Racing torsion bars

[Belatti]

Hi, does anyone knows any supplier for racing torsion springs?

I could not find any supplier on the internet and dont know any arround here.

I know F1 uses torsion "springs" (and Im not talking about torsion bars here)

Thanks in advance!

[RacingManiac]

You are talking about ride spring or something else? F1 car uses torsion BARS....not made from coil...coil torsion spring would negate one of the best reason to use torsion bars, that is the lack of hysterisis...

[riff_raff]

Belatti,

F1 cars use torsion bars for suspension springs. The reasons for this is that they are easier to package with a rocker arm suspension, they are easier to swap out during suspension set-up, and they can be manufactured without any specialized spring winding equipment.

If it's a helical torsion spring (as pictured below) supplier you want (and I don't know of anyplace where they're used on race cars) try here:

http://www.asraymond.com/spec/catalog.asp#spec_torsion



Good luck,
Terry

[Belatti]

My bad, lack of technical english or plainly stupidity

I was trying to talk about the torsion bars F1 cars use instead of springs and not torsion springs.

As you Terry mention, torsion springs are not used in any racing aplication I know.

So, topic name changed, the question remains the same

[RacingManiac]

Its a tube, can be made from any tube. You use the torsion stiffness of a tube formula to get the required rate, based on material, diameter, thickness, and length(between where the tube is supported and where force is applied.



An example, from Univesity of Western Australia's FSAE car(circa 2007). You can see the rocker arm supported by 2 bearings, actuated by a pullrod. The torsion happens on the long unsupported portion that is clamped/fixed at the far end. The extra lever arm at that end with a turnbuckle attached allows you to set preload of the bar. Much like the adjustable spring perch on a regular coil-over damper.

[Jersey Tom]

Yep. Cut a piece of tube to length. Done.

Though heat treat, polishing, peening, etc are all good...

[riff_raff]

Belatti,

No problem. Your English is just fine.

As I noted in my post, the F1 teams design and manufacture their own torsion bars because it's fairly easy to do, they can produce them quickly, and they can have absolute control over them.

The ones used in F1 are basically hollow titanium bars with a spline on each end. Sometimes the number of spline teeth from one end to the other is different by one tooth to create a "vernier". The spline vernier allows the torsion bar to be very accurately indexed (angularly) at installation. The torsion bar is very carefully machined, heat treated, polished, honed, shot-peened and inspected because it is a very highly stressed part. Due to the extreme S-N conditions they are exposed to, sometimes they are also serialized, and their total number of load cycles are tracked, so that they can be removed from service when their fatigue life is used up.

Take a look here: http://www.erode-all.com/services-suspension.html

Regards,
Terry

[Belatti]

OK, thank you guys!

One thing I was wondering if I do design the piece myself is if any Saint Venant principle related problem can occur in the case the piece is not long enough.

[ReubenG]

If you are concerned about the stresses at the ends where the torque load is applied differing from the classical torque equations, then you are correct.

The classical St Venant torsion solution assumes the torque is applied to the flat ends of the cylinder as shear tractions, whereas in most real shafts in torsion see the load applied as contact (perpendicular) tractions via a key or splines. So the torsion equations we are taught in undergrad
are really only valid some distance away from where the loads are applied.

If you are concerned about the stiffness, then I would consider neglecting the length of tube under the splined hub (of the rocker arm) to be a reasonable first approximation - the stiffness of the hub will be significantly higher than that of the tube.

I've seen some work on bending of beams (not entirely applicable but a okay for analogy) that modeled the geometric effects of the loading devices and compared it to the classical Euler-Bernoulli bending equations. It showed that for span / thickness ratios of less than 4:1 the shear stress distribution was only in agreement with the analytical prediction (parabolic distribution) at mid way between the central load and outside support.

I am not familiar with the length / radius ratios used for torsion bars with rocker arms. Could you try to include some adjustability for length of the bar /tube under torque so you can fine tune stiffness?

[Belatti]

Thank you Reuben!

I think Ill end up building the designed piece and measuring its stiffness so I can compare what I have vs. what I have calculated.

[riff_raff]

ReubenG,

Torsion bars are suspension springs. So the torsional deflection per unit length must be, by nature, very high. And because the loading condition requires a high number of reversing load cycles, the combined working stresses must be kept low.

Torsion bar stresses must take into account torsional shear in the center beam portion, as well as contact, root fillet tensile, and root shear stress in the splines at each end. And these stress values must be knocked-down by appropriate stress concentration (Kt) factors wherever there is a change in cross-section.

Finally, with notch sensitive materials like titanium, surface finish or shot peening can have a big effect on fatigue life.

Regards,
Terry