This thread is in response to a previous thread: http://www.i-club.com/ubb-files/Forum15/HTML/001722.html

After reading this thread, I wanted to throw in my opinion. This is a summary of body roll, weight transfer, and roll bars. If you disagree with anything here, please feel free to correct me (just be prepared to back up what you say http://www.i-club.com/ubb-files/biggrin.gif):

1) For steady-state cornering there is a total amount of load transferred to the outside wheels of the car. This load depends primarily on the lateral acceleration of the car, and the angle of roll does not really matter.

2) The front and rear axles have a roll stiffness (resisting torque to roll/roll angle). This stiffness is:
K_roll = 0.5*K_spring*(distance between suspension springs)^2 + roll bar stiffness

3) There is somthing called a "roll center" for each axle. This is a point which the body of the car "rolls" about. The line drawn between the front and rear axles is called the roll axis.

4) The load transfer between the left and right wheels for either the front or rear axle is given by:

Front:
LoadTransfer={[h_rcf*Wf] + [h1*Ws*K_rf/(K_rf+K_rr-Ws*h1)]}*[ay/(tf*g)]

Where h_rcf = height of the roll center for front axle
Wf = Static Weight on the front axle
h1 = distance between the center of sprung mass to the roll axis
Ws = Sprung Weight
K_rf = Roll Stiffness of the Front Axle
K_rr = Roll Stiffness of the Rear Axle
ay = lateral acceleration
tf = track width
g = acceleration of gravity

A similar expression can be derived for the rear load transfer by switching the front and rear quantities in this expression.

Comments:
[h_rcf*Wf]
This is the load transfer due to the torque generated by the lateral tire forces. Only way you can reduce this is to lower your roll center.

[h1*Ws*K_rf/(K_rf+K_rr-Ws*h1)]
This is what the disscussion is about. This is the load tranfer relating to "roll". To examine what we can do to "adjust" this load transfer, look at this term: K_rf/(K_rf+K_rr-Ws*h1).
This is essentially a ratio of axle roll stiffness to total roll stiffness (exept for the Ws*h1). This means that the load transfer on the front will be large (and that the rear axle load transfer will be small) if the front roll stiffness is large compared to the rear roll stiffenss. The Ws*h1 is only modified by reducing your hieght of the center of mass of your car, or moving the roll axis.

5) So putting a stiffer rear bar generates more load transfer along the rear axle, and less along the front. You also increase your total roll stiffness, which will help to preserve the original suspension geometry.

6) An increase in load transfer on an axle will decrease the amount of total traction available to that axle.

7) Roll is bad because generally it will generate positive camber.

Randy

[This message has been edited by rrsettgast (edited May 06, 2000).]

In response to #6...That was my understanding in the first place.


rrsettgast: Would you mind giving me your opinion on three questions...True or False; There is a significant amount of front end traction gained from using a larger rear bar. True or False; The rear end is losing "total available traction" while using a larger bar(you already answered this). Do those two concepts come together in any consideration of chassis dynamics?

There is a significant amount of front end traction gained from using a larger rear bar.I guess the amount of traction gained depends on the increase in roll bar stiffness and tires...conditional TRUE
The rear end is losing "total available traction" while using a larger bar(you already answered this).TRUE
Do those two concepts come together in any consideration of chassis dynamics?I don't know what you are asking???

Just remember that the above disscussion in for STEADY STATE TURNING, so there are other things that happen from the time you are going straight, until the time you are set into your turn.

Randy

[This message has been edited by rrsettgast (edited May 07, 2000).]

Randy,

I see nothing wrong with your analysis.

Greg,

Your understanding of the wya the car reacts to roll bar stiffness is correct. I beleive however that you still have the mechanics of the load transfer backwards. A sway bar tansfers load from the outside to the inside. So what happens is:

less load = less grip

what I beleive you are thinking is:

more load = less grip

At least that is what I understood from the previous thread.

So on an RS a stiffer rear bar overpowers that outside rear tire causing a more neutral cornering state by reducing the load on that tire.

quote:
---------------------------------------------
Do those two concepts come together in any consideration of chassis dynamics?
-----------------------------------

YES.

Randy's model is for steady state cornering, but of course vehicle dynamics is much more complex, a car with very high skidpad numbers might not handle well if it doesn't have good "transient response".

Transient Response is a vehicle's response to changes in direction. Cars with very stiff suspension have very fast weight transfer and can change direction very quickly.

So to answer your question changes in roll stiffness (sway bar changes, spring changes, damper changes) change the speed of the weight transfer. Weight will always transfer. The question with suspension tuning and vehicle dynamics is how fast will it transfer? Also what should the weight transfer speed realtionship be bewteen the axles?

Then what the heck does #6 mean? Isn't an increase in load transfer just another way of saying that weight is loaded more on the outside tire, while using a larger rear bar? Keep flogging me with this conversation, maybe someday I'll come around! http://www.i-club.com/ubb-files/smile.gif

Greg,
You are one persistant dude.

When you go into a turn the body wants to lean, the sway bar counteracts the lean taking the load away form the outside and putting it back on the inside.

By taking the load away from the outside tire its grip is reduced so it can't take as much lateral force. With a big rear bar that is what happens and then the rear loses traction before the front cause the rear can't take as much force.

Think about what happens when you drive. When you are on the brakes a little the car turns in better cause it has more weight over the front wheels. When you are on the gas it understeers real bad cause the weight is over the back and the front has no grip.

Same thing with the sway bar. It takes weight away from the outside wheel.

Remeber Vehicle dynamics is all about the management of weight transfer.

I figured it out!

I figured it out!

You thought Randy was talking about body roll as load transfer, but body roll is WEIGHT transfer. Sway bar twisting is LOAD transfer.

So the body rolls and transfers WEIGHT, the suspension compresses on one side and extends on the other, the sway bar limits this compression and extension by transferring LOAD from the compressed side to the extended side through a twisting action.

Did that work?

Ack, I tried to respond to this last night but I was too tired. Now I'm beat cause I ran my first (woo hoo!) autocross today, but I'll put a couple of yen in on this one...

Ok, now theoretically body roll transfers the weight over the outside wheel of the car and you gain more grip because it is being pushed into the ground more. However, this is a different story according to how the suspension moves with the body and how the wheels change angle since the body is rolling so much.

Now, with a sway bar in your suspension on the outside rear tire tries to compress to get more weight over that tire and increase grip, however the sway bar twists and tries to keep the car level. The more it tries to keep the car level, the less weight it is transferring over to the outside wheel since there is less body roll.

I think that is about the easiest explanation I can do on that one - especially since I'm just starting to get the hang of it myself. http://www.i-club.com/ubb-files/smile.gif

Greg,
By "Load Transfer" I mean the change in vertical load on a wheel. So if the car is going in a straight line at a constant speed, the front wheels have an equal vertical load, and the rear wheels have an equal vertical load. As you are in a turn, the Load Transfer is the change (added to the outside wheel, subtracted from the inside wheel) from the straight line load.

Pilot,
No I don't think that is what I meant.
Hmmm. Lets see if I can explain this better. By "Load Transfer" I do mean Weight Transfer. If you look at my equations above, you will notice that there is no mention of the actual amount of "roll". What we are concerned with here is "roll stiffness". Lets look at an example of adding a thicker rear sway bar.

1) A larger rear sway bar adds to the roll stiffness of the rear axle.

2) Increased roll stiffness of the rear axle leads to an increase in vertical load transfer between the inside rear and outside rear wheels, and a decrease in vertical load transfer between the front wheels.

3) The Increase in vertical load on the outside rear tire will increase the amount of grip it has. The Decrease in vertical load on the inside rear tire will give that wheel less grip. The opposite is occuring for the front wheels.

4) Due to the nonlineararity of tire friction, the net result of 3) is that the rear wheels have less total grip, and the front wheels have more grip. Less grip from the rear wheels, and more grip up front = more oversteer.

5) SO the sway bar does not transfer load from the outside wheel to the inside wheel. It is exactly the opposite. Increasing the roll stiffness leads to a load transfer from the inside wheel to the outside wheel!

Randy

[This message has been edited by rrsettgast (edited May 08, 2000).]

http://www.members.home.net/rck/phor/ http://home.kscable.com/mc2racing/suspension.html

Just a couple sites that talk about this subject.

[This message has been edited by Scooter (edited May 08, 2000).]

"Actually, the transition from sticking `mode' to sliding mode should not be very abrupt in a well-designed tire. When one speaks of a ``forgiving'' tire, one means a tire that breaks away slowly as it gets more and more force or less and less weight, giving the driver time to correct. "

-Brian Beckman

Ok this means that #5 is incorrect. And also illustrates why the back of a car slides first during trailbraking or trailing throttle maneuvers.

Why? The more the outside suspension compresses the more the sway bar is trying to flatten the car out, by doing this it is removing load from the compressed outside suspension and transferring it to the extended inside suspension. Not the the other way around, which can not happen because the suspension can only extned so far, once it is fully extended there is no more movement form the sway bar EXCEPT that which is occurring on the outside tire where the suspension is still compressing. Therefore the sway bar is taking load from the compressed outside and trying to move it back across that axle.

Why rear sway bar does not add load to the outside tire.

The sway bar is simply two lever arms connected by a torsion spring. When discussing a rear bar what we get is a dynamic load transfer from the outside rear tire to the inside front tire. This is diagonal and explains the purposes of weight jacking.

As I discussed before, weight jacking or corner balancing is the practice of balancing a cars' weight left to right by changing the ride height of the car independantly at each corner. Example: To place more weight on the left front one must lower the ride height there, or raise the ride height of the right rear, the diagnoal.

Sway bars allow this ride height change to occur in a dynamic situation, and only in a dynamic situation. I also stated previously that a properly setup racecar that is also properly balanced (CART, F1, etc) does not use a sway bar because it doesn't need it. The fundamental action of a sway bar is negated by these cars, they instead use on board weight jacking systems. The driver can actually change the diagonal ride height from the cockpit.

One more thing, you NEVER want to have a sway bar that provides more torsional springing action than your suspension (springs and dampers).

1) It will rip the sway bar mounts off the car

2) It makes the rest of the suspension work poorly

Randy,
I see what you are saying, one tire cannot take as much lateral force as two tires.

True but is this case it doesn't work because you cannot stop weight transfer from the inside to the outside, so the inside tire does not have any grip UNLESS you are accelerating in which case you have transferred load from the front to the back. In any at the limit steady state cornering maneuver all the load gets transferred to the outside.

Like I said it is all about how fast the transfer occurs. And of course when we are driving the car we often have it braking and turing or accelerating and turning.

The only purpose to limiting body roll is to maintain wheel geometry realtive to the ground.

Pilot,
I think we have a failure to communicate. Think about the mechanics of a sway bar. The ONLY way the sway bar can "prevent" body roll is to transfer load from the inside wheel to the outside wheel, WITHOUT compression of the springs. This is what the sway bar does!!!

Think about laying under your car, and have some big dude push your car so that it rolls to the right. Think about how the sway bar will react. The bar will be twisted so that the right sway arm is bent up, and the left arm is bent down. This implies that the sway bar is transferring load directly from the right wheel and pushing up on the right side of the chassis...which will level the chassis. Opposite for the left. If you are having a har time visulaizing this, go do it!

The tire thing is due to the fact that as you increase weight on a tire, you do not get an equivilent increase in lateral grip. The increase in grip will decrease until you reach the limit of the tire. Therefore, two evenely loaded tires will have more grip than one really heavily loaded tire + one lightly loaded tire.

As for dynamic effects...I think that has more to do with the roll inertia of the car, and the damping rate of the struts.

Randy

[This message has been edited by rrsettgast (edited May 08, 2000).]

I think I have found the problem, I have once again failed to properly explain myself.

A sway bar provides spring action in two directions. I pushes up on one side and down on the other.

A sway bar increases the spring compression rate on the compressed side of the car in a turn and decreases spring compression rate on the extended side.

"two evenely loaded tires will have more grip than one really heavily loaded tire + one lightly loaded tire."

-rrsettgast

This change in spring rate effects the opposite end of the car. So a stiffer rear bar causes more equal loading of the front tires creating more grip at the front of the car.

What I am trying to get at, and failing miserably at it, is that the rear sway bar affects traction at the front of the car much more than the rear of the car.

Conversly the front sway bar affects traction at the rear of the car much more than it does the front.

So why use a bigger rear bar as opposed to a smaller front bar? To control body roll to maintain tire geometry with the ground which increases traction at both ends of the car.

More roll stiffness = better tire geometry relative to the ground = more grip.

Lastly a sway bar can not actually transfer load from the inside to the outside, it can help to increase spring rate at the outside corner which more effectively loads the two front tires.

In my previous attempts to exlpain this I tried to only talk about the rear of the car for simplicity's sake and it made a big mess, disregard all talk about load transfer from left to right being possible by a sway bar, because it isn't.

The sway bar simply uses the opposing suspension movement on both sides of the car to help increase spring rate at the compressed side.

So Randy yes you are right and I am a fool for attmepting to explain this decrease in relative grip from front to back without discussing the increase in grip at the front.

This was the cause of my explanatory confusion.

Is everybody clear now?

stiffer rear bar = more fornt grip = less understeer

Stiffer front bar = more rear grip = more understeer

Sounds Close Enough http://www.i-club.com/ubb-files/biggrin.gif You are not a fool, you are just trying to work through some complicated stuff, which is better than not working through it!!

Randy

Finally!

At least this time I'm much more educated on the subjects at hand, specifically vehicle dynamics and suspension action.

Unlike the tire post where I was trying to assimilate infromation that was in many cases new to me.

Here is a technical trivia:

Did you know?

That if your suspension can continually create more negative camber relative to the road surface as the weight transfers to the outside your tires will continually produce more grip.

This is a catch 22 becuase if this ocurrs you can flip your car! Seen race cars flip? Now you know why!

"A sway bar increases the spring compression rate on the compressed side of the car in a turn and decreases spring compression rate on the extended side." - Pilot

Oddly enough, I never thought of it that way but it is completely the way it actually is. Very cool.

So let's see... if the front sway bar size would be increased, not as much weight would be able to transfer over to the front outside wheel and it'd be stuck in the back, cauing the rear of the car to push it through the turn. Understeer. Correct?

So really the best way to decrease body roll would be to completely move the suspension up to the next level, stiffer all the way around... Geez, the only thing that kinda sucks about all this discussion is the realization that I have to buy more and more stuff before I get the car handling nice and flat. Oddly enough, I managed to do pretty damn well on my RE92's at the autocross yesterday, they're pretty chewed to hell, though. LOL http://www.i-club.com/ubb-files/smile.gif

Woo Hoo! Yes that is right!

However you should still get new tires first! http://www.i-club.com/ubb-files/wink.gifThen the suspension.

Also what 8complex pointed out about how a sway bar changes weight distrubution is what I was trying to explain by saying that the larger rear bar reduces the grip at the outside rear tire. The key is relative grip.

For Greg,
It isn't that the rear bar is overloading the outiside rear tire, it's that the rear bar is increasing the grip at the front of the car.

For Randy,
I just want to make sure we are on the same page here. Do you agree with both of these statements? If not why not?

A sway bar can not actually take load off of or incease load on the outside tire.

However, by increasing the spring rate of the compressed outside suspension it can more evenly load the front tires.

Ok, just for kicks lets check out suspension costs over hte next (?) year or so...

Current:
Whiteline Rear Sway bar
Rear endlinks (on their way)
Front endlinks (got 'em cheap *shrug*)

Future:
Tires (400-500 - I eat em fast so cheap is the way to go)
Strut braces (320-360)
Struts/Springs (800-1200) - Very much an estimate, still not educated enough to buy these yet anyway.

Man, I'm gonna hafta get a second job just to pay for car parts! LOL http://www.i-club.com/ubb-files/smile.gif

For simplicity sake lets ignore front to rear weight transfer and only consider the rear wheels.

So the increased load on the outside tire is related to the force of the springs pushing the inside wheel away from the body and sprung weight of the inside wheel and suspension. This is why a very stiff roll bar will lift the inside rear tire, since the entire weight and spring compression (ground seeking force) of the inside wheel is travelling down the sway bar to outside wheel increasing the load. A car that lifts the inside rear tire is not only supporting all the body weight on the outside tire but also the weight of inside suspension. Correct?

Just trying to get these things clear in my head.

Tim

Don't forget the human factor. I think large amounts of body roll tend to make the average driver uneasy, regardless of how much grip they may have left. I think a lot of people determine whether or not they're taking a turn too fast by the amount of body roll versus the amount of grip left. With reduced body roll usually comes more confidence, despite whether or not grip or handling actually increased. Humans have a pretty sensitive balancing mechanism and it doesn't take much to trip the "we might tip over alarms".

Boxerman,
Originally I tried to explain this ignoring the diagonal weight transfer, but you can not ignore it. Without it the fucntion of a sway bar, or any part of the suspension for that matter, does not make sense.

I have to go take an exam so I will let Randy examine your last post for correctness.

Scooter,
I am ignoring driver psychology in this discussion, but you are correct about the "tip over alarm" for many drivers.

Pilot,
I think it is better to think of a sway bar as a seperate source of roll stiffness. The bar is merely providing an alterante path for the load transfer. It is either through the springs/struts, or through the roll bar.

8Complex,
"if the front sway bar size would be increased, not as much weight would be able to transfer over to the front outside wheel and it'd be stuck in the back, cauing the rear of the car to push it through the turn. Understeer. Correct?"
Not quite right.
1) Bigger Front Bar
2) More weight transfer up front , less in rear
3) less grip up front and more grip in rear
4) Understeer

Pilot,
"Also what 8complex pointed out about how a sway bar changes weight distrubution is what I was trying to explain by saying that the larger rear bar reduces the grip at the outside rear tire. The key is relative grip."
Not Quite.
1) Larger rear bar increases load transfer in rear
2) Outside Tire grip increases, inside tire grip decreases. The increase in outside tire grip is less than the decrease in inside tire grip. Therefore less overall grip in the back.
3) Opposite of 2) for the front

Pilot,
"It isn't that the rear bar is overloading the outiside rear tire, it's that the rear bar is increasing the grip at the front of the car."
Near the limit, the larger rear bar IS overloading the outside rear tire. You are right about the front...see above

Boxerman,
You cannot neglect the interaction of the front and rear axles. The difference in roll stiffness is what determines the disribution of weight transfer bewteen the front and rear axles.

Randy

[This message has been edited by rrsettgast (edited May 08, 2000).]

Boxerman,
Your example is correct. However, it should be noted that the amount of weight transfer in the rear is dependant on the roll stiffness in the front. It is not to large of a step to get from an isolated axle to a pair of axles.

8complex,
I agree. Weight transfer can be thought of as "anti-roll" in a relative sense. However, since the body roll up front is the same as in the rear (stiff chassis), it is better to think of weight transfer as a product of roll stiffness...see equations above. Total amount of roll doesn't matter, only the roll stiffness!! If the chassis is allowed to flex, then the magnitude of roll will start to matter!!

Randy

[This message has been edited by rrsettgast (edited May 08, 2000).]

rrsettgast - The equations do nothing for me but confuse me... There was a reason I didn't take calculus in high school. http://www.i-club.com/ubb-files/smile.gif

I think I kinda understand but let's go to an extreme type of an example...

Lets say you design a chassis/car so that the roll center is at the same point as the center of gravity, when taking hard corners would the car stay flat, or would it's suspension still compress on one side and decompress on the other? I would think it'd stay flat since all the forces the wheels would be getting would be lateral at that point.

Let's see if I can sum up some points.

4 contact patches are better than three for purposes of maxiumum grip.

Some cars handle better with three contact patches for reasons of vehicle dynamics.

Handling characteristics are determined by the relationship of roll stiffness between the front and real axles.

There are four ways to change roll stiffness
1)springs
2)dampers
3)sway bars
4)lowering the CG (center of gravity)

Randy,
Something I'd like to discuss in regards to the rear bar.

Ignoring the fornt of the car:

You are saying that the action of the rear sway bar increases pressure on the outside rear tire and decreases pressure on the inside rear tire. (I agree, goes back to my statement that a sway bar increases spring rate at the compressed side) The increase in grip on the outside rear tire is proportionally less than the deacrease in grip on the inside rear tire. The net overall effect is reduced grip at the rear of the car becuase too much sideways force is being placed on the outside rear tire without enough help from the inside rear tire. (Ignoring the effect on the front of the car I agree)

So would it be correct to say: a larger rear bar increases grip on the outside tire BUT removes more grip from the inside tire than it gives to the outside tire resulting in a net rear axle grip loss?

So if this is correct Greg is right and I am wrong, my contention that pushing the inside rear back down was taking load away from the outside rear. Of course I disproved this myself by noting that a sway bar increases spring rate at the compressed corner. However I still had my affects correct becuase I still had a net decreae in grip at the rear axle. Anyway Greg's assertion that the sway bar overloads the rear tire is true given that it is related to the deacrease in grip from the inside rear tires.

Taking into account the front of the car:

The increase in spring rate at the rear of the car due to sway bar action produces more roll stiffness and enhances grip at the front of the car.

Do you agree? If not why not?

If yes would it be correct to say: A larger rear bar reduces grip at the rear of the car while simultaneously increasing grip at the front of the car?

Finally if all this holds true then I am back to the reason many race cars don't use sway bars, it results in a net grip decrease, and sway bar effects can be acheived by changes in spring rate, damper rate, and ride height.

So if money is no object the ultimate scooby suspension uses very stiff springs and dampers (with adjutsable spring perches) and only uses sway bars for very minute fine tuning of the chassis IF necessary.

Street cars use sway bars as a primary means of changing roll stifffness so that they can have cushy suspensions.

Given a small budget the fastest and least expensive way to alter your car's handling behavior is a sway bar change.

8Complex,
As long as there is a roll center above the surface of the ground, there will be weight transfer, and therefore roll. Having a low center of mass definitly helps, and having the roll axis through the CM helps.

Pilot,
"So would it be correct to say: a larger rear bar increases grip on the outside tire BUT removes more grip from the inside tire than it gives to the outside tire resulting in a net rear axle grip loss?"
YES

"The increase in spring rate at the rear of the car due to sway bar action produces more roll stiffness and enhances grip at the front of the car."
No Roll Stiffness is increased at the rear, which increases front grip....and the spring rate doesn't change, the roll stiffness changes.

"If yes would it be correct to say: A larger rear bar reduces grip at the rear of the car while simultaneously increasing grip at the front of the car?"
Yes

"Finally if all this holds true then I am back to the reason many race cars don't use sway bars, it results in a net grip decrease, and sway bar effects can be acheived by changes in spring rate, damper rate, and ride height."
Maybe. Could also be that track race cars generally do not have independent suspension, and therefore do not need sway bars?

"Given a small budget the fastest and least expensive way to alter your car's handling behavior is a sway bar change."
Maybe. You want to make sure that the struts have enough damping so that there are no resonances in your suspension system. If you get really stiff bars with stock suspension, you may get into some trouble....but don't quote me on that.

Randy
I Hope this is helping.

Randy

"No Roll Stiffness is increased at the rear, which increases front grip....and the spring rate doesn't change, the roll stiffness changes."

Roll stiffness is determined by the compression rate of the spring, the damper, and the sway bar. Therefore if the sway bar is increasing roll stiffness it is also increasing compression rate. YES?

"Maybe. Could also be that track race cars generally do not have independent suspension, and therefore do not need sway bars?"

Umm...what? The only track cars that do NOT have sway bars are those with independent suspension. Any track car with solid rear axle, or dead beam rear axle has a sway bar. Also a sway bar by its design ties in the two sides of a suspension making it not quite independent.

"Maybe. You want to make sure that the struts have enough damping so that there are no resonances in your suspension system. If you get really stiff bars with stock suspension, you may get into some trouble....but don't quote me on that"

Oops to late, you've been qutoed. Now let me qualify my point by saying that the sway bar cannot have more force acting on the car than the spring and damper setup, otherwise you wil get into trouble.

If your bars are too stiff. (minds out of the gutter please http://www.i-club.com/ubb-files/wink.gif )

1) Your car will no longer have proper independant suspension action.

2) You will rip the pivot mounts (the ones on the car body) off the car. (I've seen this, pretty nasty stuff.)

3) Your car will handle like (pick expletive and insert here)!

Also, I believe I heard somewhere that theoretically, if you have too much oversteer (too large a rear bar) then there is a theoretical speed at which your vehicle will lose all stability and most likely end up spinning.

I kinda proved this one in Gran Turismo 2 (that is, IF the physics engine figured right) with my Pikes Peak Escudo... I have so much oversteer (so little rear end downforce too) that when I hit 210, if I even so much as breathe on the steering, I end up in a wall. Very cool, LOL. http://www.i-club.com/ubb-files/smile.gif

"Roll stiffness is determined by the compression rate of the spring, the damper, and the sway bar. Therefore if the sway bar is increasing roll stiffness it is also increasing compression rate. YES?"
NO. You mean "compression stiffness"...don't you? If you go over a speed bump head on, the sway bar has no/little effect. If you hit a bump with only one wheel, the flexable sway bar bushing will likely prevent the sway bar from loading up that much. Regaurdless, the spring rate is the spring rate, and the roll stiffness is the roll stiffness....I would think

Randy

[This message has been edited by rrsettgast (edited May 08, 2000).]

Hmmmmmm... makes me wonder if you have a big sway bar on and hit a bump around a corner with only one wheel, would you get bump steer? I know I used to on my old Volvo but that was cause of how the rear axle was tied to the chassis...

Hello,

Sheesh, go away for a weekend of junkyard digging fun, and miss out on a whole conversation full of tech. I like the conversation but saw one mistake that wasn't really picked up on much...

7) Roll is bad because generally it will generate positive camber. ...
The rear end is losing "total available traction" while using a larger bar(you already answered this). - TRUE

Randy had it right, then he had it kinda wrong, then his mistake was quoted back, and he made the same mistake again.

What I mean is that assuming camber remains the same, the rear end will slip out faster (with a thicker rear bar). But since camber doesn't remain the same, ESPECIALLY in a strut-system like ours, this statement doesn't hold. The camber change in the stock setup is enough that by stiffening the rear bar, you prevent enough camber change to actually increase the available traction. You keep the tire close to flat on the ground passed the point where you would have been riding on the outside edge of the tire.

Someone else came to the correct conclusion. A swaybar is a fix for a too-soft suspension. You should be able to accomplish the same thing with a stiffer suspension. However, there are two problems with that - first, at some point, the ride quality becomes rediculous. And second, you need some suspension compliance in real-world situations to accomadate imperfections in the road. But, on the flip side, you don't want to "fix" the problem by using a too-stiff swaybar since that reduces the independant motion of the suspension (your car starts to behave like an American car with a live rear axle http://www.i-club.com/ubb-files/smile.gif

Joel

"The camber change in the stock setup is enough that by stiffening the rear bar, you prevent enough camber change to actually increase the available traction. You keep the tire close to flat on the ground passed the point where you would have been riding on the outside edge of the tire." - Joel Gat, 1.8L

How does that work out? You keep the tire flatter (which is considered to be more grippy then riding on it's edge) to the ground yet you get less traction (understeer) at the same time? Did you not type this right or am I reading it wrong? LOL

[This message has been edited by 8Complex (edited May 08, 2000).]

Hello,

You keep the tire flatter (which is considered to be more grippy then riding on it's edge) to the ground yet you get less traction (understeer) at the same time?

First, I see what you're saying and you didn't say it http://www.i-club.com/ubb-files/smile.gif You get more oversteer with the thicker bar which you are equating with less traction, right?

I think the problem is defining traction. Traction to me DOES NOT MEAN no slipping. Since we know that the tire is constantly sliding sideways over the ground (during turning) to some extent, we know that "traction" doesn't mean grip like a train on rails. It can't. Traction of course involves some sliding. So the stock setup gives you lots of positive camber, you slide a bit, and then you reach the limit of the tire on the outside block of tread and you stop sliding in a controlled maner and you loose traction.

Add the swaybar, and while more load is being transfered to the outside wheel, the wheel stays flatter. You have more ultimate grip in that you slide sideways slightly through higher cornering force - to a point where the stock setup would have already lost traction, but you're still sliding in a controlled maner.

That means that with a greater load, you still sustained traction. The sliding is part of oversteer, which is not uncontrolled slipping out of the rear, but is instead allowing more sideways sliding of the tire. I think you might start sliding earlier, but you stay sliding much later.

Does that make sense or should I rephrase it when I'm less tired?

Joel

Dammit Joel,
You caught me being all theoretical again http://www.i-club.com/ubb-files/biggrin.gif. It should be noted that the above discussion was not including the "bad" effects of roll induced positive camber. Roll induced positive camber will decrease the maximum lateral load that the tire can generate. If the chassis is rigid, then the roll angle is the same for the front and rear axles, and the roll induced positive camber is the same...NOT! There is the little detail of suspension geometry, which means that the front and rear axles my have different roll induced positive camber when subjected the the same roll angle. So which axle has more positive camber given the same amount of roll? Come on...bring it Joel http://www.i-club.com/ubb-files/biggrin.gif

As for the statement:
"A swaybar is a fix for a too-soft suspension"
NO WAY!! http://www.i-club.com/ubb-files/biggrin.gif The use of a sway bar has two results. One is like you say, to increase roll stiffness without a harsh ride. The other is to tune the front and rear axle roll stiffness' so that the desired understeer/oversteer can be attained at the limit. Hence all the crap up there ^

Randy

Ahhhhhhh... yes, I got what you're saying. I forgot to take into consideration that there is grip, controlled sliding, and uncontrolled sliding. I think I proved it during autocross yesyerday when I had Revision in the car with me and went into a hard corner in a 4 wheel drift that was totally under control. Heh, they think I'm a maniac and yet I still come in 4th, go figure. http://www.i-club.com/ubb-files/smile.gif

Hello,

Maybe I should put it this way - with a thicker bar, you start sliding outwards, which is a part of normal traction, earlier becuase the camber hasn't changed yet (at low cornering forces) and there's more load transfer. So you get more sliding with the thicker bar. But, then, as camber change comes into effect, you quickly run out of traction in the stock setup while you do not in the 20mm setup. Make sense that way?

The spin-out effect of the 20 mm bar is a different matter and usually relates to hard transitions.

Joel

[edit] Damn, I follow up my own post and 2 more posts get inbetween.

Okay Randy, I'll take you on http://www.i-club.com/ubb-files/smile.gif I'm guessing that the rear has more of a camber change under cornering forces. Why? I dunno, I'm guessing. But the front comes with a 19mm bar from the factory to counter roll, so maybe without bars, the front rolls like a mo-fo - which makes sense with all that weight up front. So with the 19mm bar, they got rid of the roll in front and what's left is roll in the back. Well, they stuck a 13 mm bar in the back to reduce roll but couldn't go any further because that wouldn't be safe and the car wouldn't understeer like a pig. So, by going to a bigger rear bar, we "endanger" ourselves with oversteer (damn lawyers http://www.i-club.com/ubb-files/smile.gif) while flattening out the rear. The combo works great because the car is fairly flat this way without loosing any independant suspension travel http://www.i-club.com/ubb-files/smile.gif

But that's all just one big fat guess. Ask some of the stock auto-crossers who moved to the thinner front bar how much more roll they get - anyone still talk with Joe Longworth? He has the 16 mm front...

The other is to tune the front and rear axle roll stiffness' so that the desired understeer/oversteer can be attained at the limit.

Can't you do the same by changing the spring rates, Randy? Stiffer springs, more dampening, etc, and you can change under/over-steer...

JYD


[This message has been edited by Joel Gat, 1.8L (edited May 08, 2000).]

"The spin-out effect of the 20 mm bar is a different matter and usually relates to hard transitions." - Joel

Very true. However, I have found that it is MUCH more controlable with rear swaybar endlinks. Currently I am borrowing the MRT links from Greg and I should have a set of Greg's famous endlinks by Sunday.

And BTW, an inexperienced driver is the one that usually sets off that spin-out effect since they don't realize that they should turn the wheel quickly rather then yank the wheel.

Oh man... sure you edit before I get my post back up...

BTW, I was thinking about it while I was folding my laundry (yeah, I'm one of those think-while-doing-other-things people)... This is true with either sway bar -

With a thicker sway bar, you are causing the weight to transfer more laterally towards the outside wheel, causing less "grip". In the rear this equates to more (controlable) slide, in the front it equates to more slide as well, but just when you DON'T want it since you are actually trying to turn, not slide.

Joel,
It's smack down time! http://www.i-club.com/ubb-files/smile.gif
Did you read the above posts? You are probably like me, and just skimmed over them...or do you disagree with the load transfer effects on tractive forces?

How is the roll more in the front, than in the rear. The body is rigid enough to assume that the body roll is equal over the length of the body.

And you can achive the same roll stiffness and roll stiffness ratios with just springs, but your ride will suffer. But the point was that replacing the roll bar changes the balance of the front and rear roll stiffnesses.

Randy
"Damn this is getting long" http://www.i-club.com/ubb-files/smile.gif

"How is the roll more in the front, than in the rear. The body is rigid enough to assume that the body roll is equal over the length of the body." - rrsettgast

I don't know about you, dude, but when I go down a curb off a driveway kinda crooked, I hear all types of creaking in my roof telling me the chassis is flexing. Hence the original reason I wanted strut braces, though I think it is more of a roll bar issue at this point. I could just see trying to get out of a speeding ticket with a great big roll cage. "Honestly officer, it is in there just for normal everyday safety. You know how nuts some of those rush-hour drivers can be!" http://www.i-club.com/ubb-files/wink.gif

BEEEEEEEEE-ATCH!

"It's smack down time! Did you read the above posts? You are probably like me, and just skimmed over them."

Damn, caught me http://www.i-club.com/ubb-files/smile.gif

"How is the roll more in the front, than in the rear. The body is rigid enough to assume that the body roll is equal over the length of the body."

Well you asked me to guess which one was rolling more... http://www.i-club.com/ubb-files/smile.gif I guessed the front without the stock bars, the rear with the stock bars. And you're wrong. Sit on the front corner of your car. What happens? The car sinks over the front side that you sit on. The front "rolls" while the rear doesn't http://www.i-club.com/ubb-files/smile.gif But, yeah, again, that's not steady state turning, is it... Though I suppose in a descreasing radius turn, the front outside gets higher loads than the rear outside, so yeah, you can have differing amounts of roll http://www.i-club.com/ubb-files/smile.gif

"And you can achive the same roll stiffness and roll stiffness ratios with just springs, but your ride will suffer."

Agreed!

"But the point was that replacing the roll bar changes the balance of the front and rear roll stiffnesses."

Damn, caught me again. I didn't read the first thread http://www.i-club.com/ubb-files/smile.gif I agree 100% with that. In fact, my first post said I agree with everything except the overlooked fact that I mentioned above that y'all agreed with. Whee!

But it sure is fun to argue. Oh yeah! That's why I became a damn lawyer http://www.i-club.com/ubb-files/smile.gif

JYD

Wow this is a long and much misunderstood discussion.

Just an idea- it would be completely refreshing to have some skid pad #'s of stock vs. new rear bar setups.

You guys kill me! I have to say though I do love you guys. http://www.i-club.com/ubb-files/wink.gif

anyway Randy back to what I posted several hourse ago, you caught me being unclear again.

What I meant to say is that the sway bar increases compression rate on the compressed side of the suspension. If you agree with that then we are on the same page.

Joel,
You are a funny lawyer!
I think you should know we got into this discussion because there was a question about what sway bars do and how they work.

Now don't forget this boys and girls: Sway bars alter dynamic ride height at the compressed corners effectively changing the weight distribution at the opposite corner.

But uh oh, sprig do the same thing, and so do dampers.

Anyway don't forget the diagonal relationship, it is the key to understanding vehicle dynamics. That is why what you do at one end of the car affects the other end.

Pilot - Yeah, I was the one that posted that original post... Now I think I can say that I know how I can balance my car out and some pretty intricate vehicle/chassis dynamics. Other then that, I have no clue about the whacko equation that was posted in the original post in this thread. (damn people stealing whole sections out of books, LOL http://www.i-club.com/ubb-files/wink.gif )

EDIT : For clarity cause it is late.

[This message has been edited by 8Complex (edited May 08, 2000).]

Hey Joel,
Sitting on the front corner of your car produces uneven roll?....or roll and pitch? But yes, the flexability of the chassis will allow some difference in roll between the front and rear....so you are somewhat right. But the sitting example is a bad one because:
1) You have pitch as well as roll
2) You will dent your car http://www.i-club.com/ubb-files/smile.gif

Randy

I have no clue about the whacko equation that was posted in the original post in this thread.

Well, I can tell you this... Randy forgot to include a discussion of Dave's Point in those equations! (remember Dave Coleman in SCC named that point after himself?) Damnit, if Randy wants to throw all sorts of equations at us (I used to do that, but now I'm always too tired to actually think before posting), he should take into account Dave's Point and how the angle of a turned wheel about the point affects it's current camber and its resistence to changes in camber. But that gets wierd http://www.i-club.com/ubb-files/smile.gif and I have no idea how to calculate what's going on! I guess it could be figured out, but at some point, there's just too many variables - like the tire deal!

Dave didn't even get into that in his article http://www.i-club.com/ubb-files/smile.gif

Goofy discussion starting on Page 2 of this thread, eh? And what do you know, that's where I got into the discussion! But let me say this - all the super minor things end up cancelling out and the end effect is that a swaybar increases relative stiffness, increases load transfer, and reduces changes in camber. Too much bar and you overcome any benefit by transfering too much load to the outside wheel. Not enough bar and you roll the car which causes too much camber which reduces traction and by far reduces predictability.

JYD who can talk about anything. BTW, the skidpad numbers would be nice. Next time I hit a test track, I'll try to bring someone's 13mm bar...

[edit] And randy, okay, pitch-smitch. No matter where you put weight, if it's not along the side at the same front-to-back location of the CofG, then you'll always pitch the car forward or back. But you get the point. Try driving up those ramps that are designed to make it easy to work under your car. Drive up one of them, not both. Damn, dude, Subarus flex like butter. I have the ramps at home now for some tranny work...

But a dent is a valid concern. I still remember the big butt-shaped dent Joe Longworth got when demonstrating the stiffness of his new shocks and springs... http://www.i-club.com/ubb-files/smile.gif

[This message has been edited by Joel Gat, 1.8L (edited May 08, 2000).]

Just for fun,

Joel,
We have discovered here in this most educational of threads that a stiffer rear bar overloads the outside rear tire because it is taking load away from the inside rear tire at a greater rate than it is transferring weight to the outside rear.

So what happens is there is a net decrease in rear axle grip because the single tire with more weight on it doens't have as much grip as the two tires when they are more evenly loaded.

However the outside rear tire has more grip of its own due to the larger rear bar than without it.

So the better solution is stiffer springs because that increases weight transfer to the outside tire without stealing grip from the inside rear tire, while simultaneously transferring load back across the diagonal to more evenly load the two front tires increasing grip at the front.

Chew that one over for a bit. ;-)

EDIT: for clarity

[This message has been edited by Pilot (edited May 08, 2000).]

I am not trying to get the whole picture and the relationship to handling but the mechanics of how the bar mechanically transfers the load to the outside wheel. It seems intuitive to me that if we took our car and sat both rear wheels on scales we would get static weights for both rear corners.

Now if we take that huge guy and have him push the body over to the left, we would increase the weight on the left wheel while decreasing the weight on the right.

Switching to a larger rear bar would increase the weight transfer to the left wheel until we use a bar so stiff that we can lift the right tire. At that point the bar is stiff enough to overcome the right side spring and no weight is on the right side rear wheel. Therefore there is no longer anymore load that can be transfered from the right rear wheel because it is completely unloaded. At this point in time it would be preferential to increase the spring rate on the right rear corner because it would keep that wheel in contact with the ground longer and would increase the amount of load that can be transferred to the other wheels.

Does any of this make sense?

Tim

I think I'm starting to get out of the mindset that 'body roll = weight transfer'. It seems that body roll is actually anti-weight-transfer, while a sway bar makes it transfer even more. Heh, reminds me of watching all the VW's and CRX's take hard corners at autocrosses and lifting that rear inside wheel. http://www.i-club.com/ubb-files/smile.gif

Pilot,
I don't believe that decrease in rear traction is because load going to the outside rear tire is less than the load coming off the inside tire (although it is because some of that load is going to the front wheels)

Rather, according to last weeks tire discussion: a tires traction does not increase proprtionately with the load on that tire rather there are losses.

Lets say you increase the load to the outside tire by 50% the grip of that tire may only increase by 40% because of these losses. I think?

My understanding after this legnthy thread tied with last weeks legnthy thread is that.

1. more rubber on the ground (bigger contact patches) is good
A contact patch of 20sq in supporting 200lbs (10lbs/sq) in has more traction than a contact patch of 10sq in supporting the same 200lbs (20lbs/sq in) because of the interaction between rubber and asphalt.

3. Sway bars redistribute load.
By removing load from one wheel and distributing it to the other three (which always happens in some proportion dependant on left to right (roll) and front to rear (pitch) we can direct the load in such a way as to give a nuetral cornering attitude, but overall ultimate traction is decreased because we are not utilizing all the contact patches to their maximum.

I just had a curious thought. If there was a way to increase a contact patch as load transfers to that corner. (a wheel that telescopes in width and widens under load) We could have our cake and eat it too.

Am I correctly understanding what we are dealing with here?

Tim

I'm not real sure that sway bar "redistribute" load, I think it is more of a change in force placement.

See when there is no sway bar the car rolls more and the forces are absorbed up higher in the car. When there is a big sway bar, the forces are actually the same, but they aren't nearly as absorbed higher up, they are more distributed laterally to the outside wheel.

boxerman - What happened to point 2 in your last post? ROFL http://www.i-club.com/ubb-files/smile.gif

Boxerman,
Traction increases due to load increases are being simplified in this discussion. also here we are also making a clear distinction between vertical load and sideways force.

Two more evenly loaded (vertically) tires can take a greater sideways force than one heavily loaded tire and one lightly loaded tire.

Essentially we are increasing the force holding capacity of the outside tire by transferring more vertical load to it, however this increase is not as much as the loss of sideways force capacity on the inside tire. (in this case the proprtionality of the grip increase due to load doesn't matter, and if that increases is not linear in that more vertical begets less and less grip enhancemet then the problem is exacerbated)

And as we have discovered here there is a diagonal component as well that helps to more evenly load the two front tires so traction is enhanced at the front of the car because the two front tires are more evenly loaded.

8complex,
You are getting good at this eh? In most of the vehicle dynamics books it clearly states that you can not stop weight transfer. The forces are always the same and get transferred at different speeds. The speed of the transfer determines the handling characteristics of the car.

Heh, I think I learn more here discussing and listening then I would ever reading a book. Arguements and counterpoints are the best way to learn this stuff, especially when I can sit there and apply what I am hearing in my head and during my driving.

BTW, when you're talking about the speed of letting go between the friction of the tires on the road and the weight/downforce pushing them downward, I think about Formula 1 cars flying through corners and then just BOOM, they let go and fly off the course. Very little warning.

Guys,
this is very educational, but if we are taking about why our imprezas understeer, I personally think that body roll has to be controlled first (solid links and bushings), good grippy tires and most important of all, an adjustable central differencial alas 22B.
i belive that a 40/60 front/rear would cure corner exit understeer, just a thought!

Jose

GOOD LORD, if we keep this up we'll might surpass the clear corner/clear bumper lens thread record holder for the most posts!

I want to toss in my take on how a larger rear bar affects camber change, relative to the chassis & ground:
First off a recap that front and rear roll stiffness can be different. This is the resistance to roll, not the actual number of degrees the chassis rolls at each end. Rather, roll stiffness is how resistant an individual end of the cars' suspension settings (struts/springs/bar together) are to the rolling forces on the chassis due to cornering (front and rear considered individually for this post).
With that said, using a larger rear bar will help prevent the rear of the car from rolling over. More importantly, the larger rear bar will attempt to prevent movement of the rear suspension during a corner. This (attempted) prevention of suspension movement (compression/extention) will keep the camber, relative to the chassis, at a "near static" (non-cornering) setting. For instance, if the car has 1.5 degrees of negative rear camber (relative to the chassis) while the car is sitting still, the larger rear bar will attempt to keep the suspension from compressing during a corner and thus try to keep the camber (relative to the chassis) from changing as well.
Since the camber is being held at a constant as best it can be (relative to the chassis), it would be assumed that more rear grip is being attained at some point during a corner. Less positive camber change = more grip. TRUE, but not for long. We must not forget that the front of the car could (and does) have a completely different roll stiffness (this is NOT the same as roll angle differences, the chassis rolls as a whole for the most part). With the addition of a larger rear bar, the front end now has comparitively less roll stiffness, when compared to the rear end. For our purposes, Physics dictates that the end of the car with more weight will decide which way the entire car (front and rear, together now) will roll. Since our engines are decidedly in the front of the car (ehcsrop), the entire car will roll to which ever side the cornering-forces force the front end of the car to roll.
The grand finale...Since the front end is "running the show", it will roll to a certain number of degrees relative to the ground (not as much as stock, ie: less now with the larger rear bar, but still decidedly a large diflection) taking the rear end of the car with it (if the chassis is stiff, which it damn well better be). Camber comes into play now since the larger rear bar is preventing the rear suspension from compressing/extending and thus preventing the rear tires from gaining any negative camber (relative to the chassis) through "camber-gain" suspension geometry. However, the front end is not being very limited and thus allowing the front camber (relative to the chassis & ground) to change to more negative degrees. The net effect: Front camber remains steady relative to the ground and becomes increasingly negative relative to the chassis, while rear camber becomes increasingly positive relative to the ground while remaining steady relative to the chassis. More grip up front, less grip out back (no pun intended).

Disclaimer: 1) One of us may have already said all of this. I read through everyones posts (as I always do http://www.i-club.com/ubb-files/smile.gif), but I am sorry if this info had been overlooked on my behalf. 2) From numerous pictures of my car cornering at autoX's, I have deduced that our front suspensions gain a bit of negative camber through the compression of the strut. That may be WRONG, but it is the sole piece of information I have based this novel length post on. Please feel free, in the spirit of this land mark thread we have here, to prove my deduction wrong.
3) I would like to give all of us a big round of applause. This has got to be the most level headed thread, the most informative thread, and the most educational thread, in a long time and for a long time to come.

p.s.- I win, most parenthesis () in one post! Catch me if you can....

[This message has been edited by Greg Sharpe (edited May 09, 2000).]

[This message has been edited by Greg Sharpe (edited May 09, 2000).]

Geez Greg, 18 sts of parenthesis in one post. http://www.i-club.com/ubb-files/smile.gif

Your theory sounds about right but I'm not convinced that we gain negative camber on suspension compression up front. This had been posted in the last post, but noone caught on.

If you go to http://202.217.201.209/B4/4WDsports/sus4.swf , you'll see how the suspension flexes with the struts. It looks like it doens't gain negative camber until WAY near the end of it's travel. (Right click on it and hit 'Play' to play it again.

Mind you, that animation is for the Legacy B4, but I would think that they'd design the suspensions somewhat the same. From what I understand (what I've heard & read), if you don't gain camber when your suspension compresses, you gain toe (out? I think out...).

Now that I take a good look at it, I'm not sure if that is front suspension or rear... I don't see a tie-rod in there anywhere... Hmmmmm...

Mr. Gasket - Good catch, I never thought about looking at an Impreza in the air for camber information. You're absolutely correct with that one.

Pilot, from your post it seems as if we are in agreement although our terminology is differrent.
8complex, I also think that we are in agreement.

Going back to a car that lifts its inside rear tire. That corner has no load (verticle) it has been transferred to the other three corners.In steady state cornering the proprtion redistributed to the other three wheels has to do with front and rear roll stiffness. (Yes/No?)
It is possible to lift the inside tire without a stabilizer bar. In that case, you are reaching the limit of the inside wheels suspension travel at which time that load travels through the chassis(higher point) rather than the sway bar.

I am using an extreme exaple to keep it simple, as completely lifted wheel obviously has less verticle load (weight) than the three supporting the car. But any time the relative corner heights are changed (pitch and roll) the weight of the car is redistributed (duh! this is how we corner balance). Sway bars will cause this load transition to occur more quickly. (tire will lift sooner)

Tim

Boxerman we are definetly playing the same game by the same rules with the same strategy!

This thread is like the energizer bunny!

this means that the design was to keep the tires at the same camber relative to the body on any suspension travel sequence

Sorry, strut suspension does not control camber well at all, at least, not intentionlly. The whole reason almost all sports cars use multi-link suspensions (like the ass-end of the current legacies) is to keep the camber constant under body roll, extension, etc. The scooby uses struts because they're cheaper, lighter, and who cares about camber control as long as it's reasonable since you run -5 or more negative camber on the racing car anyways.

Joel

[This message has been edited by Joel Gat, 1.8L (edited May 09, 2000).]

"Going back to a car that lifts its inside rear tire. That corner has no load (verticle) it has been transferred to the other three corners.In steady state cornering the proprtion redistributed to the other three wheels has to do with front and rear roll stiffness. (Yes/No?)" - boxerman

Well, it DOES have to do with both the front and rear sway bars (and I think the springs too), but it just means that the front is too small compared to the rear. The rear, after all weight is transfered, the outside wheel is still stuck to the ground and compressed a LOT, now the inside wheel is being forced upwards by the too-large sway bar to the point that it lifts.

The front bar, in conjunction with the above idea, is small enough to cause enough body roll to occour up front (where most the weight is at that point in time), even if the rear end isn't capable, to roll beyond the point that the rear suspension can take (which lifts that inside rear wheel).

Personally... I think whole thing is hilarious and I love to see people go around the corners like that. http://www.i-club.com/ubb-files/biggrin.gif

So if the front and rear sway bars are too big, the car will topple over, where as the stock car will have lots of body roll, but won't tip in the same situation.. Is this correct?

8Complex, turning on 3 wheels kind of explains why the MY00s have been doing so well at the AutoXs. That LSD rear must really be working overtime on those turns! The MY98-99 have the open rear diff, I wonder which way the power would transfer in a fully loaded turn?

Revision - Yeah, the car will flip itself over with too large of sway bars on both sides, theoretically. The weight of the car (center of gravity) is higher then the roll center (the point at which the body/chassis rolls about) so the more and more you push, eventually the car will just snap up on two wheels and over.

Hey, I wonder if anyone caught me on 3 wheels while I was there... I heard that my front outside wheel was less then an inch from the ground while I was coming around that real tough corner 3 so all the weight must've been up there.

Yeah, it does exlplain why the MY00's are so much better around corners since no matter what, they still are pushing themselves ahead... the MY98-99's have that open rear diff and are probably spinning that rear inside wheel.

Just an aside, a rear LSD will cause more understeer, however there is an advantage in traction and the understeer can be dialed out by other means not the least of which can be the driver.

Pilot - Maybe that is the reason I can leave my bar at 22mm and still get a little understeer on some hard corners I'm trying to power through. I've found that it is best that I just speed into them and brake as I start turning and get some drifting going to get through the turn quicker.

8complex,
Probably a result of squat under acceleration as well as increased traction from the LSD. That is why you find it faster to trailbrake into the turna ndthen power out. Besides that is faster anyway, you carry more speed into the turn, therefore more speed through it, and get more speed on the exit cause you didn't have to slow down due to understeer.

AFTER you buy tires then I would try the MRT anti lift caster kit. I ahevn't used it myself but it shoudl help things quuite a bit at the front.

Pilot, yeah I was really eyeing that group buy going on, but I couldn't see myself starting in my first autocross in a Street Prepared class. I'm rather lucky I went without anyway since I managed to get 4th in STS... I do think the trailbraking helped a LOT. When I had Revision ride with me he was holding on for dear life as I was coming into those corners not realizing that I could take them like that all day long. I just had a discussion with him a little bit ago about the LSD and the differences in handling it presents. I wouldn't be half suprised if he says he's picking an LSD up for his MY99 in the near future, LOL. http://www.i-club.com/ubb-files/smile.gif

BTW, I wasn't sure what trailbraking was at first, now I know. Thanks. http://www.i-club.com/ubb-files/smile.gif

Have you ever seen a picture of an Impreza flying? Notice that all the tires are still upright, this means that the design was to keep the tires at the same camber relative to the body on any suspension travel sequence, I then think that when the body rolls some degree, then the tire is some degree positive camber relatively from the road, right? turning it into positive camber.
Then if I know my cars rolls 2 degrees, then I have to build 2 degrees of negative static camber.

Jose

So what is the best way to combat lifting the inside tire? Rear strut tower bar?

8complex,
It definately also has to do with spring stiffness also, as the combination of sway bars and springs contribute to your roll stiffness. As you said before you can either transfer the load through the swaybar or up through the springs and chassis.

Revision, I would think that a heavier front bar or stiffer front springs would keep that inside rear wheel down. Stiffening any of the three other corners would help to some extent. I think the opposite corner working in diagonals will have the greatest effect.

Tim

Let's see here... Coming into a corner all the weight is shifted towards the front outside tire. Since the CG (Center of Gravity) is higher then the RC (Roll Center), the car will tilt towards the outside of the turn, forward and outside.

Now with a stiff rear bar and lifting the inside rear tire off the ground, it says to me that the front end is rolling more then the back (chassis flex). I think if you add a rear sturt tower bar, you will lift that tire even more off the ground.

A front bar will cause the front end to roll less and the inside tire won't come off the ground, however it will cause more understeer.

Now, Revision's car has an STi suspension on it and is pretty stiff, yet it almost lifts that inside rear tire off the ground during hard cornering. He has his rear sway bar set to 20mm and a set of solid rear endlinks. He also runs a pretty decent set of Dunlop tires. Hiw car is also a MY99 with the open rear diff.

I think Revision is concerned about that inside rear wheel coming off the ground, or unweighting enough, to spin that inside tire around a corner since he has an open rear diff. and not an LSD.

I'm thinking that maybe a front strut bar might help combat the rear inside tire leaving the ground, because you cause the front suspension to work the way it should, however it may add a little more understeer at the same time (not sure... haven't thought about how the front suspension would be effected if it had more rigidity).

All FWD cars lift their rear inside tire because as originally designed they understeer, so you put a stiffer rear sway bar.Then you have to soften the front to kill that same understeer. These 2 concepts work but the trade off is THAT freaking lifting wheel you have to live with.
Anyway that tire will not do any work.
BUT, in our case we need that rear tire to help us brake, turn and accelerate. Look for The Traction Circle chapter in your Suspension books, that is the main idea.
My 99RS is pretty fast at Auto-x (I race against M3 and M5), and I have soften the front all the way (with Konis) and stiffen the rear (21mm)all the way too but it has never lifted the rear wheel.
The fastest way around a corner in my car is with trailbraking and scrubb some speed that way while you keep momentum, then straighten the car pointing at the apex and then once you're sure you'll hit it then gas slowly and then smash the throttle for the next turn.

Please do not try this at home...

Jose

Mr. Gasket - so you trailbrake in your MY99, huh? I'll have to remember that and watch how Donnie drives and maybe even get Revision to kick it up a level when he's driving.

What about that inside tire spinning while you're coming into the apex, though? Have you ever had it happen to you or does it somehow just not do it?

BTW, how you described cornering is almost exactly the way that I drive, hence the reason my RE92's are scrubbed sideways across the tread. http://www.i-club.com/ubb-files/smile.gif

8complex, I have not raced the original RE 92 so I do not know how bad traction is. I race Toyo Proxes T1-s and the only problem (other than understeer at corner exit)is that as I took off the ABS fuse, when I enter a corner too hot, the inside front tire looses some traction but all 4 tires are still on the ground.
I found also that the fastest way around a corner in a Subaru at Auto-x is by going slow...
But now that we are replying so often, tell me, what mods do you have and where do you want to be?

Jose

PS. do you drive the course in your head before a race? do you walk it? Can they let you draw a diagram so you can try different lines? Are your races organized by SCCA?

I was considering taking out the ABS fuse but the ABS actually reacts pretty well in this car, I've been in other cars where it highly overreacts.

So far all I've got done is an intake mod (removed all the silencer crap) and the rear sway bar (whiteline adj. set to 22mm). I've got solid rear endlinks on the way and MRT front links in my hands currently (haven't put them on yet because I'm not quite done with my review of the rears until I get Greg's rear links to test for a bit).

I drive the course in my head, but just as I am walking it. My memory isn't the best to keep the whole track in my head at once, but I react well and look ahead pretty well also. I could draw a diagram for different lines, but I only see a single line most of the time, during the last race I had to be told by another guy how to take a corner, which helped a little (apexed it short because it was a real tight corner - I should've seen that though). The race I went to was organized by TSSCC, kinda local but follows most SCCA rules as far as I can tell.

EDIT : I forgot to say where I'd like to be... I'd like to be rich enough to buy all the suspension components I need to tame this beast and make it stick like glue. Suspension (probably adjustable), braces all around and maybe a roll bar in the (far) future if I get THAT into it (at which time it will not be a daily driver anymore), and 'm not sure about power. I'm undecided between TC, SC, or N/A. I see pros and cons of each so I can't make a decision just yet.

[This message has been edited by 8Complex (edited May 10, 2000).]

I couldn't let this thread die.

8complex said:
"Let's see here... Coming into a corner all the weight is shifted towards the front outside tire."

True when trail braking. If you go through neutral the weight will shift to the outside but keep the FR/RR distribution constant, and if you are accelerating through the turn your weight is shifting to the outside rear tire.


"Since the CG (Center of Gravity) is higher then the RC (Roll Center), the car will tilt towards the outside of the turn, forward and outside."

Again, yes if trail braking.

"Now with a stiff rear bar and lifting the inside rear tire off the ground, it says to me that the front end is rolling more then the back (chassis flex). I think if you add a rear sturt tower bar, you will lift that tire even more off the ground."

When you say the front end is rolling more. Do you mean relative to the chassis or the ground? If yes relative to the ground then the chassis is flexing. If yes relative to the suspension but not the ground something else is happening.

If the chassis is not flexing the rear strut tower bar is not doing anything. The chassis is eficiently transfering the load and doesn't need any help. The strut tower bar will only eliminate the lifting that is caused by chassis deflection (twisting) I don't know if chassis deflection is the main culprit.

This brings me to a question are front and rear suspension travel neccesarily the same for a given car or may the front have more travel than the rear or vice versa? I would think so, and this difference would also cause a corner to lift at extreme lean angles.

One other thought, I have been told that there is very specific proceedure for mounting and setting up a sway bar, that if not done correctly there is a possiblity to preload one side. Adjustable links make it easier to set-up. It would make sense that if the bar is not left to right load neutral that would also contribute to lifting a specific tire.

"A front bar will cause the front end to roll less and the inside tire won't come off the ground, however it will cause more understeer"

So will stiffer springs or a lower roll center. (prevent front roll and lifting)

"Now, Revision's car has an STi suspension on it and is pretty stiff, yet it almost lifts that inside rear tire off the ground during hard cornering. He has his rear sway bar set to 20mm and a set of solid rear endlinks. He also runs a pretty decent set of Dunlop tires. Hiw car is also a MY99 with the open rear diff.

I think Revision is concerned about that inside rear wheel coming off the ground, or unweighting enough, to spin that inside tire around a corner since he has an open rear diff. and not an LSD.
I'm thinking that maybe a front strut bar might help combat the rear inside tire leaving the ground, because you cause the front suspension to work the way it should, however it may add a little more understeer at the same time (not sure... haven't thought about how the front suspension would be effected if it had more rigidity)."

This may work, but I don't know how much the front strut towers deflect during conering. Mike Shields seems to think it is insignificant, but I haven't a clue.

BTW I have heard of people using adjustable strut tower bars to increase camber. They tighten them up enough to pull the towers closer together. (heard this works on 911s)

Tim

Tim

boxerman -

"When you say the front end is rolling more. Do you mean relative to the chassis or the ground? If yes relative to the ground then the chassis is flexing. If yes relative to the suspension but not the ground something else is happening."

I mean relating between the front and rear suspension, though looking back, I'm not sure that statement I said made much sense to say in the first place... LOL, chassis flex is something I haven't learned to predict yet.

"So will stiffer springs or a lower roll center. (prevent front roll and lifting)"

Actually a higher roll center will prevent body roll, as will lowering the center of gravity. The roll center is already somewhere just over ground level and the center of gravity is just about 40% (guess) up the car towards the front, lowering the roll center will cause the car to roll even more.

"BTW I have heard of people using adjustable strut tower bars to increase camber. They tighten them up enough to pull the towers closer together. (heard this works on 911s)"

Interesting concept. As always we prove that you learn something new every day. http://www.i-club.com/ubb-files/smile.gif

OOPS.
I was confusing Roll Center with COG. But I think you may be confusing it with center of mass. I think COG only has a height component while center of mass has a front to rear component also.

We probrably should have established definitions earlier in the thread http://www.i-club.com/ubb-files/wink.gif

Tim

I think the CG has a f/r component... it is directly in proportion to the f/r weight distribution of the vehicle. Or would that be center of mass? I've never really heard of center of mass before... Something tells me we're really going to have to establish definitions even before we finish the thread. LOL, a part 3 of the thread?

Butthead mode ON: "uh uh This is the coolest thread ... I have ever seen!"

Where can we get "blueprints" of our suspension geometry?

Oh the insanity! I love this thread! http://www.i-club.com/ubb-files/smile.gif

Lifting the inside rear tire is only a problem if you are trying to accelerate while that tire is in the air, but then it isn't a problem cause we have AWD so the power gets transferred to the front which is more evenly loaded anyway. Also a lifted rear tire under trailbraking is not a problem cause it isn't doing anything anyway.

And when you hit the Apex and the accelerator the situation is reversed.

Basically AWD rules cause you can't run out of traction unless you get one wheel on each axle off the ground. If that happens you either high centered or flipped, so I wouldn't worry about that lifted rear tire unless you are teetering on a rock or upside down.

Greg - That's what I want to know... I'd love to get my hands on full prints for the Impreza's. Hell give it to me in a solid model and we'll REALLY be cookin. http://www.i-club.com/ubb-files/smile.gif

Pilot - If I were teetering on a rock or upside down, I think I'd just have to cry. http://www.i-club.com/ubb-files/biggrin.gif

BTW, even if you have a front and a rear up, in the MY00's you have LSD in the rear and could still pull. http://www.i-club.com/ubb-files/smile.gif

I think it's dead now.

Oops sorry up it went.

Yeah, I agree... I think it is almost dead.

Though I would like to know how structural rigidity (strut tower bars, rear lower tie bars, bushings) effect the effectiveness of the suspension.

For instance, we all have cars that flex under hard cornering currently, if we were to add the braces and such, would it handle flatter or enhance the under/oversteer? I've just recently added front endlinks on the stock bar (MRT, they still deflect but are better then stock) and am amazed how much better steering response I get. It takes away a lot of that 'delay' between turning the steering wheel and turning the car. I can slalom a lot better now, much more tightly. http://www.i-club.com/ubb-files/smile.gif

8complex,
As you noticed with the endlinks chassis and component flex will cause a delay in the suspensions response. Steel will store and release energy, but absorbs very little. Steel is almost perfectly elastic. Rubber and eurathane bits will absorb energy as well as some storage and release.

By increasing the chassis stiffness I would suspect that the fundamental balance is not changed, but the transitional response would be greatly enhanced since less energy is being stored and released.

Hmmmmmmm... ok, well I guess it'll have to come as I get the money... I guess tires and then maybe we'll see from there, I'm guessing that I'll probably go with some bracing after that (I can get Cusco strut bars for 10% cheaper then any place online) and then maybe suspension... No clue when/if I'll get some power mods in but I figure eventually I'll have to. Gotta figure out what is legal for STS and what is not. I already passed up on the anti-lift kit because of that.

There is a scoober around here that has some of the usual suspension mods, as well as a roll cage. He has mentioned that the car absolutely refuses to "get loose". From my view, outside & marshalling, his car seems to be very predictable/stable/smooth.