Moving Roll CenterÆs under steady state conditions of a long corner. 1

[tensor47]

Hi All,

I’ve only recently discovered this forum, after occasionally posting on the machinery vibration forum. I’m a mech eng and live in Melbourne, Australia

I’m again designing a motor cycle powered rear engined hill climb car, all ball joints.
After reading Ortiz & others again and again I’ve come to the understanding that the force based roll centre height is important and correct in all conditions, whereas the older version of geometrical roll centers which I used many years ago, is a particular instance, which is only correct when symmetry exists.

The old geometric view was to keep the roll axis at the same height and minimize lateral movement, rel the car.
The new force based view is that the axis moves as the forces change, and quite differently from the force plane intersection point, which he sees as just part of the geometric construction to find the force based RC.

My question is where do I put my geometric axis, in one place, or attempt to move it in a way that follows the force based calc’s “resolution line” which corresponds to the wheel weight %.

On Hemipanter’s amazing site, see attachment, Goran has a section on this issue and moves his force plane intersection (of red lines) to follow the force based axis, and I am wondering why ?

 

[Windjammer96]

In more laymen terms, I believe you are asking about using migrating roll centers to tune your platform.

I've looked at your sketches and descriptions and see you are still using "above ground" roll centers (RC). The detriment is the further inside the RC travels the more roll is induced, and geometry is less effective.

We've been using migrating roll centers to tune suspensions for 15 years...way before force based software became available. We found locating static RC just below the ground, and migrating to ground level, on center with the outside tire to be most effective.

You'll find weight doesn't transfer to the outside, but actually to the inside via kinematics, ie causing the inside tire to load and the outside tire to unload...slightly. Yes sprung mass is still loading the outside tire but total wheel rate is less than conventional design.

This in turn not only allows more total grip, but also absorbs steering input disruption much easier, thus upsetting the vehicle less on turn in. This allows a deeper turn in, and quicker response manipulation in the long run.

Again, for the laymen, imagine, for each instance, a solid rod connecting each tire contact patch to the RC, and another solid rod from the RC to the CG. Imagine the cg as an extremly dense cannon ball, and the RC as the anchor point.

With the RC above ground, and traveling to the inside, there is a "pulling " force countering the CG. Gravity pulls the sprung CG down, with nothing to support it but springs, the platform rolls.

As the action proceeds, the RC "pulls" against the contact patches, at a decreasing angle, but extended arm which does cause a leverage effect, which in turn accelerates weight transfer, due to "lack of resistance".

The angle of "anti-force" or "anti-percentage" is paramount to weight transfer side to side, as it is under braking and accelerating in a longitudinal fashion.

When the RC is located below ground level, the applicable forces ALL happen in reverse. The CG is above ground the contact patches are "between" the CG and RC, in a 2 dimensional view, which should be understood before looking at a top view and visualizing 3 dimensionally.

With the RC below ground level, and migrating to the outside, the forces from CG to RC are in compression, In order for the cannon ball to move outwards, it must travel upwards due to the angle of the anchor point, the RC. Consequently, gravity as it is, the CG down force locates on the inside tire instead. Anchoring the RC at outside tire patch optimizes anti forces. Locating RC outside the outside tire patch, and above ground, actually causes the platform to reverse roll.

In more simple terms, if the RC is above ground, less movement is better. When static RC is located below ground, locating the RC under the outside tire patch under load is optimum.

Migrating roll centers add another dimension to setting up your platform, and in fact, considered "voodoo" by all but the upper levels of racing, who can afford the engineers, software programs, and testing and thus rarely used in the lower ranks. Most people simply lower the vehicle as low as the rules allow, and go with what they have. I've even seen vehicles raised, (by top teams)in order to prevent RC movement.

BTW, if you really want to upset the car during a "steady state turn", locate the RC just above ground in static, and have it below ground under load. Also, If you're not running below ground level RCs with your FWD vehicle...you're backing up!

...just my humble opinion...

Hope this has helped, or at least stirred the pot a little

Bob Hahn

 

[tensor47]

Hi Bob,
Thanks very much for your insight and very importantly, practical experience of what happens when you have done this or that.

The sketch is from hemipanters site of his Pantera mods.

I'll make some sketches of what you are saying and have a good think about it.
Can you explain the last sentence jargon please as it looks significant by I have no idea what it means.
Also, does fwd and 4wd mean the same thing to you ?

Cheers

John

 

[Windjammer96]

Your welcome.

You'll find, conventionally, the RC is located above ground, and the closer it is to the ground, the more quickly it will migrate to even outside the inside tire. This in effect induces and accelerates roll, (weight transfer). The instant the RC passes through the ground level plane, the RC will "swap sides". Depending on the geometry, can also immediately travel outside the outside tire, inducing an "unloading effect" on the outside tire or "reverse roll".

Feed back from drivers will be "turn in was fine, but abruptly washed out with a major push, ..like it was in a bind or hit the bump stops. Then, after correction, is good for a bit, then pushes on exit!"

---------------------
One of the greatest challenges for front wheel drive vehicles, is "corner exit front tire grip". As the vehicle accelerates, the front unloads. Many (incl some top teams I know) use very high static RCs and unbelievable spring rates to plant the outside tire as much as possible. The result is a 2500-3200 pound FWD go cart...

To make a long story short, they are forcing everything they can as quick and for as long as possible to the outside tire patch. Fact is, grip lost from weight transfer off the inside tire, is not entirely regained on the outside tire. Consequently LESS weight transfer is better.

However, when compared to conventional setups, locating the static RC below ground, actually reverses force applications. Less weight is transferred and at a slower rate so the inside tire isn't unloaded as much; lighter springs and bars are needed(better compliance); drivers don't require lightning fast reaction time(much more forgiving)and for people like me, MUCH easier to track tune.

BTW, we have run across purpose built, SLA racing suspensions designed specifically where the static RC is about 1" above ground, and migrate to the outside tire patch. However, we've never found this to be true with any production based platforms we've encountered.

-------------------

4 Wheel Drive or All Wheel Drive both use Front Wheel Drive.


Bob Hahn