Rear Sway Bar vs. Front+Rear

[Dr Ronald]

I took my car out to my first AutoX event a couple weeks ago, and it seemed like I was having a terrible understeer problem. I know I was probably too aggressive into the turns, and I was on cheap all weathers, but still seemed to not handle like I would have liked. I just got my car, so I haven't done anything to it except for maintenance related stuff. Anyways, my question is this: Should I go with a larger front and rear sway bar, or just rear? I am looking to get coilovers, but I can't afford them right now. I'm hoping to get up to near 300-350 hp at some point down the road. I'm going to be competing in AutoX more next year, but what I would really like to get into is road racing. Any thoughts/brands/other mods? Thanks

p.s. My car is a 2g TSi

 

[rEclipserGST]

Just a guess, but probably because you are on stock suspension. Invest in some stuff, small things go a long way. Cheap stuff works too like strut bars.

Check my mod list. I've replaced some worn out sway bar end links as well. Helps a ton I think. I still need more or a tire wall that does not roll.

 

[ilikerice]

replace all your old dry rotted bushings for some Poly bushings.. will make a huge difference as far as stability.

For your understeering problem. I would start by getting a rear swaybar first. (adjustable if you got the money). Then see how it feels at the next event. Trial 'n Error dude. Do one mod at a time and see how you like it.

Keep in mind. Each driver is different. What works for one person may not work for someone else.

I hope to get my 96 GSX into autocross as well. Right now my Yaris is kicking butt. Surprised how this little car handles. .

Good luck at the events. Best way to learn how to drive your car is at autox

 

[eatsleepboostin]

If you have a TSi, your stock sway bar should be fine. Maybe new bushing all around. But personally I go with new set of aftermarket shocks and lowering springs. It will pay off. I have stock 98 GSX sway bars front and back but with Koni adjustable and Eibach sport springs and it handles like a go cart. You will be so much happier than just getting new sway bars.

 

[RvlutionMtrsprt]

You upgrade the rear bar first. After your done with all your other suspension modifications you can tackle the front bar if you're experiencing over steer. At that point, the right front bar will bring you back to neutral.

 

[jtmcinder]

I took my car out to my first AutoX event a couple weeks ago...

(bold added)

In that case, do nothing. Just learn to drive the car as it is. In the meantime, read the rules. If you have any desire to be competitive, then the last thing you want to do is make a mod that bumps you up a class before you are ready. If you absolutely can't stop yourself from making changes, get a set of Koni Sports. These are legal in all classes (i.e., they will never bump you up) and can later be used as the core for the best home-made coilover system there is.

Example to make my point: if you touch the rear bar, you're in Street Prepared or Street Touring at a minimum. That puts you up against cars with custom coilovers.

 

[tomlaw]

The oem front sway bar should suffice and you may not experience much of a difference with the rear sway bar, especially if you are AWD.

I would say get the coilovers (Megan Racing, Tein, etc) if you are more serious or just go for springs and shocks if you are on a budget. Strut bars are a very useful and cheap upgrade as well and should net you a noticeable difference upon installation...better turn in, steering response, a bit more stable, etc.

Also please remember that tires are one of the most important things you can upgrade on your car. They effect acceleration, braking, and handling simultaneously. A decent set of tires should Not be overlooked. In general, the stiffer your suspension is the more forces that will be transmitted to the tires via lower roll rates, smaller roll moment (distance between roll center (determined by suspension geometry) and center of gravity) with lowered suspension, etc.

The poly urethane bushings will help too as well as some regular suspension maintenance as mentioned by rEclipserGST.

Remember every driver is different...

 

[jtmcinder]

Actually, lowering the car reduces the amount of weight transferred from one tire to another, all else being equal...

plus, he said he wasn't AWD...

and strut-tower braces on an otherwise stock 2G is probably the least-wise mod I've ever seen suggested, since that bumps him up a class and will have almost no benefit at all.

But, otherwise, that was a very intelligent-sounding post.

 

[Dr Ronald]

At this point, me being bumped up a class isn't too big of a deal. I'm not racing competitively, just getting some practice and having some fun. I really want to get into the modified classes before I get competitive. I suppose I might go with poly bushings instead. Another thing that I would like to do once I get the money is a new set of wheels with R compound tires on them for the track. Thanks for the help.

 

[jtmcinder]

In that case, the keys to getting a 2G to turn are tires, front camber, and total roll stiffness (in that order). The way to think about it is this: get good tires, make them happy (camber), and then keep them happy (total roll stiffness). As it turns out, you don't really have to obey the standard rule about fighting understeer by stiffening only the rear, at least not for autocrossing; just do whatever you need to to increase the total roll stiffness and the fronts will be happier.

Another advantage of following the path above is if you change your mind and want to competitive in Stock, you can. Stock allows for Rs. When you've destroyed your streets running a few more events, put Rs on your stock wheels and buy new wheels and tires for the street. If you get the competitive bug once you have the Rs, just switch the path to obeying the Stock rules by getting Koni Sports next.

But, again, I would advise you to stay away from mods that don't much (if anything at all) and force you up a class.

 

[tomlaw]

Actually, lowering the car reduces the amount of weight transferred from one tire to another, all else being equal...

plus, he said he wasn't AWD...

and strut-tower braces on an otherwise stock 2G is probably the least-wise mod I've ever seen suggested, since that bumps him up a class and will have almost no benefit at all.

But, otherwise, that was a very intelligent-sounding post.

This is true but lowering the car will increase the amount of lateral g's transmitted to the tires since there is less roll rate to absorb centripetal forces (plus many other reasons), though the lowered ride height will generally better distribute the the forces between all of the tires. Also, generally the decrease in roll moment can negate some of these effects but still remain an issue.

Sorry I did not see the that he was FWD, my mistake.

Also true but I was suggesting the strut bars in conjunction with the coilovers or springs and shocks. In my personal experience, I went from stock suspension to front and rear strut bars and newer (slightly better) tires and that made a significant difference in steering response and stability. When I installed coilovers i experienced a world of difference in handling.

 

[jtmcinder]

This is true but lowering the car will increase the amount of lateral g's transmitted to the tires since there is less roll rate to absorb centripetal forces (plus many other reasons), though the lowered ride height will generally better distribute the the forces between all of the tires. Also, generally the decrease in roll moment can negate some of these effects but still remain an issue.

Nonsense.

Lowering the car decreases the lateral weight transfer. Period. End of story. It's a simple formula: cg height divided by track width times lateral acceleration. There are no other elements to this.

Now, it is true that when lowering the car you often also do other things that end up increasing the total amount of grip available (e.g., you gain camber), such that you end up taking the same corner faster, such that the third part of the formula goes up more than the first part goes down. But that's a separate issue, isn't it? Nothing changes how much a given lateral acceleration sets weight transfer besides cg height and track. And a lower cg means less weight transfer.

The only charitable reading I can give to this is that you are completely hung up on certain dynamic issues that don't apply to steady-state modeling. I have some sympathy for this view, in fact, since I find that almost all of the texts out there are all about steady state and don't always give the best advice to autocrossers. How much of the weight transfer do you want to be controlled by the shocks is one of these, which might be what you're talking about. But the idea that some of the weight transfer magically disappears (or is "absorbed" in some way) defies basic physics.

 

[tomlaw]

Nonsense.

Lowering the car decreases the lateral weight transfer. Period. End of story. It's a simple formula: cg height divided by track width times lateral acceleration. There are no other elements to this.

Now, it is true that when lowering the car you often also do other things that end up increasing the total amount of grip available (e.g., you gain camber), such that you end up taking the same corner faster, such that the third part of the formula goes up more than the first part goes down. But that's a separate issue, isn't it? Nothing changes how much a given lateral acceleration sets weight transfer besides cg height and track. And a lower cg means less weight transfer.

The only charitable reading I can give to this is that you are completely hung up on certain dynamic issues that don't apply to steady-state modeling. I have some sympathy for this view, in fact, since I find that almost all of the texts out there are all about steady state and don't always give the best advice to autocrossers. How much of the weight transfer do you want to be controlled by the shocks is one of these, which might be what you're talking about. But the idea that some of the weight transfer magically disappears (or is "absorbed" in some way) defies basic physics.

Again this is true BUT lateral force and weight transfer are two different things though they are related (one affects/causes the other). Turning the car will induce centripetal force mostly due to the car wanting to travel in the same direction and its resistance to changing direction (inertia) and other things but for simplicity's sake we will only consider inertial effects. I do not recall saying anything about "weight transfer" relative to suspension height settings by the way, but i could be wrong. So no the weight transfer does not disappear there is less of it due to the lower center of gravity RELATIVE to the roll moment of the suspension geometry and tire compound. The distance between the roll center and cg creates a torque arm which directly effects how much the car will roll upon turn in and throughout the turn. And for the record body roll accounts for a relatively small amount of weight transfer (I know that sounds counter intuitive but it's true) but it does contribute to overall handling and stability (some of the lateral turning force is absorbed through body roll). A lot of the actual weight transfer is due to the tires (in addition to many other things like track, camber, toe, etc) actually. Stickier race compound tires will transfer more wait than all seasons period. Believe me, lowering a car with crappy tires will result in lots of understeer/oversteer and unpredictable handling with surprisingly little weight transfer as the tires are slipping and cannot transfer the weight under higher loads (sharp/fast cornering). I have first hand experience with this in more than one car.

Side note: if the cg is too close to the roll moment this can cause snappy unpredictable handling

And yes most of the texts are about steady state conditions and usually do not go into much detail about dynamic handling but I can assure you I am not making this mistake. We have to make a lot of assumptions when discussing suspension setups so there will inevitably be some error...ie assuming constant spring rates when in reality the spring rate changes dynamically under compression and expansion when internal forces and structural geometry are taken into consideration.

By no means do i claim to be some sort of suspension guru of any sort as I still have a lot to learn...suspension tuning in reality is a bit of a black art (for instance the roll center can can have a theoretical location below the cg and outside of the vehicle). A bit of background info on myself: I am almost done with my Masters in Mechanical Engineering and was part of the suspension team for Formula SAE in which we designed and built, from scratch, the suspension for the University's first Formula SAE open wheel race car. I also designed coilovers for my senior design project during undergrad. I also have some experience in using Win GEO, Pro E, FEA, etc relative to suspension design and performance. This is in addition to my personal experiences with my cars. So my knowledge is Not purely based on wikipedia articles, suspension setup threads Racecar Dynamics books. I am NOT making any implications that this is the case with you I am simply making the case for myself.

My apologies for the long response in your suspension thread Dr Ronald

 

[Dr Ronald]

Hey no problem tomlaw. I enjoy a good debate. While we're on the topic, I'm going to have to disagree with a few things you guys said. I am also a Mechanical Engineer, so I know a thing or two about System Dynamics (although I admit it isn't my specialty). A car is lowered for one primary reason: to reduce the amount of weight transfer by lowering the cg of the car. Reducing the amount of weight transfer will only reduce the amount of body roll. One other way to reduce weight transfer (and consequently body roll) is to widen the stance of the car. This will not change the lateral force, but it will add grip. If you analyze the extreme case of the moment the tires are about to lose grip, the sum of the frictional forces at each tires must equal the centripetal force experienced by the car. The centripetal force is determined only by the tangential velocity of the car assuming the mass of the car and the radius of the turn are held constant. The frictional force felt at each tire will be determined by the weight distribution. A higher normal force at each tire will lead to a higher frictional force. However, this affect is asymptotic. Therefore, it is beneficial to have the weight as equally distributed as possible. This reduction in body roll also reduces the change in camber and reduces the amount of tire roll (also beneficial). This idea should not be associated with the stiffness of the springs, because spring stiffness has no bearing on weight transfer. Adding stiffer springs increases grip, however in a different way than changing the cg. This flow chart shows how these things increase grip. Obviously it is not complete. One other HUGE upgrade to add grip is better tires, which I have left out of this chart. Better tires are stiffer (causing less tire roll) and have more surface area touching the road. I'm not trying to say you guys are wrong, just want things to be clear for other people who read this thread. If anyone wants me to show them the physics of this, let me know.

Note: I have only taken into account strict side to side weight transfer. Things such as sway bar size differences in front and rear will be analyzed differently, however these concepts will still apply.


Edit: There is one mistake (that I can see) on this flow chart. The "Faster Turn Transition" should actually be coming off of the "Decreased Body Roll".

 

[snowborder714]

Here's some good reading on how to attack and prep a car for autocross.
Autocross to Win (DGs Autocross Secrets) - ATW Home Page

I would agree with jtoby in that you want to stay away from mods that bump you up a class and don't really give you anything in return. Sadly, I bought my car with a mod (small 16g turbo) that instantly bumps me up to Street Modified, so I'm running against other highly modified/prepped cars that I have a hard time keeping up with. Part of me wishes I could start over with the car, but as the car sits, I'm more interested in road racing it.

I also agree with not modifying a thing right now. Take the car out to a few events and see how you do just by modifying the driver. Try different techniques, have an experienced autoxer ride along with you and give you pointers, and then see what you really think would help the car.

 

[jtmcinder]

Wow. Now I have two things to wrap my head around. First, there's this idea that body roll absorbs the long-axis torque on the body due to cornering, preventing it from reaching the tires, and, second, there's the idea that you can both follow Dennis (who was the master of "fix the car, not the driver") and simultaneously not modify the car. On top of a hangover, this will be impossible.

 

[snowborder714]

Obviously you can't follow all of the advice at the same time. Or can you?

 

[Dr Ronald]

jtm, I don't know if you were talking to me or Tom, but I never said that body roll absorbs longitudinal torque. Also, I'm not saying that I think the car is the problem. I will be the first to admit I am a very inexperienced autocross driver. I'm just saying that I have some money to put into my car right now, so that's what I'm going to do. I honestly don't care what class I'm racing in, as long as I'm racing and getting better. Also, do you guys know if poly bushings will bump me up a class.

 

[jtmcinder]

Yes, bushings will bump you up. And, no, I was responding to Tom (which I'll do in more detail in a while). For now, even if you say that being competitive in your class isn't important, I would make very few mods. I say this partly because you might get the bug and want to be competitive later and partly because not messing with the car is the best way to learn to drive.

On the latter point, you can start throwing mods at the car to get it to turn or you can learn how to drive an understeering pig. The first approach will make you go faster, but only when driving this particular car (because only this car will have the mods). The second approach will make you faster in just about every production car for sale today (because they almost all are set up to understeer on the grounds that this is safer for the Great Unwashed). And the second approach costs you nothing but seat time. That's why I favor the second approach.

Now, if you have to spend money now, I come back to mods that will never have to be discarded later and I also am biased against those that move you up classes without helping very much. This is why I shake my head when people suggest strut-tower braces for an otherwise stock car. Not only do STBs do absolutely nothing for a car that has no grip and, therefore, doesn't generate the chassis forces that make an STB useful, but now you're in Street Prepared or Street Touring in a car that wouldn't be competitive in Stock.

So, if you have to spend money now, get a set of Koni Sports. This doesn't change your class, will increase your total grip quite a bit while not helping all that much with the understeer (so you will continue to learn how to drive an understeering pig), doesn't require any swapping of wheels at events, and allows you to play with the rebound adjustment so that you start to get a real feel for how suspension changes alter the car.

 

[9!'clipseDOHC]

jtm, I don't know if you were talking to me or Tom, but I never said that body roll absorbs longitudinal torque. Also, I'm not saying that I think the car is the problem. I will be the first to admit I am a very inexperienced autocross driver. I'm just saying that I have some money to put into my car right now, so that's what I'm going to do. I honestly don't care what class I'm racing in, as long as I'm racing and getting better. Also, do you guys know if poly bushings will bump me up a class.

Poly suspension bushings push you out of the stock class yes. They should put you in ST as long as they aren't metal. Check out this site to download all the rules: SCCA - Sports Car Club of America

 

[jtmcinder]

OK, now I'll try to respond to Tom. And please forgive me in advance if I make some simplifying assumptions and speak plainly - I have no idea how much you (or any lurkers) know and I really can't stand exchanges that neither participant really follows.

Simplifying assumptions: there is no unsprung weight (which is OK to assume if you don't talk about bumps in the road) and the roll center is between the CG and the ground (which is almost always true for double wishbone).

So why do we care about the location of the roll center? Well, as I said above, the amount of weight transfer that occurs is determined [completely] by three things: the CG height, the track width, and the cornering g. But that's the total weight transfer and it can be dividing into two pieces. One part of weight transfer is instantaneous and the other part is controlled by and operates through the suspension. How much of the total goes to each part is determined by the location of the roll center. In a way, you can think of the roll center's location as the dividing line for the weight transfer (because, if you care about details) all cornering force operates through the roll center.

Before I go on to talking about the two components, please stop and grok (i.e., deeply understand) the idea that all cornering force operates on the chassis and, therefore (given our assumption that the wheels and tires, etc, weight nothing), on the car via the roll center. I know that it seems like the tires are pushing towards the inside of the corner at ground level, but that's not where they push on the chassis. They push on the chassis at the roll center.

OK, as to partitioning the lateral forces into the two components, the height of the roll center above the ground sets the instantaneous weight transfer. The closer the roll center is to the CG percentage-wise (i.e., the higher the roll center), the more of the weight transfer is instantaneous. The remaining weight transfer, which is simply the complement to the height of the roll center, is controlled by and operates through the suspension. This is why people call the distance between the roll center and the CG the "roll moment arm." A "moment arm" is a length through which a torque operates; this particular moment arm produces body roll in a direct way, so we call it the "roll moment arm." Note that, by definition, if the roll center is way up at the height of the CG, then all weight transfer is instantaneous and none is controlled by the suspension. Likewise, if the roll center is down at ground level, there is no instantaneous weight transfer; in this case, it will all operate through the suspension.

Stepping back for a second, why do we care about partitioning weight transfer into the instantaneous and suspension-controlled parts? After all, in a steady-state corner, when the car has taken a "set" (as they say), it makes no real difference at all. Weight transfer is weight transfer and it's bad because you have the most total grip when weight is evenly distributed across the tires.

We care about the partitioning for two, connected reasons: first, some corners are very brief and, second, the part of weight transfer that operates through the suspension takes time to occur. Thus, if you have a lot of your weight transfer operating through the suspension and have a suspension that can delay these effects ... (if you trying to keep up, this means that the roll center is relatively low, since it's the distance from the roll center up to the CG that determines the amount of suspension-controlled weight transfer, right?) ... then you can actually delay the weight transfer long enough to have it not "kick in" until after the brief corner is done.

Short recap: weight transfer is bad, since it lowers total grip; while steady-state weight transfer depends only on CG height, track width, and lateral acceleration, not all of it happens immediately; and the amount of weight transfer that can be delayed is set by the length of the roll moment arm, which is the distance [downwards] from the CG to the roll center.

OK, fine, so some weight transfer is controlled by the suspension, but how does that delay it in some way? After all, force is energy and energy must go somewhere since it can't be created or destroyed, yadda, yadda, yadda. The place that this force goes (instead of directly into the tires) are the springs and sway bars. It takes energy [work, really, but let's not be anal] to compress a spring or twist a swaybar. So, during the time that the car is rolling into its "set" for the corner, the energy that could have been causing weight transfer to the outside tires [which is bad ... weight transfer is baaad, m'kay?] is actually being stored in (I hate to say "absorbed by") the springs and swaybars.

Now we finally get to the bloody point ... this idea that lowering the car can hurt handling because it stops the suspension from "absorbing" (as Tom put it) the forces that would cause weight transfer from inside tires to outside tires. The reason that lowering can hurt is because it can shorten the roll moment arm. In other words, the roll center sometimes moves downward less than the CG when you lower the car, so more of the weight transfer becomes instantaneous and less becomes controlled. This makes the car more responsive (because less of the effects of cornering are delayed), but some of that delaying you lost translates to less total grip during quick transitions because more weight transfer is happening (and weight transfer is baaad, m'kay?).

If I were to assign a homework at this point (which would make this post not only pompous, but seriously condescending, as well), then I'd ask you to use these idea to deduce why you want the roll centers to be at different heights at the two ends of the car. Fortunately for everyone, I won't bring this up.

- Jtoby

 

[Dr Ronald]

Spot on jtm. If I were being extremely picky, I would disagree with one thing. The dampers do actually "absorb" a very small amount of energy. Or dissipate if you want to be more precise.

Btw could you tell me your credentials when it come to suspension?

Thnx for the info.

 

[jtmcinder]

I completely agree about the shocks actually getting rid of some of the energy as heat. This can delay weight transfer even more. But I was sort of ignoring the shocks for now. I was also trying to deal only with the idea that shortening the roll moment arm was bad (on the grounds that having more control of more of the forces is better than having it all be instantaneous). I actually see long moment arms as good, but some other really bad things happen when the roll center gets too close to the ground or, worse, when it goes below ground level as it does on a lowered (but otherwise stock) Evo X. It's a funky trade-off. A low roll center gives you the option of delaying weight transfer and keeping more grip. A high roll center gives you much better "feel" (since it all happens fast), which is especially useful if you stink at turning in early. [I just did your homework for you, eh?, as to which roll center you want lower: front or rear.]

Most of all, however, because 2Gs have SALA (short-arm/long-arm) suspensions, the roll center moves down much faster than the CG as you lower the car. (Hence, over-lowered 2Gs handle like complete crud, even if sprung really hard.) Thus, the whole premise of Tom's warning was based on something that isn't actually true: as you lower a 2G, the roll moment gets longer (percentage-wise), not shorter. In other words, as you lower a 2G, more of the weight transfer occurs through the suspension and less is instantaneous.

"But wait," you cry. "That can't be true! My 2G became much more twitchy and responsive when I lowered it. You don't know anything!"

Did you lower it and keep the spring rates the same? Hopefully not. Hopefully you put on some real springs, not some POS OTS Sportlines. And maybe you changed the swaybars, too. So it's apples to oranges. You're comparing a stock 2G to a lowered 2G with five to ten times the total roll stiffness. And the gains from (a) lowering the CG and (b) upping the total roll stiffness to keep the tires happy far outweigh the shift towards more of the weight transfer occurring through the suspension. [And I hope you also now see why an otherwise stock 2G on Sportlines handles like crap: more weight transfer occurring through the suspension which has about stock levels of roll stiffness and now doesn't have enough travel to stay off the bumpstops.

As to credentials: I have none. I'm a geek, but not a formally trained one. I did make a complete 3D (mathematical) model of the 2G suspension, though. And I read and reread books until I actually understand them. But then I got tired of road-racing and autocrossing issues and have switched to rallycrossing as a way to make it fun again. That's why I haven't been around much.

- Jt

ps. I'm sure that I'm ignoring the influence of shocks wasn't the only over-simplification. I'm open to more detailed discussion, too. But we had to start somewhere, yes?

 

[goodbuds]

.. i think my eyes fell out an i fried my brain reading all this.

 

[2gGst98]

I dont no all of the technical terms and all but I do no from exp that a 2g that is lowered too much does handel like crud (my car now from previous owner). but at the same time if you do more than 1 event you will see your times get lower. I droped mine almost 4 seconds on a 35 sec or so trac (pro's were at 33 sec). And 6 secs on a track that was about a min and a half. So the driver part does help the more you know your car. But I have learned alot with this thread. Alot of it I didnt understand though. Google helps