Split (decoupled) anti-roll bar? 1

[etatv]

I just had this idea and wanted to bounce it off of you guys. Have any of you tried or had any experience with a split anti-roll bar. What I am thinking of is something like a standard anti-roll bar that has been split in 2 with a rotating coupling in the middle. The coupling would rotate a certain nuber of degrees before "locking" to two halves together. I know that some high end race equipment does this with a hydraulic chamber that links the shocks. I am trying to brainstorm a way to achieve the benefits of installing a larger anti-roll bar on a softly sprung "luxury" vehical with out adversly affecting the ride.
I'm thinking that the stops could be adjusted so that under normal one wheel bump conditions, the bar would be decoupled, and in a roll condition the bar would lock together and act like a normal anti-roll bar but with a higher rate than normal (due to larger size). This would eliminate the harsh one wheel bump ride that is normally associated with larger anti-roll bars.

Am I smoking something, or is this possible?
If it has been tried, were there issues with the sudden change in spring rate when the bar "locks-up"? Possibly there is a way to make the bar get progressivly stiffer as it twists.

Thanks
Ed

 

[GregLocock]

It is not only possible, it is commercially available, I have seen a video of a Land Rover fitted with it. There's a brief article on it in the current issue of "Vehicle Dynamics International", or you might check out Delphi and Tenneco Kinetic's websites.

Kinetic also cross link their shocks, which can be a clever way of controlling the roll response. I am slightly dubious about this one - shock absorbers are the slowest way of reacting forces across the car in transients.



Cheers

Greg Locock

 

[patprimmer]

I think it's a lot easier than you are thinking.

Just use sholder bolts in the attachment points for the drop links, so there is some slack in the rubber bushes.

Correctly designed bushes and bush bearing surface for the drop links could result in a progressive application of the anti roll bar.

I am sure this is a currently used technique.

Certainly in the past, I have used bigger dia harder rubber, with big thick steel washers, on the drop link attachment to the moveing suspension member, done up much tighter than normal, so as to increase roll stiffness.

Regards
pat

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

 

[harrisj]

Ed:

I also did some work on this some years ago.

It does have advantages for off-road vehicles where you can maintain wheel contact through free articulation at low roll angles.

However, we found it a bit unnerving for high speed handling on the road. We applied the device to the front only; the front turned in nicely but as the lateral forces built up and the roll bar came into play, the vehicle understeered violently (as one would expect). Bumps in the pavement also made the handling very unpredictable.

A gradual transition from 'off' to 'on' would therefore be very desirable for road applications. In theory, you could match front and rear but I doubt you would ever keep them synchronised and the vehicle roll stiffness would go through a series of discontinuities just when you didn't want it to.

As you say, the gradual increase in roll stiffness would be a good objective. You could look at stranded torsion bars (say, 4 bars held in a square formation). These give a progressive spring rate for large angular displacements but there may be insufficient deflection in a roll bar.

Regards - John

 

[shanba]

BMW's active roll control uses this concept as well. They have a big electrically controlled coupling in the middle that is quite heavy.

 

[RKPsyche]

Freeplay is freeplay and patprimmer’s method would work as well and is certainly cheaper. You could also add freeplay at the anti-roll bar’s mounting to the chassis by using oversized mounts. The problem with all of these is that weight transfer to the outside wheel is reduced until you consume the freeplay, then its suddenly applied, causing an abrupt change in the handling.

Harrisj suggests that a gradual transition would be better, but the manufacturer has already done this. The rubber mounts on the linkages and mounts have strongly progressive rates, so the anti-roll bar is actually less effective at center than when rolled a bit. You could try adding rubber to the linkage to make it softer at the center, with a gradual, but you will have to play with it at each end to get good handling balance, since the springs have an affect on the total roll rate that is unaffected by what you do to the anti-roll bar. No mater what you do, transient response will suffer because weight transfer is being delayed.

The real solution is to make the car stiffer in roll while making it softer in warp. Anti-roll bars make the car stiffer in both roll and warp. It is the roll stiffness that has the greatest effect on handling and the warp stiffness that has the greatest impact on ride comfort. I have written an essay that goes into more detail at

http://members.cox.net/rseike/modes/index.html.

Citroen has a rally car that has split anti-roll bars that are interconnected so as to stiffen under true roll conditions (both axels tilting the same way), but not under warp conditions (tilting opposite ways). This isn’t too practical unless you have the means to build fairly complicated hydraulic systems, and is probably patented anyway.

 

[mloew]

FYI: In motorsports Citroen (on the WRC) (

http://www.google.com/search?sourceid=navclient&ie=UTF-8&oe=UTF-8&q=citroen+kinetic)

and Mitsubishi (on the Dakar) (

http://www.google.com/search?num=50&hl=en&lr=&ie=UTF-8&oe=UTF-8&newwindow=1&q=mitsubishi+kinetic)

use the Tenneco Kinetic system (

http://www.kinetic.au.com/news.html).

The system will be on a production vehicle next year in the Lexus GX470 (

http://www.google.com/search?num=50&hl=en&lr=&ie=UTF-8&oe=UTF-8&newwindow=1&q=Lexus+GX470+kinetic).

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.

 

[patprimmer]

Also, I would presume the OLD Citrone Hydro Pneumatic system introduced in 1955 on the "D" series would meet all the listed critera of controling roll, retaining a soft level ride, and not limiting wheel travel on uneven surfaces, not to mention the benifit of adjusting the ride height at the flick of a switch. The power operated jack for changing wheels was a nice little fringe benifit as well

Regards
pat

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

 

[etatv]

Thanks guys,
I didn't realize that there was that much work being done with this idea. I like Pat's idea, or maybe a center joint with some rubber in it to make it more progressive, but I think (as John points out) it might be better suited to off road situations, where you would normally not have a anti-roll bar due to loss of wheel articulation. I may have to test this on some of my off road stuff.
For the road, I really like the Tenneco\Kinetic idea. What do you think the red object in the lines connecting front to back is??? I am guessing that it is some sort of pressure accumulator/regulator, that acutally allows the articulation, but can also be adjusted to fine tune roll stiffness.
Ed

 

[GregLocock]

Well, it's obviously the clever bit. It could include some passive valves, some switchable valves and a compliant element, whether that is a gas spring or a spring and piston. I imagine it would be very similar in concept to a hydraulic engine mount, which these days can be remarkably complex.



Cheers

Greg Locock

 

[RKPsyche]

This is getting somewhat off-topic at this point, and it might be wise to start another topic on the Kinetic RFS, nevertheless…. I am reasonably certain that the device in the center is a reservoir and equalizer. There is really no point in putting an accumulator there—it just softens the action of the anti-roll bars, but in a nonlinear way that would compromise transitional handling.

The cylinders are not zero-displacement type, consequently the rod end chambers have lower displacement than the blind ends. This means that there is a net displacement of fluid during operation, which must be accommodated by the system. Secondly, all hydraulic cylinders leak a little bit, and must be replenished. Third, there is certain amount of fluid transfer past the pistons from one chamber to the other. This is not likely to be equitable over long periods of time and needs to be equalized to keep the system functioning properly. Finally, temperature changes need to be accommodated.

 

[carnage1]

why not just use really plush seats?

could variable coil springs be integrated into the drop links?

but to my point, is there a way to make a torsion bar variable rate?

 

[LMP900]

Lola's Champcar uses a split ARB link with one part having a piston and the other a canister carrying two coil springs, one either side of the piston. The springs can be pre-loaded against each other, so that when the force in the link exceeds the preload, the spring rate of the link halves. Quite neat - and adjustable.

 

[JakubMech]

You could make the installation ratio for the bar variable?
That would make the bar variable dependant on the wheel travel/position, although how much you could vary is something else.

Wouldnt halving the rate of the bar on a champcar be a bit scary? Id say it was more of adjustablility thing with the springs in contact 100% of the time (or loaded 100%).

 

[LMP900]

It only halves the spring rate of the link, not the total anti-roll bar stiffness. Then the tyre stiffness has to be factored in, which makes the step change not so frightening. A similar characteristic, but halving the total roll stiffness, is available with the monoshock arrangement as used on the Dallara F3 car (and their IRL car of a couple of years ago). It works.

 

[gt6racer2]

Couple of items ref: Kinetic system. The red block is a compliance device which allows the bar effect to be reduced for sraight ahead driving - reducing head toss. These compliance volumes also serve other lesser functions, such as temperature (pressure change ) compensation.
The bore/annular area is not an issue as the system connects bore/bore and annular to annular - hence there's no volume inbalance.
Second, the later systems do not link the shocks - those cylinders are not shocks and have no flow across the piston - and hence are a stiffness device as well as damping.

In general my preference is to stay away from non-linearities ( including switching systems ) as you lose center feel and the vehicle never feels "right". What you ideally need to do is separate the modes of the suspension -so that you can optimise more than one attribute at a time.

 

[Fabrico]

For highest performance, it is likly that the driver would need at least partial control over whatever method is used.