (how much RC migration under max combined Ay and Fx). When the design is finalized and built, an "inertia relief simulation" can be run on a K&C machine to verify the load transfer distribution. RC is not the primary specification. Front & rear deltaW is. The machine sim will include the engine torque, front & rear steer and camber angles, too. If your not doing this and a competitor is, you lose. It's the current technology ("know how"). Go big or go home.

What I understand you are saying, if your car brakes and dives entering a corner, the fall of the front suspension and rise of the rear suspension will cause the suspension motion of each end to create a different roll rate, front to rear. The opposite will happen when accelerating out of the curve. I will leave it to others, but sounds like it could be a tuning tool? Or a bad crutch?

We still don't know what sort of vehicle the original poster (and his apparent alter-ego) is talking about, or what they are planning to do with it.

Typical independent multi-link or double-wishbone or MacPherson suspensions (front or rear) will all raise or lower the instant-center height by more than whatever the height change of the suspension is. If the front end dives 30mm because of braking and the rear lifts 30mm because of braking, the front roll center might go down 90mm and the rear up 90mm (or whatever). It's affected by the length of the relevant arms compared to the width of the vehicle. Short arms = more instant-center motion. But very long arms may interfere with the rest of the vehicle, like the drivetrain, or the interior, or whatever else is in the area.

If you do that ... one would expect that the front has a longer moment-arm with respect to the forces that lead to body roll (the roll center is further below the center of gravity) and the rear has a shorter one. What's the net effect? Presuming that the body is reasonably rigid, it's still going to roll when cornering forces are imposed. Maybe it's a little more, maybe it's a little less. Who knows. Depends on a ton of other factors. How much those forces translate to the contact patches will depend on the springs and dampers.

BUT ...

What the car does in this situation (simultaneous cornering and braking) is going to depend a whole lot on the brake balance, the weight transfer, the toe change due to bump/roll steer (front and rear), the idiot behind the wheel, etc. more than what the instant-centers are doing. Maybe the front wheels lock and the car plows straight. Maybe the unloaded rear end starts swinging out. At a certain point the driver needs to be smacked around "don't do that".

This winter I have established that my own little Fiat has slight lift-throttle oversteer when cornering on slick surfaces with the winter tires currently installed. It has MacPherson front (high roll center migration) and twist-beam rear (roll center close to ground level no matter what). How much does this have to do with roll centers, versus age and condition of tires? I'm thinking next to nothing.

Millions of cars have been built with front instant-centers that move all over the place (MacPherson) and rear instant-centers that also move a lot (multilink with short arms) or which stay put near ground level (trailing arm twist-beam axle) or are a fixed and high height above ground level (solid rear axle with 4-link diagonal uppers).

Suspensions with short lateral arms are going to do nasty stuff when they are far from nominal ride height. If you need long-travel compliant suspension (off-road racing) then use long lateral arms or suspension designs that aren't dependent on this (e.g. twist-beam axle). If you don't need long travel suspension then use any suspension design and don't let it move (stiff springing and damping). Formula 1 is a splendid example of what can be done using a "bad" suspension design and not letting it move. (Upper and lower almost-parallel arms ... it's designed with aero as first priority. They race on relatively smooth tracks, so the extremely stiff springing is of little consequence)

Suspensions with very high camber change are going to do nasty stuff period. See all of the examples that I gave earlier that have rear swing-axle suspension. Also see, Ford trucks with Twin-I-Beam front suspension. If you are stuck with this suspension design then the only way to deal with it is to not let it move (stiff springing and damping). If you are not stuck with this suspension design then don't use it.

Excessive camber change in front suspension not only lets the roll center move around but it also changes the steering axis inclination, the instantaneous scrub radius, side-scrub of the contact patch, etc. The fix ... don't do that. Again, if you are stuck with a suspension design that does this then don't let it move (stiff springing and damping). If you are not stuck with it then don't design in excessive camber change up front. Means the tires will lean over with body roll. (A) so what, (B) use the right tires, (C) don't let it (stiff antiroll bars and/or springing and damping).

Very frequently the things that one is tempted to do to counteract a specific effect have worse side effects than letting the thing itself just happen and dealing with it. Body roll, for example ... at least up to some reasonable point.

It will sure help to know what the original poster wants to do, and with what vehicle, and why.

Hi Cibachrome,

I have heard simulator can do things crazily accurate. For people like me, I only tune and play chassis as hobby. I cannot access to those softwares, and not even tyre data.
From the other side, some people may want a track car that he enjoys, don’t need to be very fast but predictable. If there is a race car that fast, but Hard to drive, It still a winner car, simulation data will tell the drive what the limit is at where. But not for people just want to enjoy their car for weekend.

I have seen a some chassis with several alternate locations for the inner pivot points of the control arms (Bill Thomas' Cheetah?) This adjustment would let you play with various locations of instant center and corresponding wheel attitude/roll resistance. Get a car with this adjustment and see what works for you? If you do this, you won't need the expensive computer stuff. I bet the cost of a good suspension analysis program would buy you a car where you could drill some additional inner pivot holes?? Early Mustangs/Falcons could be redrilled to lower the upper arm pivots (memory verification needed here).
I agree with Brian in his suspension nasties. As Colin Chapman said, "Any suspension will work if you don't let it".
Actually, "bad" suspension with "bad" ackermann can work on the "right" race track.

Hello BrianPetersen,

I will try find link of than discussion.
They were either discussing around FSAE car or just general RC motion question itself.
As soon as I started read you comment I realized I forgot something important. Let me talk about what I’m thinking about the RC height during braking first.

During brake, rear end RC comes up. If we think that roll moment is between CG and rear RC, then this moment is getting smaller. Because CG will be the pivot point, so it stay at same place. But mass of the car is not only at CG, it has weight both ends. Rear end comes up during brake means weight of the rear end comes up. So roll moment shouldn’t change too much, because Rc comes up with rear end of the car body. If they come up with different rate, this will change the roll stiffness, and this is what I was worrying about before.
But after you mentioned weight transfer. I realized it’s not that simple. Because under braking, weight transfer from rear to front. So that required roll stiff at rear should be reduced at same time. And it is only rear end, there is front end doing as well.
If we want front rear grip level constant. We need control the RC motion that give right roll stiffness accord to weight transfer for both front rear, and at same time make camber curve right. This it possible?
But now, I Think about driving on track, we all know under brake rear end will be looser..

BUGGAR
I think so. This is what I’m going to do.

Brianpetersen
Forgot to say. For double wishbone, I have found long lower arm with short and steep angle upper arm keeps RC quite stable. Better than same setup with lower camber gain which has same lower arm length but longer upper arm or flatter upper arm.
It’s what 2D software telling me, before I thought long arms and less angle should keep RC stable.

If I can get this to post, it's a suspension design with variable inner pivot locations. I have the fabrication drawings for this somewhere.

re submit

• https://files.engineering.com/getfile.aspx?folder=e598c920-e554-4891-8922-5

BUGGAR. This is great thanks for that.

Happen to have one under construction right now. Front suspension is identical, except for the axle. Blue plates are for suspension geometry adjustments. I'm using centerlink steering since that has to adjust too. I have a nerdy friend who is redesigning the blue brackets to be slotted and use a trunions attached to stepper motors to adjust the suspension geometry "on the fly". (This is just what old retired Engineers do)

• https://files.engineering.com/getfile.aspx?folder=0328e785-a3cd-4aaa-9417-0

I got exactly same design for the pickups! Any choice to look the picture of adjustable device? Sounds interesting

YKAI, I have no photos of his adjustment set up yet since he's still building it. He's using an Arduino computer to run the stepper motors, which weigh about 5 pounds each so there's going to be about 50 pounds of stuff at each end of the car. I may have shot him in the foot when I explained that the outer steering link had to adjust in length as well. He said it would be easier to program a total steer by wire system, especially since my rear spindles are same as front and can be made steerable. What have I started!?

Do you have any photos of your set up?

ps: we're designing the steering links for a load of 2000 pounds, with no impact allowance. This is based on strain gage measurements and is rough. Does anyone have a better figure?

Hi BUGGAR,
I will take some picture after they are made.