Roll Center...What is desired

[Jolane]

In general, is a higher or lower roll center preferred?

I know that the LCA pivot change is a change to alleviate bump steer, but I have also seen comments on it changing the roll center. If a car is lowered 1", the roll center changes (in part) due to the relationship of the LCA pivots. It lowers the roll center.

To me, it seems like the roll center should be as high as possible, but how high is good enough? Should it be at the vertical CG of the car?

Thanks,

Joshua

 

Any recommendations on where to find good starting points for tuning suspension is greatly appreciated. I have found a lot of info on how to calculate the parameters (scrub radius, roll center, etc), but don't know what a good target for these variables are.

[2126]

From all that I've read on static roll center....from approximately 3" above ground to at or slightly below ground. I've also read that the later is not preferable on early Z's. On my Z, with 1.5" shortened struts, coilovers, and with slightly less than 4" ground clearance at the front crossmember, I'm getting ready to relocate my LCA pivot point. The hopes are to minimize my bump steer....also, when I physically graphed where my static roll center will be with the LCA mod incorperated, I should be at around 3" above ground as opposed to the at ground position with the original LCA pivot locations. But for some basic useful information.....with your Z sitting normally, look at the position of your LCA's.....if they have a slight downward angle (lower point, nearest the wheel) or parrallel to the ground, you will likely be OK. You do not want an upward angle. In this instance you will not be getting the intended camber curve during bump (suspension compression). I'm no expert, but this is what I've learned from my research...hope it helps you to some degree.

[tube80z]

The reply on RCs had a lot of good info in it. I think you can't go wrong with that advice.

 

Some other rules of thumb to help you set things up:

 

Scrub -- no more than 20 percent wheel width or tire width (which ever is bigger). For a ten inch wheel that's 2 inches (and ideally you want less).

 

Caster -- one half the measured steering axis inclination

 

trail -- around 3/4 to half an inch

 

RC lateral migration (20 percent of track width)

 

Vertical migration -- make sure it doesn't cross ground level in any real world condition. If this happens in a corner the car will momentarily lose traction. This often feels like a sideways hop and can be very unnerving.

 

In general you want to run an RC that is close to the ground with the front slightly lower than the rear. Low is somewhere in the neighborhood of less than 5 inches off the ground. It depends a lot on the rest of your setup.

 

High RCs will lead to large jacking forces, which is like a big lever trying to turn the car over as you enter a corner. They manifest themselves as a lifting that increases with corner force. Old triumphs, volkswagens, and corvairs had swing axles that were notorious for this.

 

Just remember that everything is interconnected to the strut on these cars. You can't make on thing perfect without really screwing something else up. It's all a matter of coming up with the best compromise.

 

Good luck,

 

Cary

[randy 77zt]

i moved the pivot point on my front crossmember upward and out by 3/4".then i heated the tie rod ends to take that downward kink out of them.i dont know if this is correct engineering wise (i doont have fancy software for this) but the car handles very well.with gc camber plates i run 1.2 degrees negative in street and 3.2 on the track.if you are going to use the car for any kind of competition check the rule book first.my car is just a weekend open track /driver school car.

[Jolane]

Thank you for all the replies! Each one is very helpful! These definitely give me something to compare to now.

One more question...maybe I am misunderstanding roll center. I know that the tire travels along its path through suspension travel based on the suspension linkage. From the linkage, I can find the instantaneous virtual pivot point. From this I can connect that point to the center of the tire, and the point where that vector crosses the vehicle centerline is the roll center. In real terms though, is the roll center the point at which the vehicle body rolls at that instantaneous moment? If so, why is it desired that it is so far down? Is the roll center height and the size of the sway bar proportional?

Thanks again for all the help and insight!

Joshua

[JMortensen]

Jolane, I tend to disagree with the usual idea on this one. I think the RC should be as high as possible (cause it isn't going to be that high anyway) and the control arms should be as long as possible. That's the way I've seen it done on the fastest 510's I've seen. Pivot was moved up 3" or so, and the control arms were lengthened 4 inches or more. These were some wickedly fast autoxers that had this setup, and it seems to me that that all of the really really fast 510's had the pivots moved like this.

 

I don't think that you can move the pivot up much more than about 2" on the Z, and you really can't move it in too far IIRC just because of the layout of the front crossmember. But the best thing to do would be to get the control arms as long as possible and the pivots as high as possible. The "How to Hotrod Your Datsun" book had a bit in there about getting the control arms level or pointing a bit down, and I think that is repeated over and over and over. I think it was originally intended to say "Don't lower your car so low that the control arms point up, cause then it will handle REALLY bad" and has been turned into "If you have the control arms level that is the best handling setup".

 

Then you'd need to adjust out the bumpsteer, which is a whole other issue (search on that one).

[tube80z]

Thank you for all the replies! Each one is very helpful! These definitely give me something to compare to now.

One more question...maybe I am misunderstanding roll center. I know that the tire travels along its path through suspension travel based on the suspension linkage. From the linkage' date=' I can find the instantaneous virtual pivot point. From this I can connect that point to the center of the tire, and the point where that vector crosses the vehicle centerline is the roll center. In real terms though, is the roll center the point at which the vehicle body rolls at that instantaneous moment? If so, why is it desired that it is so far down? Is the roll center height and the size of the sway bar proportional?

Thanks again for all the help and insight!

Joshua[/quote']

 

I don't want this to get too long or sound too theoretical. Here's some discussion around this same topic from Mark Ortiz, http://www.auto-ware.com/ubbthreads/postlist.php?Cat=&Board=knowledgebase. That's much better than anything I could add.

 

Basically the intersection of the virtual centers as you describe above gives you the geometric roll center. This isn't really a point in space the car rolls about. It's more or less a fictional idea that helps us to understand how things will work. We can also look at the force based version, called the force application point. In instances where there is no roll center you have to do this. If you really want to know the point in space the car rolls about you have to use ride height sensors and suspension travel sensors. This takes into account things like tire squish, deflection, etc. It gets complex very quickly. Luckily most things follow how the geometric RC works.

 

And getting back to why this is important. The RC influences jacking forces and how quickly we see force build at the tire contact patches. The higher the RC the greater the jacking forces and the quicker they build. The difference front to rear controls how we distribute the load transfer. The is entirely based on the CG and track width. This is the basis of all chassis adjustments, to alter the rate of load transfer. Read what Ortiz has to say and hopefully this will make sense.

 

I do disagree with John on where the RC should be but I only autox. For me I prefer to have them underground but you need to have extremely long arms to be able to do this without screwing up too much else. At the end of the day you have to go with what works. And for the vast majority of Z with slightly longer or stock control arms that will be an above ground roll center.

 

Cary

[2126]

tube80z, very interesting post , by Mark Ortiz......pretty deep too! Good stuff however! As most senerios go, the more you know about a subject the more there is to know!

[JMortensen]

I do disagree with John on where the RC should be but I only autox. For me I prefer to have them underground but you need to have extremely long arms to be able to do this without screwing up too much else. At the end of the day you have to go with what works. And for the vast majority of Z with slightly longer or stock control arms that will be an above ground roll center.

Here is the key bit for me Cary:

If the roll center for the front or rear wheel pair – understood in the usual way, as the intersection of the front view force lines – is below ground level and between the wheels, that implies geometric pro-roll on both wheels. The outside wheel’s linkage generates a downward jacking force in that wheel’s suspension, and the inside wheel’s linkage generates an upward jacking force. The resulting couple acts to roll the sprung mass outward, exaggerating roll. Considered in isolation, this does add load to the inside wheels, and remove load from the outside wheels.

When I read this it says that if the RC is below ground "pro-roll" is created. Pro-roll being mechanical leverage that makes the car roll MORE. So the jacking forces are still there, but now they force the suspension to roll more than it would otherwise. That is the way I always understood it to work too. Once the outside of the LCA is pointing up, side load then compresses the suspension. If it is pointing down, yes, there is a jacking force, but it serves the function of lessening roll. At some point you could get too much of this, but like I said earlier I don't think it would be really feasible to do so with anything like stock components.

 

I don't claim to understand all the terminology on that page, but that part seems pretty clear. I am SURE that Cary knows more about suspension design than I do, mine is more of an "imitate success" strategy. I do enjoy talking about it, and hope we can hash it out a bit more.

[johnc]

If you're asking these kinds of questions, you should buy Race Car Vehicle Dynamics by the Millikens. Chapter 17 Suspension Geometry will answer your questions and a section in 17.5 titled: Front Suspension Design, MacPherson Struts has everything you are looking for.

[cygnusx1]

Here is a bit of Roll Center talk from the historic Jim Cook Racing Catalog.

 

http://www.jimcookracing.com/catalog/page16.html

[Jolane]

This is a great discussion! Thanks for everyones input!

JohnM,

I also thought about the RC much as you have replied here. That is what was confusing me, there seems to be two different sides of this. I always thought that the lower the RC, the worse, since the lever arm from the RC to the CG is longer. Also, from a load standpoint, as soon as the LCA goes up at the outer edge, the mass of the vehicle is working against the suspension also with a lateral force (force of tire horizontal holding friction) in addition to the rolling load. Thank you very much for raising these points. What it all means though from a driving and handling standpoint is still unknown to me.

 

My original reason for asking these questions has to do with changing my front crossmember/front suspension arms/steering R&P. I believe I could get the LCA a few inches longer and move the pivot point up as well, if that is the right thing to do. I am considering some other changes though as well...that I will go into later if I pursue this major change.

 

 

JohnC,

Thank you for the book recommendation! I ordered two different books the other day (couldn't find anyone who stocked them). I will add this to the list also. That is the primary reason for not just reading the information, no book to read...wasn't even sure which one(s) to order.

 

Certainly there is a lot of experience with MacPherson strut suspension here. I really do appreciate the help and insight.

Joshua

[tube80z]

Here is the key bit for me Cary:

When I read this it says that if the RC is below ground "pro-roll" is created. Pro-roll being mechanical leverage that makes the car roll MORE. So the jacking forces are still there' date=' but now they force the suspension to roll more than it would otherwise. That is the way I always understood it to work too.[/quote']

 

I won't argue with what Ortiz wrote, he's probably forgotten more than I'll ever learn. What I do know is that at least in autox I have found an underground roll center can be made to work better than an above ground roll center. I'm not saying you should blindly do this with your Z though. My car is a little different than most and features spring rates that are equal to the wheel rates. It reacts very quickly and dropping the role center made it more driveable. My front roll center is one inch below ground.

 

In a race car engineering class I took that showed data aquisistion using force and position sensors to show jacking forces. Below ground geometric roll centers produced smaller jacking effects and atually transfered weight onto the inner tire at the instant of turn in. They explained that below ground roll centers produce a more stable car that will react predictably to dynamic load conditions. I'd also like to add the fastest autox car in the world runs underground roll centers. So I can't be all wet

 

Above ground geometric roll centers produced higher jacking forces. They explained that this can be useful to speed up the response of a car and the faster loading of the tire contact patch will build heat into the tire quicker. They also shared the current practice in F1 is to use high roll centers and jacking forces to change ride hieght in corners but we're getting way of topic on this one.

 

I've tried both and this is what I prefer. That doesn't mean it's right but it doesn't mean it can't work either.

 

Cary

[tube80z]

If you're asking these kinds of questions, you should buy Race Car Vehicle Dynamics by the Millikens. Chapter 17 Suspension Geometry will answer your questions and a section in 17.5 titled: Front Suspension Design, MacPherson Struts has everything you are looking for.

 

I think Race Car Vehicle Dynamics is a great book, but it really doesn't tell you anything about struts other than give you basic definitions and tell you they suck. Before I bought that I'd look at some more basic books like Staniforth's Race and Rally sourcebook or Haney's book on tires. Those at least have some useful info for the DIYer trying to make their car go faster.

 

Cary

[johnc]

but it really doesn't tell you anything about struts other than give you basic definitions and tell you they suck

 

BS. It says nothing of the sort. As with all suspension types they discuss in detail in Chapter 17 they give the benefits and drawbacks of each. And MacPherson struts are basically a SLA suspension design with an infinately long upper arm (sometimes referred to as a slider). They state the obvious regarding the infinately long upper arm, "This is one of the major compromises of the strut type suspension, especially for performance applications." (page 632) The Millikens make similar, and correct, judgements about all suspension designs discussed.

 

Now, knowing that MacPherson struts are a form of SLA suspension, most of the discussion regarding SLA suspensions also apply including the discusion of instant centers, roll centers, degrees of freedom, motion path, camber curves, anti dive, squat, wheel path, caster change, wheel loads, etc.

[tube80z]

This is a great discussion! Thanks for everyones input!

JohnM' date='

I also thought about the RC much as you have replied here. That is what was confusing me, there seems to be two different sides of this. I always thought that the lower the RC, the worse, since the lever arm from the RC to the CG is longer. Also, from a load standpoint, as soon as the LCA goes up at the outer edge, the mass of the vehicle is working against the suspension also with a lateral force (force of tire horizontal holding friction) in addition to the rolling load. Thank you very much for raising these points. What it all means though from a driving and handling standpoint is still unknown to me.[/quote']

 

It's a little more complicated than that. There are three components of lateral weight transfer.

 

1. unsprung weight transfer -- this is a function of unsprung mass*lateral acceleration*non-suspended CG height/track width

 

2. geometric weight transfer -- this is a function of sprung mass*lateral acceleration*roll center height/track

 

3. Sprung mass weight transfer -- this is a function of sprung mass*(sprung mass CG - roll center)/track width

 

Geometric weight transfer is the source of jacking effects.

 

My original reason for asking these questions has to do with changing my front crossmember/front suspension arms/steering R&P. I believe I could get the LCA a few inches longer and move the pivot point up as well, if that is the right thing to do.

 

Lengthening the control arms is a good thing. Remember the TC rods and tie-rods too.

 

Cary

[tube80z]

BS. It says nothing of the sort. As with all suspension types they discuss in detail in Chapter 17 they give the benefits and drawbacks of each. And MacPherson struts are basically a SLA suspension design with an infinately long upper arm (sometimes referred to as a slider).

 

Okay, I was a little harsh on that one. But you can get all this for a lot less in other books. It's still a good one to have but it doesn't need to be one of the first purchased.

 

Cary

[JMortensen]

Cary, I know your car that you're building is completely tube framed, all custom fabbed stuff. I know you are also planning on extremely stiff springs. That would counter the roll, and make the roll center not so critical I think. But in a "normal" Z chassis with "usual" spring rates, the roll center makes a bigger difference than it will on your car. We all know that you can't put 800 in/lb springs in a stock tub because the tub itself becomes the spring. With springs not playing as critical a role in countering roll, then RC becomes another way to deal with the issue.

 

Trying to keep this rolling...

[johnc]

Race Car Vehicle Dynamics doesn't tell folks what to do to make their car handle better. In that respect it is often a big disappointment. What the book has attempted to do is describe vehicle dynamics in a fairly understandable way. Its just a step below a full-on engineering treatise (which was all that was available before this book was published).

 

Treat it like a college reference book. The kind of reference that was boring as hell reading and made you certain that college would never teach you anything practical, but years later you kept referring back to when trying to understand some "practical" problem like, "Should my roll center be higher or lower?"

[johnc]

From my own experience testing my 240Z with below and above ground roll centers (yes, I did back to back testing at Buttonwillow but not exactly for roll center positioning). Keeping the roll centers above ground made the car more responsive at turn in and the car would take and hold a line in high speed corners (above 80mph). It was easier and more predictable to drive at the limit.

 

With below ground roll centers lap times were the same and sometimes a bit better then the above ground roll centers because the lower speed corners could be driven a bit harder. But, the high speed corners were scary. Even after adjusting out much of the bumpsteer the car was darty and less predictable above 80mph. The car lost compliance and bumps that would normally be absorbed by the springs were now being transmitted into the sprung mass and affecting the chassis. Not a big deal at low speeds but very unnerving at high speeds.

 

The below ground roll centers would work for a qualifying lap but I'd bet a lot of money that you couldn't run a race without going off the track or wrecking.

 

This testing was done to see how I could get the front and rear suspension of my car past the "no change in suspension mounting points rule" for the SCCAs SM2 class while running 26" tall tires with an under 5" ride height. Take what I posted for what its worth. I wasn't a specific test of roll center location because other changes were involved, but the testing revealed enough to make the conclusions I did above.

 

I ended up just using 25mm bumpsteer spacers, 25.2" tall tires, and a little higher ride height.