[FlatOut]

Are anti-roll bars actually any good? Or is it just another of those partially-true ideas which we all believe, because it is what we have been led to believe. What is the truth about these steel bars?

Further to other topics questioning implicitly-held beliefs about the motor car and suggesting its development is more than half a century behind our requirements, here is a little bit about a length of steel bar which is is strapped under many 'modern' cars - the anti-roll bar.

Just as you ask many keen motorists why they think wider tyres make a car seem to corner better and they answer 'because there is more tread on the road' (see this thread http://www.team-bhp.com/forum/tyre-alloy-wheel-section/126787-wide-tyres-more-contact-more-grip-overrated.html) which isn't really the reason, so many think that without an anti-roll bar strapped between the legs of their front suspension, their car would heel over and corner poorly.

The business of anti-roll bars and wide tyres is linked. Always looking to redesign something cheaper to maintain competitiveness/make more money, manufacturers have largely discarded double-wishbone suspension in favour of the MacPherson strut. Where the former maintains a wheel's geometry throughout suspension travel, a cheap strut is unable to do so. Levels of grip soon drop off if there is much body heel, so enter the anti-roll bar.

The simple steel bar (often around 20mm diameter) connects opposite wheels together through short lever arms linked by a torsion spring (the bar). It is usually mounted to the subframe at each end of the bar just before it bends round to form the lever arms which connect to the strut or lower link. Trying to alter the height of one side of the suspension to the other exerts a force which tries to maintain the suspension at the same extension or compression - so it has a tendency to keep the car more level through a corner. Moving both wheels up and down together results in no force, the bar rotates in its mountings rather than twisting.

For almost no money it appears to work wonders, and allows softer springs. The tyre width needs to be increased if it is to work effectively and on smooth roads with plenty of grip the effects are impressive. Rather than having to use a wider track, double wishbones and better quality components, you can get away with strapping this bar between the two sides and the job is done. With a bar across both the front and rear suspensions, you can help tune the balance of a car by altering the degree of understeer or oversteer by altering the relative strength of the bars.

But - and it is a fairly big but - as usual in this world, not least in engineering, you don't ever get something for nothing. Vertical load is removed from the inside wheel and extra vertical load placed on the outside tyre. This tends to reduce the grip from the inside tyre, which is one reason wider tyres are fitted, since their shorter, wider contact patch reduces the area of tread which is subjected to a slip angle and with shallower sidewalls can resist lateral loads better.

Turn the smooth, grippy road into a slippery wet one and the situation can change drastically. The tendency for one tyre to provide most of the cornering work leads to less predictable grip in the wet, additionally the grip of a wider tyre lets go more abruptly and wide tyres are more prone to losing all grip through puddles and water flowing across a road in heavy rain.

If the road becomes bumpy there is an altogether different problem which compounds a car's grip - and a lack of comfort. Bumps and potholes on roads are usually encountered by either the left or the right wheels. So with an anti-roll bar, if the left wheel drop into a pothole, there is a force trying to push the right wheel up. When a bump is hit by the right wheel, it tries to lift the left wheel up. And so on. This leads to a very disturbed ride called 'roll-rock' where occupants' heads are snapped one way then the other. Much more seriously, the forces acting on the suspension are disruptive to grip and composure as one wheel then another tries to respond to the force transmitted across the axle, then the rear does the same if there is a rear bar too. The springs and dampers start having to deal with these forces as well as the forces from an uneven road.

Add a corner and this situation is compounded, add rain and a poor road can turn into a really tiring drive unless you go very gently. A side-effect of these anti-roll bars is that makers pay less attention to fundamentals such as minimising the centre of mass or creating a 'low centre of gravity', so that roll couples may be excessive and the viscious circle continues. The use of lower profile tyres to offset the higher lateral tyre loadings means there is less give when bumps are encountered, so the suspension is working harder all the time both dealing directly with bumps and the reactions of the anti-roll bar to them.

Time and again I find myself on a winding, bumpy road and being held up by slow drivers in modern cars with wide tyres and little roll through corners. I am often in an ancient piece of French machinery which is on skinny tyres and which rolls a bit through faster corners. Add rain and the situation becomes even more obvious. Many recent cars have had over-specified bars, simply because the road test on Top Gear or for a magazine gives a better result on the track. Add over-wide tyres to the equation and it becomes a viscious circle when roads are less than perfect. If you happen to regularly drive over roads which are fast but rippled or bumpy, as I do, then you suffer. Ride comfort drops off as rapidly as stability and grip is reduced - wet roads become something to be treated with real caution. The dramatic loss of grip in the wet can become dangerous for less-experienced drivers, since dry grip often appears to be very good.

There is a replacement for the anti-roll bar. A good quality suspension layout, low centre of mass and good dampers/shock absorbers works wonders and a viscious circle is transformed into a virtuous circle where everything is working in harmony. Cornering will be as fast in the dry and much faster in the wet, in utter safety. Poor roads will seem to have been miraculously resurfaced and driving will become more fun as the chassis works with the road, not against it.

The problems arise most with luxurious, tall off-road vehicles. They need long-travel springs to allow them to travel off road with ease, but cannot be allowed to wallow and heel like a yacht on the road since they are so top-heavy. Manufacturers of these vehicles as well as other top-flight cars have bought into the work of an Australian ex-college lecturer. He owned an old Citroen and was awed and inspired by its hydraulic suspension to come up with something which he fitted to a pick-up to allow it to clamber over a tree trunk and generally perform miracles off-road.

He sold his technology to US giant Tenneco a few years ago and the 'Kinetic' suspension has been bought and used by Toyota as well as McLaren in its MP4 12C supercar. McLaren reckon its cornering ability improves by 25% in the dry and on a track with the Kinetic suspension system fitted, so in the real world it could be even greater. The Citroen rally car which Sebastien Loeb won so many World Rally Championship titles with was also equipped with this suspension - and so successful it was banned from the WRC.



http://www.caranddriver.com/features...lained-feature

http://www.youtube.com/watch?v=1GfLTWQyF4A

http://www.youtube.com/watch?v=vxzWMO7uaJ8

[Sankar]

Flatout, what is your take on the Torsion (Twist) beam rear suspension? Rear twist beam setup, like McPherson strut, has been replacing many multi-link rear suspension setups because of its simplicity and cost.

[FlatOut]

Originally Posted by Sankar

Flatout, what is your take on the Torsion (Twist) beam rear suspension? Rear twist beam setup, like McPherson strut, has been replacing many multi-link rear suspension setups because of its simplicity and cost.

Don't anybody confuse this with anti-roll bars, but the rear torsion beam axle provides a very cheap and effective means of rear suspension. I think it was VW who first pioneered its successful use in the early 70s, and it has appeared under almost every other manufacturer's cars since then. It allows for softer rear springs than otherwise would be needed and makes the necessity for an anti-roll bar much less. But for a driver's car - and you may shoot me down for this opinion - I would avoid one. Bump steer from the rear is never good, and the tyre contact patches tend to shuffle rather to much for my liking as the suspension works. But I have driven many thousands of miles with one under the back end and VW seems to have improved them as the years have gone by.

[alpha1]

From what I understand an anti-roll bar effectively tries to minimize the suspension movements during cornering. Isn't this exactly what we seek to achieve (reduce the car body roll)?

However, during a bump in a straight course, anti roll bar will force both the suspension arms to move similarly. Which means a bump raises the left wheel, will also raise the right wheel, and thus the small "rocking" motion that you described.

Since most cars are designed for paved roads, I am sure the designer's don't worry about the small ridges and troughs on the roads. You might then argue - what is the point of suspension then? Which is a valid point! However, we have not eliminated the suspension, only downplayed it - it is effectively a trade off - you want body roll or you want wheel articulation?


Of course the hydraulic suspension comes close to what we really want from a suspension. But then hydraulics come with its own maintenance issues ...

[1100D]

Its not true that the Anti-roll Bar is undamped, the dampers attached to the lower arms to which the arms of the anti-roll bar is attached does provide the damping.

But yes, agree to the point that on uneven surfaces it does transfer lifting forces to the side thats going on a smoother surface, reducing cornering grip on less than perfect roads. Track usage is different as the surfaces are smooth.

[FlatOut]

Originally Posted by 1100D

Its not true that the Anti-roll Bar is undamped, the dampers attached to the lower arms to which the arms of the anti-roll bar is attached does provide the damping.

But yes, agree to the point that on uneven surfaces it does transfer lifting forces to the side thats going on a smoother surface, reducing cornering grip on less than perfect roads. Track usage is different as the surfaces are smooth.

Is is an undamped spring - the car's road spring dampers do mitigate its worst effects but they are tuned for something quite different.
Saab were a famous Swedish car maker who used to compete a lot in rallying - and won year after year. Their cars only ever had anti-roll bars fitted to the most powerful model of the 900 range. All the others managed very well without (they didn't heel untowardly) and the cars were known for their handling and comfort. The 99/900 model range lasted from 1967 to 1993 with little change, by which time GM had bought into the company - GM platforms were used from then on.

Originally Posted by alpha1

From what I understand an anti-roll bar effectively tries to minimize the suspension movements during cornering. Isn't this exactly what we seek to achieve (reduce the car body roll)?

Yes, this is correct. As one wheel moves relative to the other, the anti-roll bar twists up (it is a torsion bar) to resist this movement. This beneficial effect is quite obvious, the question is, what are the downsides to this very crude piece of suspension?

Originally Posted by alpha1

However, during a bump in a straight course, anti roll bar will force both the suspension arms to move similarly. Which means a bump raises the left wheel, will also raise the right wheel, and thus the small "rocking" motion that you described.

If there is a bump across the road which causes left and right suspension to move together then the effect is zero, the bar simply rotates. The rocking motion and unsettling of the suspension occurs when a bump affects first one wheel then the other side, or just one side.

Originally Posted by alpha1

Since most cars are designed for paved roads, I am sure the designer's don't worry about the small ridges and troughs on the roads. You might then argue - what is the point of suspension then? Which is a valid point! However, we have not eliminated the suspension, only downplayed it - it is effectively a trade off - you want body roll or you want wheel articulation?

Suspension engineers are concerned with many conflicting factors, not least that the springs must be able to keep the car from scraping the road when fully loaded. Beyond that, it is all about maintaining the maximum force of each tyre on the road, whatever the circumstances. Small ridges, troughs, every sort of road is designed for - otherwise a suspension will fail to keep the wheels (and body) under control on a less than perfect surface which is seen as totally unacceptable in a modern car.

But stiff springing is a modern trend, partly led by 'clinicking' of consumers, who have mentioned they don't like cars which heel over through roundabouts. (We joke it is the school run mums!) Many cars today have a jarring ride which never settles down and becomes tiring on a journey.
Your trade-off point is spot on - and wheel articulation is vital to comfort and the ability of a car to hold the road as speeds rise. There is no such thing as a smooth road as regards suspension - the smoothest road starts to work the suspension as speeds rise. If you doubt this, replace your dampers with solid bars then try to drive on a smooth road above 40kph! (Private road, if anyone tries this!!)

A car which heels excessively quickly is undesirable as the mass of the body (and its load) may introduce instability. Equally, if a body follows the road's bumps rather than letting the suspension articulate, then as speeds build the body's rapid movement will start to affect the adhesion of the tyres - that is to say, going quickly on a poorer road can cause the bucking body to start to bounce the car off course.

The interference of the anti-roll bar (even in a straight line on a bumpy road) adds additional loads to the suspension which reduces the load on the tyres, reducing stability.
Through corners which aren't smooth and flat, an anti-roll bar which doesn't allow much articulation will cause the car to lose adhesion more quickly, especially on a wet road. And especially so if the road is anything other than billiard-table level and smooth.

It is, like everything else in life, a compromise. But living where roads are fast but with poor foundations and lots of bumps, many modern cars' handling falls apart on these roads when taken more quickly, their springing and anti-roll bars too unyielding for the real world - set up to impress journalists on a test track.

Originally Posted by alpha1

Of course the hydraulic suspension comes close to what we really want from a suspension. But then hydraulics come with its own maintenance issues ...

I'm not sure if you have had experience with hydraulic suspension (ignoring the fact that almost all vehicles have four separate sealed hydraulic units controlling the damping), but in my experience there is less time and cost involved maintaining it than there is with steel suspension, over 200,000km. The suspension which causes the most problems in my experience is airbag suspension, which is so failure-prone that there are kits available to convert back to steel.

McLaren and Toyota have pointed to the future with damping and anti-roll combined into a very simple, very neat system which massively reduces the compromise made by a rigidly located steel bar strapped between left and right suspension. That BMW, Range Rover, Mercedes-Benz and others are attempting to remove/decouple the rigid anti-roll bar suggests this is a general recognition of its short-comings. The benefits for off-road vehicles are particularly spectacular, so it is these which will most likely standardise the system first, followed by faster cars. However, good suspension does not sell cars as well as far less necessary gizmos since its effects are not instantly seen or felt. It is the same problem with ABS.

[alpha1]

Originally Posted by FlatOut

I'm not sure if you have had experience with hydraulic suspension (ignoring the fact that almost all vehicles have four separate sealed hydraulic units controlling the damping), but in my experience there is less time and cost involved maintaining it than there is with steel suspension, over 200,000km. The suspension which causes the most problems in my experience is airbag suspension, which is so failure-prone that there are kits available to convert back to steel.

My only experience with Hydraulics is via lifting machinery.
But of course the maintenance issues are not as bad, as I made it sound in my earlier post

Actually I never understood: what is the problem with body roll? It causes nausea?

[Jeroen]

Originally Posted by FlatOut

I'm not sure if you have had experience with hydraulic suspension (ignoring the fact that almost all vehicles have four separate sealed hydraulic units controlling the damping), but in my experience there is less time and cost involved maintaining it than there is with steel suspension, over 200,000km. The suspension which causes the most problems in my experience is airbag suspension, which is so failure-prone that there are kits available to convert back to steel.

I have a lot of experience having (Co)owned and worked on various Citroens, DS, SM and some of the later models. So I've got quite a bit of experience driving, maintaining and fixing their hydropneumatic systems.

If it works, it works very well and you get an amazing ride. If it doesn't, you're in for it. Pretty complex to fix. There is a lot of information available on the net on how to maintain and fix them. True Citroen enthusiast will tell you that the systems are reliable and easy to maintain. Like true Alfisti will tell you rust isn't that big a problem.

I just never got into them. They were interesting, but I'm a firm believer that the French are only marginally better at putting cars together then the Italians. And they suck, to date!

So from a design point of view these Citroens might have been cutting edge, from an engineering point of view they failed to impress.

And yes, I've owned 2CV's as well. We used to call them throw away cars. At one point in time they became so cheap that if they broke down, you just got yourself another second hand one. They were also known as most unsafe car in the Netherlands. They were banned from taking your official drivers test, as they were so slow the examiner could not judge if you could keep up in traffic.

Once at speed you could corner through any bend at ridiculous speed, but it sure took a long time to get up to any speed.

As suspension goes, I'll stick to my Jaguar and Alfa Spider. Very different and that's what gives a car character. Whether you enjoy it, is just a matter of personal taste.

Jeroen

[FlatOut]

Originally Posted by Jeroen

I have a lot of experience having (Co)owned and worked on various Citroens, DS, SM and some of the later models. So I've got quite a bit of experience driving, maintaining and fixing their hydropneumatic systems.

If it works, it works very well and you get an amazing ride. If it doesn't, you're in for it. Pretty complex to fix. There is a lot of information available on the net on how to maintain and fix them. True Citroen enthusiast will tell you that the systems are reliable and easy to maintain. Like true Alfisti will tell you rust isn't that big a problem.

Well, to put the opposite case, I have owned many old Citroens and the one thing which has never caused any bother was the suspension systems. In fact, amazingly little has gone wrong with them in any respect, given the tens of thousands of miles covered and old age and generally high mileage of them. (The exception would be those made with all Peugeot parts on a Peugeot platform from the 1980s. When I refer to Citroens, I don't mean b@stardised Peugeots!)

I had to change a HP pump on my DS a few years ago - it was leaking a little air. However, it was the original, it took under a couple of hours start to finish and the car was over 40 years old, having covered over 400,000 miles.

I've yet to meet an Alfa enthusiast who is in denial over rust. Don't quite see how you can say that, unless it's taken out of context and the implication is that there isn't anything which can't be sorted by a specialist, or that once sorted properly it isn't a problem. How many cars older than 25 years old in Northern Europe do you know of which haven't rusted badly? The one main problem with many older cars is rust. Citroen enthusiasts are similarly candid - after all, any enthusiast knows their machine better than most. If there is a recurrent reliability point, it is that the last thing to cause problems is the suspension - provided it hasn't been tinkered with by someone who isn't used to using any tool other than a hammer.

Regarding your Citroens, the suspension systems must have either covered in excess of 300,000km or have been sorely neglected /damaged by a lack of correct maintenance. Which is hard when the requirements are oil replacement/filter clean once every 5 years and new/recharged spheres as and when - usually every 4 years or 60,000km or so - and a very quick and easy job as you'll know. I'm not sure what you are finding so complex on the suspension when it comes to fixing a damaged system unless it's a diravi steering column replacement - which I've never heard of failing. Took one out of a car once for spares, bit of a time-consuming job.

Originally Posted by Jeroen

I just never got into them. They were interesting, but I'm a firm believer that the French are only marginally better at putting cars together then the Italians. And they suck, to date!

So from a design point of view these Citroens might have been cutting edge, from an engineering point of view they failed to impress.

From an engineering point of view the Citroen suspension system is regarded by most engineers as as fine as you could get (whether or not they appreciate the subtleties) even if just on the most easily understood and basic level of component tolerance. In the 1950s the company achieved tolerances to 1 micron, several times better than any aircraft hydraulic engineering and many times finer than modern engineering workshops. Rolls-Royce and Mercedes bought licenses to use Citroen's suspension for very good reason.

You are correct regarding the erratic and often poor assembly quality in the 1970s and 80s - I cannot think of any mass-produced makes which were consistently good back then apart from German cars and even they rusted with the best of them, if not as catastrophically badly as the W210 class Mercedes from the 90s and early 2000s.

Originally Posted by Jeroen

As suspension goes, I'll stick to my Jaguar and Alfa Spider. Very different and that's what gives a car character. Whether you enjoy it, is just a matter of personal taste.

Both can be lovely cars when working well - I've owned them. My favourite Alfa was probably an incongruous little SudSprint with the smaller engine - a very sweet little car. Just make sure you replace the camchain tensioners on the AJ V8 with the best available, if they're not already done.

[Jeroen]

Well, you just have different experience and hang out with a different crowd of car enthusiasts.

But then again, none of what's in this thread is fact. Personal experiences, mine or yours, can not be extrapolated to real life statistical facts. So at best they are opinions, neither true or false. Thank God for that. I hate facts, I love opinions.

Jeroen

[1100D]

Originally Posted by FlatOut

Is is an undamped spring - the car's road spring dampers do mitigate its worst effects but they are tuned for something quite different.


The interference of the anti-roll bar (even in a straight line on a bumpy road) adds additional loads to the suspension .

People who engineer the suspension systems do take into account the loads the anti-roll bars exert and the damper parameters are arrived at with the damping requirements for the anti-roll bar into to the equation.

Unfortunately for us, the cars common people use will not be able to leverage the complexity of the interconnected hydraulic linkages. That would push the costs too high. Apart from enthusiast cars, most hydraulic suspension Citroens (coincidentally indicating a citrus fruit) I have seen since mid 80's had been squatting on the floor without ability to lift their backs up. Thankfully our Alto's and Maruti 800's fare much better!!

[FlatOut]

Originally Posted by Jeroen

Well, you just have different experience and hang out with a different crowd of car enthusiasts.

But then again, none of what's in this thread is fact. Personal experiences, mine or yours, can not be extrapolated to real life statistical facts. So at best they are opinions, neither true or false. Thank God for that. I hate facts, I love opinions.

Jeroen

Well, it's a shame we don't discuss the facts regarding manufacturers such as Toyota replacing the steel anti-roll bar with hydraulics. That's what the original post in this thread was about.

If you want to discuss Citroen's own use of hydraulics or suspension in general then start a new thread. That is something quite different.

[SPIKE ARRESTOR]

So what comes first as a design criteria, the load transfer or the anti roll bars? How does the design proceed for an anti roll bar, what are the deciding parameters?

Spike

[FlatOut]

Originally Posted by SPIKE ARRESTOR

So what comes first as a design criteria, the load transfer or the anti roll bars? How does the design proceed for an anti roll bar, what are the deciding parameters?

Spike

You would have to ask someone who works in an ride and handling design department, although most cars today in Europe come with very firm suspension settings - which prevents the common, cheap layout of MacPherson struts with a torsion beam behind from moving around too much and exposing their weaknesses.

It will be all factored in within the software which works these things out - you feed in the vehicle data from which it will work out the 'optimum' settings which are intended to work as well as possible for most driving styles for most roads. As with many digitally-processed designs, it works well in the average situation but there may be glaring gaps in a design's ability. The roads I drive on frequently, and the nature of my driving seems more likely to show up these 'gaps' than the average shopping-trip run. (There is a similar case with sailing boat hull design - there is increasing talk of CAD hulls which on paper are very efficient having some hidden nasty traits.)


I found this on another motoring forum a little earlier - it's a scan of CAR magazine with LJKSetright penning the words. He was pretty famous, to the extent his death was on national tv (in Britain anything motoring-related is seen as unworthy by UK media - unless it is a big crash or it's F1) and his obituary was in most national newspapers.
Not sure if it's readable, enlarge the browser as necessary - I've attached a text resume of what the great man had to say.

[Rehaan]

Originally Posted by FlatOut

If the road becomes bumpy there is an altogether different problem which compounds a car's grip - and a lack of comfort. Bumps and potholes on roads are usually encountered by either the left or the right wheels. So with an anti-roll bar, if the left wheel drop into a pothole, there is a force trying to push the right wheel up. When a bump is hit by the right wheel, it tries to lift the left wheel up. And so on. This leads to a very disturbed ride called 'roll-rock' where occupants' heads are snapped one way then the other.

My mother has been complaining about this side-to-side rocking at the rear of the VW Vento... i wonder if this is related / the cause?

Originally Posted by FlatOut

...as well as McLaren in its MP4 12C supercar. McLaren reckon its cornering ability improves by 25% in the dry and on a track with the Kinetic suspension system fitted, so in the real world it could be even greater.

Did you see the Edmunds suspension walkaround of the MP4-12C?

Check it out : LINK

Originally Posted by An Excerpt

The missing front suspension element is...a front stabilizer bar. The MP4-12C hasn't gotone.Go ahead. Scroll back a couple of pictures and have another look. I'll wait.

What the McLaren has fitted instead is a complex series of tubes. That hose (green) connects the two opposite sides to form a hydraulic stabilizer bar that can be computer controlled.But to make it work they must cross overone another: the compression valve on this side connects to the reboundvalve on the other side, and vice-versa.

Imagine you're in a sweeping bend. Naturally, the body wants to roll toward the outside, and this causes the outer shock to compresswhile the inner one grows. With this system, the pressure rise in the outer shock is re-routed across the car so it can repel the extension of theinner one, and vice-versa. But this only works if thehoses are"cross-wired", rebound to compression.

Here'sanother cool bit. There are valves, accumulators and electronics between the two interconnected hoses, and their presence allows the roll stiffness to be continuously variable, evenduring a single corner. For example: the system can be programmed to roll-in softly then firm-up as g-forces rise.

cya
R