Do you really need a front anti roll bar ?

[edmandsd]

With a l/weight car powered by a K or bike engine is the front anti roll bar really necessary ?

Has anyone ever discarded it and made top wish bones that mount at the front (as opposed to the back) of the antiroll bar mount through the nosecone ? This would give you an upper and lower wishbone of similar dimensions. You could then have a drop link front anti roll bar if required.

[Rob Walker]

The ARB is necessary if you do not want to run with very very stiff front springs. One of our club members prefers to run with 375lbs front springs and no ARB. The majority prefer a balance between ARB and sensible spring rates.

[paul jacobs]

I'm not so sure that heavier front springs are that bad Rob, there appear to be quite a few advantages and very few disadvantages.

 

Now, having said that, I have to say that all my experience with harder fronts are with heavier Vauxhall engined cars, but I would have thought that it is only a matter of degree, eg say 400lb fronts on a Vx and 350lbs on a K.

 

You could be forgiven for thinking that the ride would be rock hard, and the front end would be thrown all over the place every time you went over a bump, but that doesn't seem to be the case. As long as the rear end, over which you sit is reasonable, say 180lb, then a hard front isn't too noticable [as the actress said to the Bishop!]

 

Benefits include smaller or no front a/r bar therefore "purer" suspension inputs, less weight, less dive when braking, more stable suspension geometry, better turn in and a generally taughter car.

 

Frankly I'm not that convinced that the front a/r bar actually does a lot for the roadholding generally anyway. A couple of years ago, I was at a test day and one of the Vx engined HPC cars had a front a/r bar link break and the driver, who is an experienced track pilot, didn't notice the difference until someone pointed it out in the pits. I have to say that everyone showed surprise.

 

The easiest thing to do is to disconnect it and see what the car feels like, it certainly won't be a huge difference and it won't bite, it'll all be pretty gradual.

[edmandsd]

Thanks Paul, Rob - I'm also not convinced that the front arb adds a lot.

[Peter Carmichael]

In case you don't want to read to the end, the correct answer is probably a moderate set of springs and a moderate ARB. I just want to quash the notion that ARB's are evil, which is nonsense.

 

The suspension is a sprung/damped system which copes with four modes of displacement. Pitch, roll, heave and warp.

 

Any spring/damper arrangement has a characteristic of response to a step input - frequency of oscillation and decay. For a car we can define desirable oscillation and decay characteristics for each mode. For heave (a high speed crest/compression), we want a 1.2-1.4hz oscillation and 0.75-0.85 damping ratio. For pitch, we want a higher frequency, say 3 hz, so the car settles quickly under braking/acceleration. For roll, we want a higher frequency of again ~2-3hz so that the car responds quickly to steering inputs. For warp (one wheel hitting a bump) we want a slow frequency response that soaks up the bump and doesn't disturb the car.

 

The frequency and decay in linear modes (heave) are characterised by the mass, spring stiffness and damper constants.

 

The frequency and decay in rotational modes are characterised by the "distribution of mass away from the axis of rotation", the spring moment, and the damping moment.

 

There was a significant differentiation made in the last two statements. The mass per se is not important in the pitch and roll modes. Lets take pitch as an example and picture two similar 660 kg Sevens (with drivers).

 

15kg at each corner are unsprung and of the remaining mass, 55% (330kg or 165 per corner) is over the rear axle. 270kg is over the front axle. We can choose spring rates and dampers to give a 1.4 hz well damped ride in heave for the 600kg sprung mass. The difference between the two cars is that one has been modified to mount the engine 4 inches further back in the chassis and has had a saddle fuel tank fitted further forward in the chassis. Because of this, from side on it less resembles a dumbell and it has its major masses concentrated in the centre of the car. This means that for the same springs and dampers, it has a higher pitch frequency. Without detail analysis I am not going to guess what happens to the damping, but it doesn't stay the same.

 

In roll, you have less opportunity to centralise masses, but you can adjust the roll axis to make the rolling moment of inertia greater or smaller giving the same effect.

 

My point is that you have a sprung/damped system. Just because you have the spring rates and damping right for one mode of oscillation does not mean they are right for any other. The most common scenario is that the roll mode needs more spring stiffness and this is what anti-roll bars are for. Just because the ARB is an additional spring without additional damping does not mean that the damping characteristics become worse.

 

An ARB has one certain side effect. It affects the warp mode as well, raising the frequency. This makes the ride quality worse.

 

These sprung/damped systems also have a frequency response. Typically they are low pass filters. If you park a car on a slope (0hz), it tips over. If the disturbance of the suspension is momentary, you expect a small response of the sprung body. This also has an effect on how the car feels to drive. A steering, braking or accelerating input will be ignored if it is a momentary disturbance and the suspension response filters out disturbances of such high frequency. The stiffer you go, the more the chassis will respond to small momentary disturbances and be able to respond to small momentary inputs.

 

The filters have desirable characteristics for mopping up the undesired disturbances and letting through the wanted inputs. When you are driving along a road, the bumps and crests develop a frequency depending on how fast you go. A soft sprung car will hit a resonant frequency at lower speed than a stiffly sprung car. The flip side of stiff springing is the larger forces involved. The limit for a car is that it only has its own weight holding it onto the ground, so a stiffly sprung car has less ability to stay on the ground than a softly sprung car. The smoother and faster you go, the stiffer the suspension can be.

 

So the flip side argument is that if you don't fit ARBs and you try to achieve the same roll stiffness with the springs, you will have less grip over bumps.

 

All roads are bumpy. Tracks are bumpy too - a crest or a change in surface is enough to count as a bump. A kerb is certainly bumpy.

 

So can we please have less of the "anti-roll bars are bad" dogma and move on to taking a balanced view of the entire suspension system, which means non-extreme springing and a balancing contribution from an sppropriate ARB.

 

Remember, Formula 1 cars run ARBs. They wouldn't do that if there was any technically better solution.

 

[as an aside, you can provide decoupled suspension allowing independent tailoring of suspension characteristics in each of the four modes. To do this you need a lot of hydraulic components adding complexity and weight. In theory it would be possible to have the ride of a Rolls Royce and the cornering of an F1 car all from passive components.

[Marius]

Jesus....

 

teeth.gif

 

Marius

[Eugene]

The original Elan had no anti-roll bar and handles great... I did a lot of research on roadholding and handling a while back and the info can be seen on my website - link below. I have gone for stiffer springs and a light bar, although I would have liked a lighter one. The biggest difference is with bumps and pot holes as a change in wheel will be reflected in the other via the bar, so causing the car to move unnecessarilary...

 

Small Boy - with Loud Toy!

See Eugene here

and

Lotus@Herts hereid=green>

[Andrew Russell]

Marius, I think you can just call him Peter on these pages.....

 

Andrew

[edmandsd]

My car's corner weights in kg's (without me in it - 80kgs) from d/side front anti clockwise are as follows 110,118,101,99 (Total 428kgs) if that makes any difference ?

[Peter Carmichael]

Dave,

 

You have to look at where the individual masses are compared to the roll axis...

 

...erm. Or suck it and see, just like everybody else. I wasn't trying to come up with a recipe I was just trying to say that an ARB should not be deprecated as a matter of course.

[edmandsd]

I hadn't completely disregarded the use of a front anti roll bar and appreciate your techical advice in this regard Peter.

I'm always considering options in relation to every aspect of the car and this is a good arena in which to do so.

[wg_mulholland]

Peter,

My understanding of the reasons for f1 cars running front and rear arbs are for speedy tuning of the front rear roll couple. The gould hillclimb car runs no rear anti roll bar- traction is paramount.

 

Also, the differences between low and high speed damping are adjusted to deal with the diferences between roll and pitch and step input- roll bars are a last resort as they transfer load from the inside wheel in cornering and overload the outside- to the detriment ofoverall traction.

 

Big rollbars are often a sign of desperate problems at the other end of the car.

[Peter Carmichael]

WG,

 

You have phrased a classic ARB myth. "ARBs rob you of grip."

 

How exactly do you counteract roll without having more force on the outside wheel than on the inside? Do tell.

 

Your last comment points to a truth. Increasing the roll stiffness with anti roll bars at one end of the car transfers grip to the other end of the car. The overall grip can be made to go up because the contact patch can be maintained flatter against the road surface.

[wg_mulholland]

Peter- to quote Caroll Smith (tune to win p38):

 

"The roll couple will be resisted by the suspension springs and by the anti roll bars. The greater the resistance the less roll will result" note- less roll, not more force- the same force is provided by a soft pring deflected 2 inches as a hard one deflected 1 inch. Hence no change in load transfer from increacing spring resistance to counteract roll. Hope that clears that up.

 

 

"Because (a roll bar) is a direct physical connection between the wheels, increacing roll resistance by the bars will both decreace roll and increace lateral load transfer (at that end- and hence reduce grip at that end)" Granted, if you put an equally stiff bar at the other end then you will not change the lateral load transfer picture, but you will start to feel that the car has become very slidy- it will have lostroll sensitivity due to the lack of independance and increacedweel rate in bump (during roll).

 

Also, it must be remembered that the anti roll bar works almost directly onto the wheel- and so it is undamped- how stiff you have to go before this becomes a problem, I don't know.

 

Bill

 

Edited by - wg_mulholland on 22 Feb 2002 18:36:08

[Morls]

Just as a thought, there is a further complication / consideration;

 

Given that the resistance to roll is, in an ideal world, controlled seperately by the ARB rather than the springs, then are the dampers not also damping the forces stored in the ARB?

 

Does this not spoil the idea of seperating resistance to roll from control of pitch / absorbtion of bumps?

 

Assuming that ARBs are "A Good Thing", then would controlling the rate, rather than just the degree, of roll through an Anti Roll Damper not also be a good thing?

 

Is this done already on anything?

 

 

 

Mark

[wg_mulholland]

If catch your drift then:

 

A third spring system isolates roll from dive- and as all three can be damped then this effectivly provides alternative roll and dive damping (as it does wheel rates).

 

However, in the successful 3rd spring systems I've seen, there are additional (very low resistance) arb's so this seems to defeat the point in the damping sense. Also, at least two of the systems ran without third spring dampers- which seems even crazier! They worked well though....

[wg_mulholland]

I've just been rereading peters posts, and it seems that we have both come to the sam conclusions from different stating points.

 

I believe that roll bars are neccessary due to the different springing characteristics needed in roll and dive.

 

However, I prefer to keep them on the light side of medium (and the springs stiff) due to their un-damped nature,the loss of sensitivity, and the loss of idependance inherit in their use.

 

Peter seems to be saying that he prefers to use anti roll bars to cover the neccessry differnce in spring rates in roll and dive, and that he finds it faster to live with te three problems and exploit the better characteristics of stiffer arbs in achieving a better roll/dive spring ratio.

 

It is very much a robbing peter to pay paul situation. As peter suggests, total load transfer (sum of front and rear)is unnafected in both cases. However, changing arb rates does ffec the load transfer detrimentlly at the end you stiffen. So if you have a good balance, and you want to try stiffer, you need to stiffen both ends and tune untill you find the balance again.

 

 

[wg_mulholland]

When I typefast on he internet, my computer seems to miss out the odd letter- any ideas why? (all ten fingers are in place).

 

Bill

[Peter Carmichael]

Bill,

 

How exactly do you counteract roll without having more force on the outside wheel than on the inside? Do tell.

 

"The roll couple will be resisted by the suspension springs and by the anti roll bars. The greater the resistance the less roll will result" note- less roll, not more force

 

Same force not less force nor more force, regardless of whether an ARB is involved. So we agree then.

 

Your second quote I also agree with. Roll stiffness is used to tune the relative grip of the two ends of a car. By increasing the roll stiffness either with ARBs or with stiffer springs at one end of the car, you borrow grip from the other end of the car.

 

The bit I disagree with is:

Also, it must be remembered that the anti roll bar works almost directly onto the wheel- and so it is undamped

 

Well I would only amend that by thinking of the ARB as a contributor to a system that contains some damping. It is impossible to exercise the ARB without the wheel dampers moving and providing damping. The question is whether the combination of springing (from springs and ARB) and damping in roll is appropriate for restraining the roll moment of inertia with a desired frequency and an appropriate damping ratio.

 

My hypothesis is that without the ARB, the roll mode is hideously overdamped and therefore adding an ARB is meet and optimal.

 

I am not an advocate of dominating the roll situation with ARBs. Balance is the key. I just wanted to gild the reputation of ARBs as they are an essential component in the chassis tuner's armoury.

 

Edited by - Peter Carmichael on 22 Feb 2002 19:18:43

[wg_mulholland]

Quote

"Well I would only amend that by thinking of the ARB as a contributor to a system that contains some damping. It is impossible to exercise the ARB without the wheel dampers moving and providing damping. The question is whether the combination of springing (from springs and ARB) and damping in roll is appropriate for restraining the roll moment of inertia with a desired frequency and an appropriate damping ratio."

 

Yes, I agree. But if we look at a stiff arb/soft springs and stiffsprings/soft arb, then if they are set to have equal amounts of roll, and similar dive damping, then the stiffer sprung car will have a higher amount of roll damping- hence m calling the rolbar "undamped".

 

However,if we don't stiffen the damping when we increace the spring rate then the roll damping stays the same a the thick bar, but the warp/dive damping is proportionally reduced.

 

Which is best- highe or lower roll damping? I'm not sure its that simple as roll is usually(!) a low speed damping charactristic (like dive), but warp is high speed.So the trade off is really between roll and dive. And as I said before, I've seen very successful cars running with very low dive damping.. But the converse is also true!

 

[AMMO]

Eugene

 

I popped in to Elan restorer Mick Miller's today for a cuppa and a chat. Had a squint at a car on stands. The Elan has a front ARB mounted to the bottom of the front shocks. They have always had this according to Mick. No rear ARB though.

 

The Elan also shares front brake disc, calipers, uprights and trunnions with the earlier Sevens.

 

AMMO

[Lawrence_Z]

PC quote

My hypothesis is that without the ARB, the roll mode is hideously overdamped and therefore adding an ARB is meet and optimal.

 

Peter

does this theoery suppose equal high speed and low speed damping?

 

 

Lawrence

[Peter Carmichael]

I was under the impression that you mostly want roll to be controlled by the slow speed damping.

[Peter Carmichael]

Yes Bill. Again we agree.

 

I still think you are using the term "undamped", when really it is just "less damped". My suggestion is that "less damped" is a really good direction to be heading in because for all reasonable spring rates and damping ratios for the other modes, the roll mode is overdamped.

 

In none of this have I suggested that a big fat ARB is the way to go. I am very clear that it is a matter of degree. My point all along has been that stiffening up the springs can have worse consequences than mixing an appropriately sized ARB into the solution.