Roll Center...Again

[yawpower]

Hi Everybody,

I am currently working with a production based road race car with SLA suspension front and rear. Class rules do not allow the suspension pickup points to be moved, and so the geometry is a mess.

The rear is fine, with the roll center at about 2.25" and very little movement in roll.

The front is really ugly. The roll center is .075" below ground static. At 1 degree roll, it is .031" above ground, and has moved 272" laterally towards the outside of the turn.

At 2 degrees roll it is 36.9" below ground, and has moved 2300" inches laterally, this time towards the INSIDE of the turn!

My first step in these projects is to do some basic weight transfer calculations to get a starting point for spring and bar rates. Normally, I have some control over pickup points, and I make sure that the roll center is stable. That being the case I ignore the lateral movement because it is quite small. That's definetely not the case here!

So...how do I calculate the percentage weight transfer through the roll center when it moves laterally?

I cannot imagine that the transfer moves as much as the vertical measurement would suggest. Has anyone ever considered this? Every suspension book that I have (Including Milliken and Milliken) assumes that the roll center is on the centerline of the vehicle.

Paul Yaw
Yaw Power Products

 

[mloew]

Paul,

I am surprised that you have only found references to centerline constrained roll-center calculations. For geometry based roll-center calculations, the roll center is generally assumed to be the intersection of the lines connecting the contact patch center and the instant center on each side. When the body has rolled, this intersection will only be on the centerline for pure trailing arm type suspensions. Obviously, for force-based calculations (and in reality) the roll center will be the point at which there is no transnational movement.

Best regards,

Matthew Ian Loew
"Luck is the residue of design."
Branch Rickey

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[GregLocock]

1) depends what you want, both sound a little low, and the height difference is also on the low side.

This has the advantage in that it will slow the weight transfer down, making the car less nervous. I'd stick the rear 2-3" above the front, and probably leave the front roughly where it is.

2) No it is not really bad. Plenty of circuit racers have sub zero front roll centres. If your roll centre height is at zero then all of your dynamic weight transfer is directly tunable via springs sta bar and shocks - of course that may push your spring rates up too high. The fastest way of telling the rear suspension what the front is up to is via the roll centres - that's probably why the rear RCH tends to be higher.

It is worth keeping an eye on the lateral location of the roll centre - it should move around smoothly, at least, and ideally I think it should remain within say 4" of the centre, unless you have a theory!

If you can, work with force based roll centres, not geometrical ones.



Cheers

Greg Locock

 

[jb56]

To start with - intersecting line kinematic roll centers have way too large an error rate. For example, I can have a 4 inch high by 6 inch to the right roll center using this method and have 2 differant cars with totaly differant a-arm angles not to mention lengths or anything to do with pin heights and angles.

From my experiance, pickup points and force application points is where things happen when running a car at it's limits.

What I start with is to find each side of the suspension's centerline roll center and instant center at static ride heights then re-measure the car with what it does in the turn. A simple shock travel indicator can be a big help here. Then I use various formulas found in Milleken's, Gillespie's, and Rowley's books to estimate my weight transfer. Once I have my weight transfer I then run through some formulas to calculate jacking effects both at the instant centers and through the a-arm and pin angles. This isn't that hard once you have an estimate of the laterel weight transfer and the angles and lengths of the arms and instant centers.

It can be a long involved process but I've had a lot more consistency on differant cars doing it this way. I pay particular attention to very accurately designing camber gains and losses while minimizing any upward jacking effects.

I can't give you a simple answer to any of this. It's taken me a long time, a lot of reading, and a lot of testing.