Board logo

Front Anti - Roll Bar Question
omega0684 - 18/6/11 at 06:58 AM

Morning All,

How do you know what length a front anti roll bar needs to be? What material are they made from? Angle of bend in them, torsion rating etc?

Thanks Again!

Alex

[Edited on 18/6/11 by omega0684]


Richard Quinn - 18/6/11 at 07:46 AM

Length and angles tend to be to suit the car so that the ARB can be mounted in brackets/clamps on some convenient part of the chassis and the ends are appropriately located for attaching drop links to the wishbones. The effective length and angles will have an influence on the way the ARB reacts in torsion so it really needs to be designed as a "package" for your car.
Why do you want / need one?


omega0684 - 18/6/11 at 08:00 AM

quote:
Originally posted by Richard Quinn
Why do you want / need one?


getting quite a bit of front end roll on the track.


matt_gsxr - 18/6/11 at 08:00 AM

quote:
Originally posted by omega0684
Morning All,

How do you know what length a front anti roll bar needs to be? What material are they made from? Angle of bend in them, torsion rating etc?

Alex

[Edited on 18/6/11 by omega0684]


same length as that piece of string!


The design sequence is:
1) work out how much force you want the thing to apply (at the wheel) per degree of body roll.
2) design something that supplies that force and fits. Length, material properties (mild steel, or something more exotic, solid or hollow), arm length (adjustability), etc are all free parameters as they inter-relate.

or

Copy someone who has already done the design work.


Mark Allanson - 18/6/11 at 08:08 AM

Or, being really practical, go to a scrapyard and look at the REAR suspension of many modern cars.

You will find a small ready made ARB, modify to fit, adjust to suit after driving for a bit. It will cost you in labour but very little in money.


Ivan - 18/6/11 at 09:32 AM

I agree with Mark above - don't over think it - try a cheap bar like he suggests and if it is too stiff or not stiff enough once tested you know what to go for the second time round.

The proviso is that you approach the limits carefully when testing (on track not a public road) in case you have induced chronic over/under steer.


procomp - 18/6/11 at 10:23 AM

Hi

If the basic car is rolling too much then you should be looking at the spring and damper package first. An anti roll bar is there to add fine control. If you fit an anti roll bar strong enough to eliminate the major rolling then it will be far to strong and take away a great deal of the independence of the suspension setup.

Thing to remember is that no kit manufacturer supplies dampers that are specifically tailored for there car. They are just off the shelf wrongly valved dampers for a car far heavier than the application they are being used for on the average Kit car.

Cheers Matt


omega0684 - 18/6/11 at 01:31 PM

thanks for the advice guys,

i have just converted from GAZ shocks to Protechs, but I'm also running softer springs as I thought that the ones supplied from DAX were very hard. the track day was a tester day for the new shocks and although the handling of the car certainly improved as we made a few tweaks here and there, I have no idea whether it was correct or not, the car just felt better to drive as the day progressed,

Matt: I have been meaning to book in a session with you for a while but work commitments are prohibiting progress and the weather is pants so I can't drive it over to you in the rain (no roof) I'll have to book a days holiday & when the weather brightens up i'll get the car booked in with you'll have to give it the once over!

Bang Tidy!

Al


907 - 19/6/11 at 05:06 AM

quote:
Originally posted by Mark Allanson
Or, being really practical, go to a scrapyard and look at the REAR suspension of many modern cars.

You will find a small ready made ARB, modify to fit, adjust to suit after driving for a bit. It will cost you in labour but very little in money.



Hi Alex.
I agree with Mark.

The MX5 one looks handy @ 34" wide.

I believe an ARB is made from spring steel, but the drop links could be home made.
If rod ends were used then adjustability could be built in.

Cheers,
Paul G

MX5 rear ARB
MX5 rear ARB


Ivan - 19/6/11 at 06:52 AM

What Procomp says is correct (of course) but you have to consider your major use and needs - if ride comfort is important then possibly don't go for harder springs, instead a roll bar will help - just don't go too stiff.

The other question to ask yourself is "Is the roll seriously impacting on your fun?"

The more you want to orient the car towards track use, the less usable it will be on the road and the less time you and yours will want to spend in it so take your choice. Also hard springs make handling on what are often bumpy public roads a nightmare - I am sure that many of the hedgerow episodes we have seen on this forum can be closely linked to setups and tyres more suited to the track than bumpy public roads that are often damp and have paint and manhole covers thrown into the mix.

I am going through this very process with my Cobra and have decided to get it as sweet handling on the road as I can by dropping it to 4.5" from 6" at the front and to 5.5" from 6.5" at the rear, not changing the spring rates but upgrading shocks and fitting a light anti roll bar at the front and of course fully sorting camber, toe etc. This way I will retain most of the comfort when cruising but improve track day predictability and fun a lot (I hope - time will tell). I have accepted that this way I will never be the fastest on the track but hey - track days are about fun not winning - If you want winning then race don't do track days.

Also softer springs (not too soft) help acceleration by allowing more weight transfer to the rear which is the other thing Cobras are about, and give a lot more warning when one is close to the limit on corners so suit my skill level better as well.


Sam_68 - 23/6/11 at 12:01 PM

I'm firmly with Procomp on this one.

Perhaps surprisingly for some people, the primary purpose of anti-roll bars is NOT to reduce body roll; it's to balance the proportions of front:rear roll resistance, thereby controlling diagonal weight transfer and so understeer/oversteer balance.

Anti-roll bars are a crutch for the incompetent suspension designer and, if it's essential to fit them, should be kept as light as possible.

Most roll resistance should be from the main springs and, if properly damped, there's no reason a car with a CoG as low as a 'Seven' should have unacceptable body roll or unacceptable ride quality.

...but to answer the OP's original question, the calculations to work out anti-roll bar stiffness are available in Allan Staniforth's 'Competition Car Suspension' (or a number of other books, but Staniforth's also has a chapter on weight transfer that is the best basic guide I've come across to explaining their true purpose.


The advice to go to a scapyard and pick up whatever random OEM-fitment anti-roll bar falls to hand is very bad indeed - it's a quick way to screw up your handling (and possibly a quick way to the cemetary), if you don't know what you're doing.


MikeRJ - 23/6/11 at 01:33 PM

quote:
Originally posted by Sam_68
Anti-roll bars are a crutch for the incompetent suspension designer and, if it's essential to fit them, should be kept as light as possible.



Is that why virtually every modern production car has at least one and commonly two ARBs?


Sam_68 - 23/6/11 at 05:13 PM

quote:
Originally posted by MikeRJ
Is that why virtually every modern production car has at least one and commonly two ARBs?


Yes.

Although in their defence, most modern production cars are have a weight distribution that is heavily constrained by packaging (Lotus engineers tried to design the Elise without ARB's and just couldn't make it work due to the high, rearward-biased CoG dictated by the transverse mid engine installation, for example) and bean counters.

There's no excuse for heavy ARB's on the sorts of cars being discussed on these forums, though light, adjustable ARB's are a useful device for fine tuning.


britishtrident - 26/6/11 at 06:03 PM

On road cars the use of anti-roll bars is to a large extent driven by ride quality considerations.
Also with cars that have the majority of weight at one end then often the only alternative is to deliberately under design the suspension at the lighter end ie as on the front end Hillman Imp or the rear of original Mini.


Antnicuk - 26/6/11 at 07:31 PM

The caterham R500 runs a front ARB, not sure about the rear. That is one of the lightest road going cars and one of the best handling so they cant be that bad for the car.


ss1turbo - 28/6/11 at 07:54 AM

I understood ARBs to be a compromise to gain roll stiffness but without increasing the spring rate to the point of non-compliance. Make the spring rates too hard and it becomes unsafe on the road and not too good on the track either. Damping is normally underrated IMHO and people confuse stiff damping with stiff spring rates - Mr. Chapman had it right (soft springs, hard damping which reduces rapid weight transfer which causes loss of control). The ARB brings back the roll stiffness without losing the wheel-to-road compliance.

If you're sure it's a lack of ARB, then try the cheap 'n' cheerful test - strap another one of the original type to the one fitted (available from various sources) and see if you really need a thicker ARB, or just need to revisit your spring/damper settings.

I'm no expert on suspension by the way, but those I know who have spent far too many hours researching all seem to refer to Stainforth's "bible" and have got the results they want from their experimentation so there must be something in that book!

There is a point where spring rates become too hard (loss of traction is the main bugbear) in order to try and keep roll stiffness and a few people I know have dialled back their spring rates, then increased damper and ARB rates instead and have gained several seconds (and greater confidence due to increased stability - although it may still roll) that way as they also thought that the only way was stiffer springs. It's all about balance...and one of the best gauges for that is your backside..


britishtrident - 28/6/11 at 11:13 AM

Interestingly fitting a rear anti-roll bar on some suspension geometries (ie those with a lot of roll steer) can have the opposite from expected effect. This was the case with a lot of cars with leaf springs at the back.


Sam_68 - 28/6/11 at 11:48 AM

quote:
Originally posted by ss1turbo
I understood ARBs to be a compromise to gain roll stiffness but without increasing the spring rate to the point of non-compliance.


They are, but they're a bad compromise (in that they have some badly negative side-effects) and they're only one of many possible factors that you can juggle to reach an overall compromse. If you know what you're doing (and provided you aren't too heavily constrained by other requirements, as Lotus were with the Elise, for example), you can reduce the requirement for an ARB to little or nothing, which is why I made my statement that they're a crutch for incompetent suspension designers (and I find a certain irony in the fact that both Antnikuk and MikeRJ have taken me to task for that comment whilst owning cars designed to function very well indeed without ARB's, by one of the best designers in the business!).

As an example (and I should stress that I am not suggesting this as a solution - just as an example of one of the many factors involved), if your geometric roll axis is coincident with your CG, you won't get any roll at all, regardless of how softly sprung the car is (but you'll then have some work to do to overcome the jacking effect and achieve acceptable wheel geometry if you use independent suspension).

The Lotus principle of soft springs/firm damping is often taken too literally, though: on very lightweight cars like Sevens, far too many people see this as an instruction to run the dampers rock-hard, whereas many cheap adjustable dampers are really too stiff even at the bottom end of their adjustment range.

Of course, if you have a basically flawed design and you're not in a position to re-engineer it from scratch, you're sometimes stuck with heavy ARB's as a quick-fix bodge solution, but I'd always try to look at them as a last resort.


britishtrident - 28/6/11 at 12:26 PM

On classic Lotus models the damping never was as firm as most seem to think, the springs were very soft with loads of pre-load and the dampers were "effective" rather than firm. All classic models used front anti-roll bars but they were relatively soft.

One of the advantages of front anti-roll bars from the road car designers point of view is they allow the front suspension to be softened lowering the suspension natural frequency. Having a lower frequency on the front suspension than the rear gives a more level ride ie it reduces porpoising as the car passes over a bump.


Antnicuk - 28/6/11 at 09:50 PM

Thanks for the reply Sam. I first tested my car without any ARB's as thats how the car came and others on the JP forum said they have run successfully without them but i found the more power and grip i had the more roll i had (now at over 400 hp and 8 inch wide slicks,) fitting adjustable ARB's has helped that, certainly from pictures of the car and the feel and controllability. Unfortunately i dont know if it/we are going any faster but then i am not a good enough driver to be consistent enough around a track to make some tweaks, time, more tweaks, time again as my lap times would be different every time without any tweaks But i like the way it feels now and corners much flatter. At my next track day i will disconnect both roll bars and see if i can notice any major differences. I spose when JP designed it they only 100 hp and road tyres, anyone know?


ss1turbo - 29/6/11 at 01:04 AM

quote:
Originally posted by britishtrident
On classic Lotus models the damping never was as firm as most seem to think, the springs were very soft with loads of pre-load and the dampers were "effective" rather than firm. All classic models used front anti-roll bars but they were relatively soft.



A bad choice of words on my part originally - "effective" does get interpreted as "firm" where damping is concerned.

Sam_68 - I agree that with a blank sheet of paper, a suspension system can be effectively designed that negates the need for them. If you then increase the work that (previously optimised) system needs to cope with (i.e. a 4-fold power increase) then that system may not then be suitable and requires revisiting the drawing board in an ideal world - and i'm not talking just about changing spring/damper rates.

Most of us are tied to a certain layout/construction of suspension whether due to space, availability or cost constraints such that ARBs can allow that tuning/tweaking of a system that would otherwise be unsuitable. Agreed - they can have an adverse effect but I think thats more down to the original design rather than the ARB itself (as britishtrident mentioned) - those with decent location seem to benefit whereas those where the wheels do "float" certainly don't!

There are more variables than I care to think about but as Antnicuk says - inspriring confidence in your steed makes a big difference. A much stiffer setup than i'm prepared to live with may (probably will) allow for quicker times and higher cornering speeds but i'd rather enjoy a corner slightly slower but feel more in control than hanging on to something skipping around a corner considerably quicker. I deliberately don't do racing/sprinting etc as I know where i'll end up and I prefer to enjoy my cars on the road.

Suspension is always a compromise one way or another; it just depends on which aspects one is happy to trade.


Sam_68 - 29/6/11 at 11:58 AM

quote:
Originally posted by AntnicukI first tested my car without any ARB's as thats how the car came and others on the JP forum said they have run successfully without them but i found the more power and grip i had the more roll i had (now at over 400 hp and 8 inch wide slicks,) fitting adjustable ARB's has helped that, certainly from pictures of the car and the feel and controllability.


Bit of a thread derailment, for which I apologise to the OP, but:

Don't I recall that you were having problems with excessive dive/squat with the car?

Maybe I got the wrong end of the stick, but when we were discussing your dive/squat problems, my recollection is that you were otherwise happy with the other handling characteristics of the car (including roll), and I didn't pick up on the fact that you had fitted front and rear ARB's to deal with earlier problems with excessive roll.

At risk of stating the obvious, even the stiffest anti-roll bar will have no effect whatsoever on increasing resistance to dive/squat, so if I had a car that was rolling too much and pitching too much, common sense would suggest that you need to increase overall spring rates, not bolt on an extra that only functions to reduce one part of the problem?

To forestall the obvious objection that increasing overall spring rates will automatically result in an unacceptable ride quality...

As I'm sure most people will be aware, when considering the response of dampers, it is normal to consider high speed (in terms of damper shaft velocity) damping and low speed damping separately. On really top-notch adjustable dampers, you will get separate adjusters that allow you to vary these separately, but even with cheaper dampers it is often possible to re-valve the damper or swap the piston design to tailor the high speed and low speed responses to suit your requirements.

Obviously, pitch and roll movement happens relatively slowly, hence is resisted primarily by the low speed damping response. Damper velocity when dealing with surface imperfections, however, is usually at the high speed end of the range, and it's usually excessive high speed damping that's the worst culprit in wrecking ride quality... you'll usually find that the effect of increasing spring rates doesn't have as dramatic an effect on ride quality as you might think, provided the damping is correctly matched to the new springs.

As an example, the spring rates on the Elise S2 were increased over those on the S1, yet the ride quality noticeably improved, because new, properly valved (Bilstein) dampers were fitted that had a better match of high-speed response to the springs than the off-the-shelf Konis that Lotus had been forced (for reasons of development budget) to fit to the Series 1 cars.

If you're happy with the roll, now that you have fitted ARB's, and you're happy with the ride quality, but you're still experiencing too much dive/squat, the logical conclusion may be to remove the ARB's, increase the overall spring rates to compensate, and re-valve/adjust the dampers to match the new springs (perhaps with proportionately slightly less high-speed damping relative to the spring stiffness, to compensate for the slight reduction in ride quality that the increased spring rates will impose)?

quote:
originally posted by ss1turboSuspension is always a compromise one way or another; it just depends on which aspects one is happy to trade.

Absolutely!

But simply bolting on whatever conveniently shaped ARB you manage to pick up from the local scrapyard, as was suggested by a couple of posters above, is very unlikely to result in the optimum compromise, unless you have an unnaturally large amount of pot-luck!


mikeb - 29/6/11 at 12:38 PM

Interesting thread this one.

You often see Caterham front ARB's for sale, they come in a range of thicknesses.
From recollection the orange bars are the softess.
If I didn't have the time/knowledge to do all the calculations this might not be a bad starting point for a front ARB.

Just fitting a front ARB will give you sharper turn in but more understeer than your orginal set up with less roll. Caterham R300's run front and adjustable rear ARB.

We use them in racing to fine tune the handling balance as Sam said, on some circuits teams will run no bars front and rear. I have some great data showing where time is won and lost with a ARB/no ARB set up for Snetterton.

For my Haynes build I was looking to makesure I could fit a front ARB before I finish my chassis, I'll then run it stock with no ARB's, apparantly this is good for road driving. However on the haynes due to limitations of the sierra upright at the extremes of roll, the camber angle of the front wheels is less than desirable, the quick fix is to limit the amount of roll by fitting an ARB, the better fix would be redesigned front upright/wishbone geometry.

Good luck and if you do go trial and error take it steady.


Antnicuk - 29/6/11 at 02:30 PM

Sam, I was having excessive squat but i have slowly dialed a lot of this out by changing springs in 25lb increments until i was happy with the results. Included in this was playing with the adjustment on the dampers. The main problem i have is I only get on track every 6 months or so if i am lucky. The set up i have at the moment is a little firm for the road but not to the point of being too skittish. Any more and i think it would be. but you are right, i was expecting a much bigger compromise in ride comfort when i started increasing spring rate but it wasnt anywhere near as bad as i thought. I went up over 100lbs in the end.


From pictures and feel of the car, with just a front ARB, but the back still rolled a lot, although it was very controllable (i dont have much to compare it to though) , following advice on here i disconnected the front arb and it showed immediately, there was a lot more front roll but the back was still the same meaning the front arb wasnt controlling the rear very much. This is when i fitted a rear one. Both roll bars are adjustable, and although i have only tested once on the new set up at Cadwell a few weeks ago, it felt much better and cornered flatter, the rear one has allowed me to slacken the front one off a little and i had to soften the rear dampers a little to get it feeling nice.

I think as mentioned above, in an ideal world, where designing from the ground up with some very trick and probably expensive shocks, money isnt an issue and the car has one use, then arb's may not be required, but i still use my car on the road occasionally and i am constricted by the stock set up which i dont want to divert too far away from.

To get this thread back on topic and for the assistance of the OP, if you find a roll bar from a scrappies which is roughly the right width, it is easy to weld a bracket on each arm of the arb with several holes in it, this makes it very adjustable quickly and easily.

To make my rear one, i got a capri rear arb as my axle is from a capri, i cut the bushes off the ends and as above, welded small brakets on with 4 holes in and made corresponding brackets on the chassis for them to attach to. Quick, easy and cheap, the poly bushes cost me more than the ARB! I'm sure some on here will not like this but this is called Locost Builders

here is my rear


Sam_68 - 29/6/11 at 05:57 PM

quote:
Originally posted by Antnicuk...From pictures and feel of the car, with just a front ARB, but the back still rolled a lot....

...i disconnected the front arb and it showed immediately, there was a lot more front roll but the back was still the same meaning the front arb wasnt controlling the rear very much.


A roll bar can't limit the roll angle at one end of the car independently of the other - unless your chassis has the torsional stiffness of steamed lettuce. The roll angle is the roll angle for the whole chassis!

What you need to recognise, however, is that different roll resistances front and rear will cause diagonal weight transfer (ie. the car 'leans' more heavily on one outside corner) leading to a combination of pitch and roll.

What you were doing by disconnecting the front ARB was increasing the roll angle (because the overall roll resistance was decreased) but reducing the the pitch change.

With the front ARB connected, you were being tipped sideways and slightly backwards.

With it disconnected, you were being tipped more sideways, but not so much backwards.

You were being deceived into thinking that you had more roll at the front and the same at the rear. In fact you had more roll at both ends but the reduced squat exaggerated your perception that the front was rolling more.

It's a funny thing, the inner ear!

Fitting the rear anti-roll bar has allowed you to reconnect the front anti-roll bar whilst maintaining the reduced difference between front and rear roll stiffness, but you will be paying a price in terms of mechanical grip.

I stand by my suggestion that you may well find that the better compromise would be to either remove or reduce the strength of the ARB's whilst increasing the spring rates to compensate (to deliver the same front and rear roll stiffnesses you had with the ARB's before), but then back off the high speed damping to maintain ride quality and mechanical grip over imperfect road surfaces.

In line with the general principle that ARB's should be the last resort, you should explore this and only fall back on ARB's when you are totally convinced that the ride quality can't tolerate further increased spring rates regardless of how well-adjusted your high-speed damping is.

Then we get onto the role of low speed damping (which is to modify transient cornering response; ie. turn-in and corner exit behaviour), but that's another bag of frogs altogether...

[Edited on 29/6/11 by Sam_68]


Antnicuk - 29/6/11 at 06:04 PM

thanks for that, i will give it a go, although i am already running 300lb front and rear which some have said is 120 more than i should ever be running in my car .................


procomp - 29/6/11 at 06:04 PM

Hi

And the above is a perfect example of how things end up when you just start adding spring poundage and ARB's to over come a simple problem. Lack of rebound to compression ratio on the basic damper. Always check the damper out first before starting with ARB's and major spring changes.

Cheers Matt


Sam_68 - 29/6/11 at 06:49 PM

quote:
Originally posted by procomp
And the above is a perfect example of how things end up when you just start adding spring poundage and ARB's to over come a simple problem....

Indeed - very easy to get into a vicious circle with spring/ARB rates if your dampers are not right!

And as a general rule, it's the high-velocity damping that usually causes the worst problems. Separate low velocity damping adjustment can be a useful tuning tool, but it's an expensive luxury and there are usually other ways to achieve the same effect.

But, to be fair, dampers have naff all effect on roll angles in steady state cornering, so it's usually fairly easy to separate the spring rate/roll stiffness-related issues that govern roll angles and basic, steady-state handling balance from the damping-related issues that govern ride quality, traction and transient handling, with a little basic testing. A cornering circle marked out on smooth tarmac can be a very powerful tool...