Torsion bars

[rubinhothebest]

Hi.
I'd like to know more about torsion bars, their physical caracteristicas and so on, because I don't believe they have the same behaviour as a coil spring. What are the differences between them?
Thanks for helping.

[JGM]

A torsion bar is simply a coil spring which hasn't been coiled up. They are the same thing, only the packaging is different.

[RWC]

a torsion bar is a simple straight rod,held one end & twisted the other.

Imagine a quarter turn section of a coil spring.
When you press down on the high end,it TWISTS the other end.So a coil spring could be said to be an infinite number of torsion bars working together.

[Ray Bell]

That is correct, but I believe that there are some differences in result caused by the coiling, that damping requirements are therefore different.

Any shock absorber experts around who can give us the ins and outs on that?

[ma]

One of the differences is that coil springs cause a bending load on the shock, raising internal frictions. Torsion bars do not have that problem. Hyperco will be introducing a product at the PRI show in December that will take care of that problem, as well as hollow coil springs.

There may also be a difference in the natural frequencies, at least as concerns the designs currently used (short & small dia torsion bar vs a longer & thicker dia wire coil spring of the same effective rate), but I'd have to go back to the textbooks to verify that one.

[wuzak]

Ma, I'm not sure about the coil springs causing bending moments in the dampers. Can you please explain it some more?

The bending moments that you talk about must surely be largely dependant on the system of mounting of the springs and dampers, and of the geometry of the suspension as a whole.

In F1, for example, damping and spring units are often (always?) independant of one another. This because the springs and dampers are acuated through a push/pull rod through a rocker arm and into the springs and dampers. The springs and dampers have separate actuating rods. All of these joints are ball joints to eliminate any trnsmitted bending loads.

I know that motorbike fork dampers have had problems with binding in the past. But this was more to do with the forks taking all of the braking loads as well as the suspension loads.

[KC]

I would think that torsion bars are progressive springs, meaning the more they are twisted, the harder it is to twist them. Coil springs can be made to be constant rate or progressive. It depends on just what you need.

[ma]

Wuzak:

A coil spring transfers its force to the end coil at the point of closure. How well that load is distributed around the closed coil depends on many factors, but in general the load stays nearest the point of closure. As the closed coil is being forced to stay perpendicular to the spring axis, the load will not be distributed evenly, and therefore it is not averaged at the spring centerline.

If you compressed a coil spring and allowed the closed coil to tilt as it desires, you would see that the end will tilt up to about 4 degrees.

I have data that shows that a typical coil spring will produce a side load of approximately 20% of the coil loading - ie - 1000# of axial loading produces about 200# of side load. Hyperco's device will reduce this by about 92%.

If you go look at an ild formula car that is still using long, low rate springs, you will most likely see that the spring rubbs on the shock body. It rubs because of this side loading.

Manufacturers such as Hyperco try to reduce this problem by compressing the spring a bit with a special machine,allowing the closed coil to tilt as it desires at that load level, and THEN grinding the closed coil flat. Unfortunately, this only means that the load is somewhat centralised at only that particular load level.

If you look at most modern McPhearson strut setups, you will see that the spring is not concentric around the strut. The manufacturers were aware of this problem a long time ago, and found that skewing the spring a certain amount reduced or eliminated the problem. The strut bending loads induced a lot of friction and compromised ride quality.

The F1 boys went to torsion bars for that same reason - it eliminated the bending-load induced frictions in the shocks (compared to the traditional coilover setup), and therefore allowed better mechanical grip. For ease of packaging, the torsion bars are usually placed at the pivot of the bellcrank that actuates the shock.

[rubinhothebest]

Thank you very much, though I have some difficulties with technological terms.
I still have one doubt: from what I knew, springs and dampers have dependant behaviours, because the forces that act on one, act on the other.
So, what are the torsion bars made from? Are they from the same materials or do they use harder ones profiting their advantages?

[ma]

The biggest problems with the use of torsion bars in F1 currently (according to a couple of "inside" sources) is longevity (one race), and some strange harmonics. These 2 problems are primarily due to the extremely short nature of the current bars.

Because of packaging problems, the length of the bars are kept really short - about 6-7 inches overall. The diameter is also small to get the rates desired, but the angle of twist is high, causing premature fatigue. The diameter, and the surface finish, is critical, as a .002 diameter change is significant. Any surface roughness will decrease the life expectancy significantly.

So far, the materials tried aren't really different than what has been used for coil springs, but experiments are being performed on alternatives.

[desmo]

All the F1 teams appear to use steel as their material for torsion bars. A multiphase steel referred to as 'maraging' steel is used by at least one team.

[TomNo]

O.K., what you had said is enough, I think. But what about materials, which are used for torsion bars? Which kind of steel? If you know, you can say it in EN standarts. And what is about physics I will write propably tommorrow. (We had just talked about in school.) Tom

[f1parts]

Torsion Bars can be Titanium, Maraging Steels or other through hardened alloy steels. The length can be (but not necesarily I'm sure) longer than the 6-7" suggested and wall thicknesses of the outers is often less than 2mm. (the torsion bar is usually an inner bar fixed to an outer tube and then pre-loaded with levers). The levers that opeate these bars may also be of similar materials.

[KC]

Materials are based on appications. NASCAR stockers are so heavy they predominantly use carbon steel coil springs. F1 cars are very light and extremely weight conscious are probably use exotic titanium and titanium alloys. The top guys in sprint cars use 24" titanium hollow torsion bars for their front ends and shorter ones for the rear. I have seen coil over setups that use the normal looking spring over damper and then a separate smaller spring on the damper shaft to increase max compression resistance. Many CART cars use coil overs for wheel deflection and a short torsion bar for pitch damping. The permutations are nearly endless. Materials can be just about any material that can be worked into the shapes needed to produce springs. Chevrolet has used composite materials for suspension springs on the Corvette.

[TomNo]

So as I wrote yesterday, I prepare for you simple physics of torsion bars, that are scanned papers A4 (links are below). I expected, who is more interested into engineering, he know, and who isn't, for him is better this simplier version (not so much).
Because of limitations of attachements I had to place my work to external site. For slower connections I made lower quality JPG images, but I think, it's enough.For better connections (nearly all) there are more quality GIF images.
I'm not sure with terms, which I had used there, if is here anybody, who can say right ones, I will be happy.

Links to the files >>

for slower connections -
http://www.mujweb.cz/www/torsionbars/bar1.jpg >> first part
http://www.mujweb.cz/www/torsionbars/bar2.jpg >> second part

higher quality images -

http://www.mujweb.cz/www/torsionbars/bar1.gif >> first part
http://www.mujweb.cz/www/torsionbars/bar2.gif >> second part



Tom Novotny

PS: All questions write to this forum or email it to tom.novotny@altavista.com