Hey Man...Now dig This...my cars roll center is off the chart man...it's like beyond the solar system...and, space being curved, it's cocurvolinear with itself...man...that really blows my mind... ....

But more seriously, there are some far out dudes on this other site, and one of them said something egregious, and it got me thinking about something I'd just read, by somebody who's nearly every word is decidely not egregious, and I thought I'd just mention it, and...

Ok, remember back, back, little further, no - too far, closer, there...yes, and there, yes, Bill Mitchell, Erik Zapletal, etc...and the force based moment center thing...and then Jay Novak wrote some stuff on the DSR board...and then I made up my mind to pretty much ignore such things in the future because it gave me a headache for not good enough reason as to be worth it...

And then years went by...and there was dust and cobwebs...on my car...and in my mind...and as I was making a particular part that had a particular effect on a particular parameter, I asked myself: What ever happened to Erik...and then one day he was in RCE again...and then again...and then a third time...how fortuitous... .....

And in that time...on a page upon which his words did dwell...he doth wrate: "The point we are trying to make is that roll centres (sic) frequently zoom off into space , often out past Pluto, but that is nothing to worry about....The problems only come if we assume that roll centre position is an important indicator of performance, which it clearly is not."

Heavy man? You bet it is. Of course you want to offer up the wonderful swinging axle as your first retort. But you don't, because you know that a solid axle with the same roll centre (Hehe) height doesn't behave the same way.

So what's the answer? I don't know. I don't even know what the question is. There is a roll center somewhere. You can purposely put it somewhere, or it can wind up wherever it is as a result of other decisions you make. There's probably a region of space where the whole works is reasonably more or less comfortable. It can be in there.

So about n-lines, and jacking forces, and such...I don't have much to say. But I'd note that one of the sharpest FSAE minds in the world told me that if the car is stiff enough then jacking forces result in so little change in ride height as to be insignificant. As is the case with so many things when you're stiff enough.

Scott, who hopes you can all keep the rubber side down, because the most important thing to keep your roll centres from doing is iinverting themselves...

 

Scott, you know you're more than welcome to post over there. The tech doesn't show up that often, but it's usually good when it does. Personally I've decided that roll centers are not worth worrying about, but I reserve the right to change my mind.

 

Im all ears and eyes too... lay it on me!
I like 60's jargon, and shakespeare too.
Convince me

 

I love this thank you

 

...so...good things happen when you're stiff enough? tee hee

 

I can't do that. I don't have the mental horsepower. But if I could defer to an authority figure ("That's a cop out man!"), here's something that Jay Novak wrote on the DSR board some time back ( http://dsrforum.yuku.com/topic/2802?page=6 ):

------>

I have a few comments on this issue.

1. The term roll center is a misnomer in that the vehicle does not necessarily roll about a geometric position that is defined by the geometry of the suspension.

2. The commonly accepted geometric roll center as defined by the control arm geometry is a very reasonable way to estimate the weight transfer characteristics of the suspension in question. Is it extremely accurate? I think that it is very reasonable approximation. I have experience, through extensive laboratory forced based measurements and additional track testing, that really indicated that it is a very good approximation of weight transfer throught the suspension points. Forced based measurements that I have conducted have always indicated that the "force based roll center" is ALWAYS higher than the "geometric based roll center"

3. I think that we all understand that higher roll centers transfer more weight through the geometry than do lower roll centers. This is certainly true & what is equally important are the actual & real jacking forces associated with these higher roll centers. The thing to remember is that once the roll center is below a reasonable level, that is best related to track width & COG height, then jacking forces become negligible & other factors are far more important

4. Does assymetrical suspension geometry such as used in oval track race cars affect the weight transfer? ABSOLUTELY it does. I can guarantee you that the transfer forces do change due to assymetry. Then remember that EVERY car has assymetrical suspension once it hits the ground & starts moveing around a race track.

5. Does "geometric lateral roll center migration" affect weight transfer? Yes it can, but ONLY when the roll center is fairly high above the ground in the 1st place. Think of it this way. "Geometric roll center migration" can be huge when the geometric roll center is very glose to ground level. This is simply due to the fact that the geometric lines are very close to parallel to the ground & the intersections move all over the place. This is esentially meaningless.

6. Here are my rules:
a. define your tires. This means that you need to understand what are the real optimal camber / lateral load / slip angle characteristics are. This does not mean that you need a full blown tire model. Just find out what the most commonly used tires are run at & get some idea of what the real world needs & works. Get the tire data if it is available then try to understand it.
b. define what your optimal track width needs to be. Sounds simple but is much more complex than 1st looks might indicate.
c. define your static camber and approx. camber gain
d. define where you want to have your roll center height. Limit the lateral migration if possible (& it usually is possible).
e. think about all the other variables all the time. Pay attention to caster, mechanical trail, scrub radius, anti-dive/squat etc.
f. Do not go crazy obsessing about the details. Make sure that you have enough adjustment built into the design then build it & go testing, then test some more. Do not be afraid to make changes. PAY ATTENTION TO TIRE TEMPS every single session on the track, record everything you can think of.

enough BS. Build it & go racing.

Thanks ... Jay Novak

*****

And also this one ( http://dsrforum.yuku.com/reply/19557#reply-19557 ):

---->

RC geometric modeling software can give what appears to be large variations in the lateral position of the RC. This is generally caused by having low roll centers that can cross the ground plane. When the RC is very close to the ground you may note large differences in RC lateral position with very slight roll or vertical height change. This is due to the fact that the software calculates the RC by defining the intersection of multiple lines, some of which can be very close to the ground plane. This can cause large lateral position changes in the calculated geometric RC lateral position.

This situation is NOT relavant to the real world of force based weight transfer. The RC calculated from geometric modeling is an APPROXIMATION of what is happening with the force based weight transfer of the suspension geometry.

I have conducted numerous tests in the Ford Kinematic & Compliance Laboratory on race cars in an effort to better understand (among other things) the weight transfer characteristics of race cars. There are cases where there can be significant non linear increases of weight transfer as a fuction of lateral load. This can be caused by high roll centers and or assymetrical geometry changes caused by the forces (lateral migrations of the RC are usually a result of this)

Here are some points of interest:
1. The forced based RC is ALWAYS slightly higher than the geometrically calculated RC.
2. The RC always moves higher with increasing lateral force, even if the RC is geometrically below ground statically.
3. Cars can have significantly different weight transfer as a function of the direction of the lateral force. This is usually caused by assymetrical suspension geometry.
4. I have NEVER measured a car where the force based RC was below ground, even with cars that had geometrically calculated RC that were below ground.
5. High roll centers combined with inappropriate geometry can cause non-linear weight transfer. This is usually also associated with significant lateral migration of the geometric RC.

I do not want everyone to think that these points I raise are some sort of gospel. This info is the result of much testing & discussion by engineers from Ford & many NASCAR, CART & F1 engineers.

Here is a question for those of you who like to think about this subject.

Where is the roll center of a car (either front or rear) when there has been close to or up to 100% of weight transfered to the loaded tire?

Thanks ... Jay Novak

*******

Scott, who recommends searching that site for some extremely valuable giveaways....

 

At the 100% loaded tire.


I agree with Jay whole heartedly. And everything I've ever experienced shows me that within a certain ruleset - getting the CG as low as possible yields much high benefits than having a higher geometric RC.

 

+1. I read that stuff a while back on Scott's recommendation and that's what sealed the deal for me on worrying about roll centers. I've got more important things to worry about, like trying to figure out what the car needed for the surface at LAP. I missed the part about the force based RC always being above ground. Kinda makes sense for 30,000' view. All the forces acting on the car are above ground.

 

Ok Scott.. Im gonna hire you as my crewchief. Get your behind out here to california, and lets get cracking.
I am more than sure you have the mental horsepower!
Great Job.
Those sites reminded me of a book called "The Physics of Baseball"
In had to read it 4 times before the lights went on.


Paul... who is now convinced Scott is a supergenius will try an coax the secret of cold fusion from him.

 

Ha ha...the guy I'm currently crew chief for is having an extraordinarily hard time getting his kaa down off the jackstands. He is constantly bugging me about it, but it's not entirely my fault. Our teams project engineer is slowing down the works. And our machinist is lazy too. And our driver isn't hungry enough. And our Captain Money Bags is leaving a bit to be desired as well. And our team principal wastes too much time on the internet and reading things like Colonel Jozef Beck's Final Report.

Scott, who thinks we're all lucky to be living in this era of plenty info wise...I'm super indebted to the generous amongst us...

 

I worry about RC hieght and migration. A dose of sugar isn't bad for any of us from time to time, but as a steady diet it can ruin us. So, how far out of whack do roll centers have to be before we begin to chase problems associated with poor RC location (s)?

Depends...tire compliance affects RC migration, spring and dampers affect RC migration, bushing compliance affects RC migration, body flex affects RC migration, and, static RC location - geometry - affects RC migration. Which ones are more important? Will a 10% increase in spring rate tether RC to Saturn instead of Pluto for the duration of a given rip through a give track?

I worry about RC location but I also know that most of us drive fairly pedestrain fare that may benefit from unorthodox tuning techniques that seem to run over RC location.

I guess the outcome matters more...how fast one can go...but will you go slower if the RC bursts past Pluto? Dunno...current F1 cars place RC locations in a funny place...but aero is most important in F1...do we really care about Pluto?

But, do we care if RC drops below grade or do we care if it migrates so high it casues jacking???

EDIT thought I would throw this out there...why have roll centers at all? Get rid of them...but we don't get to blast past Pluto...

...and re #5...higher RC will load tires faster than a lower RC location, that is true and very effective in transient. However, a lower RC postion will, in steady state, transfer more load on to the tire. Steady state suffers with a lower RC - these are relative assumptions based upon the length of the virtual lever arm between CofG and RC...a longer arm has more leverage but it takes longer to transfer weight. A short arm will transfer weight faster but has less leverage so the resultant weight transfer in steady state is less. Jacking occurs of the RC migrates above CofG...unless we have lots of aero.

Re where's the RC on a 100% loaded front tire? RC moves in the opposite direction of CofG...so away from a loaded tire.