Suspension thread

[fastback33]

I figured maybe we should make a thread where we can sticky it add everyone can ask their questions here, on specifics components adn theories, i know i have a lot of questions so hopefully others will contribute too and if not, ohwell.

Okay, this has been on my mind for a while now, but how do you determine how much dive and sqaut and obviously anti-X that is required? When you determine your front dive geometry are you taking into account bump steer as well? Also how is bump steer determined and calculated?

Jersey Tom, i await your response, as well as everyone else!

[Jersey Tom]

Beats me!

Anti- geometry is one thing I haven't played with much... yet.

[RacingManiac]

The amount would stem from I think how much braking/acclerative force and tire grip your vehicle would have, which will give you how much load transfer occurs vs your spring rate(or really, dive/squat rate). As to how much, I can't say I am 100% on how the Anti-x works in my head, all I know is that they mess up your kinematic calcs especially if you are trying to design for roll center location. Historically for our FSAE cars we've incorporated some anti in the cars, but they are mostly carry over design. In 2006 though we've had about 25% anti-dive on the car and it puts enough loading into our thin a-arms that were bending them and we had to up the tubing size. We then since reduced in 2007 and eliminated that completely just to make the chassis jig easier to manufacture.

As to bump steer, most kinematic software will be able to spit out that number for you, or you can also do that fairly easily I'd think in any 3D CAD software

[Shredcheddar]

Free body diagram.

[fastback33]

Oh, good, so it is one of the things that is relatively unimportant. With respect to the design.

Now for my big question, kind of. Say i am not working with any known tire data (how hard is that too get by the way?) I have read to start the design of the car at the wheels and then work your way in as well as you also want to start at the roll center and....another factor i can't remember at the moment. But this is where i am. Any suggestions? Tom, i know you are designing your own car right now, do you have any pointers for a beginner?

[RacingManiac]

Tire data, can be hard to get. Tests are expensive to run in a test lab, and expensive to get without using a test lab(you can try to resolve the tire forces from data acq if you have strain gauges on your suspension links and/or wheel hub).

You do want to start from the wheel, if you have tire data the simplest you can do is try to see what the tire behavior is at different camber angle, slip angle, and at different normal load. Design your kinematics to suit such that at your anticipated peak lateral force, with load transfer and wheel load that your corner is at the "optimum" value for the tire you are using. Without tire data it would be sorta difficult....from the tire data it'll also tell you stuff like the ackermann behavior of behavior of the tire, which will dictate how your steering geometry is going to be laid out....

You might want to get somesort of a kinematic software which will make your life quite a bit easier since they can spit out most of the kinematic curves. Susprog3D is probably the simplest one to learn and use, also not very expensive...

[fastback33]

We, have a copy of mat**b (i censored it just in cae ) running around with some of the tire data for FSAE. I was just curious as to how you would do it without the tired data.

Thank you by the way!

[RacingManiac]

BTW, the process I've mentioned is basically a trial and error kind of process. Perhaps not entirely scientific but would be easier than starting from nowhere(and basically how we started the 2006 car, which was my first FSAE car as a design leader). If you have simulation capability you can also integrate all the curves you generated and perhaps run iterative sweep to try to find the optimum design.

[Jersey Tom]

You can certainly get a suspension that's "pretty good" without tire data. There are plenty of things that are universal... like low compliance, controlling bump- and roll-steer, etc. While you may not hit an optimization point on the first go around, you can also avoid totally f'in yourself on kinematics with some common sense.

Really so long as you have a wide range of adjustment (particularly regarding force control, and static alignment) without sacrificing rigidity I'd say you'd be just fine.

[RacingManiac]

Good point, I tend to try to package the most toe base and upper and lower ball joint seperation I can within the wheel package I use. Just to get the best installation stiffness I can. And from those points you pick the inboard points that gives you the desired camber/bumpsteer/whathaveyou curves you want....

[Ciro Pabón]

Thanks, Tom, cool. There is any kind of free software you could recommend, just to learn the basics? Where can you find an explanation of the theory? Is there a theory or it's some kind of micro-simulation (I mean, a general simulation of springs, levers and axles or whatever).

[Jersey Tom]

Hmm. Well.. you never know it all, I'm still figuring this crap out, but for learning the basics with an emphasis on kinematics...

Some things either to keep in mind or that might help:

1. There are very few universal truths... with kinematics, tires, setup, etc. I'm always wary anytime someone says "this is how you HAVE to do it."
2. No one knows it all when it comes to this stuff. Don't take what people say for fact until you prove it to yourself (if you have the resources to). Not from me, from Carroll Smith, Milliken, Rouelle, or whoever. Prove it to yourself.
3. The Carroll Smith "___ to Win" series are still really good for learning basic, practical. Milliken's RCVD is an excellent REFERENCE book afterward.
4. In my opinion, a lot of your kinematic design (for SLA suspensions) comes down to managing dynamic swing arm length. Your SAL defines your instantaneous bump- and roll-camber curve simultaneously! This is big. SAL is a really key concept and how it changes through suspension travel as you change a-arm parallelism and relative lengths.
5. This stuff isn't that complicated if you break it down into simple bits.
6. There are two ways of designing something, in my opinion. One is ground-up, which I consider "tinkering." Ie, you have some design.. "Ok let's try raising roll centers an inch and see what happens. It might make the car handle better." Or "Let's try moving the UCA chassis-side points down and see if that makes the camber curves better." The other way is top-down, which I consider "engineering." Ie you SPECIFY how you want the car to react, and everything else just kinda falls into place. Ie "Ok I want the static roll axis inclination to be X, height above the CG to be Y, and the front and rear camber curves to look like (draw your favorite curve)." That defines a LOT of the variables as to where stuff has to go. It does the work for you!

That said you can get an idea of how the linkages work in simple CAD software. You can make a 4-bar linkage sketch in Solidworks and play with control arm lengths and crap and see what happens.

So yea. That's for the "position-control" bit in kinematics, assuming an infinitely rigid suspension. Then there's the "force-control" bit of kinetics which is equally as critical.

[fastback33]

What are the acronyms: SAL and UCA?

[Jersey Tom]

Swing arm length, and upper control arm.

[Carlos]

Shame to see Metaldyne file for bankruptcy. We'll probably never see Tom Wagner's X Suspension System in production now. Here's a discription of this innovative front suspension http://www.autofieldguide.com/articles/030705.html