Axel/Suspension Geometry Rule of Thumb?

[kincaid05]

I'm designing a custom car, just for fun, and I need to settle on a suspension geometry pretty soon. One thing that's holding me up is the type of joints and alignment to use on the drive axels (half shafts).

The transmission is a 914 and the track width at the rear is ~56" to wheel centers. I still haven't decided on a rear upright, i may make a custom one. But any suggested uprights would be investigated.

I'm looking for a rule of thumb for where to put the axel joints relative to the suspension pickups. Like the steering rods, and the rule for reducing bump steer by placing the joints in line with the arms, i'm looking for a similar rule for the axels.

One limitation is that i can't put the pickups too close to the transmission or i wont be able to fit the transmission into the frame (trans is not a stressed member). This means that the in board u-joints would be a few inches more inboard than the inboard suspension pickups. I could imagine that i'd want to apply the same offset to the outboard u-joint from pickups, to make all the links travel in the same arc.

Basically I dont see how one could design as system which could utilize two u-joints and no cv joints without binding, and still have the flexibility to get good suspension geometry.

Also, does anyone know off hand what kind of joints they used in the 914 rear axels?

Thanks a lot!

 

[NormPeterson]

You have to let one of the axle joints or some other design feature allow some plunge. Most live axle suspension arrangements wouldn't work very well either if this couldn't be done.


Norm

 

[BrianPetersen]

I wouldn't suggest using any "rules of thumb" in this case. You need to work out the situation and the geometry properly. And honestly, I don't understand the question that you are asking. Maybe by putting thoughts out there, you can re-ask the question you meant to ask.

Halfshafts ordinarily allow not only angular motion but also have splines or joints designed to allow the shaft to extend or retract; this way the suspension hub has complete freedom to move around within the bounds defined by the travel limits of the joints or the spline. Figuring out how long the shafts need to be in order for this to be true throughout the travel of the suspension ... is up to YOU. There isn't a "rule of thumb". You have to go through the design and check what's required. (There are exceptions, of course; e.g. Jaguar and early Corvette suspensions in which the double-jointed shaft had a fixed length because it also acted as one of the suspension links. Heaven forbid one of the joints failing with that design, though! And it requires the shaft and its mountings to carry axial loads. Not all differentials will cope with this.)

Universal joints or CV joints don't like being "stationary" while transmitting power, so ordinarily in an independent suspension design with double-jointed shafts, the transmission will be mounted slightly ahead of the axle centerline so that nowhere in the normal travel of the suspension are any of the joints absolutely straight. But there isn't a hard-and-fast rule. If you position the diff centerline an inch or so ahead of the axle centerline, that should be enough. If you want to be persnickety ... get the required minimum angle recommendation from the manufacturer of the joints, and go from there.

Conventional (non-constant-velocity) joints will only work without transmitting a lot of torsional vibration if the design of the suspension keeps the hub centerlines parallel to the differential output shaft centerline throughout the travel of the suspension. This is only possible with suspension designs such as pure trailing arms or parallel and equal-length upper and lower arms ... but this is not good for handling characteristics. Even in cases where the suspension design does give camber change (normal), there have been many cases where a "cheat" has been to select the location of the transmission and the normal ride height of the car so that the transmission output shaft is at the same height as the hubs, so that the two joints have equal angles at normal ride height and goes away when the ride height is away from normal, and either not worrying about the consequences or putting some sort of vibration damper (that gives some torsional flexibility) in the drive shafts. Failing this ... Use CV joints. Every modern front-wheel-drive car has double CV joints on each shaft.

If you are dealing with a defined transmission and defined hubs, then there isn't a lot of choice in the location of the joints in the cross-car direction. This transmission is for a longitudinal engine, so it must be on the centerline of the car.

Porsche 914 = glorified VW Beetle underneath and the tranny is similar/same, so the shafts ought to be similar. Plenty of places sell parts for air-cooled VW's and they should be able to help you out.

 

[kincaid05]

Im familiar with the wide range of suspensions and axels you mentioned. Especially the 71 vw ifs. I understand that the cv joints allow a plunge within range and it's imparitive that the suspension usage not violate that range.

I guess i was just hoping for the perfect solution. Since the design is from scratch, there should be no comprimises. I always considered CV joints to be less than ideal due to their complexity. If it were possible to create a suspension whos axel moves with the control arms, yet not bearing any forces like the control arms, then i'd try to strive for that solution.

So on the inside i'd be using the VW/Porsche CV or some beefed up version. Is it imparitive that i use an identical or symetrical cv on the outside? or would it be best to again use the cv designed for the minishaft on the outboard? i was reading some other threads on this site about vibrations of improperly positioned u-joints. The CV doesnt have this issue?

Also, I was kind of concerned about the small offset between my trans drive line and wheel center lines, but now you guys have put that concern to rest.

The car i'm designing is microscopic. I figured it would be best to strive for f1 style suspension. How are their axels and joints configured? Though they do have that beautiful stressed trans case <3

 

[kincaid05]

Also, any ideas where i can find a reasonably priced pre-fabbed upright for a 1500lbs car? Or any recommended donar cars. I considered the corvette. I'll have a look into that now. It could take half a year to design a proper upright from scratch though so i'd rather not.

 

[BrianPetersen]

What do you mean by an "upright"? To me, that just means something - anything - that is straight up and down. (Utility pole? Sign post?) Do you mean the damper and coil spring assembly, commonly called a "coil-over"? Or do you mean something else?

CV joints don't have the angular alignment issues of traditional U-joints.

I think you will find the most economical and suitable half-shaft solution to be one that comes already assembled with the inner and outer joints. Since you are dealing with a VW transmission, I'd strongly suggest staying with the VW half-shaft assembly that is meant to go with that gearbox. This way you know the joints are designed to work together - and it will also make life a whole lot easier should you ever break one or wear one out.

F1 cars use rather ordinary-looking halfshaft assemblies with CV joints. They're probably optimized to the hilt and made of unobtainium in the interest of lightness, but the arrangement appears conventional.

Your bigger problem with a mid/rear engine car is going to be containing the oversteer.

If you are designing the rear suspension from scratch and it is rear or mid engine and rear drive, then a multi-link is the first choice. (Every serious open-wheel race car uses some variation of this.) You can orient the links to tilt the wheel in during bump travel but not tilt it out (much) during droop travel - in front view, upper links shorter than lower links and not parallel. You can arrange for the suspension to toe-in a little during bump or droop travel to hopefully keep the oversteer under control during body roll. You can very easily arrange it for anti-dive during braking.

The little smart car uses a de-Dion axle (a rigid axle but not containing the diff - the drive is through CV-jointed halfshafts from the transmission). I mention this because that car is designed to absolutely contain oversteer. The rear wheels are cambered in slightly, and the rigid axle maintains this no matter the body roll during cornering and no matter the load inside the vehicle. It has way fewer links than an independent design. You might want to stick your nose underneath one of those cars to see how they did it.

 

[kincaid05]

Thanks, all your suggestions on the links sound inline with my own reasoning. One thing i have not invested much thought in is toe-in on bump. Thanks for the heads up.

You mentioned that oversteer is uncontainable. Does this mean inevitable? If i understand correctly a car can be considered oversteering when the rear end has broken loose and the nose is pointing in more towards the center of the curve, sometimes requiring counter steer to maintain the cars trajectory. Correct?

This scares me because i've been concentrating on moving the passengers and misc. mass forward in the car to acheive approximately 50/50 if not slightly heavier in the front weight distribution. This sounds like it's only going to induce more oversteer. Is the ideal weight distribution with slightly more in the back then? Or am i not correct in assuming that a heavier front will cause oversteer?

I do know, however, that either the front or back can be detuned to balance the oversteer or understeer effects of the other end. It doesnt make the car perform better, but does make it more "controllable" and less prone to serious error situations. In otherwords if i get in an oversteer condition, i could detune the front to understeer to give a neutrally sloppy steer But i could be completely wrong, and obviously that's not an ideal fix.

To answer your question. By upright i meant the rear axel carriers. The vertical section connecting the outboard aarm ends and supporting the wheel bearings.

Keep sending your suggestions, i'm all ears. Thank you.

 

[BrianPetersen]

I didn't say the oversteer was "uncontainable", I just said that it's going to be your biggest challenge to overcome with the engine in the back. If you do stuff to transfer weight forward, that will help *reduce* oversteer. The oversteer will be trickiest if there is too much weight hanging out beyond the rear axle and too much weight on the rear in general. If you can get 50/50 weight distribution, that will help a lot ... but it will be a challenge to achieve this with the engine in the back!

You are entirely correct about the ability to "de-tune" the front suspension in order to preserve understeer. In your original post, you asked about a "rule of thumb". The only "rule of thumb" is that the end of the car with the most weight, is going to need the most sophistication in order to preserve grip! A front wheel drive car can get by with a simple rear suspension, but it needs sophistication up front to avoid understeering like a pig. A traditional front engine / rear drive car can have both ends comparable. A rear engine car needs sophisticated rear suspension and sometimes primitive front suspension in order to be controllable, and a brief review can show some successes and failures:

Original Corvair (double wishbone front, swing axle rear) - had problems.

Second generation Corvair (double wishbone front, multilink rear) - was quite good. The rear suspension on these cars is not unlike that of the first generation Corvette.

Hillman Imp (swing axle front, semitrailing arm rear) - supposedly decent in its day, but this was done by using simple rear suspension and extremely primitive front suspension.

Porsche 911 has always had MacPherson front, but started out with semitrailing arms in the rear and they oversteered madly; the new ones have multilink rear and are easier to control, although they also have electronic stability control.

smart uses MacPherson front, de-Dion rear to ensure that the rear wheels are always cambered in. They don't oversteer - but they also have electronic stability control just in case.

 

[GregLocock]

Could you sketch out your approximate inetentions?

If you want a short spindle double wishbone then use a FWD spindle with a double wishbone adapter, or say a Mustang Cobra IRS.

i don't know of a tall spindle double wishbone that is especially suitable.

If you want a strut, ie some sort of Macphersonny thing, the again any FWD spindle is a good start.

But a sketch is the place to start.



Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[hemipanter]

I might add that the splines in the axle does not slip very easily under high torque, so a locking of the suspension movement to a certain degree will result, if the camberkompensation is great.
Goran

 

[BrianPetersen]

Should also note that MacPherson as rear suspension has gone out of favour, even on front-drive applications. The camber change with wheel travel is exactly the opposite of what is wanted to maintain grip at the rear (i.e. you want camber gain in bump travel but MacPherson has less/none compared to unequal upper and lower links, and you don't want camber loss in droop travel but MacPherson tends to have a lot, leading to a tendency for lift-throttle oversteer).

 

[patprimmer]

What's wrong with a bit of lift throttle over steer. It can be used to very quickly wash off speed and continue to turn in when in trouble mid corner.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

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

for site rules

 

[BrianPetersen]

For an experienced driver and on closed-course conditions, that's one thing.

For the average driver who overcooks the exit ramp from a motorway in wet conditions and lifts the throttle in response, and then the rear starts sliding causing the driver to lift off completely making the oversteer worse - or stabbing the brakes - that's quite another.

"Your honour, the subject of this class-action lawsuit is the design of the suspension on this vehicle ..."

 

[kincaid05]

Thank you all. Your advice has been very helpful.