9 replies to this topic

[John Oliver]

This is one of the topics that disappeared a few weeks ago, but if anyone still cares, I finally found a picture of a 3-spring suspension layout on the web that I can link to.




You can see that when the vehicle rolls (one wheel in jounce and the other in rebound), the rocker attached to the 3rd spring just pivots, but if the vehicle body moves vertically or in pitch, the 3rd spring is compressed or extended.

(Thanks, desmo, for tipping me off on how to embed this image!).


[This message has been edited by John Oliver (edited 03-21-2000).]

[BRG]

At first sight this reminded me of Citroën 2CV suspension, an unlikely origin for anything to do with F1!

Looking closer, it potentially has interesting characteristics, depending on the relative spring rates and the relative leverages of the rockers. If one wheel runs over a bump, that side’s spring will be compressed as usual, but it will also partly compress the centre spring, assuming that the other wheel does not move. But if the centre spring is highly rated, it might cause the rocker to exert a pull on the other side’s spring instead, effectively decompressing it and thereby lifting the whole car slightly.

I am sure that good engineers can figure it all out. But it seems to introduce more variables and unless they have useful benefits, would it be worth the extra parts and weight?

Going to the other extreme, there was a fashion a year or so ago for mono-springs on the front suspension of some single seaters, certainly in F3 (and I think Tyrell used it in F1). I could never see how this worked, unless the cars were just so stiff that it didn’t matter.





------------------
BRG

"all the time, maximum attack"

[desmo]

Here's an article written by the designer of the concept:
www.gtf1.com/suspension/reactive/reactive.htm

There must be something to it. Look at my post on the McLaren rear suspension. It uses a variation of this concept and Newey knows his stuff.



[This message has been edited by desmo (edited 03-21-2000).]

[PDA]

bear in mind that in that drawing, the pivot point under the centre spring is actually a torsion bar which provides roll resistance. I see that Lola are using this type of springing in their ALMS car. The theory is that the centre spring is pretty strong and carries most of the aero loads, allowing the "side" springs to be much softer as they only have to carry roll and single wheel bump loads. They should be able to run across kerbs at chicanes better (Mac seems to)

[Marco94]

The article on Gruers site is probably taken from Race Car Engineering. It's the type of magazine that carries a lot more articles like this. Shame there aren't any credits where they are due.

[John Oliver]

BRG, your comments about single-wheel bump are correct, but are true of any suspension system. In fact, with a conventional anti-roll bar, the single-wheel bump rate can be significantly higher than the two-wheel bump/pitch/heave rate. It is my understanding that this issue is precisely what led to the development of the 3-spring system -- high single-wheel bump rate decreases the ability to put power down, so the 3-spring system reduces single-wheel bump (and roll) rate while still allowing the car to be stiff relative to pitch/heave -- see PDA's comments. As far as extra parts, the long drop links are there anyway, as they connect to the anti-roll bar in a conventional layout, and the lower pivot replaces the anti-roll bar, so the only extra parts are really the 3rd spring/shock unit itself and its upper mount.

These systems are in use by most teams for at least some tracks -- the 3rd spring can be installed or not at the team's discretion, based on track characteristics. They've also shown up in CART and sportscars.

The monoshocks that were in vogue several years ago were essentially the opposite of this concept -- the spring/shock controlled pitch & heave, but the unit was nearly infinitely stiff in roll. The only roll flexibility at all came from compliances in the components and/or a stack of belleville washers on either side of the single rocker to allow it to move laterally. I talked to a fellow who did some consulting work for some F1 teams in that time, and he confirmed that these systems were very difficult to tune for limit oversteer/understeer balance and were falling out of favor.

PDA, the pivot under the center spring is not necessarily a torsion spring. One common implementation of an anti-roll bar is to have a T-shaped bar in that location that is twisted when the vehicle rolls, but if the rocker for the third spring has a bearing where it connects to the pivot, then that bar will not be twisted. In fact, allowing it to act as a torsion spring will pretty much defeat the whole purpose of the system, which, as you correctly point out, is to reduce single-wheel bump and roll rates.

[PDA]

The diagram shows a three spring system linked to a conventional anti roll bar. The anti roll is provided by the T bar, which is, in effect, a torsion bar. Tyese T bar anti roll bars are very useful as they can be easily adjusted. Although it is not allowed in F1 cars, CART cars have adjustment levers in the cockpit so that the driver can alter roll stiffness on the move - an essential feature, especially on ovals, where they need to adjust as fuel load goes down.

Luc Pellerin's system which has free pivoting provides roll resistance by the roll moment about the pivot, which works, but is not so easily adjusted. As referenced above, this is in the Race Car Engineering, November 1999 issue, a journal that anybody regularly accessing this thread should consider.

[John Oliver]

I've never understood why they no longer allow driver-adjustable anti-roll bars in F1. It seems like a really useful tool for managing the changing characteristics of the car with fuel load, tire wear, etc. Maybe they figured that with everyone making so many pitstops now, the characteristics of the car don't change as much as they did in the days before refuelling?

I would agree that the center pivot of the 3-spring set-up looks just like a conventional T-bar anti-roll bar, and it could be used as such, but I would contend that the rocker doesn't HAVE to be rigidly fixed to the vertical part of the T, thus taking the anti-roll bar out of the system.

[PDA]

The third spring has to be attached to part of the springing system, in this diagram, the antiroll bar, otherwise it is not suspending anything. It has to be a pivot which connects the two side springs. It does not have to be a torsion anti roll bar. If it is not, then some other means of attaining roll stiffness is required, and it has to be adjustable.

[John Oliver]

If the 3rd spring rocker is not a T-bar anti-roll bar, then the roll stiffness is obtained from the 2 primary springs -- the "conventional" ones. These spring rates can then be set to achieve whatever roll stiffness is desired, and additional heave/pitch stiffness can be added in with the 3rd spring. Thus the car can be made vertically stiff to resist large aero downloads without increasing the single-wheel bump rate and compromising traction.