Roll Center; Exactly what is it???

[K-fab]

Dump!,
Thanks for that booklet - covers just about all of it.

In the last suspension setup that I designed, I took into account pretty much everything you've described - and let the RC end up where ever it ended up. I designed for camber and caster changes, scrub issues and scrub radius around the kingpin inclination.

Getting rid of bump steer was probably the hardest part of the design - single seat car with a very wide a-arm setup - front of the a-arms mounted almost together at the nose of the car while the rear of them mounted out near my knees. I had to make a pretty complicated set of bell cranks to get the motions all working together w/o inducing toe changes as the suspension cycled.

I also addressed scrub by making the a-arms as long as possible, so the travel I was getting (about 11") worked closer to the horizontal aspect of the arc that it traveled. - as you're aware, the more your suspension moves downward (on the arc), the more horizontal movement you get. - 45 degrees = the same amount of verticle and horizontal travel in the X-Y plane (x would be parallel to the bottom of the chassis, y would be parallel to the plane of the motion of the arcs the suspension moves).

I know of a five link design that actually induces quite a bit of camber as the suspension gets to the bottom of it's travel and makes the contact patch move outward. The top of the tire will tilt inward near the bottom of the travel, keeping the contact patches of the rear wheels fairly constant in distance thru the suspension's movement. - makes for a pretty stable car.

I guess, in retrospect, the questions about the RC of the car have been answered for me. - Don't really worry about it... Make the suspension work to it's best ability - camber, caster, toe, scrub radius around the kingpin inclination, bump steer and the likes and let the roll center end up where it ends up!

Thanks for the info.

[SPONSOR]

Membership in the largest desert racing community has its advantages - Participate in Forum Discussions - Send and Receive Private Messages - Maintain Public Photo Albums - Access to Groups - User Profile in our Social Network - Increased Access to more Sub Forums - Reduced Online Advertisements Join our community today

[ntsqd]

I haven't yet read all the way thru each post, but so far what I've seen is Roll Center and Roll Axis being confused.

Roll Center is the POINT in space that either the front of the rear suspension allows the sprung mass to rotate about for that suspension ONLY.

Roll Axis is the LINE btwn those two points of rotation that the whole sprung mass rotates around.

So if you're talking about the whole vehicle then it's Roll Axis, but if you're talking about the front or rear suspension only then it's Roll Center.

Not sure if that will help anyone else, but for me it's an important distinction when I'm trying to visualize what's going on.

[Outcast]

Actually, live axle suspensions have their own roll axes, apart from the vehicles roll axis. Which is why you get roll over or understeer in those suspensions.

Thus you have the Roll Center, which is the exact instant center that the sprung mass will rotate about at the given position of the individual front or rear suspensions. As stated, the Roll Center, itself, is an Instant Center meaning it moves up and down, and even side to side, depending on the geometry, as the suspension cycles.

Then you have the vehicle roll axis, which is the line connecting the two roll centers. This basically depicts how the entire spring mass will roll, again at a particular position of the suspension. Remember, since the Roll Centers move, so will the vehicle roll axis.

Finally, with live axle suspensions, you have the individual suspension roll axis. This depicts how the axle will "steer" relative to the longitudinal vehicle axis as the suspension articulates. This effect is called "roll steer". Depending on the orientation of the suspension's roll axis, the axle will steer different directions, either in to or out off a turn, if you will, as the body rolls. Like the two previous variables, the suspension roll axis changes as the suspension cycles and it is quite possible for a suspension to vary from roll under steer to roll over steer as the suspension cycles. Generally, neutral or roll understeer geometry is preferred over roll oversteer as it provides a more consistent feel to the vehicle when cornering. Let it be said that some of the more revered designers in the off-road racing industry place a pretty high priority on suspension roll axis and create their geometries around acheiving desired roll steer characteristics and also minimizing change in roll steer as the suspension cycles.

As to why trucks, with live axle rears have a swaybar on back and not the front, what you will typically find is that with most vehicles, especially off-road racing, the roll center of a linked, live-axle rear suspension will be higher, often by quite a bit, than the roll center of the independent front suspension. And while roll center height definately comes into play in how much body roll will occur, remember body roll is dependent on the suspension's roll rate. Meaning shock geometry and wheel rates factor in as well. Thus, a truck may have a sway bar on back, even though the rear suspension's roll center is higher but the actual roll rate of the rear suspension may, in fact, be less than the front because of the different wheel rates (shocks on trailing arms versus A-arms). Also, be aware that increasing the roll rate at a particular end often leads to less cornering traction for that end because more weight is transferred between the inside and outside tire. Thus, a swaybar is added to the rear to help make it easier to swing the rear around in turns so you can "back it in" or "flat track" around turns.

[Samco Fab]

Dump and K fab, you guys mention scrub, or from what I gather it is the trac width change through out the wheel travel. You mention that with less scrub you get more stability through whoops, and I would imagine in ruts and in many other places.

Do you adjust this scrub with your camber gain pulling the bottom of the tire out to widen the trac width under droop? I know the longer the control arms the better it is, and limiting wheel travel sooner would also help, as the trac width change gets more dramatic the farther down you get per inch of wheel travel. I could see mounting the spindle at the same ht. as the lower control arm would swing the contact patch of the tire out more as well, without too short of an upper arm.
Also does anyone use a specific program that could be purchased to design a arm suspension? How simple are those programs? I am semi computer illiterate.

[DUMP!]

Yes, To obtain no scrub or minamal scrub you will gain negitive camber at full droop. Also I believe this to be a helpful side effect. I would much rather have a little more neg camber at full droop if the vehicle was to nose dive hard on one corner or the other. With the neg camber the vehicle will have a better chance of driveing out of the nose dive situation rather then hooking an edge and causeing a roll over.

As for computer programs, I started with Autocad (2D) about 8 years ago. I've also used a program called Velum, also (2D). About two years ago I started to teach myself Soilidworks (3D). I'll never go back to a 2D program. There is no program that I know of that you just plug in the numbers and it magicly spits out the perfect design. That is a misconseption that most people are under. All of the design programs I mentioned require the user to create the design. The information you get out is only as good as what you put in. Cream in, cream out, crap in, crap out so to speek.

Dump

[K-fab]

I had (have???) a copy of Solidworks floating around - they make it look so nice and easy in the demos... I need someone to walk me thru the basics again...

ACAD is good for static design, but Solidworks - NICE!!! Make your parts, put 'em in motion and see where stuff hits.

[Scott_F]

I would add that it is not possible to eliminate side scrub, or track width change, in a long travel A-arm design. If it was attempted, the camber would be way too negative at bump and droop.

[Ryan_P]

Yes, To obtain no scrub or minamal scrub you will gain negitive camber at full droop. Also I believe this to be a helpful side effect. I would much rather have a little more neg camber at full droop if the vehicle was to nose dive hard on one corner or the other. With the neg camber the vehicle will have a better chance of driveing out of the nose dive situation rather then hooking an edge and causeing a roll over.
Dump

What about positive camber at droop, only a few degrees, to keep the outside tire contact patch more level in a corner?

Speaking of Solidworks, I was given a copy of the program a couple nights ago and basically taught myself. It's not very difficult, if you are computer oriented and have patience, you can teach yourself. I was making spindle snouts with flanges in a couple hours after installing the program.

-Ryan

[DUMP!]

I would add that it is not possible to eliminate side scrub, or track width change, in a long travel A-arm design. If it was attempted, the camber would be way too negative at bump and droop.

Exaxtly what was ment when I said in my earlier post that "Like anything with these cars it is a series of compromises." The suspension picture in my earlier post has .843 of scrub per side. I would consider that to have greatly reduced any effects that scrub could induce into the handleing of the vehicle. I guess that it all depends on what your defanition of "too negative" is. Below is a chart of the actual numbers for the design pictured earlier.

****** Camber / Scrub / Wheel Travel

Bump / 9.5 deg neg / .002 / +14.5"

Ride / 1.5 deg neg / .000 / 0.0"

Droop / 3.0 deg neg / .841 / -8.5"

It's not that it isn't possible to acheive 0.0 scrub. It's a matter of what you are trying to gain with your design. An example would be , if you didn't care much about scrub, you could no doubt develope a design that would have much more then the 23 inches of travel that the design refered to above has.

I guess my point is that the first thing that any designer must do is sit down and figure out what his design criteria is, what is it he is trying to acheive with his design and then go from there.

What about positive camber at droop, only a few degrees, to keep the outside tire contact patch more level in a corner?

Speaking of Solidworks, I was given a copy of the program a couple nights ago and basically taught myself. It's not very difficult, if you are computer oriented and have patience, you can teach yourself. I was making spindle snouts with flanges in a couple hours after installing the program.

-Ryan

I don't think positive camber is good any time. Having said that, I'm not sure I understand your question. Did you mean the inside tire contact patch? If so, the inside tire is the unweighted tire in a corner and basicly along for the ride at that point. I don't beleive it is offering much in the way of grip and is unimportant as to the attutude on the contact patch.

Maybe this will illustrate why I believe neg camber is more important.

Imagin you hit a jump in such a way that the rear end kicks and comes down hard on the right front corner (bad things always happen on the pass side). With neg camber it is more likely it will come down on the tread of the tire and roll out with just a sore neck or messy shorts. With pos camber you take the chance of it comming down on the sidewall of the tire. This could cause a sidewall puncture because the sidewalls are constucted differantly from the tread. Or it could peal the tire off the bead or out of the beadlock depending on your tire PSI. Both cases could result in a roll over.



Dump

[partybarge_pilot]

What about positive camber at droop, only a few degrees, to keep the outside tire contact patch more level in a corner?

They all ready have that, it's called I-beam suspension. I'm with Dump on this one, at speed the inside tire is just there to help hold down it's corner.