Does Roll-Center height have much effect?
 

How much effect dose the height of our "roll-center" have on body roll.

What happens if we lower the front a bunch and the rear stays high relative to the front?

I guess I am trying to quantify the effect some way in my mind.

That is, if one has there car set low in the front effectively putting the RC below the ground compared to Raising the spindle up on the strut say 35mm. Can these differences be felt in handling? Are the effects sufficient enough to require resetting the front rear sway bar settings or spring rates?

On the back of the 930's Porsche raised the inside arm attachment point 10MM. Seems that would raise the rears roll center. How much impact might this have?

What fs we instead lowered the inside rear arm attachment say 20mm to increase camber gain. Would this require stiffer springs to counter act the lowered roll center?

How big if an impact or issue is "roll center" height?

Roll center is the coupling point between the sprung and unsprung masses. Lateral force can be translated to the roll center with a corresponding moment about that roll center equal to the force times the distance between the center of mass and the roll center. A high roll center transfers lateral force quickly, transmitting that force mostly through the suspension members themselves, rather than through the springs. A low roll center has more of a rolling moment for the springs to react. A high roll center does cause "jacking" where the suspension raises up over the contact patch of the tire because of the higher application point of lateral force- like sliding a heavy box across a wood floor and catching the edge of the carpet.

Roll centers only matter inasmuch as the suspension moves. Many race cars are so stiff that the suspension does not move much anyway. Poor roll center location can be countered by corresponding sway bar settings.

Roll centers and anti-sway bars have similar effects as far as cornering dynamics go. Changing the roll center height or sway bar settings on the front or rear changes the roll stiffness. Changing the roll stiffness changes the distribution of total lateral load transfer. Stiffen the rear for quicker rotation.

This has ben discussused before,
the difference in roll center heigh and center of mass height is the lever arm that rollong forces work on. If the roll center and center of mass are at the same height there can be no roll.

on 911 when the car is lowered the static roll center goes down faster than the center of mass, which can lead to increased roll unless stiffer springs are addded.

the rear roll center is usullly held to be a bit higher than the front, connect the 2 and you have the roll axis.

roll center moves around a lot, depending on what the suspension is doing, you can only minimize negative effects, never eliminate them.

If you want to quantify the suspension dynamics that you're describing, my recommendation is to get the either or preferably both of the following books.

Competition Car Suspensions by Allan Staniforth or Race and Rally Car Source Book by the same author. In Competition Car Suspensions he walks you through the basis of how spring rates, role centers and all of the other suspension factors react. I was able to create an excel spreadsheet based on this book that does the math to answer the type of question that your describing.

http://ecx.images-amazon.com/images/...500_AA240_.jpg

The second book covers much of the same, but also delves into more detail in regards to spring rates, wheel rates and stuff like that.
https://images-na.ssl-images-amazon....7L._SL110_.jpg

I've read a number of other texts on the subjects and believe that these two are head-and-shoulders above the rest in regards to covering the subject. By the way, Staniforth has been a regular contributor to Racecar Engineering and/or Race Tech magazines over the years.

Thanks Bill Flieger and John. Input is very much appreciated.

I do think I would benifit from more reading. Wish my math was better

I feel I understand the lever between the roll center and center of gravity.

I feel I understand what lowering the COG does.

What widening the track dose...

I kind of understand that the orientation of suspension components may transfer forces es to create jacking or what ever the opposite is. Just not how significant this is.

However, If Porsche thought it worth the effort to move the spindle up 18mm on the RS to restore roll center, tilt the front A arm set up on the 930 to reduce dive, and raise the inside attachment point 10mm on the rear arm of a 930 to reduce squat-- these small differences must make contributions.

I have no feel for how important or significant such is and if they are worth working toward or if just doubling or tripling the wheel rates makes them mostly unimportant.

The best.

I'm sure that all of the above work great, the question is are you building a car for the premium pro level circuit use or are you building a prject where time and money is of no consequens or are you doing the best you can on a budget.

You do what you can to optimize what you have on you time and money budget. Lots of stuff that you do to make a better track car ruins it for the street. Youreally need to get you prioroites straigh before spending time and money.

You can have a great reasonably priced dual use car by just selecting reasonable compromises.

911s are well known in this regard, calculating to the nth decimal place just isn't going to happen because most of the knowledge base is empirical in nature. I suppose it's great fun to talk about it and it's even occasionally enlightening. But it's not going to be too helpful in componenet selection.

decide on how you want to use the car
decide on a budget & time frame that you can live w/
realize that it's not going to be a real RSR, just you own personal recreation that is consistant w/ you budget and use
if you do want a real RSR than that is a totally different topic.

Bill,

I just like to know how things work. I understand this dose not always translate to real world mod's. Still it is very interesting to me. I know I am not going to win any races, DE's, or TT's. More an intellectual pursuit that alows me to understand how things interact and work.

Thx, K

911st; I agree with Bill that much of the knowledge in the 911 community is empirical, but there he and I diverge (in a friendly way). I think that one of the characteristics of many of the car communities is that very often people do things because "so-and-so said that it's the right thing" or because "I saw it on so-and-so's car". If you were to ask that person why they did it, they will usually say that they learned it from expert-A, and expert-A learned it from B, and so on. The risk is that if you were to somehow trace it back to the original source, you quite likely may find that original-Alpha did it merely because he thought it looked cool, or because it seemed to work well with some other modification, when in fact that other modification somehow didn't get passed down across the generations. This philosophy makes me a little bit non-PC, especially since those people who blindly follow others when it comes to Porsches end up buying a lot of expensive parts.

Anyhow, blindly copying someone else's performance modification can often turn into a wild goose chase. An extreme example is that all of the F1 cars run some sort of barge-boards. F1 cars are fast, so if I want to be fast I should put some barge-boards on my 911. Everyone has heard that Ross Brawn and Sergio Rinland have both said that their barge boards make their cars faster. Well at the end of the day, I think that we'd all agree that barge boards would be silly on our 911's and not provide any performance improvement. They key point is that barge boards work well with flat bottomed formula cars with expressly limited aerodynamic modifications -- and not with Porsche 911s!

So getting back onto our Porsches, there is a very common belief that lowering our cars makes them faster. There are countless people on this BBS who insist on setting the car right down on the pavement just because the adjustments on the car allow them to. The reality that I believe you already appreciate is that just because a little is good, a lot isn't better. Lowering 911's past a point will mess up the relationship between the front and rear roll centers, reduce the suspension travel in bump and put the front suspension in an undesirable portion of it's travel in regards to camber gain. Take a car like that out onto a race track and it most likely will be quite a handful, and quite likely slower then a less modified car.

I'd go back to suggest making a spreadsheet after reading the books that I suggested. This will help you to understand how the different pieces go together as you construct the spreadsheet. If you'd like you can PM to me your email address and I'll share with you the copy of the spreadsheet that I made. It models the changes in tire loading of steady state turn, but does not tell you anything about the following aspects of the handling which are also key:

- Camber changes
- Swing-arm length
- scrub
- toe (aka: bump-steer in either front or back)
- Contact patch shape
- The nut behind the wheel

As designed and in spite of it's weight distribution, Porsches (and the 911 in specific) are designed to be relatively forgiving, so that they can be driven quickly, and safely for extended periods of time, such as in an endurance race. They can also be reoptimised into more of sprint configuration which makes them faster and less forgiving. If you do the modifications poorly you risk just making the car less forgiving.

John,

Thank you for the good words. I know I should be working mostly on tightening the nut behind the wheel.

I am currently really interested in getting the tires at the best angle to the track without making the suspension any stiffer than needed.

One example of "the rules" noted is "the camber between the front and rear should be 1/2 deg less in the front. Thus, others were telling me if I could only get one deg on in the front, the rear should thus be -1.5 deg.
Ok, my car still has stock (but track tuned) suspension and I lean about 4 deg and lose about 3+ deg of camber doing such. Why give up rear traction with a lesser neg camber setting just because the front will not allow needed range of adjustment?

That rule probably works for a car with mods to allow full range of camber adjustment front or rear and that has a very stiff suspension that allows very little body roll.

I suspect that rule comes from the idea that with about 6 deg of caster when the wheel is turned it adds about 1/2 deg of additional neg camber with wheel turn. I set both my ends up with the max camber possible where I could still get balanced and proper alighment. Now I just ware out the outside of the fronts tires and my rears are closer to ideal.

Further, I have seen many pictures of 911's at full tilt lifting one wheel in the front because the roll stiffness is less than ideal in the rear. That puts the outside front wheel and even more of a disadvantage relative to effective front camber. Thus, we might need even more neg camber in some cases up front than the rear if we are going to run a front stiff setup to achieve max traction on the inside rear tire (low tire stager).

Now what interests me is I read about the changes to the rear of RSR arms where they shortened the inner attachment for faster camber gain. With this the car was said become somewhat twitchy. However, Porsche did almost the same thing when they put the 930 to together. However, they raised the inner attachment point 10mm on the 930 which must be what made it viable for a street car. The expert say it was to provide anti-squat. However, I suspect that this did more. It probably flattened the first part of the camber curve (and toe curve) so the bad habits of the RSR set up could be tamed. It also may have changed the rear orientation to side loads so the rear would experience less downward jacking pressure from side loads. I have heard several people report that a Turbo Look with the same suspension as a 911 normal corners much flatter. Could this be small part of the reason?

ER has a cool custom spring plate arm that moves the outside link attachment forward (same idea as the 930 but made the outside arm longer instead of the inner shorter). They note it better matches the rear camber curve to the current tires. (I saw this added to a 930, would this not make the camber really agressive?)

So I am wondering, what if we just lowered the rear arm inner attachment point on a normal 911. This would not change the camber curve, but it might tilt the camber curve for more gain faster. This would then allow for the need of less static neg camber. Would not be that expensive to do this but it might help with rear braking (tires more square to ground with weight shift) and reduce inside rear tire ware of the car is street driven. ???

However, this might intoduce more roll or dowanward jacking forces or roall at the same time. Thus, the reasion for my starting this post.

Also, I found a camber curve here for the front and could see that raising the spindle puts the front back into a place of the camber curve where there is less camber loss. However, if one is running really stiff suspension it looks like we benefit less than .25 of a deg. On a stock car it might be almost a full deg of less camber loss with compression when cornering.

If my goal is to get my tires at the best angle to the track possible and keep the suspension as compliant as possible for driving to the track, raising the spindles 35mm (16" rims) is a mod worth looking at.

Knowing this stuff helps me make more informed decisions. Besides, for me it is just fun from an intellectual point of view.

We raised the pick up points on my 911, handles great but how much did we really gain?
We installed a bung on top of the torsion tube for a raised pick up point, similar to some 935s.
http://forums.pelicanparts.com/uploa...1247155339.jpg

Is there some anti squat gain with these adjustable boxes or are they just for bump steering the rear?
http://forums.pelicanparts.com/uploa...1247155556.jpg

Ted; If you lowered the car by roughly the same amount that you raised the suspension mounting points, then I believe what you really accomplished is to put the suspension geometry back where Porsche intended it, and lowered the center of gravity of the car. By doing these two things, you also reduced the roll couple at the rear which would make it act similar to having a stiffer sway bar.

John,

Your sig line from Colin Chapman was exactly what I was thinking.

Ted,

Interesting.

I am not an expert and have a lot to learn but what I am thinking that a significant amount of any dynamics like roll center placement and such are minimized with a full race suspension like yours.

As an example, a stock 86 and newer has about a 149 lb rear wheel rate supporting about 900 lbs per wheel at rest. Thus, will be using about 6 inches of spring travel to support its weight before any cornering load is imparted.

A race 911 might have a rear wheel rate of say 600 lbs supporting about 700 lbs . It will be using about 1 inch of spring travel at rest. On top of this the COG will be much lower imparting less twisting load on the car.

Thus, on a more stock car, suspension dynamics like camber curve might be more important and relevant.

On a full race car camber curve with only .5" of suspension travel is not going to have much effect.

I am guessing what is more important on a full race car like yours is having the range of adjustment to get the ideal camber setting at the tire, the right shock shock rate keeping the tire on the road, and most usable front to rear balance to put the tire hold where needed.

I suspect you might even have more susp give at the tire if running 40 psi than in the springs.

Still, it is a beauty to look at.

Just my guesses and lots to learn.

Sorry, yes I suspect you could get less anti squat from rasing the forward suspention pickups on a 911. However, if you lowered the car 4" and only rased the suspention 1" it might not be the main advantage. The adjustability gained from such a setup might be more important.

The primary effect of raising the inner pickup relative to the outer is to toe the loaded wheel inwards while cornering, giving a form of the rear-steer introduced in the more complex suspension geometry of the later cars (the toe curve then nearly matches a 993s over a limited range). This increases rear end stability. Ideally as you lower the car I believe you want to raise both rear pickups significantly if you can (with coilovers) to restore anti-squat. You'd move the inner pickup roughly 10mm further to achieve the desirable toe-curve.

In front I believe you raise the spindles as much as practical with your wheel diameter to raise front roll-center height back towards or above stock, reducing both roll and roll-center movement.

Tires vary widely; a V710 and an R6 have very different camber requirements, so actual setups will vary.

Ted remember the flatter you run a strut arm the less gain you get during travel, in other words bump wouldnt change as fast, Kevin

Nothing to add here except you guys are making me feel pretty dumb. :) Good summaries and discussion.

Just to put all of this into perspective, back to the original question:

Roll center height has a big impact on the drivability of the car. For example: perhaps the major change between the 997 GT3 and 997 GT2 suspension was roll center height. In going to the GT2 they lowered the front roll center and raised the rear; before these changes the GT2 was said to be nearly un-drivable in testing.

They have copied those changes on the 997 GT3 mk2 and the claim is that's a large part of the magic in that car... Sure, you've got some marketing mumbo-jumbo in there, but the point is that roll center is certainly a big enough variable to be worrying about. $.02

However, a sufficiently stiff anti-sway bar can cover up bad roll center placement and achieve the desired lateral load transfer rate. If suspension/chassis/wheel/tire geometry necessitate a certain suspension design with bad roll center placement, anti-sway bars can help.

Would not stiffer springs do the same thing as stiffer sways?

As John has in his sig line above:

"Any suspension -- no matter how poorly designed -- can be made to work reasonably well if you just stop it from moving." -- Colin Chapman

Very uninteresting info. Thx.

Now I better understand what Flieger was saying as to sway bars. It is not so much about making the suspension stiffer, the sway bars allow you to change the front to rear bias overloading the effect of a less than ideal roll center distribution.

I suspect roll center might be more of an issue on a softer suspension than on a very stiff race suspension. With stock wheel rates around 100-150 lbs side loading the suspension in a way that poor geometry adds to compression or restricts it will have more effect and will be more noticed.

Another example of poor suspension geometry would be lowering the front of a SC/Carrera to far. This can put the front in a virtually flat part of the camber curve. On a stock suspension, raising the spindle can put the front back to a part of the camber curve where there is almost 1 deg less of camber loss in a turn for a major improvement.

However, put full race coil overs on said car and the improvement in the camber curve is only about .25 deg as the range of motion is so much less under load.

Just trying to learn...