Roll Center

[Mike131313]

Alright I did as you all typically ask and I checked the archives on this topic and didn't find the answer to my question. So, here it is, I am just getting back into suspension design after about a 1 year hiatus and seem to have forgotten everything I know. I was reviewing a relatively simple book just to get my bearings straight and I run across something that stumped me. In Paul Van Valkenburgh's book "Racecar Engineering and Mechanics" he made a comment about the roll center. Now I am very familiar with the standard methed of the intersecting lines of the two a arms and the contact patch to locate the roll center, which is the method he describes. He then goes on to state that as the "radius approaches infinty" thus becoming parallel a arms the "the camber gain in bump goes to zero" (assuming equal length a-arms).... "since the chassis rolls to the outside of a turn the wheels camber in the wrong direction." So this to makes sense too what I get hung up on is what he says after this which is, "this can be eliminated by reducing roll, through moving the roll center closer to the center of gravity...." Now it is my understanding that when your a-arms are parallel you locate the roll center at the ground, this makes sense too. but if the a-arm radii is to approach infinity then how can you manipulate the roll center at all. Is there a new method of locating the roll center if your a-arms are parallel that I am unaware of? Or did the reader [me] just assume to much in his statement, for instance the author did say "approaches" and not "equals"? Any info would be appreciated and sorry this is such a basic question.

[Denny Trimble]

Pick a point where you want your roll center to be.

Draw a line from one tire contact patch through this roll center.

Make both your a-arms parallel to this line.

Voila! I think...

[Brian Smith]

nope.. if your arms are parallel then your rollcenter will be at the ground no matter what angle they are at. Think about it, your instant center is impossible to find so the rc would be too. I think Paul meant for you to put your arms near parallel and put your rollcenter near cg heigth.

does anyone do this in fsae or know of any other cars that try this? This would give you a heck of a lot of jacking and dosent seem worth it.....

[Denny Trimble]

OK, sorry. Try this:
Build in .001" error in each pickup point so your instant centers can be located 1mile east of the statue of Liberty. Your a-arms won't be parallel, they'll be a whopping .000000001 degree apart.

These manufacturing errors are inevitible; but my little theory should work because of them.

What makes you think the RC defaults to ground if you can't figure its location out?

Does this pass your inspection Brian? http://fsae.com/groupee_common/emoticons/icon_smile.gif Or am I missing something else?

Back to making chips...

[Brian Smith]

The line connecting the contact patch to the IC would be parallel to the arms if the arms were parallel to eachother. I was thinking that the IC would be infinately far away and this puts rc at ground but the ic would also be infinately high, but at a slope equal to the arms. Its arms parallel to eachother and the ground that have rc at the ground.

[Brian Smith]

....however in Competetion Car Suspension, p.67 "Wishbones, equal and parallel.The rollcenter is on the ground," but I assume he means wishbones parallel to the ground. Also there is a picture 0n p.61 of a front view suspension with equal and parallel arms (parallel to ground). It has rc at ground.

[fade]

i doesnt seem like he saying to move the roll center while keeping the arms equal length and parallel. to me it seems as though hes saying if you have no camber gain in the suspension then to keep the tires from reaching positive camber minimize the body roll by reducing the roll moment arm.

[Tim Heinemann]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Brian Smith:
nope.. if your arms are parallel then your rollcenter will be at the ground no matter what angle they are at.
<HR></BLOCKQUOTE>

That's wrong: The roll center is the point of intersection of the two lines from the contact patches through the according instantaneous centers which are the points of intersection of the lines through the according a-arm joints (and it's just a "common outlier" that the roll center is on the middle of the car in case of symmetrical suspension!). Thus the roll center will - regarding parallel wishbones- only be at the ground for the special case of the wishbones also beeing parallel to the ground (however of cours in this case the lines don't have a single but infinite points of intersection so it's just a definition derived from the symmetry of the car that RC is in the middle ones again. For asymmetric cars such as oval racers which you Americans mysteriously like this definition simply wouldn't work).

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>

Think about it, your instant center is impossible to find so the rc would be too.

<HR></BLOCKQUOTE>

Nope: If two lines are parallel they will intersect in the infinite hence the line through the contact patch and the instantaneous center will simply be parallel to the A-arms and IC will therefore also be in the infinite.



Tim

[Brian Smith]

thats exactly what i said in my second post.

[Tim Heinemann]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Brian Smith:
thats exactly what i said in my second post. <HR></BLOCKQUOTE>

Jip, I read it just after posting mine.



Tim

[1975BMW2002]

How about this simple rule. If the arms are parallel, or even .000000001 degrees from parallel, you can assume that the roll center will lie on a line that is parallel to your arms coming from teh contact patch. so if thye aim up at 10 degrees going to the center of the car, draw a line that is at 10 degrees from horizontal coming out of yoru contact patch.

coincidentally, the roll center is not always inthe center of the car, even with a car that has symmetrical suspension. It is where the lines from both wheels intersect. At rest in a symmetrical suspension car it is at centerline. However, once you take a turn, and the arms on one side are pointing a little above rest, and the other side is a little below rest, the roll center moves.

This is all stuff that I have reasoned from reading VanValkenburg, Smith, and some SAE book that we use for our vehicle dynamics class.

If any of this is wrong, I'm screwed.

Does anyone know how to do the force calculation method? I'm kind of interested in it. I've considered deriving one, but it would be easier if I could just do it someone else's way. The only way I've dreamed up is very iterative. I hate matlab.

just my '02 cents worth.

Bill

[Jeff Curtis]

First you guys need to understand what a roll center is. All you need to know is that the each tire or corner of the car develops a lateral force. This lateral force is transmitted into the chassis by what? The linkage that connects that tire to the chassis. The percentage of this lateral force that exerts itself through rigid linkage (a-frames) is what forms non-rolling weight transfer. The rest of course is transmitted though springs and bars making up roll-induced transfer.

Nevertheless, a roll center or better yet instantaneous centers are just imaginary points where the forces from the tire travel through rigid linkage of the suspension and couple with the chassis's forces.

As a rule of thumb you want two things in geometry (in terms of instantaneous center positions.) You want to be able to control you camber curves in the (roll, ride) gradient of you car. And you want the geometry to be stable. Having IC's crossing ground plane, causes the forces in the linkages to switch directions, making all the slop and lash in your linkages rear their ugly head.

At the same time I will note, that one would probably want to keep their IC's as close to the ground plane as possible with out ever crossing it. Having excessive forces being transmitted though you linkages creates extra added friction in your joints, which greatly reduces grip and road holding ability of the chassis over irregular road surfaces. Not to mention the excessive jacking of the chassis under lateral forces.

Lastly it is my opinion on an FSAE car. It is best to have the rear roll center just slightly above the front. This induces a small amount of corner entry oversteer which I believe is ideal for a tight auto-X course.

I hoped this helps a little. If you have an questions please ask.

Jazzy