Matlab for tire data/suspension design

[fastback33]

So, i have mat lab with some tire data and am curious how I am supposed to get the data into a readable/understandable format. Secondly how do i design to a specific parameter? When do i want my maximum slip angles for grip to take place etc., etc.? Any advice would be great!

[Belatti]

So, we are two that are in the same discovery road!

Another ear wide open here

[Jersey Tom]

What is your background? How much suspension design have you done before? How comfortable are you in vehicle dynamics?

You can go a long way and design a good suspension without ever touching tire data, and if it's relatively new stuff I'd avoid overcomplicating things. Otherwise you can easily get way too caught up with tire data while missing the big picture.

With regard to answering your initial question, you have to be a little more descriptive here. What tire data do you have? Is it tabular, or empirically fit? Where is it from? Do you know if its any good in terms of accuracy? Or could it potentially be junk? What channels do you have available (ie FX, FY, FZ, MX, MZ, RL, RE)? Is it at a range of appropriate loads? Cambers? Speed? Inflation?

Again, if I were you and this is a first time thing I'd start off without the tire data. Absolutely critical things to consider...

• Using a wheel that's very rigid. Avoid cheapo 3-piece wheels. Unsprung mass be damned, toe- and camber-compliance will really hurt grip and handling. It's hard to quantify the effect of unsprung mass on grip (in fact, I don't know of anyone that can do it at the moment. Myself included. Most people prefer handwaving). But you can put real eye-opening numbers on camber compliance.
• Likewise with the above, designing the upright and sizing the bearings to really minimize compliance. You'd be amazed how much load goes into joint fasteners and get's resolved into bearing races.
• Really work to minimize bump- and roll-steer. You want the driver to steer the car, not the car to steer itself (and be a function of your overall springrates!) Though there are certain times you can use roll-steer to your advantage it is a pretty niche area.
• Work to have a linear installation ratio rate, unless you're comfortable with the effect of progressive or digressive TLLTD on handling!
• Size your installation ratio for available spring and damper rates.
• Select a good damper. They are expensive, but worth it.
• Size ARBs so they actually make a difference on your TLLTD.
• Make everything so it can be changed out QUICKLY. Driver-adjustable bars are great.
• Ultimately, IMO, most of the actual A-arm pickup geometry comes down to controlling dynamic FVSAL

[Belatti]

Thanks Tom!

Please would you complete this glosary?

TLLTD:....
FVSAL:....

[Jersey Tom]

Total lateral load transfer distribution

Front view swing arm length

[fastback33]

Okay, i will take your advice. Though the data i have is from the TTC, so i know it is useful because generally every team uses it or has it available to them.

But basically this is how i feel comfortable starting:
1. String calculator
2. once i have my instantaneous roll centers set, i design them and my center of gravity, to get as close as possible to one another.(btw is that how i design for roll steer?)
3. I understand bump is very important under braking, so how do i design for this?
4. Roll steer ( I don't know if i have read about this in tune to win? <---on my second time reading it btw)
5. Try to get everything as low, and as close to the center of the car as possible so as to reduce the secondary moment of inertia?

One last thing, i feel i can't get a straight answer from anyone i ask....
and this very much relates to #2, but basically I know that the further my roll center is from the CG the larger moment arm i will have. BUT, After Reading TTW it seems like carroll is telling me something different, I forget which page it is on but the chapter is #4 (suspension Geometry).

As you can see, i have a lot of questions and the more i learn the more it becomes evident that the less i really know.

[Jersey Tom]

Ok, so you're designing a FSAE car then. You hadn't mentioned that. Not sure what you mean by "string calculator."

In any event... gotta crawl before you can walk, before you can run. Gotta understand the basics before you go off the deep end with tire data. Since timing is critical on these cars, I'd leave tire data out of the design loop as a critical path item until you've gotten the suspension basics down. Would be good as a separate study for a sophomore or junior. It should be pretty easy to get the stuff working in Matlab. In fact, there's documentation on the DVD that explains how to do it!

Additionally, just because a bunch of teams have or use TTC data doesn't mean its perfect or even good. Real world test data seldom is. In fact I've seen all the TTC data and I know very well it has some definite weaknesses and you can get yourself in trouble if you just use it "straight out of the box."

With regard to the other points...

You do GENERALLY want the CG to be as low as possible, to minimize load transfer. Sometimes not! For the acceleration event you probably want a quick way to raise the CG as high as possible to get every ounce of load transfer to the rear wheels.

Why do you want the roll axis close to the CG height? (Personally, I wouldn't want it anywhere near). Keep in mind, body roll is NOT inherently bad. IMO more important than roll axis height (assuming its in a reasonable range) is roll axis inclination and how much that changes, since that will effect handling balance.

Bump is very important under braking? What do you mean? Bump-camber rate? Bump (compression) stiffness in the dampers? General force variation going over bumps?

Roll-steer is just the amount of steer (toe) change you get per unit body roll. Likewise bump-steer is the amount of steer change you get per unit vertical wheel travel. I like to minimize both, so the slip angles are purely defined by driver input (and chassis sideslip, and yaw rate..) and the car isn't trying to steer itself around.

With regard to your last question, you should probably find the exact C. Smith line before asking. Your total amount of load transfer is a function of the CG height above the ground. The FRACTION of that that goes through the springs, is a function of your roll moment arm divided by that CG height. But again, keep in mind... body roll is NOT inherently bad. In fact you probably do want some amount of it on these cars or else (a) your kinematics wouldn't do anything (b) your dampers wouldn't do anything.

Biggest thing to keep in mind is you are an engineer. That means specifying and designing a system to behave like YOU want it. Not like Carroll Smith, Milliken, or Claude Rouelle want. No one (certainly including myself) has the complete picture or understands everything. Don't take anything at face value. Don't use it unless it makes logical sense to you.

[Belatti]

Fastback,

is this what you have read in TTW?

[fastback33]

Yes! That is what it was i believe. However after rereading the suspension section on "jacking" I found what I was looking for. The situation is more complex then i had initially thought it to be. Basically what I was trying to understand is why you want the roll moment, and why you do not want the RC exactly over the CG. So, someone please correct me if im wrong, but....I've found you want the RC somewhere below ground, and not on the CG as, the closer the RC is to the CG creates some kind of jacking force on the outside tire, which is basically lifting that wheel/tire off the ground or possibly unloading it. This makes the inside tire do a lot more work then it should. So basically having a high roll center/moment is not a good thing, but more so, when looking at the RC you need to look at it where it is relative to the ground plane and NOT the CG. So I had it all wrong, i think, someone will hopefully reassure me...soon.


JT - It isn't necessarily for FSAE but more so for my self and designing my own car(s). Which is why im uncertain if i should ever use tire data, and how common is it in a racings eries for them to use the data for their design?

[Belatti]

So, someone please correct me if im wrong, but....I've found you want the RC somewhere below ground, ...

...

Which is why im uncertain if i should ever use tire data, and how common is it in a racings eries for them to use the data for their design?

The STATIC roll center should not be below ground. That way the rolling moment grows too much. You have to place it somewhere between the Cog and the ground. How to go with the position the roll center takes dynamically is something Im on with.

Here, racing teams do not use tire data and I could get some tire data test report, so Im looking for a team who wants to hire an engineer and invest some time and money to design next years car

[Belatti]

... I could get some tire data test report, so Im looking for a team who wants to hire an engineer and invest some time and money to design next years car

Well, I have some problems with some of the tire data graphics I got from the tire manufacturer. Some of them I havent seen in the books I read.

Here, a list of the seven graphs I have:

1- Self Aligning Moment (Mz [Nm] vs alpha [°])
2- Residual Moment (Mzr [Nm] vs alpha [°])
3- Pneumatic trail (t[m] vs alpha [°])
4- Lateral Force (Fy [N] vs alpha [°])
5- Cornering stiffness (Kd [N/°] vs Fy [N])
6- Aligning stiffness (Kz [Nm/°] vs Fz[N])
7- Camber stiffness (Kgamma [N/°] vs Fz[N])

In black the curves I cant get well.

Any help is welcome.

[Jersey Tom]

All of those plots have meaning and are valuable one way or another (aside from #2, there's no way I know of to separate RSAT from slip angle [alpha]. Makes more sense as residual moment vs gamma).

However, some of them are redundant. If you have #1 and #4 at various loads and cambers you can generate all the rest... they don't present any new data. Of the ones in black, #5 is probably the most valuable in terms of evaluating balance change in sub-limit handling and on entry/exit maneuevers.

[Belatti]

Thank you Tom!

Like if, as you say, having #1 and #4 at various loads and cambers you can generate all the rest... can you tell us how, please?

As I said, I have got nothing in the bibliography I own and Im messing up myself a little bit with these.

[Jersey Tom]

#2 - I wouldn't bother with.

#3 - Pneumatic trail is just (Mz / Fy)

#5 - Cornering stiffness is the slope of Fy vs SA @ 0 SA. Generally done by a 1st-order fit between +1 and -1 degrees SA. The slope is cornering stiffness, the offset is a combination of plysteer, conicity, and camber thrust (which you can then use for #7)

#6 - Aligning torque stiffness likewise is the slope of Mz vs SA @ 0 SA. Evaluated the same way as cornering stiffness.

All of the above is meaningless in and of itself.

The real question is, what are you trying to do with the data? For that matter, who is the tire manufacturer?

[Belatti]

Thanks again Tom!

The real question is, what are you trying to do with the data? For that matter, who is the tire manufacturer?

I´ll have to define better my scope.

My intentions were to design a suspension geometry or to be able to aid a car setup selection using tire data. Something I have heard millions of times happens in F1 and other major racing series and that we (in Argentina) are far from doing, except for a couple of prestigious designers that even are F1 consultants.

The 90% of racing series here use a control tire from a local manufacturer that has a very basical and artisan-made product. Sizes vary a lot from tire to tire and there is no machine testing data availiable. Development is made on-track with teams when some change is introduced in the tire.

The other 10% uses Pirelli P-Zero and thats all I can say, at least here

A simulation in Matlab using the curves I have got seemed a nice way to start working by myself and spending 0$