Roll stiffness: Springs vs anti-roll bars?

[Captain Snooze]

Question time!!! Ready?
If somebody wants to reduce roll during cornering they can increase the spring stiffness or increase roll bar stiffness (or both but I want to keep this as simple as possible). What is the practical difference between the two options?

[Calum]

Ride quality is effected by the spring rate change more then the sway bar change.

[Xauterus]

For road cars, sway bars are used more to stop roll but keep a plush ride. On a track car much stiffer springs can be used due to the better surface. Sway bars are then used for tuning.

[7thgear]

Originally Posted by Captain Snooze

Question time!!! Ready?
If somebody wants to reduce roll during cornering they can increase the spring stiffness or increase roll bar stiffness (or both but I want to keep this as simple as possible). What is the practical difference between the two options?

they can also..

1. reduce the weight of the vehicle, particulary removing weight away from the centerline of the vehicle

2. change geometry, remember that ROLL is a factor of geometry

take a pencil and hold it up vertically, gently, with your fingertips. FIrst hold it in the center, and move it side to side. See how the pencil stays vertical? That's because the weight above and below the ROLL CENTER is equal.

Now shift your finger grip higher, so that 70% of the pencil is below and 30% is above. Move it side to side, see how the pencil swings, thats because there is a difference between roll center and the center mass of the object.

now imagine you flip this upside down and you get what goes on in almost every single production car out there.

SPRING STIFFNESS therefore, does not REDUCE ROLL, it FIGHTS IT.


so one other way you can reduce the cars tendency to roll is to change it's geometry, that way you get to keep a softer spring for better mechanical grip and a more pleasant ride

[atledreier]

Also note that lowering the car even further can mess up the roll centers, and actually make the car roll MORE, depending on geometry. Haven't studied the geometry og the '86 yet, so won't comment on this actual car.

[-max-]

Springs stiffen each corner seperately. Keeping each corner independant. Sways work against each side being independant.
Ideally No sway bars would be best, but the sping rates would be astronomical. So, like everything, it is a compromise.
The new Mclaren MPC12 uses no sway bars. Each corner is truely independant and stiffness controlled by computers and electro-something shocks.

[fatoni]

Originally Posted by 7thgear

they can also..

1. reduce the weight of the vehicle, particulary removing weight away from the centerline of the vehicle

2. change geometry, remember that ROLL is a factor of geometry

take a pencil and hold it up vertically, gently, with your fingertips. FIrst hold it in the center, and move it side to side. See how the pencil stays vertical? That's because the weight above and below the ROLL CENTER is equal.

Now shift your finger grip higher, so that 70% of the pencil is below and 30% is above. Move it side to side, see how the pencil swings, thats because there is a difference between roll center and the center mass of the object.

now imagine you flip this upside down and you get what goes on in almost every single production car out there.

SPRING STIFFNESS therefore, does not REDUCE ROLL, it FIGHTS IT.


so one other way you can reduce the cars tendency to roll is to change it's geometry, that way you get to keep a softer spring for better mechanical grip and a more pleasant ride

that doesnt make any sense. if there was more roll and now there is less roll, it reduces roll. it doesnt change really change the weight transfer if thats what you mean. and the guy says he wants to keep it simple so i dont think changing the cars geometry considering that very few race teams even do that.

Originally Posted by -max-

Springs stiffen each corner seperately. Keeping each corner independant. Sways work against each side being independant.
Ideally No sway bars would be best, but the sping rates would be astronomical. So, like everything, it is a compromise.
The new Mclaren MPC12 uses no sway bars. Each corner is truely independant and stiffness controlled by computers and electro-something shocks.

that is absolutely not true. sway bars are a good thing to have. they are used in just about every single street car and they are even on f1 cars. how much effective stiffness do you think sway bars have? removing that stiffness with spring wouldnt make the rates astronomical. for instance, a typical miata sway bar wheel rate is going to be from 84 lbs/in to 157 lbs/in. as for the mclaren, using shock to add stiffness is a compromise too and in many competitive cases illegal to use. even if it wasnt, you are comparing a 250k car to a 25k car.

[Element Tuning]

Every car/chassis is different and speaking in generalities doesn't always apply to a particular setup. Roll stiffness isn't the most important objective, traction is. The more grip you have, if you can balance each end is most important. High cornering speeds are extremely important but so is being able to accelerate as early as possible.

Now I will probably make a new post discussing this chassis and how it works on the track. On my FRS race car we are running slicks, Element Tuning custom valved and sprung BC Coilovers, and ditching the rear sway bar dropped lap time by 2 seconds.

[jamal]

The Mclaren has some sort of a hydraulic anti-roll system I think.

The answer comes down to testing, but the BRZ isn't a big tall sedan like the Impreza, which usually benefits to a significant bar upgrade. These cars aren't going to need as much. Trying to limit roll with geometry is not a good idea (short of widening the track). The shock and spring and swaybar are supposed to be doing that.

Element: any testing between stock suspension and the BC stuff on the same tires? I'd think just going from the Michelins to slicks would drop lap times by the seconds.

[Element Tuning]

Yes we ran on stock suspension with slicks for two races. I pretty much knew after the second race a larger sway bar in the rear was going to be a big mistake as I was already starting to unload the inside rear wheel causing the Torsen diff to go open.

For reference Mike Skeen who is more than qualified to extract everything out of the car stock ran just about 8 seconds slower than our car at oem power. The two second gain was came from giving more independence to the rear suspension allowing me to hit the curbs and it kept the inside wheel down therefore preventing the differential from going open.

[jamal]

Not surprised. I've been involved with the setup of some other rwd cars where we did the same thing. Nice having a car where you can dial out understeer with the right foot.

[Matt Andrews]

2 seconds where?

that is a ton of time on just about all east coast tracks. I wonder how much of that could be made up with a clutch pack diff. Tuning around a torsen sucks.

A lot of this comes down to driver preference and what makes them comfortable. I like rwd cars that roll in the back. generically, guys that come from karting seem to like a really stiff rear.

[fatoni]

Originally Posted by Element Tuning

Every car/chassis is different and speaking in generalities doesn't always apply to a particular setup. Roll stiffness isn't the most important objective, traction is. The more grip you have, if you can balance each end is most important. High cornering speeds are extremely important but so is being able to accelerate as early as possible.

Now I will probably make a new post discussing this chassis and how it works on the track. On my FRS race car we are running slicks, Element Tuning custom valved and sprung BC Coilovers, and ditching the rear sway bar dropped lap time by 2 seconds.

something worth mentioning is that disconnecting the rear sway bar on a lowered car allows the inside wheel in a turn to droop further. Since the inside front with swaybar still attached will not droop, rake changes. This means the front stays low, the rear jacks up in a turn. The effect is slight but it changes the roll axis, camber gain and a few other things. For autocrosser, this compromise is usually worth it as it will greatly reduce wheel spin and reduce oversteer during low speed transitions. For track use however, this will usually result in terminal understeer that will have you pulling your hair out trying to get rid of. In short, if you can get your autocrosser to work with the rear bar, keep it hooked up. For track use, save yourself the headache and leave it on. also what track were you testing on? 2 sec is a lot of time for a sway change (unless the additional understeer is adding that much more confidence to the driver)

[jamal]

Probably worth the trade off if you gain corner exit speed. Element might also have enough rear spring rate to not need the bar.