Hi all!

Something occupy my mind for some time. It is influence of anti-roll bar on lateral dynamics of car. (in LFS)

I expect to stiffer anti-roll bar affect that increase lateral force which wheels can stand in curves. For example if I have oversteer, i make stiffer rear end ARB and get better control of rear end. Ofcourse if i make to much stiffer i will get understeer.

Actually it is hepend. Long time a go i recieve some "WR" setup for XF GTI - BL GP track (demo days ) and I feel much over steer so I make stiffer rear ARB a car vas mor in control. After short time I drive much better time!

BUT....

Now I drive LX4 - AS Club R and somethinh strange happend. Increasing stiffnes of rear ARB makes car oversteer much more than before, and stiffer ARB on fornt doesn't decres understeering

So i make two run, tryin to drive on same conditoins - same line, speed, throttle (aproximatly), with 2 same set, but one difference. In 1st run front ARB was on max and rear on 0, and on 2nd run oposite.

And i saw this.

Run 1



Run 2



Pict 1 show clear understeer. Pict 2 show on all 4 wheel "red" force but it is also clear owersteer.

Edited

But if we look on vertical force on tyres wi see that are in 1st case vertical force on tyre (as U say tyre load) bigger then in 2nd run! And tyre load on rear (outer ofcourse) wheel is bigger when is rear ARB stiffer. So stiffer ARB incress vertical force on outer wheel resulting weight transfer during turn. Thats I expected but...

How can wheel with bigger vertical load loose grip more rather than same wheel when have smaller tyre load?

I hope we can agree that is lateral force same in both case, cause it depend of CoG position between front and rear wheel, and it is same on both case....


What do U others think about this?

Without reading the whole post I think you've missed the important principle. Increasing your anti-roll bar decreases grip but increases the agility of that end of the car. Soften rear/stiffen front for understeer vice versa for oversteer.

Yeah, you got it completely wrong...

low ARB = can take more load on the outer wheel = more grip but slower cornering
high ARB = can take less load on the outer wheel = less grip but more agile

About the higher bar: I can only assume that you get that only because you're already sliding, thus you are over the limit... You'd have to look at the bars in the exact moment it starts sliding, I think you'll see it's smaller then...

As ajp71 says increasing ARBs at once end of the car decreases grip at that end, so they are working fine.

About the loads, higher ARBs always reduce the difference in left and right loads when cornering, which helps increase grip (thanks to tyre load sensitivity).
Stiffening just one ARB transfers more load to that end of the car (front or rear) when cornering, so infact decreases grip at that end of the car (again, thanks to tyre load sensitivity).

So if you want more grip, increase BOTH front and rear anti-roll bars. If you want to change handling balance, increase the ARB at the end you want to lose grip.

Quote from Blowtus :

http://en.wikipedia.org/wiki/Anti-roll_bar

OK i bilieve.... But...

Quote from Bob Smith :

As ajp71 says increasing ARBs at once end of the car decreases grip at that end, so they are working fine.

About the loads, higher ARBs always reduce the difference in left and right loads when cornering, which helps increase grip (thanks to tyre load sensitivity).

Stiffening just one ARB transfers more load to that end of the car (front or rear) when cornering, so infact decreases grip at that end of the car (again, thanks to tyre load sensitivity).

But still... How tyre with biger load can't stand bigger lateral force. This two thing should be proprtional...

Or I missing something again?

http://www.engin.brown.edu/cou ... iction_files/image009.gif could'n fin better image now...

N=mi*T

(mi-friction coeficient)

Tyre phisics is bit different (deformation,trail...) but not so different.

The reason for this is something called "tire load sensitivity." When you stiffen the bar and increase the vertical force a bunch on one tire, the lateral force goes up, yes, but the friction coefficient of that tire goes down.

In physics 101 it's taught that Coulomb friction is how everything works. I.e, the friction coefficient (let's call it "mu") is constant between two surfaces, such as a chunk of wood on smooth tile. This is pretty close to true when dealing with hard materials, but it's not correct for tires or other things made of rubber even on hard surfaces.

Imagine the two rear wheels. If Coulomb friction actually held true for tires and their friction coefficient (mu) with the road was 1, the amount of side force each of the tires could produce would be the same as the vertical force or load on them. So in a case where you have little weight transfer (your soft rear ARBs) on an axle with a weight of 1000, you might have vertical loads like this:

Inside tire 400
Outside side 600

The maximum cornering forces are also 400 and 600 (multiply the vertical load by mu, in this case 1). So total cornering force at the rear is 1000. If you then stiffen the rear end so you get a lot more weight transfer, you might have vertical loads like this:

Inside tire 100
Outside tire 900

If mu stays at 1 then you still have the same overall cornering force on that axle of 100 plus 900, or 1000. So really, if tires worked that way ARB's wouldn't do much of anything at all to the handling balance.

Tires have that "load sensitivity" mentioned earlier. This is just a fancy phrase meaning that the friction coefficient (mu) varies with the vertical force. I.e., it doesn't stay constant at 1 or some other number all the time. If you increase the vertical force, mu drops. The cornering force for that tire will of course increase, yes, but it does not increase in proportion to the vertical force increase. If you double the vertical load, the cornering force almost doubles too, but not quite. The inside tire produces less force since it's less loaded, but produces just a bit more than it otherwise would since its friction coefficient goes up. However, the outside tire loses more than the inside tire gains, so in the end you have less overall cornering force on that axle. Harder springs or ARBs will then make things more slippery on that axle.

If mu changed with load something like this:

Load
100 1.12
400 1.08
600 1.05
900 1.00

And we look at our soft and hard rear ends again (that doesn't sound right ) we will have different overall cornering forces for each one.

Soft:

Inside tire 400 load means mu = 1.08, so cornering force for this tire is 400*1.08=432
Outside side 600 load means mu = 1.05, so cornering force for this tire is 600*1.05=630

Total cornering force at the rear with the soft rear setup is 630 + 432 = 1062.

If we switch to the hard anti-rollbar we have a different situation because of this load sensitivity:

Inside tire 100 load means mu= 1.12, so cornering force for this tire is 100 * 1.12 = 112
Outside tire 900 load means mu = 1.00, so cornering force for this tire is 900 * 1.00 = 900

That gives us a cornering force of 1012 at the rear (900+112). This is actually less than the 1062 we get with the softer set up, so we tend to get more oversteer when we stiffen up the rear.

This phenomenon is pretty much the only thing that makes chassis tuning at the limit possible at all in most cars

And always we must remember that:
soft = grip
stiff = agility/handling

It would be too easy if stiff would be fast, it is not even quite lot of setup makers do think this way

Setting up a car for race is art form of itself, you have to make compromises, you need to evaluate which ability you can allow to be less, which one has less overall effect to race performance, for qualify or hotlapping you must think different aspects again, how to make car go as fast as possible for race that is not 100% same.

Antiroll bars are just one part, when you adjust them don't change them more than click or two and do lot of laps to tell what changes did affect.But if your tire pressure are way off it is no use to fine tune antiroll bars.

I'm currently building XR GT setup for blackwood, it is endurance racing setup so car have to perform flawlessy quite few laps.
I have been doing it quite long time already, yesterday did 1.34.72 without even pushing, there is plenty of more left (especially if driver would know how to drive, I don't) and setup still has few problems that I need to take care of, currently my antiroll bars are set to F15kN/R12kN, still car is rather stable and agile, most of opponents are using Gti so that creates own needs for setup, I have to be able to enter to corner at much greater speed that is efficient and still maintain my exit speed, gotta love RWD, can do some thing FWD will never do

Quote from AleksejBASOwarrior :

How can wheel with bigger vertical load loose grip more rather than same wheel when have smaller tyre load?

Short answer: more vertical load on the outside wheel means less overall grip, not more. Reason: the outside wheel gains some grip but the inside loses more.

Long answer: see Todd's post.

Quote from bbman :

Yeah, you got it completely wrong...

low ARB = can take more load on the outer wheel = more grip but slower cornering
high ARB = can take less load on the outer wheel = less grip but more agile

Lowering ARB means that on that end of the car less weight is transfered from inside to outside. This means the outside wheel can take less load, the inside wheel can take more load and the overall grip on that end of the car increases. And more grip is the same thing as faster cornering.

Quote from Bob Smith :

About the loads, higher ARBs always reduce the difference in left and right loads when cornering, which helps increase grip (thanks to tyre load sensitivity).

I think you meant lower?

Quote from Bob Smith :

Stiffening just one ARB transfers more load to that end of the car (front or rear) when cornering...

No, ARB's can't influence the weight transfer between left and right or front and back. This is defined solely by CoG position, G-Force track and wheelbase. ARB's do influence the way the forces are distibuted, just not from front to back or from side to side.

Example:
LF 70 RF 170
LR 130 RR 230

can become
80 160
120 240

Note that in both cases the overall weight distribution is
left 200
right 400
front 240
back 360

You could look at it as diagonally transferring weight i.e. lf+rr changes by the same amount as rf+lr but in the opposite direction.

DISCLAIMER: all the above does not take physical movement of CoG due to roll and pitch into account.

Quote from J.B. :

No, ARB's can't influence the weight transfer between left and right or front and back. This is defined solely by CoG height, G-Force track and wheelbase. ARB's do influence the way the forces are distibuted, just not from front to back or from side to side.

Example:
LF 70 RF 170
LR 130 RR 230

can become
80 160
120 240

Note that in both cases the overall weight distribution is
left 200
right 400
front 240
back 360

You could look at it as diagonally transferring weight i.e. lf+rr changes by the same amount as rf+lr but in the opposite direction.

Ah, thanks, was always a little grey in that area. I haven't coded the suspension part of my setup analyser yet, hopefully my understanding of all this will improve once it is.

A question though, is it directly acceleration that causes weight transfer or is it the combination of the forces on the vehicle. For example if you have a wing up very high on a car, above the CoG, even without acceleration the drag should cause rearward weight transfer (so long as the car is moving)?
If so then surely weight transfer wouldn't be caused by the accelerative force (e.g. linear driving force - resistance forces) but each component affects weight distribution in it's own right due to work at a different vertical distance from the CoG height?

Any clarification there would be great.

The main purpose of the sprung links (coil springs, leaf springs, sway bars) of a car is to keep the chassis from bottoming out. The ideal would be to set the car up with coil springs just stiff enough to prevent bottoming out under the heaviest of braking maneuvers, and acceleration. The ARBs (we call them sway bars in American slang, perhaps elsewhere too) should be just stiff enough to keep the car from bottoming outin turns. Excessively large sway bars can actually lift the inside tire off the ground. There are very few instances where this is desireable. Obviously the coils will help in turning as well. Also, you must consider braking/accelerating when turning.

Playing around with different ride heights and spring rates takes a while but will help in the long run. On the whole, whichever end of the car sits lower, will have the most grip, and a lower car will have a better center of gravity. The softest combination of springs and ARBs will offer the best overall grip.

It should also be said that the tires also act as springs. Higher pressures simulate stiffer springs, and also reduce grip. The best pressuer setup will see temperatures progress evenly from the inner edge to outer. Camber angles will cause the inner edge to run hotter, but not by more than a few degrees. Camber angles should only be used to even out temperatures by putting the tires contact patch parallel with the road under typical cornering. In LFS, we can see forces on the tire. I like to see even forces during cornering.

For the most part on production cars, Larger sway bars are added to limit body roll since the suspension geometry causes radical camber changes as the car rolls (I believe Macphereson strut setups cause this). It's not ideal, but it must be done if you are blessed with one of these cheaper to produce suspension designs. Racing oriented designs correct this and allow the car to roll up to the point that the chassis comes within milimeters of the ground (double wishbone setups help by having adjustable chassis pivot points). I will only say a word or two on dampers, and that is that as novice tuners, we should keep stocklike values unless we have radically raised or lowered spring rates. Adjustments to dampers should be proportionate to spring changes.

Spending time in the pursuit of increasing your tire's contact patch will increase overall grip. This is done with camber changes to even out temps from inside to outside and pressure changes for middle to outer. Once the tires are cornering flat and operating within an acceptable temp range, small spring, ARB changes will have a large effect on handling.

A chassis tuner that uses adjustments to lessen grip is not doing there job correctly. The right way is to try to add grip to the opposite end.

Note that you can't change the overall force on the outer tires versus the inner tires with anti-roll bars. You can only change the distribution of forces between the front and rear. The total force on the outside tires is a function of cornerning force, center of gravity, and the distance between the tires from side to side (track), (and track camber).

Quote from Bob Smith :

If so then surely weight transfer wouldn't be caused by the accelerative force (e.g. linear driving force - resistance forces) but each component affects weight distribution in it's own right due to work at a different vertical distance from the CoG height?

Any clarification there would be great.

Yeah, you're right, every force that acts on the car will transfer weight. So that would be accelerative forces, aero forces and also real mass transfer due to the movement of the CoG. This is also the reason why especially with aero cars you differentiate between low-speed handling and high-speed handling. It's just that when you talk about the effects of ARB's you don't mention aero forces to keep things simple. Which doesn't mean they're not significant of course. In fact it would be interesting to have a look at the forces view in LFS, once at a constant high speed and once at a constant low speed to see how much the aero forces do.

@ jtw62074

Now this is good explanation. Thanx!

I know for tyre load sesitivity (thats why truck never can have same tyre performance like car - thaths what I heard) but just not thinkin that way.

Quote from JeffR :

Note that you can't change the overall force on the outer tires versus the inner tires with anti-roll bars. You can only change the distribution of forces between the front and rear. The total force on the outside tires is a function of cornerning force, center of gravity, and the distance between the tires from side to side (track), (and track camber).

Almost, but not quite. See my example above. The total force on the front tyres is a function of longitudal force, center of gravity, and the distance between the tires from front to back (wheelbase).

Quote from J.B. :

In fact it would be interesting to have a look at the forces view in LFS, once at a constant high speed and once at a constant low speed to see how much the aero forces do.

"how much the aero forces do"? In terms of load transfer? If so, that's what I want to add to my aero analysis pretty soon. I have nearly all the necessary data thanks to .bin outputs in LFS.
With the front wing below the CoG height and the rear wing above the CoG height, it just means each wing actually has more effect then what it at first seems to (from just negative lift), I'd imagine the two would nearly cancel out anyway for a balanced aero setup.

Ah, good stuff. Do you also have data to do this for for the wingless cars? What exactly is LFS .bin output?

owersteer? lol

Quote from imthebestracerthereis :

owersteer? lol

owersteer, its how we pronunce it up where i live in aberdeen, we got owersteer an un-ersteer., lol

[edit] Spam post, im sorry

Quote from J.B. :

Ah, good stuff. Do you also have data to do this for for the wingless cars? What exactly is LFS .bin output?

Press "O" (for Output) when in the garage. A binary file labelled CARNAME.BIN will be created in the RAF folder within LFS. This contains tons of data about the car in the garage. It's in binary format, hence the extension, and has been available since test patch U15. It was a feature added by request for myself and Colcob for help with our analysers, I created a reader for the format while Scawen was developing the format.

It appears that the aero properties of the car body are pretty much not set, so the lift coefficient for the body of all cars is always zero, and I assume that the aero drag/lift centres for any car are also at the CoG position. I did post a link to download the reader in the U test patch thread, which has since been removed from view. I have included the reader in my new setup analyser, a preview of which is available in the unofficial add-ons forum if you're interested.

Thx. I remember the requests for this feature by you and Colcob (where is he these days?) but completely missed it's inclusion and your tool even though I love reading patch logs. Will look into it.

Quote from imthebestracerthereis :

owersteer? lol

illepall (only not thumb then middle....)

I'm not really a setup maker or whatever, but when I get a setup, I tend to change the bars to my likings, because that's one of the biggest things I notice about a set. Now I was wondering it this can effect straight line speed?

ahh I see <- the whole gang

[/spam]

Quote from sgt.flippy :

I'm not really a setup maker or whatever, but when I get a setup, I tend to change the bars to my likings, because that's one of the biggest things I notice about a set. Now I was wondering it this can effect straight line speed?

Not directly, but as it can affect to exit speed from corner that naturally does have some effect to speed what you get on straights.