Torsion

[manchild]

I was wondering if instead of double wishbones single longitudinal torsion suspension arms could be used on F1 car. Why? - having in mind simplicity of construction and obvious decrease in number of parts – reliability, less weight, lower center of gravity etc.

I’ve imagined it with torsion bar/axle transversally mounted at the lower front end of the gearbox (bottom). Torsion anti roll bar would also be located near that point to ensure lowest center of gravity. Oscillating arms could be molded with air inlet for brake cooling right on top of it (top of the axle) so that it would pick the coldest air possible (air entering oscillating arms at front end and exiting from it into disc.

Shock absorbers could be placed longitudinally at each side of the engine sump, even moved all the way to monocoque but in each case fixed on one side to engine. I also thought that perhaps oscillating arms could be located as near to gearbox as possible and absolutely longitudinally with extremely short driveshaft.

This would also enable to locate brakes nearer to longitudinal axle of the car. Also some weight could be lost because the external part of the driveshaft could be shorter since there would be no need to fit brakes inside the rim.

If necessary, third shock absorber could be added transversally (above the diff) to connect oscillating arms via universal joints.

Shoot… kill me with comments

Picture is not 100% a scale, especially parts I’ve added!

[akbar21881]

how to control the lateral movement and geometry of the tyres?

[manchild]

how to control the lateral movement and geometry of the tyres?

Well, I wasn’t inventing anything here just wondering if typical suspension with transversal torsion bar could be applied in F1. Same goes for lateral movement – I suppose that it would be controlled the same way it is done on such suspensions on passenger cars.

There is axle on which oscillating arms are connected independently with torsion bar going trough it and connecting oscillating arms on its ends. Bar can be made as two shorter bars fixed to chassis on one end or as a single bar connected with chassis in the middle. The oscillating arms have no lateral movement event though this could be achieved if the axle is bended on each end for a certain degree and same goes for torsion bar.

In general, behavior of such suspension is much different than behavior of double wishbones. Here the wheels have no lateral movement – they are always spanned the same only the distance between them changes. That also means that tyre always touches the surface with same part, except in turns when lateral deformation naturally occurs.

[RH1300S]

What you are proposing is a trailing arm suspension system. I like the idea that the torsion bar and dampers get mounted low down and have compact packging (pretty hot place to but the dampers 'though ).

On the sketch the obvious problem is the bending loads being fed into the drive shaft. You could over come this by angling them outboard towards the wheels.

Problems I can see:

Very little chance of tuning the camber curve to suit the tyre's needs

The height of the roll centre will be fixed (again reduces the designers options for roll axis tuning front/rear)

I can visualise some weird stuff happening when the braking forces (stopping the caliper rotating as well as the directional force) are fed into the arm and then to the chassis - would need a very beefy arm at the outboard end.

Because of the geometry of the arms and driveshaft - this is also effectively a swing axle (you could not locate a CV joint at the outboard end) - so the wheels will move in an arc.

In fact angle the control arms outboard and apart from the spring/damper mounts you have the same arrangement as the Alpine A110

[manchild]

On the sketch the obvious problem is the bending loads being fed into the drive shaft. You could over come this by angling them outboard towards the wheels.

Like this?


I was aware if it but drawn more extreme version with better cog and less weight.

Camber could be easily adjusted if each oscillating arm would be made of two elements connected via adjustable joint (rotated transversally) or you could have oscillating arms twisted for various degrees (wishbones used on F1 are not universal - they are made in various lengths depending on specific track demands etc.)

Perhaps a pipe instead of shaft for outboard axle could deal more successfully with forces with only greater aero drag since the diameter of pipe would have to be wider than shaft’s?

I can visualise some weird stuff happening when the braking forces (stopping the caliper rotating as well as the directional force) are fed into the arm and then to the chassis - would need a very beefy arm at the outboard end. .

Caliper rotating?

Because of the geometry of the arms and driveshaft - this is also effectively a swing axle (you could not locate a CV joint at the outboard end) - so the wheels will move in an arc.

I didn’t consider CV joint on the outboard end – the driveshafts would actually be only those shafts with joints located between diff and the arms.

In fact angle the control arms outboard and apart from the spring/damper mounts you have the same arrangement as the Alpine A110

hehe

[RH1300S]

The biggest problem must be the limitation of roll centre placement - without doing some sketches, I can't quite imagine where it is (or could be), but it will be quite high - approx. level with the gearbox output. I am very sure that F1 cars have very low roll centres, so this may not give good transient behaviour in the car and will force a compromise at the front of the car to modify the r/c height there.

A agree, you can engineer static camber adjustment, I can't visualise how you can get dynamic camber curves.

Funnily enough, I can imagine a system like this being quite good for a mid engined off road buggy with quite long travel suspension........

[manchild]

Here is the sideways sketch. I’m not sure I understood right roll centre placement – were you referring to location of torsion bar?

Regarding dynamic camber – wouldn’t another short torsion bar connecting two-element oscillating arm enable dynamic camber?

[RH1300S]

You dynamic camber idea looks terrible You are attaching the wheel via a spring, so the cornering load will twist the spring and put the wheel into positive camber. That is looking at the sketch as drawn - otherwise it certainly is out of the box thinking

The geometric roll centre is harder to explain. I may attach some pics one day .

When the car rolls in a corner the roll happens along the longtitudinal axis of the car. But where is this axis? The axis of roll is drawn along a line connecting the "roll centres" of the front and rear of a car. The roll centres are "made" by the suspension geometry and can be moved around by adjusting the geometry.

Stick a pencil lenthwise through a box and rotate the box - that is it's roll axis. You can stick the pencil through lower at the front than the back and the box will roll in a skewed fashion along it's axis (which is why some cars lift a wheel cornering). If you marked the box somewhere on it's side face to represent it's CofG and pushed the box around the pencil by it's CofG - by putting the pencil axis higher or lower the box is easier or harder to "roll" (more or less leverage between r/c & CofG). Put the pencil above the CofG and the box rolls into the corner instead of out

It is hard to explain how the roll centre is derived, but an easy one to explain is a beam axle (not De-dion) or a swing axle - these both have a roll centre at the centre of the differential.

For a wishbone setup (or any other, but harder to visualise) - you have to imagine that you project the arms past their mounting points to a point where the angles converge (you make the wishbones into virtual swing arms by doing this) and then project that point back to the contact patch of the tyre, do this for both sides and where the two lines from the tyres cross over (hopefully in the centre line of the car) this is the roll centre. Parallel arms give a r/c height at ground level BTW).

The position of the roll centres and the roll axis is critical to making the car handle well. Hard to explain, but the distance of the r/c height to the CofG at each end of the car is important in this due to the roll couple (pushing the box around a pencil by it's CofG) - if this is wrong as the car rolls into a corner and transfers it's weight diagonally front/rear then the weight transfer can be un-predictable and hard for the driver to cope with.

So, being able to chose exactly where the r/c is at each end is mighty important and I expect you will find that race teams fiddle with this quite a lot during a season.

BTW placing the roll axis on or above the CofG in the hope of getting zero roll or inward tilt causes other problems due to "jacking" which un-loads the tyre (another discussion ).

[manchild]

You dynamic camber idea looks terrible You are attaching the wheel via a spring, so the cornering load will twist the spring and put the wheel into positive camber. That is looking at the sketch as drawn - otherwise it certainly is out of the box thinking ).

Not attaching it via spring, there must be longitudinal axle on which rear element of the oscillating arm would rotate as much as torsion bar swings.
Limiting its motion to prevent positive camber can be done simply by shaping nearing sides of the oscillating arm so that only one-way motion is allowed.

I get now what you meant when you mentioned roll axis
Sorry for making you write the “novel” about it

Still, I don't get why an F1 car couldn't behave well (mechanically speaking) if so many other passenger and tuned cars do with transversal torsion rear axle?

Since you mentioned A110 let me mention R21 Quadra with transversal torsion rear axle and 4WD.

...just came up with this… that green marked anti-roll bar (its bended longitudinal ends) could be a bit longer and make additional “camber bar” obsolete 'cause those ends could replace it

[RH1300S]

Still, I don't get why an F1 car couldn't behave well (mechanically speaking) if so many other passenger and tuned cars do with transversal torsion rear axle?

Easy answer (I hope ) - passenger cars have much higher CofG heights and higher R/C needs (think about it relative to wheel diameter as well) - F1 cars need very low R/C heights.

[manchild]

Here is the version with solution for dynamic camber as I’ve suggested with some simplifications including third shock absorber whose purpose would be taking care of camber related motion. Oscillating arms are shorter (stiffer) since torsion axle has been moved backwards.

[RH1300S]

Never say die eh Manchild?

Nice to see you are manufacturing your own engines these days

The whole thing looks like a structural nightmare, although I find it strangely interesting. I can't quite get my head around what happens with the dynamic camber - if it is a passive system, it looks to me as if you would get positive camber in roll, a bad thing to have.

Shortening the swing arm will create more wheelbase change than the longer arm.

[Guest]

...just trying to see all aspects of idea and develop them (visually)

Those are not my engines but this year's V10 (bought cheaply) with limited revs for 2006 (I can't afford to produce V8 in time)

[manchild]

That was me

[Guest]

I did actually post something like this a while ago.....
with not much suspension movement anyway ,I had proposed to have a single longitudinal link bolted by means of blades to the bottom of the car.
This would be a incridibly stiff thing and you could have it done perfectly for AERO,as there would be the possibility to do away with the keel altogether + the upper wishbones .You ´d basically run the front suspension as a beam axle but in reality that´s the case anyway.If you´re smart enough you could use the blade attachment to control rideheight + Spring stiffness as well.....doing away with a lot of parts....
ok you would still need the pushrods to work the dampers + a steering rack...which brings me to the subject of just why nobody uses a hydraulic steering actuation.(not servo assistance!),as this could remove the need for one link in a very critical flow path....