Front Roll Center Migration

[Whiskey11]

The bottom line is that for just about any evil you could care to mention in the suspension world, there IS a fix. Some easier than others, but the fix exists, none the less. If you REALLY want to obsess about RC location, camber curve, and CG, call Griggs or A47, they both have SLA kits... That will fix the evil strut problem once and for all.

HEY HEY HEY, Cortex has an SLA kit as well and I would argue quite heavily that the uprights he uses are superior to the Griggs/A47 kit.

Here are pictures from the Performance Autosport AI car:

And of course some YouTube love of their AI car going around VIR. About 30 seconds in the video of the front suspension shows up int he lower left:


To my knowledge these kits are custom made and you have to call to get them from Filip. The uprights are what he uses for the Fox/SN95 cars as well and they are pretty pimpy looking and have a solid track record from what I've read:

 

[Norm Peterson]

I would still like to experiment a little with how I constrain the chassis. As you suggested, track width is not really a constant in the real world, so it's not really correct that my model fixes the contact patches. However, if I took off that constraint, I would no longer have a single solution for the roll center. For a given chassis roll angle, the chassis would be able to travel on an arc, and each location would have its own solution for the kinematic roll center. Unfortunately, I don't know how to proceed from there. A reasonable approximation might come from comparing the wheel rate due to the springs to the wheel rate due to the ARB, which might be what you were implying earlier. I'm guessing it will lead to finding some sort of weird force-kinematic hybrid roll center with unknown usefulness.

Google "kinematics and compliance model".


Norm

 

[Norm Peterson]

The more interesting question is, perhaps, the change in camber due to the cornering forces involved combined with the soft bushings in the front suspension. This could, I expect, be modeled by allowing some variation in the LCA attach point locations as well as some variation in the LCA lengths.

The thing is, though, I'd expect those changes to yield second-order level effects on the resulting angles rather than first-order effects. But if you're at the point where a half inch of change makes a substantial difference in the instant center location, then those second-order effects start to become important.

From a structural engineering/finite element analysis point of view, you'd model bushings as elements between the control arms and the chassis with relatively soft material stiffness and let the LCA pivot axis move around.

I think if you're getting to the point where bushing compliance is that significant of a factor (which might very well be the case), the analysis really isn't worth it at the amateur level. It might not even be worth it at the semi-professional level. Testing and/or actual suspension measurements would be needed at that point. I think even developed software like Adams/Car would struggle to produce useful data with compliance issues like that.

It might be worth running a few rough compliance estimates if only as a sanity check on being able to neglect them. If you're getting 1/16", you can probably ignore the resulting camber change, if it's more like 1/4" (over 0.5° camber change), maybe not.


Norm

 

[ddd4114]

I understand that ddd4114 and kcbrown are very interested in the engineering side of handling. I actually quit talking to Sam Strano when he tried to tell me that trial braking wasn't useful to keep more weight on the front tires and improve grip at turn in.

However I continue to feel obligated to tell new people to the sport that what ddd4114 and kcbrown are discussing is at the highest level and may or may not have any appreciable difference on the track, much less the street.

Terry Fair was dismissed slightly above as being "old school racer" and not open to the science or engineering of designing a proper suspension. Well Jason at Vorshlag can talk numbers with any of you, the difference is that they know what matters and what doesn't.

ddd4114, you are a test engineer. Just curious, if a particular spec is +/- 0.10, do you worry about 0.001? I think that is what Terry is trying to tell the both of you.

For example, here are the points required by NASA Time Trial rules for various modifications:

UTQG treadwear rating of 120-200: +2
UTQG treadwear rating of 50 to 130: +7
UTQG treadwear rating of 40 or less: +10
Hoosier A6: +13

Here are the points for tire width increase equal to or greater than:

10mm +1, 20mm +4, 30mm +7, 40mm +10, 50mm +13, 60mm +16, 70mm +19, 80mm +22, 90mm +25, 100mm +28, 110mm +31

Here are the two modifications that seem related to this thread:

Bump steer kits or shimming of the steering rack +2
Alteration of ball joints/dive angles +2

Why are the points for tires so much higher than for improving front suspension geometry?

So, just curious, do you think that NASA Time Trial drivers will accept an uncontrollable car to get their one fast lap, especially since if they go four off or spin their time does not count for that session?

If I haven't made even a little bit of a point yet, here is some CMC undercar video from TWS:

http://vimeo.com/23297940

I have read all of the Carroll Smith "to win" books, and even back in 1978 he gave the tire designers credit for overcoming a lot of race car geometry issues.

Again, for the new guys, having your roll center underground doesn't mean that it will catch and flip the car like a driveshaft! Seriously, that is what this thread seems to imply!

I guess I didn't explicitly say it here, but I have no intention of getting a bump steer kit or extended ball joints. If the roll center moves all over the place, that's fine, but I want to know because it will affect how I add roll stiffness to the car. I'm already taking points for springs, so I want to get a good idea about what rates I'd need. The offerings for "race" springs are all over the place, and I wanted to see why. If I'm going to take another +2, I'd much rather use anti-roll bars instead of ball joints.

The reason I keep discussing this is because I think it's interesting. I've never done much with a MacStrut car with a live rear axle, so this is new to me. I understand that stuff like this isn't for everyone.

 

[ddd4114]

ddd4114, you are a test engineer. Just curious, if a particular spec is +/- 0.10, do you worry about 0.001?

Without getting too far off topic...

If the spec is +/- 0.10, I might worry about 0.01. It sounds ridiculous as a generic example, but it depends what it is. Sometimes you can measure a difference, sometimes you can't. Sometimes it matters, sometimes it doesn't.

I get your point, and probably the roll center is 0.01 to the +/- 0.1 roll stiffness spec.

 

[NoTicket]

HEY HEY HEY, Cortex has an SLA kit as well and I would argue quite heavily that the uprights he uses are superior to the Griggs/A47 kit.

Here are pictures from the Performance Autosport AI car:

And of course some YouTube love of their AI car going around VIR. About 30 seconds in the video of the front suspension shows up int he lower left:


To my knowledge these kits are custom made and you have to call to get them from Filip. The uprights are what he uses for the Fox/SN95 cars as well and they are pretty pimpy looking and have a solid track record from what I've read:

Putting an SLA kit on the front of an S197 is like filling your in ground pool while your house is burning.

 

[Whiskey11]

Putting an SLA kit on the front of an S197 is like filling your in ground pool while your house is burning.

Wut?

 

[NoTicket]

It will still have a stick axle in the rear.

Also you will have a nice pool but your house will still be cinders.

 

[Whiskey11]

It will still have a stick axle in the rear.

Also you will have a nice pool but your house will still be cinders.

So screw the front suspension since the rear suspension "sucks"? Right? Because I have yet to hear someone say "Man, my front suspension has SO MUCH GRIP!"

There are always ways to improve the performance of a car and even though the rear might be a limiting factor, you can tune around it pretty easily.

 

[NoTicket]

Oh come on. There are tons of amazing handling cars with strut suspensions. The struts are not the weak link in handling is my point.

 

[Whiskey11]

Oh come on. There are tons of amazing handling cars with strut suspensions. The struts are not the weak link in handling is my point.

There is no comparison between a factory tuned MacStrut modified to "race" and a race setup SLA. The SLA is going to keep your tires flatter to the pavement, period. Less negative camber necessary to keep the tires happy means better wear and more grip when it counts.

 

[NoTicket]

I'm sure that's true. What about a race tuned mcstrut vs an SLA? Those Porsche cup cars seem to do just fine with struts.

I am not arguing that an SLA is not more effective than a strut based setup. It clearly is better. It is the best suspension setup (if you add in pushrods it is the ultimate setup).

However, what I am arguing is that for far less money you can reach the limits of total car traction potential through the use of struts unless you go full retard and swap out the rear to an IRS multilink setup.

Once you are dropping the amount of money required for an SLA setup you would be a lot better off switching to a different car.

 

[Norm Peterson]

What about a race tuned mcstrut vs an SLA? Those Porsche cup cars seem to do just fine with struts.

I wonder how much suspension travel they expect to need for coping with body inertia loadings (roll and pitch) . . . vs suspension travel needs for coping with wheel displacements caused by road contour (bumps and such) where it's the displacement that's "fixed". I'm sure it's significantly less than what most people would choose for a dual-purpose street/open track car.

A "race tuned McStrut" still has McStrut geometry. Being stiffer and better damped doesn't fix the camber gain curve or the RC migration.

As an intermediate solution between the OE S197 and Corvette rear suspension arrangements and certainly consistent with a SLA conversion, I think a case could be made for better 3-link geometry (longer links, particularly the upper), along with a decambered axle.


Norm

 

[Whiskey11]

I wonder how much suspension travel they expect to need for coping with body inertia loadings (roll and pitch) . . . vs suspension travel needs for coping with wheel displacements caused by road contour (bumps and such) where it's the displacement that's "fixed". I'm sure it's significantly less than what most people would choose for a dual-purpose street/open track car.

A "race tuned McStrut" still has McStrut geometry. Being stiffer and better damped doesn't fix the camber gain curve or the RC migration.

As an intermediate solution between the OE S197 and Corvette rear suspension arrangements and certainly consistent with a SLA conversion, I think a case could be made for better 3-link geometry (longer links, particularly the upper), along with a decambered axle.


Norm

Well you know what Cortex does for their cars (Torque Arm + Watts link). Seems to solve a few problems, probably creates some too.

 

[Dubstep Shep]

Well you know what Cortex does for their cars (Torque Arm + Watts link). Seems to solve a few problems, probably creates some too.

I love my Griggs TA and watts setup. I can't think of any disadvantages besides some additional unsprung weight and the cost.

I have an article written by Bruce Griggs that's pretty good. Basically he favors straight axles on these cars over an IRS, and this article explains why. I'll see if I can copy paste it to here later today.

 

[Sky Render]

Speaking of Cortex's SLA conversion...

http://www.motoiq.com/MagazineArtic...e-Martis-KN-Engineering-Ford-Mustang-RTR.aspx

 

[Dubstep Shep]

Speaking of Cortex's SLA conversion...

http://www.motoiq.com/MagazineArtic...e-Martis-KN-Engineering-Ford-Mustang-RTR.aspx

That car is damn near what I want mine to end up like. I'm loving how well it's put together.

 

[ddd4114]

Since it's been a year, I figured that I would resurrect this thread and post an update about what I've learned since creating it...

Since I was a little worried about ride quality when making this thread, I first bought a set of Eibach Pro-Kit springs, Koni Yellow dampers, and a MM adjustable panhard bar. I kept the same Steeda camber plates that I had been using. My ride height was reduced by ~1" in the front and ~1.2" in the rear, and I was able to increase my static camber from -1.2 deg to -1.8 deg. On the street, I didn't notice any difference in body movement, and the car seemed to scrape on anything that wasn't perfectly flat. It was a pain in the ass to get out of my development because the entrance is on a slope. The ride was definitely a little stiffer, but it wasn't annoying at all. The biggest difference I noticed was that the Koni Yellows were WAY better than the OEM shocks over bumps. With the OEM shocks, the car would bounce all over the place, and the car was much more controllable with Koni Yellows. My wheel hop problem was also completely eliminated.

However, on the track, I didn't notice any improvement in lap times. I messed with the rebound settings a few times, but whatever I did, the car seemed somewhat unpredictable transiently, and my best lap times were within 1/2 second of what I ran the previous year (on multiple tracks). I also noticed that the OEM diff would break loose constantly on corner exit, and it didn't do that with the OEM suspension. Overall, I was a bit disappointed.

For my next local event, I moved from HPDE4 to TT, and I had to add 100 lb of ballast to remain in TTB. However, I also replaced the springs with MM Road and Track springs. I kept the same MM panhard bar and the same Koni Yellow shocks. The springs dropped the car an additional 0.2" in the front and an additional 0.6" in the back. They also brought my static camber out from -1.8 deg to -2.2 deg. On the street, the car was much easier to drive. The ride quality wasn't much different, but the car didn't scrape on everything anymore.

On the track, the springs made a HUGE difference in lap times even with the extra 100 lb of ballast. My lap times repeatably dropped by ~2.5 seconds at my local track, Mid-Ohio. While the car still noticeably understeers, it's not nearly as bad, and I can use the throttle to manipulate the car's balance a lot more easily. Here is a comparison between the springs (Black = Eibach, Red = MM):

While I suspect some of the improvement could be from better ultimate grip, I think a lot of it is driver confidence and car response. The biggest differences can be seen in turn 1 and the entrance to Madness. I definitely didn't gain that much grip in those turns, so the large improvement in cornering speed is simply a result of the car being easier to push without encountering drama. The same can be seen on the entrance to Thunder Valley. The difference at the end of the trace (Carousel) seems to be from taking a better line, so I doubt that had much to do with the springs.

Overall, this experience has shown me that controlling body movement and improving balance offers much more improvement than I would see with simply maintaining a higher front roll center. While I definitely still think that a higher front roll center would benefit the car, the compromise isn't worth the improvement in response and lower CG. If what Terry and others say about real coilovers is true, I imagine a real suspension setup will be worth another 1-2 seconds per lap.

 

[Norm Peterson]

Do you have any lap times at Mid-Ohio running the chicane just before the keyhole?

I don't recall seeing what tires and wheel sizes you're running - can you post those?


There's no way I'm ever going to keep up with you under acceleration with only 4.6L, but on a good lap I've run pretty close to your red trace through most of the turns.


Norm

 

[ddd4114]

Unfortunately, no, I haven't run the club course layout since I was in HPDE1. The black trace is a 1:42.2, and the red trace is a 1:39.2.

For both of those, I was using 275mm NT01's. My car weight was about 3730 for the black trace, and it was about 3830 for the red trace.