Nature of anti-roll bars

[detail]

Help me, please, to understand, I can't find anything relevant in the inet.

I thought I knew how the anti-roll bars work. But I guess there is something new to learn. Correct me, please. Anti-roll bar connects the wheels of one axle and redistributes the force between theis ammortisers.

But one guy tells me that there are such bars mounted to the dependent suspensions of some trucks. I looked attentively to a scheme of a dependent suspension and can't understand where to attach an antiroll bar.

I'm not joking. Please, explain me this.

[SoloNijN]

Help me, please, to understand, I can't find anything relevant in the inet.

I thought I knew how the anti-roll bars work. But I guess there is something new to learn. Correct me, please. Anti-roll bar connects the wheels of one axle and redistributes the force between theis ammortisers.

But one guy tells me that there are such bars mounted to the dependent suspensions of some trucks. I looked attentively to a scheme of a dependent suspension and can't understand where to attach an antiroll bar.

I'm not joking. Please, explain me this.

With the antirollbar you adjust the stiffness of the (doh) antirollbar, so, with low settings the cars rolls over more than with a higher (stiffer) set

[Bob Smith]

Redistributes force? I wouldn't really say that.

They do indeed connect together the wheels from one axle, and offer resistance to the wheels moving apart, reducing body roll (and the subsequent weight transfer).

As described, I can't see how an anti-roll bar could be attached to a non-independant suspension system (solid rear axle for example) and still be of any use.

[Fonnybone]

Whenever a question like this arises, i always go there first:

http://auto.howstuffworks.com/question432.htm
Here's another part:
http://auto.howstuffworks.com/car-suspension6.htm
Some more stuff here:
http://www.edmunds.com/insideline/do/Columns/articleId=104166#2
That's just using "anti-roll work" in google btw...

Anti-roll cars do what they say, they keep the car from rolling.
You have 3 rotation axis, pitch, roll and yaw. Roll is rotation
around the longitudinal axis (a line from the front to the rear
intersecting the center of gravity, cog) of the car. It means
the car 'rolls' around this virtual line. This is usually the case
of a steady-state cornering situation, once the car settles
into a turn. Anti-roll bars are springs that try to counter this.
They will force the car to remain flat when it tries to roll.
There is usually 2 anti-roll bars, one connecting the 2 front
and another connecting the 2 rear wheels.

As for solid-axles, this is what actually generated the need for
an anti-roll bar. Solid axles usually have one big virtual axis they
rotate around, the driveshaft. As the shaft rotates, it tries to
rotate the whole rear axle also, this is where various form of 'bars'
(Panhard,Dion,etc..) were implemented to both keep the axle from
rotating but also to locate the axle as the leaf springs usually
found in such setups, don't constraint the axle laterally (from side
to side). Considering your axle remains parallel to the ground, the
torque then tries to rotate the whole chassis around the driveshaft
instead since it's easier. This rotation is very similar to roll but is
instead initiated by the driveshaft instead of the chassis. Don't
mistake 'anti-torque' bars with anti-roll bars. The first usually
secures the axle to the chassis to prevent the driveshaft from
rotating the chassis, whereas the anti-roll bars are secured on
the chassis with each end connected to a wheel by a pivot to
prevent inertia (centipede force) from throwing the whole car
into the ditch, more on-topic, to prevent the car from leaning
into a turn.

Independant suspensions don't NEED anti-roll bars by definition and
in fact hinder an independant suspension's work which is usually mentionned
as "anti-roll bars make independant suspensions less independant".
Some very respected 'car guys' have expressed this, but it's still widely
used to find a good confort/handling compromise. In racing, it's also very
usefull for fine tuning of the handling as they can effectively change the
front/rear grip ratios giving you more/less understeer/oversteer with
little or negligeable effort. Sometimes, there's even a manual adjustement
IN the cockpit allowing the driver to tune the bars while racing. If you
forget about the ideal conditions, which racing rarely provide anyways,
they are also usefull at adapting to changing conditions like adding
understeer in the rain or conpensate for wear/damage.

[Bob Smith]

Sometimes, there's even a manual adjustement
IN the cockpit allowing the driver to tune the bars while racing. If you
forget about the ideal conditions, which racing rarely provide anyways,
they are also usefull at adapting to changing conditions like adding
understeer in the rain or conpensate for wear/damage.
Like F11 ;)

[Forbin]

...prevent the car from leaning
into a turn.

Small correction: cars lean out, bikes lean in. :)

[qwejibo]

Redistributes force? I wouldn't really say that.

They do indeed connect together the wheels from one axle, and offer resistance to the wheels moving apart, reducing body roll (and the subsequent weight transfer).

Generally, I don't think that body roll has any effect on weight transfer - a car with 100% rigid suspension(thus no body roll at all) would still transfer the same weight as one with soft springs + bars. You could test this in LFS by driving in circles at the same speed/G load with both soft + stiff suspensions, weight transfer (from one side to the other) should be identical.

[skiingman]

As described, I can't see how an anti-roll bar could be attached to a non-independant suspension system (solid rear axle for example) and still be of any use.

As Fonnybone pointed out, antiroll linkages were developed for and remain more necessary on live axles. :)

My Wrangler comes stock with a 30.5mm front ARB. It needs it. It breaks the endlinks every couple of years.

[skiingman]

Generally, I don't think that body roll has any effect on weight transfer - a car with 100% rigid suspension(thus no body roll at all) would still transfer the same weight as one with soft springs + bars. You could test this in LFS by driving in circles at the same speed/G load with both soft + stiff suspensions, weight transfer (from one side to the other) should be identical.

They do indeed connect together the wheels from one axle, and offer resistance to the wheels moving apart, reducing body roll (and the subsequent weight transfer).


This is not true. I highly suggest you read a layman's treatise on the physics of this. Tune to Win by Carroll Smith is old but very good.

Hint: If what you said were true, two vehicles with similar ride heights would have similar drag strip performance with any spring setup. Clearly not the case.

Stiffer springs result in less weight transfer. Outside of considerations in dynamic camber control, softer antiroll linkages are better.

To be clear, increasing the stiffness of the anti-roll linkage INCREASES weight transfer. Tire adhesion available is not linearly related to the vertical component of force on the tire (twice the force doesn't result in twice the grip), so less weight transfer is always better.

[qwejibo]

This is not true. I highly suggest you read a layman's treatise on the physics of this. Tune to Win by Carroll Smith is old but very good.

Hint: If what you said were true, two vehicles with similar ride heights would have similar drag strip performance with any spring setup. Clearly not the case.

Stiffer springs result in less weight transfer. Outside of considerations in dynamic camber control, softer antiroll linkages are better.

To be clear, increasing the stiffness of the anti-roll linkage INCREASES weight transfer. Tire adhesion available is not linearly related to the vertical component of force on the tire (twice the force doesn't result in twice the grip), so less weight transfer is always better.

Weight transfer is created by lateral + longitudinal forces generated by the tyres. The only factors in the equation are the cars weight, CoG height and track/wheelbase.

Springs react to the change in vertical tyre load by moving to a position which will support the changed weight. Stiff springs will have to move less than softer springs to produce the same force. The magnitude of the weight transfer is unaffected.

What is affected is how long it takes for the weight transfer to occur. 100% rigid suspension will transfer the weight instantaneously - OTOH, soft suspensions take time to move to their new position. This is why stiff suspensions feel more responsive, because the car will react more quickly to weight transfers.

You state that "Stiffer springs result in less weight transfer" - if this is true, then to increase the grip of the front end of a car during cornering, would I stiffen the front springs?

[vpr01]

Roll bars do two things.

1) protects the driver from having his head smushed to a pulp if the car flips over.

2) provides extra rigidness, which improves handling.

[tristancliffe]

Two different type of roll-bars mate.

Roll Over bars are to protect drivers heads...
Anti-Roll bars are to do with chassis balance etc

[vpr01]

i always thought they were the same things. well at least i know now!

[skiingman]

Weight transfer is created by lateral + longitudinal forces generated by the tyres. The only factors in the equation are the cars weight, CoG height and track/wheelbase.

Yeah, well, for some rather obvious reasons that isn't completely true. Big hint: The centroids don't stay in the same place as the car rolls. My previous hint about drag cars should have made that rather obvious. Soft springs don't just help the launch, they help the whole run down the track compared to a vehicle with a similar static rideheight and stiff springs.

Also, there being antiroll linkages, weight transfer is affected by another set of forces. Importantly, I'm talking about weight transfer between the left and right tires. Other frames of reference are pointless.

What is affected is how long it takes for the weight transfer to occur. 100% rigid suspension will transfer the weight instantaneously - OTOH, soft suspensions take time to move to their new position. This is why stiff suspensions feel more responsive, because the car will react more quickly to weight transfers.

This is all well understood.


You state that "Stiffer springs result in less weight transfer" - if this is true, then to increase the grip of the front end of a car during cornering, would I stiffen the front springs?

Only if you could subsequently stiffen the rear springs in a rather equal proportion.

The fastest car will have the stiffest suspension system that can still manage the bumps in the road. This is why third spring systems exist.

This particular misconception of yours is completely off topic anyways. Importantly, anti-roll-bars INCREASE weight transfer.

[J.B.]

Big hint: The centroids don't stay in the same place as the car rolls. (...) Importantly, anti-roll-bars INCREASE weight transfer.

OK, so you are taking CoG movement into account. An anti roll bar makes it harder for the CoG to move. Why would this increase weight transfer?

[Bob Smith]

OK, so you are taking CoG movement into account. An anti roll bar makes it harder for the CoG to move. Why would this increase weight transfer?
Finally, sense. :)

[Letin-Klanen]

the anti roll bars actualy removes grip on a suspension system like wtcc.

in rain they use almoast no anti roll bars in front for gaining grip.
on road cars its the other way, because the suspension is so soft

[qwejibo]

Yeah, well, for some rather obvious reasons that isn't completely true. Big hint: The centroids don't stay in the same place as the car rolls. My previous hint about drag cars should have made that rather obvious. Soft springs don't just help the launch, they help the whole run down the track compared to a vehicle with a similar static rideheight and stiff springs.

Also, there being antiroll linkages, weight transfer is affected by another set of forces. Importantly, I'm talking about weight transfer between the left and right tires. Other frames of reference are pointless.

I think you overestimate the effects of other elements on weight transfer. Apart from the cars weight, CoG height and track(or wheelbase in accel/braking) anything else such as CoG movement, rollcentre movement, etc. generally only account for a few % of the total weight transfer. F1 engineers and the like probably worry about those few %, but for most ppl it isn't going to be that important.

You have made the statements that:

1. Stiff springs reduce weight transfer
2. Stiff anti-roll bars increase weight transfer

Can you explain why you think these statements are true?

[skiingman]

You have made the statements that:

1. Stiff springs reduce weight transfer
2. Stiff anti-roll bars increase weight transfer

Can you explain why you think these statements are true?
I would hope I could, or else I had better not make those statements.:)

1) Stiff springs keep the rollcenter (or centre, if you will :) from moving around as much. For a variety of reasons this is beneficial, all of which are rather complex and not worth discussing in such a forum. More to the point, when the body rolls it invariably increases weight transfer from inner to outer wheels. This isn't a good thing for obvious reasons, and since "a few percent" less weight transfer might mean a third or a half more load on the inside wheel, I don't think its worthless enough to simply write off. Note that all of this is more or less equally true for "stiff springs" as well as stiff ARBs, with the exceptions explained below.

2) Any antiroll linkage incearses the weight transfer from the inner to the outer wheel. This should be intuitively obvious. Stiffer rear bar = less rear grip -> more oversteer or less understeer. The actual action is dead simple to understand, but probably less so in words than in pictures or demonstration. The stiffer the bar, the more resistance there is to droop of the inboard suspension while the outboard suspension is being compressed. A simple torsion bar of some sort is the usual way to effect this action, racecars in LFS have magical ideal torsion bars with infinite adjustability.

If you could get away with it, no roll bar would be great. You'd minimize the weight transfer between the wheels, and you'd make testing/tuning a lot simpler. For a variety of reasons, that isn't usually possible. For one, my point number one above is often a big factor. Despite the increased transfer due to a big bar, its better to keep the body from rolling too much for a variety of other reasons. Not the least of which is transient response. Also, the weight transfer due to a high CG allowed to roll can yield worse effects than the same system prevented from rolling. Ergo sedans with very big bars on the non drive wheels lifting inside wheels in corners. Guess what, drop the bar, and the inside wheel will get back on the ground. This should be more obvious proof that rollbars increase weight transfer, to the point of pulling wheels right off the ground. That requires more than 100% weight transfer.

As it is, rollbars are a great tool. Since most cars don't have third-spring setups, the wheel rates need to be soft enough to absorb road inconsitencies AND stiff enough keep the rideheight in the appropriate place with sometimes large downforce. This as well as low CG and big track leads to small bars for singleseaters.

On the other end, look at a big sedan. High CG, lots of unsprung weight. Stiff springs will be liable to prevent the wheels from staying on the ground over bumps. Big anti-roll bars can keep the transient response tight with softer springs and also keep the thing from nailing the bumpstops when in cornering. If it rolls enough to hit the bumpstops, the springrate is no longer soft, it is near infinite. Sedan suspension design is usually far from ideal in camber control and modern camber sensitive radials really suffer if they get positive camber dynamically. Keeping the body flat with big antiroll bars is one effective, if not ideal, way to do this. Longer (and often better designed) suspension links and a lower CG would be better, but thats often not allowed in the rules and is obviously more development and cost intensive.

Carroll Smith has been quoted as stating at FSAE competition that ARB rate should be no more than 10% of the wheelrate (both measured in force/degree of roll) in a situation where the design is not very limited. Current FSAE competitors have often disagreed with this logic, but I don't know how wise that is.

The main thing is to understand that there are advantages and disadvantages to increasing rollbar stiffness. If you can increase springrate instead, its probably a better idea. Obviously, sometimes you can't.

I know the LFS differential modeling currently means people driving fast drive locked-diffs, but try this out for kicks: Grab the stock F08 (or FOX) setup. Select, say, the viscous diff, set to a low rate. Turn downforce down to reduce the masking of mechanical grip. Try this setup with both front and rear bar set to maximum. Note the massive wheelspin when attempting to put power down on corner exit. Retry with bars set to just about nothing. Note the better power put down on corner exit.

Since everyone uses locked diffs now, that kind of effect is somewhat masked. When locked diffs start giving the disadvantages they should to the extent they should, I think you'll find that setups change quite a bit.

[skiingman]

OK, so you are taking CoG movement into account. An anti roll bar makes it harder for the CoG to move. Why would this increase weight transfer?

Explained at length above, but long story short, the uber-antiroll bar will lift the inside wheel straight off the ground. 'Nuff said. If you transferred much more weight than that, you'd roll the feck over. :lol:

[J.B.]

Explained at length above, but long story short, the uber-antiroll bar will lift the inside wheel straight off the ground. 'Nuff said.

I'm sorry but this is simply not true.

Let me try to explain. The centrifugal cornering force causes a torque around the longitudinal axis of the car. This torque has to be equalized by the outside tyre pushing harder aginst the track than the inside tyre. This results in what we call load transfer. Putting in an ARB will not change this. (Do a free body diagram)

The wheel lifting phenomenom is a result of the way the front and rear roll stiffnesses interact with each other. It is not the direct result of a single ARB.

Imagine a car, CoG in the middle between the axle lines, cornering hard, with the inside rear wheel in the air. If you were now to magically remove the front end of the car, the inside rear wheel would drop to the ground. The rear wheel was only in the air because the roll stiffness at the front was a lot lower than the roll stiffness at the rear. If the roll stiffness at the front had been the same as at the rear, the wheel would not have been in the air. It was in the air because of the way the front and rear axles interacted with each other.

The combination of soft front / hard rear roll stiffness results in increased load transfer on the rear axle and at the same time decreased load transfer at the front axle. This is because the overall load transfer from left to right has to stay unchanged to equalize the cornering torque. ARBs can be used to increase load transfer at one end of the car, but only at the expense of decreasing load transfer at the other end of the car. Overall load transfer from left to right cannot be changed by adjusting roll stiffness.

[5th Earth]

There are so many wildly varying opinions flying around I'm not even going to try sorting them all out, and make some sweeping statements based on comments I've seen.

The probability that any of the following are wrong is feasable.

1: Antiroll bars make an independent suspension less independent: True. However, the benefits of decreased left-right weight transfer (due to decreased CoG movement) and decreased suspension movement (allowing less camber and better tire contact patches) mean that ARBs, in moderation, are beneficial to a car's handling on a reasonably smooth track. The tires stay flatter, and the inside tire does more of the work. For serious offroading, of course, you should probably remove the ARBs entirely, but for road racing, by all means use them.

2: ARBs can be fitted to a non-indepedent, aka live axle suspension: True. Fix each end to each wheel, fix the middle to the chassis, and think about it: it works. If you're trying to use one on-road, you absolutely want one so you can have some semblance of stability, but again, for off road you don't want it. (incidentally, live axles have certain distinct advantages for extreme off-road situations, but that's another discussion)

3: ARBs decrease grip: False. This is a common misconception, one that I held myself for a long time, which stems from the fact that stiffening the ARB on one end of a car changes it's understeer/oversteer tendencies, i.e. stiffening the front ARB causes understeer. What is actually happening is that having different front and rear ARBs causes front-rear weight transfer in response to lateral forces. A stiff front and soft rear will, in response to a lateral force, cause weight to be transferred from the front wheels to the rear wheels, specifically the front inside to the rear outside (in the opposite situation, this is why "hot hatchbacks" are famous for cocking their rear inside tire in the air in a hard corner--the hard rear ARB and the soft front ARB used in FFs means literally all weight from the inside rear is being transferred to the outside front, helping with the understeer). Total grip remains the same, but you lose some in the front and gain it in equal proportion on the rear. Changing both ARBs by the same amount will have no effect on front-rear balance in response to acceleration forces, and any change in grip is due to other factors.

[skiingman]

Let me try to explain. The centrifugal cornering force causes a torque around the longitudinal axis of the car. This torque has to be equalized by the outside tyre pushing harder aginst the track than the inside tyre. This results in what we call load transfer. Putting in an ARB will not change this. (Do a free body diagram)

Except, it will change this, because the CG doesn't stay in one place. That axis its rotating isn't where the CG is, or else things would be rather ugly.

The wheel lifting phenomenom is a result of the way the front and rear roll stiffnesses interact with each other. It is not the direct result of a single ARB.

Agreed.

Imagine a car, CoG in the middle between the axle lines, cornering hard, with the inside rear wheel in the air. If you were now to magically remove the front end of the car, the inside rear wheel would drop to the ground.

Yes. As it likewise would if you magically removed the rear ARB. Where was I wrong again? Increasing the roll stiffness at one end makes that end transfer more load and produce less grip.

The rear wheel was only in the air because the roll stiffness at the front was a lot lower than the roll stiffness at the rear. If the roll stiffness at the front had been the same as at the rear, the wheel would not have been in the air.

This isn't true in all cases. If there is enough available grip and a high enough CG, both inside wheels will lift without any outside input like a curb. (umm, a variety of examples exist, karts are a common one, its becoming more common at FSAE events where the cars have to be tilt-tested to almost 2Gs to pass tech) This will happen regardless of ARBs, but stiff ARBs don't help. Maybe I'm wrong on that point, but I don't think so. They certainly don't help if there is an outside force like a curb, but thats a different issue entirely.

It was in the air because of the way the front and rear axles interacted with each other.

In the case you describe (FWD racing car it seems) I agree totally. Many RWD touring cars are setup just the opposite, lifting the inside front for much the same reason.

The combination of soft front / hard rear roll stiffness results in increased load transfer on the rear axle and at the same time decreased load transfer at the front axle. This is because the overall load transfer from left to right has to stay unchanged to equalize the cornering torque. ARBs can be used to increase load transfer at one end of the car, but only at the expense of decreasing load transfer at the other end of the car.

Its really important that 5th Earth reads this carefully. The larger ARB does effectually decrease grip at its end of the car.

Overall load transfer from left to right cannot be changed by adjusting roll stiffness.
If the movement of the CG was minute and unimportant, that may be true. The mere act of preventing the CG from moving away from the center of the circle the car is traveling in can/does have a noticeable effect. It was for this reason that "sway bars" were initially used. Not for tuning of handling characteristics. Cars that have a high CG and lousy suspensions tend to run brutish ARBs at both ends. As in, the crap Detroit produced in the 50's/60's when they started including "sway bars" as standard equipment.

Its a great tuning tool, isn't it?

None of this changes the facts that real racecars abide by in contrast with LFS:
-They don't have infinitely adjustable ARBs, blade types are more or less stiff or soft. Slider types are better and more adjustable.
-Adjustable ARBs have limitations and aren't used when not necessary.
-Need independent action of the suspension. LFS tracks (it seems to me) have some big bumps across the track, but few pothole type bumps that effect only a wheel. For instance, it seems at BLGP that there is no bump driving onto the concrete at T1, just a change in grip. Most real racetracks aren't all that smooth. Maybe they are in Europe? Wouldn't surprise me, some of the roadcourses here look like rural roads in comparison to the beautiful european tracks.

5th Earth, the "live-axle" for offroad is a good debate. You nailed that one, except I tend to think given the appropriate cash a live-axle would get stomped by an appropriate IFS/IRS system even in rockcrawling. Off-road racing has some of the coolest and most bizarre suspension designs out there. Rockcrawling is brutal on parts (two u-joints last summer just barely picking through rocks with my silly D33) and beefy as heck live-axles are readily available. Any mass-produced independent suspension is going to suck for rockcrawling, but let it not be said that it would be at a disadvantage to a live-axle if properly designed.

Here's a bit of silly trivia: No matter where a car is built, you'll find the engineers specify the "swaybar" size in mm's in the specifications sheets. However, few examples I've found are built to an actual "metric" size. 25.4mm (one inch) swaybars are rather common, as are various sizes in 1/8th inch increments expressed in mm. One has to wonder why this is.

I still stand by my statement that if you could get the CG very close to the road and the track wide enough, the swaybar would do you more harm than good. I think some examples of that exist. I hate driving FV8 with beefy ARBs, same for FOX.

[Vykos69]

Here is a nice pic attached, taken from this really cool webpage about suspension: http://www.chris-longhurst.com/carbibles/suspension_bible.html

You see relly good, where it attaches at the suspension, and that it basically distributes the spring/damper forces between the wheels of one axis, making a car less rolling in turns.

[SladiVadi]

Thanks Vykos! :D

[skiingman]

Here is a nice pic attached, taken from this really cool webpage about suspension: http://www.chris-longhurst.com/carbibles/suspension_bible.html
.

Excellent link, all sorts of nifty rare stuff. Thanks.

[J.B.]

skiingman, it seems we are in agreement :thumb:

One more thing worth mentioning is that the reason (among others) that you lose grip on the axle that has more load transfer is the load sensitivity of the tyre:

The grip of the outside tyre increases with increasing load and the grip of the inside tyre decreases with decreasing load.The rate of increase/decrease isn't the same in both cases though. The higher the load on the tyre, the lower the increase rate. This means that the overal grip of the axle drops with load transfer.

In one sentence: when increasing load transfer from inside to outside, the inside tyre loses more grip than the outside gains:

And nice link Vykos. :thumbsup:

[skiingman]

skiingman, it seems we are in agreement :thumb:

One more thing worth mentioning is that the reason (among others) that you lose grip on the axle that has more load transfer is the load sensitivity of the tyre:

The grip of the outside tyre increases with increasing load and the grip of the inside tyre decreases with decreasing load.The rate of increase/decrease isn't the same in both cases though. The higher the load on the tyre, the lower the increase rate. This means that the overal grip of the axle drops with load transfer.

In one sentence: when increasing load transfer from inside to outside, the inside tyre loses more grip than the outside gains:

And nice link Vykos. :thumbsup:

yeah, I said:
"Tire adhesion available is not linearly related to the vertical component of force on the tire (twice the force doesn't result in twice the grip), so less weight transfer is always better."

which is not nearly as thorough and well defined as your explanation above.

That link really is cool, I had some fun reading about that RFH LeMans hydraulic suspension.

[5th Earth]

For instance, it seems at BLGP that there is no bump driving onto the concrete at T1, just a change in grip. Most real racetracks aren't all that smooth. Maybe they are in Europe? Wouldn't surprise me, some of the roadcourses here look like rural roads in comparison to the beautiful european tracks.

All true. It's worth noting that, for my South City setups, I run softer ARBs, preciesly because the track is bumpier, including some very noticable bumps from manhole covers and the roughness of the brick-road chicane. And for my RallyX setups I run very soft ARBs.

5th Earth, the "live-axle" for offroad is a good debate. You nailed that one, except I tend to think given the appropriate cash a live-axle would get stomped by an appropriate IFS/IRS system even in rockcrawling. Off-road racing has some of the coolest and most bizarre suspension designs out there. Rockcrawling is brutal on parts (two u-joints last summer just barely picking through rocks with my silly D33) and beefy as heck live-axles are readily available. Any mass-produced independent suspension is going to suck for rockcrawling, but let it not be said that it would be at a disadvantage to a live-axle if properly designed.

Well, the main point in favor of live axles besides durability is that, with a live axle, when one wheel is pushed upwards by a bump, it exerts a downward force on the other wheel as the axle rotates around its central point. Ignoring the effects of non-flat tire contact, which is valid with very soft ground, large rocks, or extremely low tire pressures (all hallmarks of offroading), it's easy to see that more downward force means more grip. More grip is always very desirable.

An independent suspension with cross-linked pneumatic or hydraulic springs will have the same effect, and this system has been used to good effect on some Land Rovers, but the system (at least on Land Rovers) is also famous for being very unreliable and hard to repair. It's also undrivable on normal terrain without switching back to conventional springing. You're probably right about it being better in the case of unlimited funds, but rockcrawling isn't yet such a big sport that most people have that kind of money and engineering skill.

I still stand by my statement that if you could get the CG very close to the road and the track wide enough, the swaybar would do you more harm than good. I think some examples of that exist. I hate driving FV8 with beefy ARBs, same for FOX.

Well, there's always the matter of personal taste. I can certainly see that softer ARBs on a formula-style car would certainly be no great disadvantage, and the advantages may be enough to justify it. In a perfect world, of course, the CoG and the roll center would coincide, and then there would be no body roll at all, but nobody has ever made a car that had a CoG that low. I still think that in the absence of high-tech systems (pneumatics, Bose active suspension, etc.), ARBs are essential for high-centered cars like typical road cars.

[Vain]

Isn't this (http://www.early911.co.uk/historic/assets/images/ba3ir73.jpg) a great example of bad ARBs? Looks like the rear ones are too soft and allow 'big' rolling, while the front ARB is too stiff and the front axis can't roll enough to touch the ground with both wheels.
This is both an example and a test wether I understood it all right. ;)

Vain

[ColeusRattus]

Isn't this (http://www.early911.co.uk/historic/assets/images/ba3ir73.jpg) a great example of bad ARBs? Looks like the rear ones are too soft and allow 'big' rolling, while the front ARB is too stiff and the front axis can't roll enough to touch the ground with both wheels.
This is both an example and a test wether I understood it all right. ;)

Vain

Could also be a problem of frame stiffness...

[5th Earth]

Could also be a problem of frame stiffness...

It's a Porsche, I'd say it's not a good setup, but given RWD, heavy in the back, and powerful, it doesn't really shock me. Accelerating out of a corner with a lot of grip could do that.

Oh, and Vain, you're right. Soft rear ARB + stiff front ARB could very well cause what that picture shows. Assuming the suspension has been modified to act that way, the guy is trying to compensate (overcompensate?) for the Porsche's tendency to oversteer.

[skiingman]

Oh, and Vain, you're right. Soft rear ARB + stiff front ARB could very well cause what that picture shows. Assuming the suspension has been modified to act that way, the guy is trying to compensate (overcompensate?) for the Porsche's tendency to oversteer.

Yeah I agree totally. That kind of setup is popular on front engined sedans in order to minimize power on oversteer to get good grip on the way out of the corners. On something with the engine behind the rear axle, its only more necessary.

I think if you look closely at that pic you'll see the thing really is all twisted out of shape. Look at the plane of the rear spoiler and that of the front bumper. :O Could probably use better a beefier cage, but stiff isn't always fast and that setup looks rather vintage anywho.

[Fonnybone]

Hehe, i went to get my European car magazines, there was a series of
articles called Suspension Guide (not a buyers guide..) a few years back.
December 1999 contains Part 2 - Controlling body roll. I found the articles
very informative and pleasantly objective including both scientific and
vulgarised explanations. It is in one of those articles that i learned about the
relation between load and grip on a tire. That alone seems to be the main
idea behind most suspension tuning. Like stated earlier, as the load
on a tire increases, it's grip also increases, but up to a certain point where
grip falls down again. It is not a linear relationship, that means there's an
optimal operating zone, and therefore there are also overloaded or unloaded
situations, both reducing grip until there is virtually none left.

I find most of the disagreement is on the wording more than the actual
understanding of the situation. One thing i find very confusing in your posts
is the liberal use and exchange of COG and ROLL AXIS (a result of the
ROLL CENTERs, also referred to as 'centroids' in some posts), although
i'm sure you guys know the difference, i think it's important to mention to
others that they are 2 different things and 'roll' is actually the result of their
interaction.

There are 2 ROLL CENTERs, one for the front and one for the rear, each a
result of the suspension geometry (the front of the car rolls around a
specific axis while the rear rolls around another axis).

The ROLL AXIS is the line (axis) formed between the front and rear ROLL
CENTERs. In short, the car body rolls on an axis that is defined by a line
from the front roll center to the rear roll center.

The COG is the FORCE acting around that ROLL AXIS, which is the line
between the front and rear ROLL CENTERs. :) This is the actual rolling force.
It can be easier to simply think of this as the car's weight. Consequently,
it's location is basically an average of the car's parts/weights, a more
appropriate name would be 'center of mass'. I don't think there's a way
to calculate gravity's exact position even less it's 'center' :p

Anyone who had physics will tell you that Force(N) = Mass(kg) * Acceleration
(ms²). So 'weight' is a FORCE and not a MASS, as the common misconception
goes, your MASS is the same on the moon as it is on earth. What DID change,
is the FORCE you exert on the ground, this is what we call weight.

Something else i want to mention is that all this is not about weight transfer
as much as it's about contact patch management imo, for the reasons
specified above, namely the load/grip relationship of a tire, what we want is
not to reduce body roll as much as we want to even out this 'weight' (force)
over the 4 tires as not to overload any. The 'reduced' weight transfer is only
another way of saying that the COG moves less. Reduced body roll being the
visible indicator. This (COG moving less) is good for 2 reasons:

1- A moving COG creates torque around the ROLL AXIS which is what we
notice as body roll. This torque is the actual 'weight shifting'. The result is
an increase in the outer tire's load and a decrease of the inner ones. Again,
in theory, this isn't a problem unless the outer tire are OVERloaded.

2- Supposing that the COG is perfectly along the center of the car, which it
usually is close to, as the body leans, the COG becomes off center and tends
to 'fall to the ground', this actually ADDS itself to the centripede force, the
one that 'pushes you sideways'). Think of an object on top of a ball, if you
put it perfectly on the center, it will stay there, relatively static. If however
you place it slightly off-center, the object will fall off. Gravity is the force
that was at work there and it also applies to the COG . Obviously, the COG
would be the object and the ball would be the ROLL AXIS.

I still like to think that equalising the tire load is the actual goal. I just find
it simpler to understand the effects on handling. A leaning body doesn't
slide, a fully deformed and overloaded contact patch does. There is nothing
dramatic about tranferring weight in itself if it wasn't that our current tires
do NOT like to be overloaded.

To close the loop, anti-roll bars work to reduce the COG's movement (in
amplitude, not in acceleration though) and therefore optimizes the tire's
contact patches by spreading the load on all 4 tires to prevent overloading
them. At the extreme, if you overload all 4 tires, no ARB will help you, oh
and you're a n00b, learn to drive :p I'm simplifying by saying all 4 wheels,
but it's actually 2 bars each optimising grip on 2 tires. This is why ARBs
are also usefull for tuning as it effectively changes the front/rear 'grip'
bias. This is where the light starts dimming....

I told myself to keep it short, but i guess i just can't help myself ...

[CombatWombat]

This is not true. I highly suggest you read a layman's treatise on the physics of this. Tune to Win by Carroll Smith is old but very good.

Stiffer springs result in less weight transfer. Outside of considerations in dynamic camber control, softer antiroll linkages are better.


From Carroll Smith's Tune To Win:

"The greater the resistance of the springs, the less roll will result--but there will be no significant effect on the amount of lateral load transfer because the roll couple has not been changed and there is no physical connection between the springs on opposite sides of the car. The same cannot be said of the resistance of the anti-roll bars. In this case, because the bar is a direct physical connection between the outside wheel and the inside wheel, increasing the stiffness of the anti-roll bar will both decrease roll angle and increase lateral load transfer." 38.

I've no idea why I just quoted that.