Live rear axle Vs. IRS

[cps13]

Hi,

I have building a caterfield style kitcar, I'm not too hot with my welding abilities and therefore got a pre-assembled chassis.

The chassis has all the correct support bars/fixings for a live rear axle. From reading up on it, it looks to me like IRS is the better option so I was wondering if it would be worth getting the chassis altered to allow IRS.

Does anyone think this is the correct way to go or would I be wasting money getting the chassis altered?

Thanks!

[singlecoil]

IRS is better at coping with uneven surfaces than live axle, but other than that it isn't really a matter of big importance. Converting a chassis is going to be far from straightforward, and isn't going to be the end of your problems. I think you would be better leaving it as it is.

[anonymous-user]

last time it was discussed here, the conclusion I read was that a IRS was better on a uneven road, on a smooth rack its debatable. my Phoenix runs a live rear axil, my thoughts are that a good live axil will beat a bad IRS,and vice verse, Some of the hubs I've seen on IRS look real heavy, I don't think I'd change if you have the chassie already

[Steve_D]

The amount of work and subcontracted welding out weighs any benefits.

Enjoy your live axle and get used to the driving experience then go indy on your next kitcar.

Steve

[juansolo]

If it's set up right, live axle is actually my preference! Though I may be odd.

[cps13]

Thanks for you answers - I think i'll be staying with my live axle!!

[Chris71]

You have to really know what you're doing setting the geometry for mounting points on an IRS conversion, I wouldn't trust anyone to do it unless they have serious credentials and prior experience with that particular chassis TBH. Great as a home project (where you only have yourself to blame!) but I'd be very annoyed if I paid someone what would presumably be quite a lot of money for a one-off conversion only to find it had taken on some nasty handling traits.

The only exception I can think of is if there already is an IRS version of the chassis and you can get the factory to convert it to a proven factory spec.

[Comadis]

intersting discusion:

all 3 live-axle sylva´s which i have driven (and 1 still own), also the 2 fisher fury´s from friends dont feel like the typical live-axled kitcar (which i also have driven) and are absolutely compareable with a friends IRS westfield.

so there are some "secrets" to get a live-axle performing well even on uneven and bumpy roads.

[Kevp]

I understand IRS is more power sapping than a live axle. On a live axle the drive shaft runs gently to the 1 UJ at the diff nose. The IRS has 2 UJs at a steeper angle, requiring more power to turn. The diff also has a UJ just to take out the stress from misalignment & movement of engine/gearbox. Could all be a load of b... though.

[Mr2Mike]

Kevp said:

I understand IRS is more power sapping than a live axle. On a live axle the drive shaft runs gently to the 1 UJ at the diff nose. The IRS has 2 UJs at a steeper angle, requiring more power to turn. The diff also has a UJ just to take out the stress from misalignment & movement of engine/gearbox. Could all be a load of b... though.

A live axle will typically have just 2 UJ's in the propshaft whereas IRS will have the 2 UJs in the prop in addition to 4 CV joints in the driveshafts. CV joints are not particularly efficient when running at an angle, so I think there must be some truth in this though I suspect the difference is not very large.

[Sam_68]

Mr2Mike said:

A live axle will typically have just 2 UJ's in the propshaft whereas IRS will have the 2 UJs in the prop in addition to 4 CV joints in the driveshafts. CV joints are not particularly efficient when running at an angle, so I think there must be some truth in this though I suspect the difference is not very large.

The driveshafts on an independent suspension certainly do absorb more power, but IRS also transmits power more effectively to the tarmac (especially if the tarmac is less than perfectly smooth). It's no good having more power if the tyre contact patch isn't as consistently loaded and you can't put it down on the road...

I'm a big fan of beam axles (albeit de Dion rather than live axle, by choice) but being realistic you have to accept that a good, properly set-up IRS will beat an equally competent live axle set-up every time, these days.

The live axle Sylva rear suspension is superb, of its type, but the IRS Sylva is noticeably better still. I strongly suspect that Comadis' friend's Westfield was badly set up (probably with springs/dampers too set stiff for the car; sadly far too common on this type of car, because much of the knowledge base has been inherited from live axle 'Sevens' that need stiff springs to control the axles).

Having said all which, I firmly agree with the early advice on this thread, that it's not worth the cost and complication of changing to IRS. If it bothers you, it might be worth considering the half-way house of changing to a de Dion, possibly in conjunction with better location (more carefully thought-out trailing arms and perhaps a Woblink or A-frame in place of the Panhard rod - the normal Westfield/Locost type live axle set-ups have very short, parallel trailing arms that are really pretty awful, geometrically), but to be honest even that would be a lot of effort for relatively little advantage.

[procomp]

Hi

There are some untruths there. If you look at any of the very well setup live axles they will be running in the region off 100lb - 140lb springs tops.Certainly not stiff. The trick is to get the dampers with the correct valving to control the unsprung weight. Also the sylva axle setup is amongst the worst out there. the 4 link arm configuration it runs is what lets it down as it severely restricts the axle travel in roll acting like a very stiff rollbar. Hence the reason the rear chassis brackets are well known for detaching from the chassis. As well as it not being able to incorporate anti squat in to the geometry as well as a more traditional 4 trailing arm setup which can incorporate over 100% if needed.

All IMHO

Cheers Matt

[Sam_68]

procomp said:

If you look at any of the very well setup live axles they will be running in the region off 100lb - 140lb springs tops.Certainly not stiff.

Those figures are a bit on the light side. Road use would often be at the upper end of the figures you quote; race use could typically be closer to 180-200lb/inch at the rear, though it does depend on the precise car and set-up, and driver preference, of course.

I run 180 front, 140 rear on my Sylva, set up for road, track day and occasional hillclimbs, and I wouldn't want to go much lighter without using really decent (and really expensive!) dampers, like Penske, where I could play around with digression and not just linear adjustment of the high- and low-speed bump and rebound. And having £4K's worth of springs and dampers would kind of defeat the object, on a £3K track slag.

Back in its circuit racing days, it was on 230lb fronts and 180 rears.

But 'stiff' is relative. Even if we accept your figures for spring rates, you've got to bear in mind that, for example, the IRS on an Elan (which has a kerb weight about 50% heavier than a typical live-axle Seven, of course) has a standard spring rate of 68lb/inch rear (and 75lb fronts, for what it's worth).... so about 30% softer than the lightest springs you're suggesting for a live axle car, on a car that's 50% heavier!

I stand by my assertion that a well designed and set-up IRS can run much softer springs.

procomp said:

The trick is to get the dampers with the correct valving to control the unsprung weight.

Agreed.

procomp said:

Also the sylva axle setup is amongst the worst out there. the 4 link arm configuration it runs is what lets it down as it severely restricts the axle travel in roll acting like a very stiff rollbar. Hence the reason the rear chassis brackets are well known for detaching from the chassis. As well as it not being able to incorporate anti squat in to the geometry as well as a more traditional 4 trailing arm setup which can incorporate over 100% if needed.

It's designed to add to roll stiffness. Amongst other reasons (and in conjunction with appropriately located roll centres) this allows the Sylva to function effectively without a front ARB. Front ARB's are the work of the devil, and a crutch for incompetent suspension design, since they introduce lots of unwanted side effects (not least by partially removing the independence of your nice, sophisticated double wishbone front end by linking the two sides together).

It does rely on being rubber bushed to work properly, though; the only failures I've heard of personally are where people have misunderstood how the system operates and have rose-jointed the links.

Out of interest, why would you be worried about anti-squat with a beam axle set-up? the main problem on an IRS suspension with squat is that it makes the rear wheels adopt unfavourable camber angles (which reduces traction) under hard acceleration. This doesn't apply to a beam axle, therefore it's not really of much importance.

If you think the Sylva rear end is amongst the worst out there, then you're misundertanding how it works and the clever subtlety of its design.

...but ultimately, proof of the pudding and all that: if the Sylva is so terrible, how come it's won more kit car championships than all the other marques put together, and how come the handling is so well balanced?

Edited by Sam_68 on Thursday 26th March 18:31

[juansolo]

Just for comparison I find 250/150lb to be ideal on my Zetec Westie with ARBs. On the old one without ARBs I had to go to 300/175 to handle slicks. But having it soft at the back helps with traction and bumps.

All IMO as handling, and what makes good handling, is all very subjective.

[procomp]

Hi

You seem to be missing the point. The increased roll stiffness is there because the axle is physically unable to move independently of body roll.Which simply light loads the inside wheel. The asymmetric longitudinal Watts linkage you refer to is what causes this problem any body roll is simply trying to rotate the axle casing along it's length. Which is why so many have broken chassis brackets of the chassis. Relatively common when the cars are used on track days / competition. Also why there is no rear wheel steer because the axle is held in a virtual fixed position. I believe it was Mark Hales who commented on it's complete inability to work. When he did a 7ns shoot out in one of the magazines.

Front ARB's are hardly the work of the devil. They just like a set of dampers need a little attention to the detail. A good one will allow the car to be fully tunable for varying conditions and layout. And also allow the car to run softer front poundage springs whilst still being fully in control of the body roll / camber gain. All in all there are very few down sides to a well designed front ARB.If one dose as you say remove the independence. Then it is simply far to stiff and incorrectly designed.

Regarding the sylva in the 750 Mc championship. If you actualy take a good look at many of those cars that did the winning you will find that NON are particularly STD. Having helped out with many of those drivers over the years with setup and handling issues you soon see just where the problems are in the rear axle setup. Regarding the Roll centres I think a bit more research is needed there. Also regarding Balance. That is the one thing that the sylvas where most notorious for not having due to being far to short in the wheelbase. Perhaps more suited to sprint and hill climb with tighter corners but certainly not on the more open circuits which is where it has always had problems. Rear anti squat Again more research is needed there and an understanding of weight transfer on corner exits.

There is however no dispute that a well designed IRS will certainly out do a good old live axle / dedion. But the question is. Is there room in the back of the average size kit car to actualy fit wishbones of the right length. Hence the reason virtually all kitcar manufacturers have made such a bad job of designing IRS rear end setups with very poor camber control.

Cheers Matt

[Sam_68]

procomp said:

You seem to be missing the point. The increased roll stiffness is there because the axle is physically unable to move independently of body roll.

Not at all. The axle is allowed to move in roll by compressing the walls of the rubber bushes. It is designed to do this and obviously the resistance to deflection offered by the bushes offers resistance to roll... effectively, the rubber bushes act as anti-roll bars in themselves.

You (and Mark Hales, and many other people who misunderstood how the system works and suffered breakages after Rose jointing their Sylvas) are looking at it as if it is fully Rose jointed with no compliance.

As an aside, there is sufficient play/compliance even in many Rose jointed systems to allow apparently 'solid' linkages to work very well. Take a look at a Mumford link some time and try to figure out how it works in roll... it doesn't stop it working very well indeed on many Clubmans and 750 Formula cars.

procomp said:

All in all there are very few down sides to a well designed front ARB.If one does as you say remove the independence, then it is simply far to stiff and incorrectly designed.

Half of the load of a single wheel bump at one wheel will be transmitted to the other wheel with a conventional anti-roll bar... it really is that simple!

I agree that ARB's are a useful tool for fine tuning, but there are other tools and if you feel you must use an ARB they should be kept as light as possible and the basic design of the suspension should accommodate that.

And if a suspension design is good enough, you won't necessarily need one at all (as the Sylva proves)

procomp said:

Regarding the sylva in the 750 Mc championship. If you actually take a good look at many of those cars that did the winning you will find that NON are particularly STD.

I've got one in my garage that was good enough to set second pole at Silverstone and be a consistent front runner in the hands of a novice driver. It's suspension is completely standard.

And whilst many cars have non-standard modifications, I can't recall any successful live-axle Sylva that has had the rear suspension re-designed to remove the longitudinal Watts links that you find so offensive (I do recall a V8 engined Striker that was hillclimbed in Scotland a good few years ago that had been converted to long trailing arms, but it wasn't a success - though that's not necessarily the trailing arms' fault, of course).

procomp said:

Regarding the Roll centres I think a bit more research is needed there. Rear anti squat Again more research is needed there and an understanding of weight transfer on corner exits.

Weight transfer and the influence of roll centres thereon is my speciality. I've been researching and modelling it for perhaps 15 years (part of which has been a geometrical analysis of the Sylva, hence my understanding and respect for the rear suspension design).

Edited by Sam_68 on Thursday 26th March 21:34

[pigeondave]

I have been following this thread and would like to know what the Strikers use on the rear? Panard Rod, Watts link or this new ish Mumford link. Im guessing that the Mumford link is a fairly new concept.

[procomp]

Hi

We are obviously looking at this from two totally different points of view. I spend many hours working with drivers and there setup. every one of them who has either modded the actual setup or modded the chassis has found significant improvements in grip levels and traction. I think the fact that you are modeling around roll centers means we will never agree on geometry setups.

The car you refer to at silverstone i dont suppose for a moment that is one of Mark Arden's. As he was the only so called novice i can think of who had success in the 750 kits. That was during the 99/2000 championships. I assume the car must have been re bodied since it raced as the petrol filler neck joined / protruding through the bodywork would mean it wouldent pass scruitineering.

Cheers Matt

[Sam_68]

pigeondave said:

I have been following this thread and would like to know what the Strikers use on the rear? Panard Rod, Watts link or this new ish Mumford link. Im guessing that the Mumford link is a fairly new concept.

The Striker uses a Panhard rod at the rear. The Mumford link has been around a good few years, now, but only tends to be seen on pure race cars because the lack of roll compliance means it's not really suitable for road cars with larger supension movement.

And, of course, there aren't that many pure race cars around using beam axles any more. You'll come across it on some 750 Formula cars and the odd old Mallock-style Clubmans. I think the Pulsar kit car still uses it, too - the prototype certainly did and the photos on the company's website here appear to show that it still does, from what I can see, but for some reason they describe it as a 'specially designed articulated Watts linkage'!?

procomp said:

The car you refer to at silverstone i dont suppose for a moment that is one of Mark Arden's. As he was the only so called novice i can think of who had success in the 750 kits. That was during the 99/2000 championships. I assume the car must have been re bodied since it raced as the petrol filler neck joined / protruding through the bodywork would mean it wouldn't pass scruitineering.

Petrol filler is as raced, and no, it wasn't Mark Arden.

Both the car and myself are showing our age, so it was probably well before your time - to give you some idea, the guy it was second pole to was Martin Stewart, IIRC! From memory it was 1990 or 91 and back in those days the protruding filler cap wasn't an issue.

procomp said:

I spend many hours working with drivers and there setup. every one of them who has either modded the actual setup or modded the chassis has found significant improvements in grip levels and traction. I think the fact that you are modeling around roll centers means we will never agree on geometry setups.

I was doing your job (chassis set-up) as a sideline back in the late '80's and early '90's. My interest in geometry and roll centres led on from that, when I decided I wanted to develop a better understanding of why cars behaved as they did rather than just looking at immediate cause and effect of specific handling issues.

We may or may not agree on geometry set-ups; we haven't actually discussed them. But you were saying that the Sylva rear linkages were physically incapable of roll movement, which is just plain not true - they are capable of more than enough roll movement (the chassis would ground before the suspension ran out of roll compliance), provided you leave the rubber bushes in the suspension as it was designed to work.

I ask again, though ('cos I'm genuinely a bit out of touch with the kit car championship, these days), what are the mods to chassis and set up on live axle Sylvas to which you refer? Do they include replacing the longitudinal Watts links with twin trailing arms? Are they still using Panhard rods?

Edited by Sam_68 on Friday 27th March 17:33

[Comadis]

i didnt say that my friends IRS westfield is bad, but my live-axle sylva is compareable to it.

also the wessi does not have the typical seven-like hard suspension. its relatively soft.