24 replies to this topic

[TheScaleF1Project]

Hi guys,

I'm new here, so hello!

I was wondering if anyone could help me out with a small tech question that has been confusing me - mainly because I've read too much about it and confused myself!

With F1 suspension, there seem to be 2 trains of thought: torsion springs and coil springs.

My question is, I have read in several places that torsion bar springs are being favoured for packaging reasons. But, (see the first photo - FW24) in this case the springs are integrated into the damper anyway.... so, even if torsion springs are used (see the second photo), the dampers still take up the same space.

Can anyone tell me - am I right, or have I missed something obvious?!

Thanks!

Martin




Coil Spring and damper as part of the same unit



Torsion spring (bottom right) - in fact taking up more space because of the connecting arm from the bell crank.

Of course this is not taking into account the rotary damper used by Ferrari and Sauber... but leave that out of the equation for the moment!

[Lukin]

At a guess I'd say compare the two extended damper lengths though. Even for F1 (whose springs could quite easily be substituted with a solid block of ali) need a certain free and loaded length for the spring.

The torsion bar allows you to reduce the damper length. That seems like one advantage to me.

Im not 100% on something though. Unless your spring maker can make springs who's loaded lengths are all identical, you have to play around with spacers or change spring perch's. This wouldn't be an issue with a torsion bar would it?

[Powersteer]

F1 teams are going maximum on aerodynamics so making everything as compact as possible it their priority so its torsion.

[TheScaleF1Project]

Ok thanks for the replies.

Am I correct in assuming that to adjust the springs on both torsion bar spring and coil spring systems you have to swap out the spring - i.e. change it for a spring of higher or lower rate? There is no such thing as an "adjustable" spring? (the use of bump rubbers aside)?

Also, kind of on this topic - adjustable dampers. Are 2 way adjustable dampers adjustable for both bump and rebound? whereas 1 way adjustable dampers are adjustable for only bump or rebound (or both, but at the same value)?

So.... what are 3 and 4 way adjustable then?!

Sorry for the "newbie" questions, but having searched on t'internet for ages I can't find any specific answers to these questions....

Many thanks!

Martin

[Powersteer]

A torsion spring can be designed to be adjustable without changing the spring itself. Design a torsion bar in a way that it could be mounted in shorter or longer variation will change the load. Think a one way adjustable shock absorbers are rebound adjustable.

[LandOfSnow]

3- or 4-way adjustable probably means separate high and low speed adjustment on bump and/or rebound.

[jgm]

It is perhaps worth remembering that a coil spring is a torsion bar in a different shape - they both achieve their springyness for the same reasons - by twisting the spring metal.

[murpia]

A few advantages of torsion bars over coil springs on dampers:
No side loads on the damper shafts = less friction.
Easier to change, they just pop in and out on splines and are 'set' so no ride height change is required.
Less friction as no spring loads are handled by sphericals (current practice is to mount the torsion bars concentrically in the rocker pivots, unlike the photo above of the old Arrows).
Cheaper to make to the required tolerance.
Regards, Ian

[TheScaleF1Project]

Thanks for all the replies! Some very useful information.

The info about the torsion bar springs being mounted in the pivots makes more sense re: packaging now.

Cheers!

Martin

[Powersteer]

Problem with torsion bar when compared to a coil spring is that is it heavier. The bar is mostly mounted in the wrong end of suspension movement so more strength or metal is needed where as a coil spring is mounted at the right end of suspension movement. If you take a Formula One torsion and where they always mount it, on the push-rod rocker pivot point, when the suspension move, the most minimal movement of the rocker is its pivot point so a torsion bar would have to be build a lot stronger. If it was mounted on the highest ratio of that pivot point it can be made a lot lighter. One way to be creative is to gear it, increase movement ratio. This way a torsion bar can be made as narrow and long as a coil spring but cleanly mounted.

[murpia]

Are you sure torsion bars need to be heavier?

For equal wheel rates the energy contained in either a torsion bar or coil spring will be equal for the same wheel deflection. Now if you design the torsion bar properly then it will store this energy at the maximum material stress permissable for sufficient fatigue life. It may or may not need to be a tube to achieve this. This makes it a very efficient use of material mass.

If you store the same energy in a coil spring you may also achieve the same maximum material stress and have an equally efficient spring, but you might not because you can't easily make it a tube and you are contrained by also needing to achieve the correct free length and correct coil bound length (which will affect the number of coils and the coil diameter and the coil spacing).

Regards, Ian

[Engineguy]

A lot depends on the lever arm you use with a torsion bar. If the torsion bar is at the pivot of a short arm (A below, on a 6" arm) it can be pretty light. If the torsion bar is at the pivot of a longer arm such as the (13") upper wishbon (B below) it will be twice as heavy. If the torsion bar is at the pivot of a really long arm like the (20") lower wishbone (C below) it will be 3 times as heavy. In addition the lower wishbone will gain weight to withstand the severe bending load the torsion bar puts on it... and its going to do odd things to any compliance bushing you have at the pivot.



Porsche 911s traditionally had the torsion bar at the lower wishbone pivot, and the wishbone was pretty heavy. When they removed the torsion bar on the 935 (replaced by a coil on the strut), they were able to use a very lightweight lower wishbone made of scarey small tubing.

[TheScaleF1Project]

Originally posted by Engineguy
A lot depends on the lever arm you use with a torsion bar. If the torsion bar is at the pivot of a short arm (A below, on a 6" arm) it can be pretty light. If the torsion bar is at the pivot of a longer arm such as the (13") upper wishbon (B below) it will be twice as heavy. If the torsion bar is at the pivot of a really long arm like the (20") lower wishbone (C below) it will be 3 times as heavy. In addition the lower wishbone will gain weight to withstand the severe bending load the torsion bar puts on it... and its going to do odd things to any compliance bushing you have at the pivot.



Porsche 911s traditionally had the torsion bar at the lower wishbone pivot, and the wishbone was pretty heavy. When they removed the torsion bar on the 935 (replaced by a coil on the strut), they were able to use a very lightweight lower wishbone made of scarey small tubing.

So in essence it's simple moments from GCSE physics - the force necessary to twist a torsion spring reduces the longer the cantilever (ie the pushrod) is.

So theoretically, if the torsion bar was mounted in the pivot, the force necessary to turn it would depend on the size of the bell crank...?

That would be one way to make them adjustable (after a fashion) - build an adjustable bellcrank (or - if they were geared as powersteer suggests, adjust the gear ratios...)

Thanks for the replies and help!

Martin

[Powersteer]

Originally posted by murpia
Are you sure torsion bars need to be heavier?

By design. A shock absorber needs rigid mountings, what better way to mount the coil spring around it combining the two's need into one chassis mount. Sachs cleverly design the rotary shock absorber so it would work just like how a torsion bar does, twist, which makes for a far more impressive packaging design and combining the two's torque values into one mounting.

Its also a ratio issue as I have mentioned before, a coil spring can be mounted on the higher ratio of a suspension system making it a lot more sensitive to wheel movement. Gearing a torsion bar could work. Like jgm said, a coil spring is a torsion bar, and for me its just mount on the more efficient are of the suspension.

Originally posted by Engineguy
A lot depends on the lever arm you use with a torsion bar. If the torsion bar is at the pivot of a short arm (A below, on a 6" arm) it can be pretty light. If the torsion bar is at the pivot of a longer arm such as the (13") upper wishbon (B below) it will be twice as heavy. If the torsion bar is at the pivot of a really long arm like the (20") lower wishbone (C below) it will be 3 times as heavy. In addition the lower wishbone will gain weight to withstand the severe bending load the torsion bar puts on it... and its going to do odd things

Yeah exactly. Awesome post by the way.

[murpia]

Originally posted by Engineguy
A lot depends on the lever arm you use with a torsion bar. If the torsion bar is at the pivot of a short arm (A below, on a 6" arm) it can be pretty light. If the torsion bar is at the pivot of a longer arm such as the (13") upper wishbon (B below) it will be twice as heavy. If the torsion bar is at the pivot of a really long arm like the (20") lower wishbone (C below) it will be 3 times as heavy.

Sorry, I still stand by my original argument regarding strain energy. The above is not quite correct. Just because a torsion bar may be required to be stiffer or softer does not mean it is heavier or lighter. If designed correctly it will just be a different arrangement of the same mass (larger or smaller tube o.d. / thinner or thicker walls / longer or shorter as required).

Feel free to post some design calculations and prove me wrong.

Regarding the mounting points, an F1 tub lay-up will be optimised to the loads, in whatever form they take. So the damper mounts will be weaker than if they were spring + damper mounts, but still the required stiffness.

Regards, Ian

[Powersteer]

Originally posted by murpia
Feel free to post some design calculations and prove me wrong.

His post already cover the basics.

Originally posted by murpia
Regarding the mounting points, an F1 tub lay-up will be optimised to the loads, in whatever form they take. So the damper mounts will be weaker than if they were spring + damper mounts, but still the required stiffness.

So it needs strengthening, mount the spring and damper in different places then two seperate areas need to be beef'ed up, instead of one making it more complex.

[Greg Locock]

The energy stored per unit mass can be higher in a torsion spring (because no-one makes coil springs from tubes!), so I think there is a good, fundamental, argument that given appropriate geometry a torsion spring could be lighter. The equations are easy to find, Bosch Blue Book has them for a start, so someone can easily nut out the theory. I vaguely remember that designing the optimum (lightest) coil spring for a given wheel end load and rate is not exactly straightforward, either.

I'm (probably) not going to do it.

[Powersteer]

Originally posted by Greg Locock
because no-one makes coil springs from tubes

And if they do?

[murpia]

Originally posted by Powersteer
So it needs strengthening, mount the spring and damper in different places then two seperate areas need to be beef'ed up, instead of one making it more complex.

I don't think the typical F1 monocoque designer cares much about complexity, just about maximising performance. Stiffness is often the requirement, not necessarily strength.

Just one question - do you accept the strain energy argument regarding torsion bar designs for the same wheel rate at different motion ratios?

Regards, Ian

[Powersteer]

Originally posted by murpia
Just one question - do you accept the strain energy argument regarding torsion bar designs for the same wheel rate at different motion ratios?

Sorry, but what do you mean strain? On the arm or the spring?

[murpia]

Originally posted by Powersteer
Sorry, but what do you mean strain? On the arm or the spring?

Strain energy stored in the spring.

1) For equivalent wheel rates at the same wheel displacement the strain energy stored in either a coil spring or a torsion bar will be equal regardless of the motion ratios involved.

2) It is lighter to use a torsion bar to contain this strain energy because it can be made from a tube to an exact geometry and therefore approach the limits of the chosen material.

3) A coil spring is heavier for the same energy because it also needs to achieve the correct free length and correct coil bound length (which will compromise the number of coils and the coil diameter and the coil spacing) and because it will have 'dead' coils at each end to allow it to mount. It also (probably) can't be made from a tube.

Point 1) may be the hardest to get your head round, but just note that whatever coil spring motion ratio you choose you are trading stiffness for displacement - the soft coil spring a long way from a rocker pivot has to move through a greater distance than the stiff coil spring close to a rocker pivot. But those 2 springs generate the same torque at the rocker, the same torque that the torsion bar has to generate. Torque x (angular) displacement = energy.

Regards, Ian

[Greg Locock]

If someone makes coil springs from tube I'd love to see how they design the ends. I'm always a bit suspicious of what goes on there, it seems to me the bottom coil thinks it is in bending, while the next coil up is in torsion. But, yes, if you can make tubular coil springs I think they could in theory be as light as torsion bars, ignoring practicalitites, because the energy storage method si the same. I don't seem to have goaded anyone into doing the algebra!

[Powersteer]

Originally posted by murpia
Point 1) may be the hardest to get your head round, but just note that whatever coil spring motion ratio you choose you are trading stiffness for displacement - the soft coil spring a long way from a rocker pivot has to move through a greater distance than the stiff coil spring close to a rocker pivot. But those 2 springs generate the same torque at the rocker, the same torque that the torsion bar has to generate.

I think everybody in the forum understands that so maybe I could get my head around that.

Originally posted by Greg Locock
I think they could in theory be as light as torsion bars

Metal for metal you meant here? Anyway, its just that torsion bars-beam always seem to be huge.

bar's on that thing looks thick although it could be tube.

Speaking of torsion bar:

[murpia]

Originally posted by Powersteer
Anyway, its just that torsion bars-beam always seem to be huge.

Not as huge as an unwound coil spring:



It's just a pixel count, but that black bar is the same dimensions as the piece of metal that the coil springs in the picture are wound from.

Regards, Ian

[TheScaleF1Project]

I haven't seen anything in the regs prohibiting it, but have any teams ever used adjustable gas springs as opposed to torsion springs (either coils or bars)?

M