Front Suspension Capstone Project 3

[OGR34]

Hey guys,

I have been assigned a Capstone project involving the design of a front suspension system for an Intermeccanica Speedster (shown below)

The current front setup of the vehicle consists of a classic VW transverse torsion bar suspension. My partner and I intend to replace this outdated setup with an SLA double wishbone type suspension. Links, springs, dampers etc. will all be purchased. The suspension geometry will be designed by us and the springs and dampers will be selected to achieve engineering requirements. Ultimately we would like to achieve a balance of good comfort and performance however, we're not quite sure of what engineering requirements we would need to achieve this.

Currently I have very little background with suspension systems and have started reading some automotive textbooks, namely Race Car Vehicle Dynamics by Milliken. The book is very descriptive and reading it has allowed me to become familiar with most of the key words of terms of a front suspension system. However, in terms of starting a design, I have absolutely no idea where to start! The amount of parameters in suspension geometry is quite overwhelming and right now I'm trying to figure out what my starting point should be in the design process as well as how I can piece all the information I've learned to achieve my goal.

Does anyone have any suggestions? Any help is greatly appreciated.

 

[BrianPetersen]

Nice looking car.

Look at how other suspension systems are built. Look at both OEM and aftermarket systems. Also understand how the front suspension works and why it was built the way it is.

You're working on the front suspension. One of the things you had better do is talk to whoever is doing the rear suspension. And I hope you're doing the steering, too, because the suspension and the steering act together and the geometry has to be correct.

You are working on a rear engine car. They have particular problems regarding vehicle stability. You need to take into account the need to keep stable handling characteristics. Front suspension that is "too good" relative to the rear suspension will cause oversteer and make the car a real handful to drive.

With the traditional double-trailing-arm VW front suspension, there is no camber change relative to the bodyshell. Whatever body roll happens, the wheels tilt the same. Promotes understeer ...

The roll center is at ground level regardless of ride height, which means all roll stiffness is provided by the springing. I don't recall if there is an antiroll function.

The two factors above also mean that there is no tire side-scrub with suspension movement. Go over a bump, and the tire contact patch doesn't move sideways. This is actually a good thing. Note that a roll center above ground level inherently means side scrub, you cannot separate the two effects. Modern tires can deal with it.

Understand the steering geometry, too. Note that the steering box is off-center in the car. You might think this gives unequal geometry and bump steer - but it doesn't. The steering box is ahead of the tie-rod connection to the steering knuckles in the same relationship as the trailing arms. In side view, the arc is the same. No bump steer even though the left and right tie-rods aren't the same length.

Compare the above description to the original rear suspension - which has excessive camber change, excessively high roll center, excessive side-scrub, and toe change with suspension travel ... and bear in mind my earlier statement that you essentially want bad (ish) front suspension and good rear suspension on a rear-engine car to promote understeer ...

Maintaining the stock front suspension geometry but using better, modern springing and damping and antiroll while upgrading the rear suspension to a proper multilink would not be a terrible approach ...

You might want to compare what you have, to what a modern rear engine car has. Porsche 911 uses MacPherson in front and multilink in rear. A smart uses MacPherson in front and deDion (essentially a beam axle) in rear. The beam axle rear suspension keeps the rear wheels straight up to the ground no matter that the bodyshell does.

 

[OGR34]

Wow that's a lot of useful info, thanks Brian!

Sorry, I forgot to include the following information. When I had a meeting with my project sponsor Henry Reisner (owner of Intermeccanica), due to limited resources, the objective of the project was to keep the choice of VW or 944/911 semi trailing arm rear suspension and change the outdated double-trailing-arm VW front suspension. The main reason why the front suspension was the desired change was that customers were not too satisfied from buying a $50,000 car with an ancient set-up at the front. Changing the front to SLA will give a little more eye candy when the car is lifted by a ram as well as improve comfort and performance at the front which will hopefully give customers more incentive to buy Henry's products.

That being said, I can see it will be quite difficult to promote understeer because upgrading the front suspension only, as you mentioned, will most likely cause oversteer. Since the car is pretty low in power (~170 HP) and is mostly used for cruising, I would say comfort is more desired than performance. Is there a way that I can keep the new SLA front suspension from being "too good" compared to the rear if we decide to take that approach?

 

[BrianPetersen]

Sure. Double-A-arm gives a lot of latitude for tailoring the camber curve.

Parallel and equal-length A-arms which are parallel to the ground will give a roll center at ground level (when it is at nominal ride height) and no camber change relative to the bodyshell.

Probably the upper arm will have to be a bit shorter, but even then, in the region surrounding nominal ride height, it will have no camber change relative to the bodyshell.

You can always make the A-arm mounts adjustable so that you can tweak their behaviour; it's easiest to do this on the upper mount. Lots of older American cars with upper-and-lower-A-arms had shims under the upper mount so that the height of the front and rear attachment points could be separately adjusted.

Always be mindful of the old saw ... you can make any bad suspension system work if you don't let it move. (very stiff springing and damping) It gets much trickier if you want both good ride quality and good handling characteristics.

Formula 1 cars have almost-parallel upper and lower A-arms ... and very stiff springing and a ton of roll stiffness.

 

[GregLocock]

The way I do it is to start with the known properties of the car and then work out what needs to be improved, what can be left the same and what can be allowed to degrade. Then I look at the proposed hardware and see if those targets can be achieved with it, or what else we'd need to change.

Then is the easy bit of designing some stuff to meet those targets and getting it built.

In your case you should be able to pick up some hints from Porsche's progress with the 911, how did they turn it from a white knuckle ride (reputedly) to a fast but dull beetle?




Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[OGR34]

So Brian, essentially what you are saying is to lower the roll center (RC) at the front to have less responsiveness compared to the rear as it will have a higher RC?
Also, is it common knowledge that semi-trailing arm suspension has excessive camber change, excessively high roll center, excessive side-scrub, and toe change with suspension travel like you mentioned?
And I guess in terms side-scrub and toe change with travel, it would be ideal to match the front's characteristics with the rear?

Thanks Greg, that sounds like a very good approach and I will keep that in mind. Right now I'm still waiting on Henry to get back to me as soon as he digs through some his files for specifications.

 

[OGR34]

Scratch the side-scrub comment. Lowering the front roll center should reduce side-scrub.

 

[MikeHalloran]

Perhaps you could just tell customers that the Citroen DS series and the Citroen SM all used parallel trailing links in front, and had excellent handling and super ride.
Squeezing a couple of hydropneumatic struts into that tiny roadster body would be a challenge.

Maybe one strut, transversely, with bellcranks.
... but that would take up all the trunk space afforded by the transverse torsion bars.
... just like SLAs will do to some extent.

I've always admired the front suspension of the Merkur XR4Ti (probably copied from something else), comprising a huge anti-roll bar whose ends also served as leading links, plus Macpherson struts, all cantilevered off the firewall. I.e., basically no structure forward of the firewall, save for cantilevered engine mounts and the forward clip itself. In this case, it might allow bigger footboxes and more luggage space, but lightening the front of a rear-engined car may worsen its bad habits.

"Eye candy" is not strictly an engineering requirement, except that when the idea is presented by The Boss, it is.

Have fun with the challenge. I might take it as an opportunity/excuse to make half a dozen different mini-prototypes, quarter scale with kart power, or tenth scale with RC and electric power, to explore the possibilities. Make 'em shiny, so the one The Boss likes can remain in his office or on his desk.



Mike Halloran
Pembroke Pines, FL, USA

 

[BrianPetersen]

Remember that your starting point (the original design of the vehicle) has the front roll center at ground level and no camber change (relative to the body) with suspension travel. In reality, you will probably want the front roll center a little above ground, which will be higher than the original vehicle had it. Most modern front suspensions have the front roll center a few inches (say 2 - 4) above ground and I wouldn't think this would be any different.

The roll center isn't the greatest thing to be concerned about, it's the camber change.

Semitrailing arms pretty much inherently have a high-ish roll center, quite a bit of camber change with suspension travel, and toe-in whenever they are away from nominal ride height (either up or down), except that compliance due to side loading tends to push them towards toe-out once cornering loads get established. The situation is better (less bad?) with semitrailing arms than it was with the swing-arm rear suspension that the VW rear-engine platform originally had, but there's a reason that Porsche went away from it and no one uses it (for new vehicle designs) any more.

Don't forget that the caster and the king-pin inclination angles also affect camber change, although that is mostly at lower speeds when significant steering angles are involved (e.g. autocross). This is when things start getting messy and I bow out of the discussion ... aside from pointing out that the consequences of a bit of oversteer at parking-lot speeds with significant steering angles involved are very different from those of having a bit of oversteer when cresting the hill on turn 2 at Mosport (100+ mph) when very little steering angle is involved ...

The old American muscle cars were designed to understeer. A little bit of investigation into their front-view suspension geometry and the slopes of the control arms, and an estimation of where the instant-centers are, may prove very interesting. If you conclude that on at least some models, the camber change is actually going the wrong way with initial suspension movement, don't be too surprised. I'm not saying you should do that, I'm just saying it's worth understanding why it was done that way ... Some modern MacPherson strut geometries have a bit of that going on, too, although on those it's usually only at extreme bump travel ... The OEM vehicle manufacturers would rather scare drivers into slowing down and failing that, have them go off the road frontwards, instead of encouraging them to drive faster and then go off the road sideways or upside down or backwards ... Fewer lawsuits that way. It is something you need to be mindful of!

 

[OGR34]

Awesome info guys, there are some very good suggestions here!
I'll keep you guys updated with progress as I am still waiting for Henry to provide vehicle specifications.
After that we can finally get started with this fun little project

 

[MikeHalloran]

The Big Boss, is, by definition, BUSY.
Do not get caught waiting, for anything.
Start. Now.
You can have some ideas, on paper or in steel,
to focus your discussion, and
to give him something to mark up
and to gesticulate over
when he does show up.




Mike Halloran
Pembroke Pines, FL, USA

 

[Overrun]

If you want your redesign to look good you’ll have to get the read swing arm defanged if this has not been done. The jacking problem is well known. Less well known is rather poor suspension control of rear toe. With power on the rear tires are driven to toe in. However, when you overcook a turn and lift the engine braking causes the tire to toe out, which is very bad.

 

[MintJulep]

Since it's a redo of a vintage car it might be historically interesting, and better yet, possibly technically appropriate to look at the evolution of the Formula Vee into the Super Vee.

If it weren't so expensive, this book might be an interesting read.

http://www.amazon.com/Formula-Vee-Super-History-Preparation/dp/083062063X

 

[OGR34]

Thanks for the replies guys.
Right now I'm trying to figure out a good value for %anti-dive. I would assume there is there a general value that I should aim for particularly since this vehicle is mostly for comfort driving. If so, is there a way to determine the ideal coordinates of my side view instant center (IC) relative to the front wheel to achieve my desired %anti-dive value? I'm also aware that anti-dive depends on %front braking but I believe that the Speedster just runs at full pressure when the pedal is pressed so I'm also a bit lost as to what value should be used.

 

[golfpin]

A bit late to add to this post and I am not quite sure of the relevancy so the moderators might shoot me off the page, but the issue of the VW torsion beetle suspension immediately rang a bell and I thought the following would be of interest. In 1969 Dr Helmut Marko set a record lop time in a Formula V around the Nurburgring of 9 min. 51 secs, In that same year same track the 908 Porsche of Siffert/Redman qualified for the 1000 kms race in 8 min 2 secs. Let us look at the tech diff in the cars, FV 1300 and restrictions/rules hp 56, tyres 15 inch beetle road drum brakes. Porsche Eng flat 8 +- 280 BHP tyres largest latest etc. lap time diff 68 secs that I think is about 12.59 % how bad can that parallel link front drum brakes, swing axle back end be! Bit tongue in cheek but interesting, me, I believe in keep it simple.
Golfpin

 

[NormPeterson]

Right now I'm trying to figure out a good value for %anti-dive. I would assume there is there a general value that I should aim for particularly since this vehicle is mostly for comfort driving.

You probably want the combination of [front] wheel rate from the springs and anti-squat to be sufficient to keep the car from hitting the bump stops under heavy braking. Wheel rate may be dictated by other considerations such as suspension frequencies tolerable to the anticipated customers and flat ride, which might cut out some of the iteration.

Norm