Adjustable anti roll bars

[EFA]

Davey,

 

if its a VX in the fornt 300's will be fine, poss a bit soft if you are wide tracked - the nose will dip under braking......

 

170's on the rear won't be too lively...... What rear ARB (dia??)are you running??

 

Fat Arn

The NOW PROVEN R500 Eaterid=red>

See the Lotus Seven Club 4 Counties Area Website hereid=green>

 

[wag]

I cannot quite get my head round the effect of having different settings at each side if the ARB. Imagine that the bar is infinitely stiff and you have it set on hole 1 (short) on the N/S and hole 4 (long) on the O/S. Go round a right hand bend and the left spring compresses. The ARB then compresses the right spring more than the left so the car leans to the right, as if it is on banking.

[Alex Wong1697456877]

The resistance to roll comes from the bar, not the arms. The position of the links on the arms just adjust the leverage on the bar.

 

No matter how you set each side, the resistance to roll is always the same on both sides, provided that the links are lengthened/shortened so that no torque goes through the bar at rest.

 

Alex

 

Edited by - Alex Wong on 10 Dec 2001 00:15:00

[Davey Bee]

Hi, Arnie

Thanks for the reply, it is a wide track Vx. I've got 15mm rear adjustable arb. Something I've come across, there are two mounting points on the de-dion tube for the shocks, my car is a very late 1997, the Leda fit in the upper mount which goes through the de-dion do you know Caterham changed the position and what are the advantage/disadvantage of either fitting.

[Alex Wong1697456877]

I think it's down to shock absorber operational length. My late 97 car runs Blsteins through the lower mounting. Roger Swift's Leda shod HPC uses the upper ones.

 

 

Alex

[wag]

Alex, I am embarrassed to argue with someone of your experience. The resistance to roll must come from the arms because the bar is free to rorate in its bushes. 'Why am I here' made the point that if the arms are set on different holes, the bar twists when you go over a bump. This seems relevant but I am not sure why.

[wag]

Alex, I am embarrassed to argue with someone of your experience. The resistance to roll must come from the arms because the bar is free to rotate in its bushes. 'Why am I here' made the point that if the bar is set on different holes each side, it twists when you go over a bump. This seems relevant but I am not sure why.

[wag]

I always wondered why people got double entries.

[EFA]

Right I now have some groundbraking theory for you guys to rip to shreds regarding the earlier cars fitted with Bilstein dampers from new.

 

Earlier Bilsteins (Pre 1996 1 7/8 spring dia type)were seriously over damped and using the 170lb springs Caterham supplied would cause the rebound to be slow to the point that the damper would reduce in operational length either when compressed for longer periods:

 

For ex:

 

In a long high speed corner on a circuit or when the damper was subjected to many fast high shock movements such as hacking down a bumpy country lane.

 

In each of tehse instances the compression (particularly of the front end) would change the geometry out of the desirable range, but also (and particularly in scenario 2) the dampers whould then be hitting the bump stops causing the car to dive off the driven line and become extremely nervous.

 

When new, my car suffered both these precise sympoms. Changing to harder springs allowed the dampers to rebound more quickly to normal operational length and hence the handling was vastly improved.

 

When the 1996 chassis was introduced a number of changes weer made - most significantly to the front end where anti dive geometry was incorporated, plus the dampers were re-valved to overcome the problems above.

 

Now if like me you changes from Bilsteins to some other kind of race damper it was maybe natural to carry over the spring settings that had previously been used ( and factor them at the front for damper angle if you added wide track)

 

So I ended up with 200lb springs on Trackspax at the rear and 350lbs on same at the front.

 

Only now have I realised the Bilsteins were so over damped, Alex's car has re-valved (maybe as new) Bilsteins and anti dive geometry (which is how he gets an optimum setup from low spring rates). ALex also uses stiff ARB's to control roll but retain suppleness I would guess.

 

I've gone right down to 130lbs on the rear of my car, and the front I am contemplating how low I can go before the nose dips under heavy braking - as it is wityh 350lb springs I think it dips at the moment, hence the sump assaults my car has taken recently.

 

So now I'm lost. But could I be right????

 

Fat Arn

The NOW PROVEN R500 Eaterid=red>

See the Lotus Seven Club 4 Counties Area Website hereid=green>

 

[wg_mulholland]

OK, I've just changed my (addmittedly feeble) mind about the anti-roll bar settings argument. Alex- picture this: The car is set on 4, 4 and while going in a straight line the car goes over a sleeping policeman. bth wheels move up say 3 inches, and both roll bar ends rotate say, 30 degrees: result- no rolbar action.

 

Now think about it with a 4, 1 setting. Both wheels move up 3 inches, but the end that is on the one hole twists the bar a greater number of degrees than the one on the (further from the bar) 4 hole. Result- the bar has a twisting force in it which rolls the car.

 

I dont think that this effect will be very large, and probably the advantage of the fine rollbar settings outweighs this- has anyone noticed this effect? Or has my brain letme down?

 

Cheers

 

 

Edited by - wg_mulholland on 11 Dec 2001 14:22:03

[Alex Wong1697456877]

WG_M,

 

You have a very good point. I maintain that setting the bar unequally will result in the same roll resistance to either side. However, you're right that a equal bump to both wheels would result in different movements of the bar (albeit tiny) and hence cause some twist.

 

In conclusion, unless the links are parallel on the right and left, they will always cause a slight twisting action when both wheels move the same distance - so there are only 4 settings, even though uneven settings would cause the same resistance to roll on either side. - UNLESS, 4-1 and 3-2 are equidistant from the pivot point that is the lower end of the link.

 

blush.gif

 

Wag,

I'm afraid I disagree. The bar is free to roll in the bushes but is connected to the other wheel. The overall leverage one side causes on the other is the result of both lever lengths, and the limiting factor as to how much it can lever is in the strength (size) of the anti roll bar.

 

Then again, I could be talking complete ****!! smile.gif

 

Alex

 

Edited by - Alex Wong on 11 Dec 2001 14:56:32

[Peter Carmichael]

I have only skim read the postings so far.

 

The rear ARB is a torsion spring. The arms are essentially rigid.

 

If set differently from side to side, strange jacking effects will happen in pitch (acceleration/braking) and heave (fast crest - all four spring/dampers working in the same sense). In roll, the effects will be even side to side...

 

... unless...

 

... the rear springs are progressive. If they are progressive, the ride height changes as the roll takes place. This will superimpose a weird jacking effect from an unevenly set anti roll bar.

 

I suppose you could get all of this optimised for the offset weight distribution. Some of the effects would also be small and negligible. If anyone can actually report the difference between a 1:2 setting and a 2:1 setting then they are doing well in my book (and are probably good at self-deception).

 

The Juno bar is not affected in the same way because it has even-length levers on each side.

[EFA]

Peter, I think you are wrong. The fact that the ARB can turn freely in its mountings has been overlooked.

 

The postion from the ARB fulcrum point to the position to which the drop link is attached is when loaded, a moment. The moment is equal to the force over the length. On the side where the fulcrum to drop link length is shorter, the turning force is greater, thus meaning the moment is equal side to side.

 

 

Your theory would however be 100% correct if the ARB was unable to revolve in the bushes.

 

Basic physics dear boy.

 

 

 

Fat Arn

The NOW PROVEN R500 Eaterid=red>

See the Lotus Seven Club 4 Counties Area Website hereid=green>

 

[EFA]

And why has nobody chosen to comment on my revalation earlier???

 

Fat Arn

The NOW PROVEN R500 Eaterid=red>

See the Lotus Seven Club 4 Counties Area Website hereid=green>

 

[Davey Bee]

WOW! Arnie I wish I'd gone to school on that day. I don't know if your right, but I'm impressed.

Had my car flat floored by Sebah yesterday, I now need a track day to try it out. it would be nice if I got one in before christmas, any suggestions.

 

Thanks again Arnie and Alex for your help.

 

Dave

[Peter Carmichael]

Arnie you are correct in some of your detail but you are drawing inaccurate conclusions (particularly about my understanding that the ARB rotates in the bushes - you are insulting my intelligence and your comment is your invention and has nothing to do with me).

 

ARB attached to short lever on LHS, long lever on RHS. Now obviously the moments have to be equal and opposite but that means that the force on the smaller lever has to be higher to achieve the same moment.

 

Let's set up an example:

 

Rear suspension on 25N/mm (~142lb/in) linear springs acting directly on the axle, 400mm from the centreline of the car.

 

The rear sprung mass is 306 kg shared equally between the two wheels. That is 153kg per wheel, which is 1500 N.

 

At rest the springs compresses 1500/25= 60 mm. The spring force on each side obviously matches the loading of 1500N.

 

Attached to this there is an ARB with a torsional stiffness of 6000 Nmm/degree attached at the LHS with a 200mm rigid arm and at the RHS with a 100mm rigid arm. The ARB drop links are vertical and are also spaced at 400mm from the centreline of the car.

 

When the car rolls by 1 degree in a left hand bend, the difference in suspension height from side to side is 1*pi/180*800 = 14mm. The LHS extends. The RHS compresses.

 

The angular displacement of the two ends of the ARB have to account for a 14mm difference in the height of the droplinks, 7mm on each side. The angular displacement of the LHS is ~7/200 radians or about 2degrees. The angular displacement of the RHS is ~7/100 radians or 4 degrees in the opposite sense. Overall the bar has been flexed by 6 degrees.

 

The torsional moment in the bar is 6*6000 = 36,000Nmm.

 

On the LHS this is reacted by the drop link with a force of 36,000/200= 180N, pulling the sprung body of the car downwards.

On the RHS this is reacted by the drop link with a force of 36,000/100= 360N, pushing the sprung body of the car upwards.

 

Overall, there is a force of 180N pushing upwards on the car body. This has the effect of jacking the car up on its springs by 180/50 = 3.6 mm (two springs mean 50N/mm overall spring stiffness).

 

The overall displacement of the LHS is 7+3.6= 10.6mm extension. The overall reduction in spring force is 10.6*25= 265N. The spring force is 1500-265 = 1235N. The ARB force of 180N acts to reduce this further. The overall force acting on the LHS of the sprung car body is 1235-180 = 1055N.

 

The overall displacement of the RHS is 7-3.6= 3.4mm compression. The overall increase in spring force is 3.4*25= 85N. The spring force is 1500+85 = 1585N. The ARB force of 360N acts to increase this further. The overall force acting on the RHS of the sprung car body is 1585+360 = 1945N.

 

As a sense check, 1945+1055= 3000N, so the car is in equilibrium.

 

The righting moment on the car is 1945*400-1055*400= 356,000 Nmm.

 

Without the ARB, the righting moment at one degree of roll would be 7*25*400 = 140,000 Nmm.

 

The car has jacked up by 3.6mm.

 

In a corresponding right hand bend, again with 1 degree of body roll.

 

The same torsional moment will be developed in the ARB but in the opposite sense.

 

On the LHS this is reacted by the drop link with a force of 36,000/200= 180N, pushing the sprung body of the car upwards.

On the RHS this is reacted by the drop link with a force of 36,000/100= 360N, pulling the sprung body of the car downwards.

 

Overall, there is a force of 180N pulling downwards on the car body. This has the effect of jacking the car down on its springs by 180/50 = 3.6 mm (two springs mean 50N/mm overall spring stiffness).

 

The overall displacement of the LHS is 7+3.6= 10.6mm compression. The overall increase in spring force is 10.6*25= 265N. The spring force is 1500+265 = 1765N. The ARB force of 180N acts to increase this further. The overall force acting on the LHS of the sprung car body is 1765+180 = 1945N.

 

The overall displacement of the RHS is 7-3.6= 3.4mm extension. The overall reduction in spring force is 3.4*25= 85N. The spring force is 1500-85 = 1415N. The ARB force of 360N acts to reduce this further. The overall force acting on the RHS of the sprung car body is 1415-360 = 1055N.

 

So you can see that the roll effect is symmetrical in left and right hand bends, but that the car jacks up and down depending on which direction you are turning.

 

[EFA]

1st para teeth.gifteeth.gifteeth.gifteeth.gifteeth.gif Gotcha!

 

"So you can see that the roll effect is symmetrical in left and right hand bends, but that the car jacks up and down depending on which direction you are turning."

 

No!! The other end of the ARB twists.

 

So anyway, you're not so busy today then......teeth.gifteeth.gifteeth.gif

 

Fat Arn

The NOW PROVEN R500 Eaterid=red>

See the Lotus Seven Club 4 Counties Area Website hereid=green>

 

[Peter Carmichael]

Arnie. You are wrong. Read it again until you understand. Stay behind after school.

[Blackbirdman]

Arnie

 

The twisting of the ARB is not the only factor here. To reduce the roll of the sprung body the ARB needs to act on the body. This is done through the ARB bushes.

 

Let's do a thought experiment. Take the ARB off the car and hold the bar section in both hands with your hands at the positions of the bushes. Now get someone else to move the arms of the ARB. If both arms are moved up or down the same amount you will feel the bar rotate in your hands but there will be no vertical forces.

 

Now your little helper moves the arms in opposite directions by holding the very ends of the ARB arms. You should now feel a vertical force as you try and hold the bar still. One hand will be pushed up, the other down by equal amounts .

 

Now get your helper to move one hand half way along one ARB arm and push the arms in opposite directions. To do this they will have to push twice as hard with one hand as the other. (If they don't the bar will rotate instead of twisting). You will feel one hand pushed up (or down) harder than the other. This is the jacking effect Peter is talking about.

 

biggrin.gif

 

 

 

Is it a bike? Is it a car? No it's Blackbirdman cool.gif

 

Edited by - blackbirdman on 12 Dec 2001 16:13:42

[Peter Carmichael]

So I am suggesting there are only five settings that purely provide an anti-roll effect:

 

1. disconnected - just at one end; let's not be picky.

2. 1-1

3. 2-2

4. 3-3

5. 4-4

 

All the other combinations are distinct and will have jacking effects.

 

The jacking also affects the damping in roll. The damping is achieved from one damper acting in rebound and one in bump.

 

In the example I gave, in the LH bend:

- the LHS sees 10.6mm rebound travel

- the RHS sees 3.4mm bump travel

 

In the RH bend:

- the LHS sees 10.6mm bump travel

- the RHS sees 3.4mm rebound travel

 

On standard Caterham Bilsteins, the bump:rebound ratio of damping force is 7:1, so you will be getting significantly more roll damping in turn-in to right hand bends than turn-in to left hand bends.

[garyedwardc]

If the car is jacking up and down and you don't have Watts linkage you in different doo doo depending on which way you are turning. I suppose this could be of benefit on a predominantly right hand cornered track but sounds a bit dicey on the road to me.

[Peter Carmichael]

Yup. Different amounts of rear steer depending on whether the car jacks up or down.

[Barry H]

Fascinating. I think I understand about 10%.

 

However, with an adjustable front arb where the adjustment is only possible on one side, is there a jacking effect here also?

 

I always assumed not because the arb rotates and because it is not a function of length but of the stiffness of the adjusting arm. But now I have doubts.

 

Any thoughts?

 

Barry

 

 

Edited by - barry h on 12 Dec 2001 15:46:03

[Peter Carmichael]

With the Juno adjustable front ARB, the bushes again allow free rotation, but restrain displacement. This means that on the central portion of the bar the moment is the same end to end. Because the levers are the same length (but different stiffness) this means that the force reacted is the same (but opposite) end to end. i.e. there is no jacking effect.

 

The amount of rotation in the bushes will be different side to side, but the forces will be the same. The principles of the Juno bar are sound.

 

The slightly bizarre thing about the Caterham front ARB arrangement is that the levers are angled out from the car, so a component of the force produces a bending moment on the central portion of the bar as well. This is reacted by the bushes and does not get transferred to the other end of the bar. This has no affect on the anti-roll characteristics but explains how the forces reacted in the bushes come to be bigger than the forces at the wishbones. eg. the bushes are closer together than the ball/sockets, so to react the anti-roll moment into the body, the forces at the bushes have to be larger.

 

[...998...]

 

[Barry H]

Peter, thanks for the explanation and confirmation.

 

Barry