Does Roll-Center height have much effect?

[stownsen914]

I've used a suspension analysis program called susprog to do a bunch of design work with for the front and rear of my 914 race car. You punch in coordinates for all the suspension pickup points, and susprog lets you model it visually and gives summary of various attributes like roll center, camber change, and lots of other stuff.

I've spent many hours playing with this on my 914, so I thought I'd chime in here. At this point in my project, the front of the car is carved up and I'm about 75% done building all the pieces for the new front suspesion. Lots of welding is what remains at this point. The rear suspension will have to wait until next year I think.

I used susprog to design everything "on paper" before cutting any metal. My goals for were to get more camber gain, reduce scrub (side to side movement of the tires as the suspension moves up and down), minimize bump steer, and get the roll center where I wanted it - about an inch above ground for the front. Why an inch above ground level? From the reading I've done, it seems like most purpose built suspensions have roll centers 1-4 inches above ground. With a few compromises, I've pretty much managed to get the parameters I wanted for front and rear (again, at least on the computer).

Regarding roll center, I analyzed my old front suspension before designing the new one for comparison sake. My old one was closer to stock, but with raised pickup points for the A arms, and the spindles raised about the standard 19mm as far as I can tell. Even with the raised pickup points, the A arms were angled with the strut ball joint siting higher than the A arm pickup points on the chassis. I've heard that this isn't ideal. The roll center on my old setup was lower than I thought - lower than ground level. The roll center rises significantly as I raised the ride height using susprog. I'm surprised that the front roll center would be 6" above ground for stock, but maybe this matches my observation that roll center rises with ride height.

Also, I saw some discussion about wanting the roll center higher in the rear. This matches what I heard as well, so I've aimed for the rear roll center to be about an inch higher in the rear.

Scott

[Flieger]

The issue is not so much the calculations but measuring the angles accurately.

[subhu]

is it necessary to maintain the roll center above the ground during bumps for an off road vehicle. and by doing so the camber during bumps become very high. and wht is the limiting value for camber during bumps

[subhu]

is it necessary to maintain the rollcenter above the ground during the bump (dynamic condition) also? by doing so, the camber angle increases to -17 degree. and i add one more question also. whether the camber angle can go (change) from +2 to -17 degrees for an off road vehicle of 7inch bump and 5 inch droop.

[winders]

Originally Posted by subhu

is it necessary to maintain the roll center above the ground during bumps for an off road vehicle. and by doing so the camber during bumps become very high. and wht is the limiting value for camber during bumps

Originally Posted by subhu

is it necessary to maintain the rollcenter above the ground during the bump (dynamic condition) also? by doing so, the camber angle increases to -17 degree. and i add one more question also. whether the camber angle can go (change) from +2 to -17 degrees for an off road vehicle of 7inch bump and 5 inch droop.

This isn't an off-road forum......I would look elsewhere for information on off-road suspension dynamics.

Scott

[Flieger]

I'd love to learn about and talk about offroad but you'll need to pose your questions in coherent sentences and give background information about them.

[winders]

Originally Posted by Flieger

I'd love to learn about and talk about offroad ....

Please, not here. This thread is useful as it sits. Don't clutter it up with technical details that in no way relate to Porsche cars on pavement.

Scott

[autodoctor911]

interesting thread. I know that a formula one suspension is about as applicable to a 911 as an off=road racer, but both are extreme examples of high roll centers, dictated by the parameters of the vehicle they are mounted on, particularly F1, where higher noses have raised the rollcenter above the COG, from what I can tell looking at one. I assume jacking is limited with shock rates, roll stiffness(spring or sway bar), etc.
what is the reason that prior to the limitations set by the aero advantage of a higher noses, most race cars were designed with a very low roll center(-2 to +1")?
to minimize weight transfer caused by loads being transmitted through the arms(jacking). this low roll center design is not practical on a strut car though, with a fairly fixed strut angle. In order to get enough camber gain to compensate for the body roll, the roll center must be raised, since we have no upper arm to move the pivot point downward. I think Porsche pretty much got the roll center right at the factory hieght for the tires they came with, but to maximise performance with a tire like a hoosier A6, more camber mid turn is needed, adjusting the roll center back to stock after lowering, helps considerably at keeping the camber close to where it needs to be, But I suggest doing whatever possible to increase camber gain to more than stock, to allow statick camber to be minimised, to keep the tires in good contact under braking. this works best if you can do it front and rear equally.
Another problem with the swingarm rear suspension is that the roll center moves a lot. see picture of triumph earlier in thread for extreme example. for this reason, it may not be a good idea to raise rear roll center too much, so more static negative camber in the rear is usually needed than in the front.
I am trying to lower a 986 boxster while maintaining or increasing camber gain. I know that the 996 GT3 RS has taller spindles for this. Does anyone know what other components are needed with the spindles?
i.e., when using the GT3 spindles, can stock 996 calipers, hubs and rotors be used without modification, or are the offset for hub and rotor, and caliper transferred to a different mounting for the calipers, requiring all the parts to need to be used together.
If anyone has tried to upgrade a 996 to gt3 brakes, and been able to use stock spindles, then the swap should work the other way as well.

[winders]

Originally Posted by autodoctor911

I assume jacking is limited with shock rates, roll stiffness(spring or sway bar), etc.

Jacking forces, by definition, bypass conventional dampers, springs, and anti-roll bars so they cannot distribute the force over a long time period.

Scott

[autodoctor911]

OK, so the jacking force is not negated, but it's effect of unwanted body/suspension movement is minimized by the other factors. In fact it is the other forces limiting the suspension movement that limit the usefulness of any extra downward force that results, making the overall net downforce due to gravity and mechanical loading to be negative, but the forces themselves are actually minimal, compared to the extra aero downforce afforded by having the mounting points higher. Otherwise, current F1 cars would have traditional geometry with a roll center near ground level.
I wonder if, with the upper and lower arms at a downward angle as now is the case with current F1 cars, do they actually make the lower arms a little shorter than the uppers to get some camber gain, or is there no body roll to speak of anyways, or maybe even negative roll forces, making camber gain in the positive direction(camber loss?) favorable.

[Flieger]

911s have rear semi-trailing arms, not swing axles.

[winders]

autodoctor911,

If you want to discuss F1 suspension systems, please start a thread in the "General Car/Truck Questions & Discussion " forum. Please don't pollute this thread with off-topic posts.

Scott

[Flieger]

Well, the last post before the offroad and F1 stuff was 08-15-2011, 08:08 PM by me. So I guess the 911 guys don't have anything more to discuss. I don't see the harm in it. Lets say it is a McLaren-Porsche or an 804 Porsche. Or a dune buggy with a Porsche engine competing in a race on a dirt track. Now it is Porsche Track Racing which fits this forum.

[autodoctor911]

Well the question was "Does Roll-Center height have much effect?".
I was trying to help demonstrate, with some examples, how a race car can have a roll center that is far from ideal, whether it be because of original design limitations as in the case of modifying an old 911, or rule limitations(any race class that requires stock suspension mounting points), or where the other aspects of a race vehicles design require geometry that is not Ideal(ground clearance, or aerodynamic considerations)

This is to try and ascertain whether it is favorable to go to great lengths to have the right roll center, or how to deal with a roll center that is less than, or far from Ideal.

[Flieger]

I think that the answer is that roll center height does not have as much effect as aerodynamics, weight, weight distribution, and tire performance (althoug tire performance is affected by rch). So go for low cg and work around the roll center if you can't raise spindles or move suspension pickup points.

How important is it to have a stable roll center? If you could have the roll center stay in one position relative to the car so that it only changed position in height and only due to pitch (dive/squat), would that be worth pursuing?

[winders]

Well, with production-based race cars, Porsche went to great lengths to adjust the static roll centers to be closer to what they thought was ideal whenever they could. The rear suspension of the 935 was altered significantly to raise the roll center.

The bottom line is that adjusting the roll center is worth pursuing. Just because may not have as much effect as aerodynamics, weight, weight distribution, and tire performance does not mean it is not important.

Scott

[Flieger]

Originally Posted by winders

Well, with production-based race cars, Porsche went to great lengths to adjust the static roll centers to be closer to what they thought was ideal whenever they could. The rear suspension of the 935 was altered significantly to raise the roll center.

The bottom line is that adjusting the roll center is worth pursuing. Just because may not have as much effect as aerodynamics, weight, weight distribution, and tire performance does not mean it is not important.

Scott

Yes, if you can do it within class rules and not mess up something that is more important like cg height or aerodynamics. With serious race cars the suspension moves so little anyway...

[burgermeister]

The theoretically ideal roll center for a car where no other limitations apply appears to be near ground level. It minimizes any ride height changes caused by jacking loads (high roll centers cause the car to rise during cornering) or rake changes (if the roll centers are different front / rear). And as a practical item, F1 cars would appear to have roll centers near ground level, and that's as high performance as I can think.

For a street driven car, chances are any RCH that doesn't make ridiculous compromises for streetability is low enough that it doesn't have any significant effect (as Flieger suggested). On the other side of the coin, roll centers at axle height are the domain (and bane) of swing axle suspensions, and they're terrrible for tire wear as well as rear end stability.

I don't know much about tuning race cars, but I imagine RCH could be used to tune around certain limitations, accepting the compromises that come with that approach (or perhaps undoing compromises made for the roadgoing version of the car). But I'd also imagine there are easier ways to tune around most things?

Off roading seems like it is more of a terrain (dirt) shearing exercise than a keeping tire patches parallel to the roar surface exercise, and terrain shearing is probably not so dependent on camber. And obviously 12" - 24" of suspension travel is is more desirable than perfect geometry when traversing huge bumps at speed.

[winders]

Originally Posted by Flieger

Yes, if you can do it within class rules and not mess up something that is more important like cg height or aerodynamics. With serious race cars the suspension moves so little anyway...

I find this argument silly. Chapman said that "tongue in cheek". Why? Because it obviously matters greatly to have proper suspension geometry even if the suspension does not move all that much. Otherwise, race teams in various series and formulas would not spend so much time and money on it.

Series with production-based cars clearly have more suspension movement than most formula series. Even so, I was watching the Formula 1 race in Canada and saw shot showing the rear suspension moving several inches.

Also, on production-based cars, adjusting the roll center usually is done because the car is lowered quite a bit altering the roll center so it well below ground. Getting the roll center back up to ground level or slightly above means the car will roll less for a given spring rate. Stiffer springs means less mechanical grip and mechanical grip is very important on production-based cars. Raising the roll center is not going to alter cg much and certainly is not going to mess with aerodynamics.

Let's not forget that suspension is a system and roll center is a part of that system. If it can be improved, it should be. Not doing so is leaving time on out on the track.

Scott

[Flieger]

What I meant by it interfering with aero was meant in regards to F1. They appear to have very high roll centers, but I believe that is only because they want the chassis high to get more air to the diffuser/floor. They use (or did use) those keels to get the lower pivots down without blocking too much air.

If you have very stiff spring rates anyway then sending the load through the suspension arms vs. through springs won't be such a big difference in terms of the load vs. time for the tire. There will be extra force from the jacking, though.