Question regarding Roll Centers of a Double A-Arm suspensions

[MiquelSants]

Greetings.

As per the title, the question is as follows:

If I'm facing the car up front, the procedure to determine the IC's and finally the RC seems simple:
- extend lines from upper control arm and lower control arms, check their intersection point
- repeat the step above for the others side
- from the IC's, extend lines to the center of the contact patch of each tire
- where these lines intersect, we should find the RC's

However, on a double a-arm assembly (as depicted here:Link) , each arm (upper or lower) has 2 different attachment points to the chassis.

As such, how to draw the lines from upper and lower arms (so they intersect (if they intersect) and we can find the IC's), considering that there are 2 attachment points to the chassis?

I am aware this is a "basic" question, but I'm trying to figure out exactly how to implement this in my tools (I'm a also a programmer). I am also intent on checking the 3-dimensional coords of IC's and RC's, and not only the usual width & height locations of these.

Thank you for any help.

MS

 

[BUGGAR]

I couldn't access your link.

There are several suspension analysis programs out there (I like Vsusp, it's free). Most of these are 2d but some are 3d, considering the interaction of front and rear roll centers, etc.

But I believe there's always room for more and better. Suspension analysis is a fascinating subject and seems like an endless subject for research.

ps: I just found out about Ollie's rule for mitigating roll center migration. But some say this is not important(?!) I'm currently involved in design of Lotus 7 replicas and suspension design is so complex that I'm considering just using an infinitely (within reason) adjustable suspension system that is tunable to the track.

Bob

 

[MiquelSants]

Hello, Bob.

I checked the link. Carbible for some reason prohibits viewing of the image directly. One has to go directly to the page itself:

http://www.carbibles.com/suspension_bible.html

The image is just below "Coil Spring type 1".

I tried vSusp a while back, didn't quite fit my own needs. It presents a front view only, and I need a side and upper view.

Regarding the issue I am describing, I want to know if upper and lower arms lines come from the front/rear attachment points or a line directly from the ball joints (upper and lower) to which the wishbones are attached.

 

[BrianPetersen]

The line goes between the pivot point centerlines. For the ball joint, that's easy. For the A-arm, imagine a plane that is perpendicular to the pivot axis and which passes through the ball joint. Wherever the pivot axis passes through that plane is where the inner end of that line is in front view.

The lower arm usually has its pivot axis horizontal and parallel to the centerline of the vehicle (or very close to this), so that one is easy. The upper arm often has its pivot axis parallel to the centerline but sloping down towards the back. Sometimes, the upper arm pivot axis is not parallel to the vehicle centerline. Then things start getting complicated.

 

[MiquelSants]

Ah, bingo, Brian!

That clears it up for me.

As for non-parallel (to centerline) pivot axis, in such a case the best I can get is a good, honest estimate (ballpark figures, that is)...or increment the complexity of the calculations.

Thanks, Brian.

 

[JackAction]

You must find the planes of the control arms. Where those planes intersect, you find the instant axis of rotation. Where this axis intersect with the vertical front and side view planes - passing through center of the tire contact patch -, you find the instant centers of rotation.

In the attachment, you'll find a representation of the concept as described in Milliken's Race Car Vehicle Dynamics.

Here's how you find a plane with 3 known points (the ball joint and pivot of the control arm);
Here's how to find the instant axis location knowing the control arms planes;
Here's how you find the ICs with the known axis and either the front or side view plane;

 

[MiquelSants]

Thank you JackAction.

Good heads-up there.

 

[fpaccoret]

Bob,

With regards to your question about roll center migration it becomes increasingly important with higher roll gradients. You will find that your roll moment variation with roll can affect LLT with similar detrimental effects like nonlinear wheel rates etc. Some of this can be designed for, but for simplicities sake, stick to the basics. Vertical migration is the killer here, this can affect your total overturning moment like i said as well as affect jacking force. Anyways, i digress…

BTW im a personal fan of W.C. Mitchell software: Wingeo ~3d and allows for anti-geometry analysis.

Regards,

FP

 

[GregLocock]

A couple of points. IC is governed by the motion of the contact patch, so once you know that you don't need to use any rules of thumb about intersecting planes and wheel centres.

Secondly RCH is very dependent on ride height, and almost any car pitches as it rolls, so worrying about niceties of pure roll RC migration is chasing a rainbow.

In practice force based roll centre height is far more useful. = track/2*dFz/dFy

Cheers

Greg Locock


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

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

 

[BUGGAR]

There seems to be some controversy over the importance of roll center (moment center) migration. Just what happens when the roll center moves laterally or vertically. Does it make the vehicle handling unpredictable?

Somewhere out there, there must be a car with suspension geometry allowing a wandering roll center. How does she drive?

Bob

 

[fpaccoret]

What Greg mentioned is correct, your Force Application Point is what governs loads on your suspended and unsuspended mass.

 

[BrianPetersen]

All cars with MacPherson strut front suspensions have roll centers that wander all over the place. That covers practically everything in the last four decades that has front wheel drive plus quite a few rear-drives.

If the car is in roll while cornering, that's where a MacPherson geometry is the most whacked. The instant center for the loaded outside wheel is likely nearing infinity somewhere on either side of the car (it could easily be on the "wrong" side!) and the one for the inside wheel is likely up high and close in because of the droop in the lower control arm. (Draw a diagram - you'll see.) But the messed up instant center for the inside wheel doesn't mean much, because that wheel is less loaded. This is why the force based approach makes more sense.

Lots of cars with MacPherson drive fine. Lots of them don't, too. But ... "upper and lower wishbone" can describe the front suspension of a Formula 1 car, or a 1955 Chevy.

 

[MiquelSants]

Many thanks to all of you for the enlightening replies.

Greg,

I agree that, ultimately, force application is what determines both the contact patch motion/shape and RCH. The only thing is, determination of lateral and longitudinal forces is obviously not simple: aside from the suspension and tire variables, one has to contend with aerodynamics effects, and these are dependent on many different factors (speed, attitude/AoA, shape, etc, etc). So, best case scenario, you get an envelope of performance.

Hence my need to have a basis to work with - albeit being more of a geometric nature.

In any case, good heads-up there. Thank you, Greg.


MS

 

[CWAnthony]

Hi Bob (BUGGAR),

I picked up your described plan to build a 7 with infinitely adjustable geometry. Have you got any concrete plans for this? I've considered doing something similar myself, as a learning exercise, because my own experiences of "roll-centre tuning" in education and in industry have been less than convincing.

It would be good to discuss this and share some ideas.

Regards,

Chris

 

[BUGGAR]

Chris,

I've actually started working plans for what I call Das Boot, since it looks more like an aluminum boat than a car.

If you can pm me or whatever you do on this site, I can share some sketches of ideas.

Bob

 

[CWAnthony]

Bob,

Thanks for the reply. Sounds interesting.

I can't actually find a PM feature on this site I'm afraid (I may have missed it?!), could you email me at - chris . w . anthony [at] hotmail . co . uk ? I'm very keen to see what you've come up with.

Best Regards,

Chris

 

[BUGGAR]

Chris,

I was unable to use your e-mail. So I'll attempt to download the sketches here.

Bob

 

[BUGGAR]

Here's the other file. This forum is just as poor with its attachments as the locost one.

 

[CWAnthony]

Thanks for the posts Bob, the sketches look good. What have you used to generate them?

I've not come across Olley's rule before, I'm keen to give it a try on a CAD sketch generator tonight. Is this mentioned in Milliken's Chassis Design?

So which parameters will you be looking to change when you build the car? Roll centre migration only, or static roll centre heights too?

Have you got any ideas as to how the chassis brackets will physically accommodate the adjustability yet? This is the biggest problem I am trying to overcome in my designs, especially with regards to trying to ensure compliance levels are both reasonable and consistent between settings.

Also, what Locost forum are you on? I'll see if I can PM you in there. Strange that my email doesn't work, will have to try and sort that!

 

[BUGGAR]

Chris,

Try me at rgarner2@san.rr.com

I've got an article on Olley's rule.

I was on the locostusa forum but that forum can't accept pdf's so I gave up on them.

Bob