55 replies to this topic

[mariner]

The new BMW 5 series has double wishbone front suspension as well as multi-link rear suspension so possibly it is the end of the classic BMW front strut and trailing arm set up

Here are two images

http://www.autocar.c....aspx?im=310452

http://www.autocar.c....aspx?im=310467


What is interesting is how the front of the front end lower wishbone is arched up to clear the track rod arm. I think it must be so they can mount he rack tight against the cross member instead of stuck out on brackets.

[cheapracer]

Once upon a time the Japanese bought Euro cars to copy them now it seems BMW went out and bought themselves a Honda Accord.

[kikiturbo2]

trailing arm rear has been off the menu since E36 3 series..

[Greg Locock]

As the 5 series gets heavier (as a trend over 30 years) it'll take on more and more of the solutions that were developed for the 7 series. Bear in mind that the X5 started with a strut and has gone to an SLA. They've used double ball joint lower arms (ie virtual steer axis), if I had my time again I'd probably have argued more strenuously for a single ball joint or Jaguar style lower arm, at least for RWD.

Presumably their R&H people got fed up with trying to make a MacP work with modern tires, while giving a good ride. I suspect that's one of the reasons they always ended up at the handling end of the ride/handling curve.

Note active steer on the IRS.

Good to see they've caught up with the Falcon, 4 years late.

/late edit/

Apparently AWD is an option, that is the most likely explanation for the DBJ. I wonder why they went for a single piece lower arm for the IRS, rather than a pair of links?

Edited by Greg Locock, 28 November 2009 - 23:38.

[Powersteer]

Front anti-roll bar has a device on it probably making it variable. With the double wishbone and this, it seems BMW needed to sort out under-steer probably from diesel engine weight. The rear suspensions strut does not seem in line with the lower arms angle backwards or caster, must be a reason for this.

[imaginesix]

Front anti-roll bar has a device on it probably making it variable. With the double wishbone and this, it seems BMW needed to sort out under-steer probably from diesel engine weight. The rear suspensions strut does not seem in line with the lower arms angle backwards or caster, must be a reason for this.

Nice catch, but I doubt it has anything to do with the weight of the engine since that tends to stay pretty predictable, even through a corner. More likely it's just a part of their Stability Control system?

[Powersteer]

I was thinking that, the weight of the engine made them go from MacPhearson Struts to Double Wishbone. A soft front anti-roll bar for slower and sharper corners where chances of understeer would occur more on a longer wheelbase 4 door sedan type and becomes harder as speed increase for, as you say, stability.

[ferruccio]

As far as I can see they're still somewhat a variation of 'Macpherson struts' especially in the rear. The front employing upper wishbone and multi link lower control arms both front and rear. Nothing revolutionary in my view for BMW or their contemporaries. Just an evolution of a multi link system already in use I would say.

Furthermore the suspension design parameters for a road car is different to that of a race car. Packaging is important and stability over wildly varying conditions. The best compromise for comfort, handling, stability without taking up to much space and weight etc etc.

In a race car all you want is a an unequal length double wishbone up front and similar rear with toe control.

Edited by ferruccio, 01 December 2009 - 16:30.

[Greg Locock]

As far as I can see they're still somewhat a variation of 'Macpherson struts' especially in the rear.

No. It is a coil over, that is the only real similarity with a MacP.

[meb58]

I actaully toyed with the notion of replacing the Mac Strut front end in my 05 mini with Honda/Acura SLA front end setup. In theory it all worked out in my head and although there are/were some serious uni-body considerations I decided that I did not know enough to pursue this...after all, the change had to improve handling

I then wondered about improving upon Honda's really poor steering feel? Niether seemed an easy task.

[Ray Bell]

While I don't fully comprehend the rear suspension layout (it doesn't seem to have any means of resisting braking torque), the front is very much like the present day Falcons...

[Greg Locock]

Indeed. The X5 uses the same thing these days. I suspect that AWD is the big driver, with a front suspension it is very easy to run out of drivehsaft plunge, on lock, unless your outer CV and steering axis are properly aligned. And the suspension that gives you the most fredom to locate the steering axis is an SLA with double ball joint arms. However Audi have put everybody else off DBJ upper arms.

The IRS reacts brake via the short purple link on the right of the picture, from the spindle down to the lower arm, I think. But I'm puzzled by that, effectively that converts it back into a swing arm suspension.

I suppose if you are German you are happy to resolve all your kinematic difficulties in the rubber bushes, but it seems a funny approach.

[imaginesix]

While I don't fully comprehend the rear suspension layout (it doesn't seem to have any means of resisting braking torque), the front is very much like the present day Falcons...

Now there's a nice catch Ray. I just assumed we couldn't see everything buy yeah, there has to be some movement there from brake torque, maybe giving the rear some toe in/out? But according to everything Greg has taught us about suspension design, that would be madness.

I think we need more pics.

[Greg Locock]

Read the middle para of my previous note.

Now, while I was being snarky about using the bushes to solve kinematic horrors, you can see that it'll give you direct control over a toe change when you brake, without upsetting the lateral compliance of the lower arm. Nice. The location of the purple link is very important, as are the bush rates.

So I retract my snarkiness.

Our IRS solves a similar problem a different way, by providing the control blade, which is a direct path for the traction and braking forces. This means that those forces aren't reacted in the lateral arms so there isn't much steer effect (the little there is is easy to control). It has a downside - the forward end of the blade is very remote from the suspension , so of necessity it mounts to the body not the subframe. The big plus for us is that we have a nice big lower arm to carry the spring under the rail, whereas they have a strut, feeding all their shock absorber noise into the parcel shelf area. It means their trunk & ski hatch is narrower, but deeper, I suspect.

[imaginesix]

Read the middle para of my previous note.

Now, while I was being snarky about using the bushes to solve kinematic horrors, you can see that it'll give you direct control over a toe change when you brake, without upsetting the lateral compliance of the lower arm. Nice. The location of the purple link is very important, as are the bush rates.

So I retract my snarkiness.

Our IRS solves a similar problem a different way, by providing the control blade, which is a direct path for the traction and braking forces. This means that those forces aren't reacted in the lateral arms so there isn't much steer effect (the little there is is easy to control). It has a downside - the forward end of the blade is very remote from the suspension , so of necessity it mounts to the body not the subframe. The big plus for us is that we have a nice big lower arm to carry the spring under the rail, whereas they have a strut, feeding all their shock absorber noise into the parcel shelf area. It means their trunk & ski hatch is narrower, but deeper, I suspect.

I understand that brake torque is controlled, contrary to Ray's concern, but it was a good point he made to notice the odd linkage which does appear to allow some movement due to brake torque.

EDIT
And I still marvel that they are able to control toe change withing a few fractions of a degree with this type of linkage. I imagine a tiny bit of wear or damage would have to throw the system pretty far out of whack?

Edited by imaginesix, 03 December 2009 - 04:54.

[cheapracer]

I actually toyed with the notion of replacing the Mac Strut front end in my 05 mini with Honda/Acura SLA front end setup.

In theory it all worked out in my head and although there are/were some serious uni-body considerations I decided that I did not know enough to pursue this...after all, the change had to improve handling
.

K.I.S.S.

One of the things that truly shits me in the motoring world is the bullshit beliefs that something is better just because it is - McP struts versus Dbl Aarms or IRS Vs Live Axle for example.

There are advantages for both and it's harder to get struts wrong in comparison to some of the wrong Dbl Aarm setups I have seen, but it must be better simply because it's Dbl Aarm? (especially if it has pushrod actuated suspension ). I do see the disadvantages for a strut as the weight gets to a certain point by the way.

I decided my next car will have double beam (beam at each end) because I want to throw it in peoples face's as much as I believe it's a good thing

[DOF_power]

As the 5 series gets heavier (as a trend over 30 years) it'll take on more and more of the solutions that were developed for the 7 series. Bear in mind that the X5 started with a strut and has gone to an SLA. They've used double ball joint lower arms (ie virtual steer axis), if I had my time again I'd probably have argued more strenuously for a single ball joint or Jaguar style lower arm, at least for RWD.

Presumably their R&H people got fed up with trying to make a MacP work with modern tires, while giving a good ride. I suspect that's one of the reasons they always ended up at the handling end of the ride/handling curve.

Note active steer on the IRS.

Good to see they've caught up with the Falcon, 4 years late.

/late edit/

Apparently AWD is an option, that is the most likely explanation for the DBJ. I wonder why they went for a single piece lower arm for the IRS, rather than a pair of links?

It has to do with costs.
BMWs could out-handle many cars with their struts layout, and that also meant higher profit margins for the company.

[Greg Locock]

It has to do with costs.
BMWs could out-handle many cars with their struts layout, and that also meant higher profit margins for the company.

Of course they could. Any mildly competent company can make a crap suspension work for handling, if they are prepared to throw everything else away. Lotus made a lot of money by proving you don't have to throw everything else away.

[DOF_power]

Of course they could. Any mildly competent company can make a crap suspension work for handling, if they are prepared to throw everything else away. Lotus made a lot of money by proving you don't have to throw everything else away.

Funny because BMWs for most part, would usually out-handle MBs, Audis, VWs, Opels, Fords, Toyotas, Hondas and every american and australian car.

Does this make 90%+ of the worlds car manufacturers less then mildly competent ?!


And Lotus, to my knowledge, hasn't made money for most of its existence, that's why they kept swapping owners.
Where does this "made a lot of money" come from ?!

[Greg Locock]

"Funny because BMWs for most part, would usually out-handle MBs, Audis, VWs, Opels, Fords, Toyotas, Hondas and every american and australian car."

Which, even if it was true, doesn't contradict my statement at all. How are your English lessons going? Next we can do engineering.

"And Lotus, to my knowledge, hasn't made money for most of its existence, that's why they kept swapping owners.
Where does this "made a lot of money" come from ?!"

Lotus overall is a bit of investor's nightmare, just like Porsche. But 'we' made a lot of money retuning other people's suspensions so they would both handle and ride at the same time.

As far as projects go, we didn't do many Porsches. We didn't do much work on BMWs. But I worked on both at Hethel.

Edited by Greg Locock, 03 December 2009 - 11:20.

[cheapracer]

Funny because BMWs for most part, would usually out-handle MBs, Audis, VWs, Opels, Fords, Toyotas, Hondas and every american and australian car.

Does this make 90%+ of the worlds car manufacturers less then mildly competent ?!


And Lotus, to my knowledge, hasn't made money for most of its existence, that's why they kept swapping owners.
Where does this "made a lot of money" come from ?!

Funny because BMWs for most part, would usually out-handle MBs, Audis, VWs, Opels, Fords, Toyotas, Hondas and every american and australian car.

Does this make 90%+ of the worlds car manufacturers less then mildly competent ?!


And Lotus, to my knowledge, hasn't made money for most of its existence, that's why they kept swapping owners.
Where does this "made a lot of money" come from ?!

Your first sentence is an 80's throwback, mostly now the only falldown in most of those cars is crap (read cheap) production damping which sometimes lets down roadholding (in difference to handling). I would agree I am still to drive a Japanese car at the limit that was inspiring, not that I have driven all models.

Ozzie cars have been at the top of the tree for some time now substatially embarrassing much more expensive BMW's and MBenz have had the finest handling cars for as long as BMW have only letting weight be a bit of a handicap and a softer ride but their ability to cover distance fast is extroidanary.

Although not deep into Lotus history I think you'll find that they keep getting new owners out of respect for their capabilities when a car company decides it wants to improve themselves.

[DOF_power]

Your first sentence is an 80's throwback, mostly now the only falldown in most of those cars is crap (read cheap) production damping which sometimes lets down roadholding (in difference to handling). I would agree I am still to drive a Japanese car at the limit that was inspiring, not that I have driven all models.

Ozzie cars have been at the top of the tree for some time now substatially embarrassing much more expensive BMW's and MBenz have had the finest handling cars for as long as BMW have only letting weight be a bit of a handicap and a softer ride but their ability to cover distance fast is extroidanary.

Although not deep into Lotus history I think you'll find that they keep getting new owners out of respect for their capabilities when a car company decides it wants to improve themselves.

1] I can post Nurburging and Hockenheim laptimes that showcase otherwise.


2] There are some fine aussie cars, but they don't make up most of the market (your average Toyota makes up for most of the market).
Your typical BMWs have better driving characteristics then your typical non-BMWs, and the gaps, while no longer being what they used to be, are still there.

3] And point was BMW generally/historically made money, while Lotus didn't.

[Greg Locock]

Bit dull to repeat myself but...

"Funny because BMWs for most part, would usually out-handle MBs, Audis, VWs, Opels, Fords, Toyotas, Hondas and every american and australian car."

Which, even if it was true, doesn't contradict my statement at all. How are your English lessons going? Next we can do engineering.

"And Lotus, to my knowledge, hasn't made money for most of its existence, that's why they kept swapping owners.
Where does this "made a lot of money" come from ?!"

Lotus overall is a bit of investor's nightmare, just like Porsche. But 'we' made a lot of money retuning other people's suspensions so they would both handle and ride at the same time.

As far as projects go, we didn't do many Porsches. We didn't do much work on BMWs. But I worked on both at Hethel.

Do try and keep up old chap, it does get a bit dull repeating everything.

[dosco]

I decided my next car will have double beam (beam at each end) because I want to throw it in peoples face's as much as I believe it's a good thing ;)

As long as it doesn't use a propeller for motive force, I think I might be able to stifle my giggles.

(were you around for (edit) Franklin's (edit) exploits?)

Thanks for the pointer, Greg.

Edited by dosco, 03 December 2009 - 12:54.

[Greg Locock]

That'd be Franklin. He's probably made a mint by now.

[cheapracer]

1] I can post Nurburging and Hockenheim laptimes that showcase otherwise.


2] There are some fine aussie cars, but they don't make up most of the market (your average Toyota makes up for most of the market).
Your typical BMWs have better driving characteristics then your typical non-BMWs, and the gaps, while no longer being what they used to be, are still there.

3] And point was BMW generally/historically made money, while Lotus didn't.

1/ I can post Bathurst lap times where Holdens and Fords make BMW's look slow, so what? Whats that got to do with road cars? If you gave me the choice of any car to drive around Australia I wouldn't hesitate to choose a Falcon or Commodore and a Beemer wouldn't make the top 10.

2/ Guess again, you are quoting the total Australian vehicle market including commercial and 4WD and yes has often had Toyota as number 1, just, but when it comes to cars only market.... but what thats got to do with handling I have no idea. Which driving characteristic? Slow speed potholes? Reverse parking? My Mazda has crap strut damping but I don't change them because 95% of my driving is freeway and its nice and comfy more so than my mates overly hard similar sized Beemer 530. I recently rode in both a new Benz 600 and a 350, the 350 was amazing for comfort but the 600 was awful but it only had a 1000kms on it so it may not have 'settled' yet.

3/ As I said, I'm not into it deeply and I probably should follow what Greg says as he was an insider. I have no idea of BMW's financial history.

You are swayed by clinical performance figures and media hype, not everybody wants or needs a performance car for the road.

[dosco]

That'd be Franklin. He's probably made a mint by now.

You'd think he'd be here to gloat about the superiority of a BEAM AXLE. Heh.

[cheapracer]

As long as it doesn't use a propeller for motive force, I think I might be able to stifle my giggles.

(were you around for (edit) Franklin's (edit) exploits?)

There's a plan, you may enjoy this read for now also consider I have an aluminium extrusion factory at my disposal (rather than 2" water pipe)....
http://forums.autosp...mp;hl=beam axle

Yes I remember some of Franklin's stuff ;)


Hmmm I could call the double beam car a "Beemer"

Edited by cheapracer, 03 December 2009 - 13:07.

[Greg Locock]

Might be an idea to read Milliken's book on Maurice Olley, he goes into exhaustive details on the evils of front beam axles. As to the virtues of any given suspension architecture, I agree, the success or failure of a given design is generally more to to do with the details than with the exact choice of architecture - I suppose even to the point of admitting that a series one Range Rover's handling is at least acceptable. even with beam axles at both ends.

[meb58]

In other words Greg, Development?

[gruntguru]

The IRS reacts brake via the short purple link on the right of the picture, from the spindle down to the lower arm, I think. But I'm puzzled by that, effectively that converts it back into a swing arm suspension.

Not really - the "LCA + purple link" isn't controlling toe or camber.

[Greg Locock]

Not really - the "LCA + purple link" isn't controlling toe or camber.

I suggest you take another look.

[gruntguru]

I suggest you take another look.

Remove the blue link - no camber control.
Remove the grey "tie rod" - no toe control.
What's left? A pretty lousy swing arm if you ask me.

[Powersteer]

Purple link is to hold upright castor off the lower arm and is design in such a way for the grey link to allow toe movement.

[Greg Locock]

Yup, without the purple link the suspension can't react brake torque unless it ties itself in knots, or you replace the ball joints by long cotton reels, in which case it turns back into a sort of old Jaguar style rear suspension, which while it was good enough then, isn't really fashionable today (and would give lots of toe out in braking, which I guess is what they are really trying to avoid- the braking g.a.s (thanks Dave) really liked it when we changed from a very neutral toe setting in brake, to toe in ). The purple link is fighting the others, in toe and castor, no two ways about that.

[kikiturbo2]

3] And point was BMW generally/historically made money, while Lotus didn't.

generally making money has not a lot to do with handling..;)

[Greg Locock]

My point was the Lotus Engineering made a lot of money on the handling side. It was pissed away on other things, building cars and F1 and suchlike fripperies. Actually I don't know which bits were profitable and which bits weren't, but generally consulting/development is pretty profitable if you're already paying for all the facilities.

[ferruccio]

My point was the Lotus Engineering made a lot of money on the handling side. It was pissed away on other things, building cars and F1 and suchlike fripperies. Actually I don't know which bits were profitable and which bits weren't, but generally consulting/development is pretty profitable if you're already paying for all the facilities.

Agreed. Whilst Lotus Engineering is not as glamorous or as visible as Lotus Cars, they have a lot going on. They provide plenty of consulting work (not just handling tuning) to many auto manufacturers around the globe. I think they have a hand in almost every manufacturer. For example they provided services to the Nissan GTR project, Porsches Variocam PLUS valvelift can be traced back to them and they do a lot more for less glamorous manufacturers and their bread n butter models. It's just not publicized outside the auto industry.

[cheapracer]

Might be an idea to read Milliken's book on Maurice Olley, he goes into exhaustive details on the evils of front beam axles. As to the virtues of any given suspension architecture, I agree, the success or failure of a given design is generally more to to do with the details than with the exact choice of architecture

Note that it will be a slip joint type beam allowing independence of caster control and made of extruded aluminium so those 2 main problems (weight and twist) are resolved. Probably to call it a beam is not quite true.

As you correctly say, "details".

[Greg Locock]

Back to the Beemer.

ADAMS calls it an integral link IRS, and by eyeballing the coordinates I can get the kinematics into the realms of reality, but am nowhere near getting good curves out of it. That doesn't mean much, merely that the mechanism isn't straightforward (which, obviously, it isn't).

[mariner]

For those of us who are more visual than mathematical there is a nice three plane projection of the original M-B five link rear suspension on page 646 of Millliken. I am sorry I don't have a scanner to post the image.

The Autocar image of the BMW rear end is not clear enough to my simple brain to see if it varies in any fundemental way from the M-B layout.

[DOF_power]

generally making money has not a lot to do with handling..;)

In BMW's case it does. They made better handling then your average cars, while also turning profits. Most car companies had one or the other.

[cheapracer]

In BMW's case it does. They made better handling then your average cars, while also turning profits. Most car companies had one or the other.

Choose any 1970/80's Toyota - buckets of **** (handling) driving Toyota to the most profitable and biggest car company in the world. I wouldn't mind shares in KYB or Tokico I might add, based on profit/supplying bad handling over the last 40 years they would be at the top of the tree I guess

[cheapracer]

For those of us who are more visual than mathematical there is a nice three plane projection of the original M-B five link rear suspension on page 646 of Millliken. I am sorry I don't have a scanner to post the image.

The Autocar image of the BMW rear end is not clear enough to my simple brain to see if it varies in any fundemental way from the M-B layout.

Theres always pretty pictures here....
http://www.carbibles...sion_bible.html

[Greg Locock]

It isn't the same as the MB 5 link.

Here's how it works.

Start with a double wishbone +tierod. Set it up to give zero bumpsteer zero antidive etc.

In side view apply a braking load at the contact patch. the bottom of the spindle kicks backwards, resisted by the inboard LCA bushes, the top kicks forward, resisted by the inboard UCA bushes.

Replace the upper triangle by a single lateral link.

Front view kinematics are unchanged.

In side view the upper arm can no longer help to resist the torque on the spindle, until it develops a substantial angle.

So add a vertical link from the front of the spindle down to the lower arm. This reacts the braking torque into the lower arm

In front view, align the upper integral link ball joint with the OTR, and the lower IL BJ with the lower ball joint on the spindle.

This superimposes the link on the spindle in front view hence there is no change in front view kinematics. So, we now have a suspension that behaves kinematically like a double wishbone.

Now fiddle with the position of all the links, and the compliances of all the bushes, to give any desired behaviour. Good luck.

Incidentally if you look at car bible's description of my IRS it is laughably wrong. The guy has obvioulsy never seen one, or even looked at a photo of one isntalled in a car. Nice pictures though.

[mariner]

There I knew I was dumb thinking it was like the M-B layout.

But WHY?

I am sure this is being even dumber but given the eexistence of a sub frame member to react the single upper link aginst why then run a brake reaction link down into the lower wishbone? I am sure you can but the whole thing sems full of rubber compliance joints close together and so at high loads plus you have to beef up the lower wishbone to take all the bending loads.

Also on the very end of the front sub frame legs there are two hard bolt points beyond the big compliant mounting bush. Are these perhaps some kind of brake reaction point?

[Greg Locock]

Why do this rather than a double wishbone? Because you can play more games.

I didn't understand the subframe extensions, unless they take the load across the rocker from the main rail to the outer rocker.

[meb58]

The front layout looks very similar to late 90s honda SLR...

Regards to brake torque and acceleration torque...wouldn't we expect more compliance in the forward bushing in the UCA than the rear under braking and acceleration? I have often wondered why this bushing in this type of layout isn't a little larger...I suspect that over-design is cheaper than making two asymetric arms since the same arm can be used on both sides.

Edited by meb58, 07 December 2009 - 20:26.

[mariner]

One last questiion to Greg if he will tolerate it!!

I can sort of see that once you accept that all links and structures can distort you have compliance to deal with so you might as well build it into the bushes where you can control it and do clever games during design and development.

However does the testing of this sort of design where bushing compliance is a key part of the ride and handling response include testing with bushes that are ( in effect) 5 - 7 years old?

Most older road cars I have played with have some sort of slop of degradation in the flexible bushings so are the tests done with accelerated wear as well as new bushes?. Or maybe the newer elastomers never degrade?

[Greg Locock]

One last questiion to Greg if he will tolerate it!!

I can sort of see that once you accept that all links and structures can distort you have compliance to deal with so you might as well build it into the bushes where you can control it and do clever games during design and development.

However does the testing of this sort of design where bushing compliance is a key part of the ride and handling response include testing with bushes that are ( in effect) 5 - 7 years old?

Most older road cars I have played with have some sort of slop of degradation in the flexible bushings so are the tests done with accelerated wear as well as new bushes?. Or maybe the newer elastomers never degrade?

We assess handling after the car has completed durability, by which point the body itself is starting to crack through from fatigue in a few places (that is, it is equivalent to a full vehicle life). The criterion is that it is safe to drive, not that it delivers as new handling. Obviously the bushes are damaged in different ways during 3 months of abuse, than in a typical real car where ageing is important, so that may not be entirely representative. There's also a test where we drill all the rubber out of all the bushes, again the car has to be safe to drive.

On the other hand bushes are a serviceable item, I don't remember anyone claiming that bushes have to last the life of the car necessarily. If it clunks I would expect you to fit new ones, or accept a degradation in handling.