Where to start suspension design

[Ericyu]

Hello,
I am new to suspension design, so please get me some help.
I want to know where to start, how to determine the parameters? and witch program is for this job, and where can I get some useful knowledges? I am in China, some of those suspension books can't be found here.

Thanks

[Ericyu]

Hello,
I am new to suspension design, so please get me some help.
I want to know where to start, how to determine the parameters? and witch program is for this job, and where can I get some useful knowledges? I am in China, some of those suspension books can't be found here.

Thanks

[exFSAE]

See if you can dig up Carroll Smith's "Tune to Win" to start with. Good basic stuff.

To do a legit suspension design, you need tire data. If you don't have tire data, its a crap shoot, but there's some basic rules to follow. C. Smith would have 'em.

There's a wide variety of suspension software available on the market. But again if you don't have specific targets in terms of tire loads / inclinations / slips to aim for, I don't see much value in spending the $500-1500 or whatever on software. May as well just CAD it up and do some simple linkage analysis.

Keep in mind, moderately successful cars have been built with this level (or less..) CAD resources..

http://i292.photobucket.com/albums/mm19/exfsae/Rearupright.jpg
http://i292.photobucket.com/albums/mm19/exfsae/coolshit.jpg
http://i292.photobucket.com/albums/mm19/exfsae/goode.jpg
http://i292.photobucket.com/albums/mm19/exfsae/iso.jpg

Though that may have been the car with minor driver fitment issues (aka only the smallest guy on the team could fit in it.. not sure how we got around that at comp. Was before my time!)

[Chris Lane]

You really need some form of CAD software, whether it be 2D (AutoCAD) or 3D (Solidworks, ProE, Inventor).

You really need Microsoft Excel to do your basic suspension calculations.

You need some form of kinematics software as well to do 3D analysis (Try Susprog 3D or Optimum K. Both companies do SAE discounts as far as I'm aware).

You need to read some books. Try Carroll Smiths books as exFSAE suggests, or Staniforth's book 'Competition Car Suspension' which is a great beginners book to get you started. The cheapest place to get these is most likely Amazon. Buy them on credit card and have them shipped to you.

Try to find someone local who has done this before and hassle them as much as you can with questions. Its invaluable!

Stick at it mate, doing suspension design from the ground up with no prior knowledge is really tough. I've *just* been there myself and my first suspension system is being manufactured at the moment. It is really nerve wracking (I hope I have everything right), but it'll be worth it when its all built and running.

Good luck mate, and don't be afraid to ask questions!

[exFSAE]

+1 to wrapping your head around suspension design being a PITA. It took me 3 years to get to the point where I confidently THINK I know what I'm doing.

First, we had to figure out all the "bits and pieces." Camber, caster, toe, slip angle, slip ratio, spring rates, damper rates, wheel rates, cornering stiffness, load sensitivity, Ackermann effects, a-arm configurations, bump steer, roll steer, roll camber, pitch camber, compliance, mechanical hysteresis, etc.

With the turnover rate we had, usually our suspension guys would barely get beyond that. They'd move a pickup point an eighth of an inch or so and see what happened. We had people making design decisions based on crap like "well we must want more Ackermann.. more slip angle means more force!" Our suspension design was just guess and check.

After that, we took some stabs at designing around those parameters. Stuff like "Well lets put the roll centers at the ground so the chassis doesn't jack" or "Let's put the roll centers higher like on the 04 car so it handles well" or crap like that. Eventually in 07 I got my hands on some tire data, and not knowing what to look for just started plotting shit. Took the first stab at designing around tires saying "These things are so camber sensitive [wrt cornering stiffness], and with these smooth tracks there will be basically no single wheel bump.. let's try a really short VSAL to keep the tires at 0 inclination mid corner". At that point I was at least starting to look at the picture, but not the whole movie (to paraphrase Claude).

Finally now that I've seen more and more tire data and done a full suspension.. I've got a completely different perspective on suspension design. I don't design for the sake of keeping roll centers stationary or VSALs long or what have you. It's all about the tires. You want max braking, max drive, max cornering, max combined, with smooth predictable transitions. Take anything Smith or Milliken or Haney says about tires with a grain of salt. There are few if any global truths about tires. Some like camber, some don't, some are more heat tolerant than others, some will hold together pretty well if driven nicely, some will just shred regardless. And yes, you can have too soft a compound even on these cars.

Anyway. With some DAQ logged data handy you should have a rough idea of accelerations seen (or even by just previous comp skidpad and accel times and backing it out). So you know at any point roughly how much total WT you'll have.. you get to then play with stiffness and distribution and kinematics so at any point in time you've got the tires loaded and pointed the way you want them. If that means running some weird geometry that no one else does, so be it! Good in fact! One of my half serious mottos was "If you see it done on a previous years car.. don't pay any attention to it at all because it was probably done wrong or not for the right reason." If that means roll centers move around a bit, who cares [so long as your suspension isn't doing really dumb stuff, like inverting itself.. seen that on an old car..].

And that's just a snapshot of pure mechanical grip. There's much more on actual handling, response gains and linearity, etc, that I won't begin to get into.

Plus there's ride dynamics if you want to get into that..

OR.. if you don't have multiple years to really figure this stuff out.. get a Carroll Smith book, get a rough feel for what generally works, and go build it. In any event I'll stop here and wait for Bill Cobb to show up..

[Anvit Garg]

Are there teams that are successful without the tire data?

I do not understand why everyone emphasizes the use of tire data to start design (The tire data analysis seems to make more sense in the tuning for maximum performance phase):

What happens if you switch between wet or intermediates or dry? Dont they all have different points of optimum slip versus etc?

It doesn't sound right to me to only design the car around one tire. Do FSAE teams only design their vehicle to be adjustable to the maximum of the optimum specifications of just one tire?

Is it common for a team to design with maximum adjustability in mind instead? (Have adjustable mounting points of all your suspension)

Please don't flame, I am just asking questions because I am still in the learning phase.

Thanks

[Wesley]

The success of a racing car comes down to one thing - optimizing the tire contact patch. You can make a thousand horsepower, have a thousand pounds of downforce, have the lightest car ever, and if you can't get your tires to stick properly, it's all for nothing.

Understanding what the tire wants (in terms of camber sensitivity, load sensitivity, cornering stiffness, etc.) will mandate what your suspension setup will be. Choosing a tire is like choosing an engine - if you want to change, it's going to take some redesign.

Rain tires are making a compromise anyways, and the water is going to kill performance enough that not optimizing for rain tires is going to make little difference in the feel. You're still in the rain. You can tune the suspension for rain settings, but the car is designed for a dry track.

It all comes down to the tire data. You design a frame and drivetrain around an engine, why not design a suspension system around a tire?

[exFSAE]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">(The tire data analysis seems to make more sense in the tuning for maximum performance phase) </div></BLOCKQUOTE>

It's a racecar. What other phase is there than designing for max performance? http://fsae.com/groupee_common/emoticons/icon_smile.gif

Regarding designing for different tires, it depends. Certainly if you design your car around say a Michelin radial and then switch to a Hoosier cross-ply you will not be getting the most out of the tire. From a Hoosier cross-ply to a Goodyear cross-ply may not be as big of a difference in camber and load sensitivity, but it will likely change the balance and feel of the car.

If you're going to be at an event that's more than likely going to be dry and sunny (Cali, Aus) then you can design, setup and tune your car around one tire and be fine. If you're potentially going to be running where its wet (Detroit!) then you should at least get a day of wet testing in before hand and feel the car out and make whatever changes you need.

Unfortunately there's no wet-test data for FSAE rain tires.

That all said, you can certainly be successful without tire data if your drivers blow everyone else away in terms of skill, experience, and seat time. Driver is a massive variable at this level.

[flavorPacket]

it doesn't come down to tire data, it comes down to laptimes and competition score.

obviously force and moment testing can give you an idea of where to go, but those models are horribly simplistic. From our experiences with private testing, we have become much less dependent on tire data because of how difficult it is to use correctly. It'd behoove you to take your data with a grain of salt.

In fact, having designed 3 fsae suspensions, I suggest to you that if you just make everything adjustable and get the car done as soon as possible, you will be faster than if you spend a month completing that last bit of jacking or yaw center migration analysis.

get your tires hot, wear them correctly, make sure the driver is comfortable.

[screwdriver]

Hmm, I think it's possible to do a suspension that gets you through the comp without either tyre data or books.
I'd start drawing up a geometry in Google Sketchup and test it with the Sketchy-Physics plugin. A combination of optical inspection and taking measurements from screenshots should be enough to come up with a geometry that has somewhat desired values (dynamic camber, caster, toe and ackermann).
From there I'd try and implement adjustability for either of these values.
A brief check to see, if the extreme settings produce results within your desired range and done.
Off to the CAD software into construction.
Get something rolling quickly as to be able to test as much as possible.
Take tyre-wear and -temperatures as a gauge. Using those two indicators, you should have a good setup soon - even without any DAQ.

After that just adjust the car to what the driver thinks is best.

[Anvit Garg]

What is the better 3D kinematic program for FSAE? I have the OptimumK demo, havent tried Susprog3d or the other few.

During the Rouelle seminar he explained how his program is much better than Adams.

[PSUAlum06]

Tire data is very useful when used properly but as with a lot of things it's easy to get lost if you don't know what you're doing. I think the main benefit teams would get from test data is just general load, camber, and slip angle ranges where a tire works well. Differences between track and lab surfaces and conditions prevent a good one to one comparison.

But as far as tips go, if you can just get your car on the ground before competition, get a fair amount of seat time, and get your car sorted out you'll be ahead of the game.

[Chris Lane]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Anvit Garg:
What is the better 3D kinematic program for FSAE? I have the OptimumK demo, havent tried Susprog3d or the other few.

During the Rouelle seminar he explained how his program is much better than Adams. </div></BLOCKQUOTE>

Really depends on what you want/need.

I haven't personally used Susprog yet, but from what I've seen it is a more design oriented package. Optimum K on the other hand is much better at analysis and perhaps a little lacking on the design side.

I used Solidworks and Optimum K for my design, but if you wanted to analyse a car that's already built, then Optimum K is the clear choice.

I would say Optimum K is my choice, but as I said I haven't really had much exposure to Susprog.

[Wesley]

Agreed with getting as much seat time as possible - we had our car finished by March, so we thought we'd be good, but a series of engine explosions (due to shady engine sources) set us back many times.

[flavorPacket]

I'd recommend either optimumK or suspensionSim if you're looking to spend money. nothing else comes close to these 2 in terms of ease of use and bang for your buck.

[benjo]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Anvit Garg:
What is the better 3D kinematic program for FSAE? I have the OptimumK demo, havent tried Susprog3d or the other few.

During the Rouelle seminar he explained how his program is much better than Adams. </div></BLOCKQUOTE>

Depends on your budget mate.. OptimumK is good with the FSAE discount it's only 500$, otherwise SuSprog3D isn't bad and that's pretty cheap also.

Adams from what I know is very expensive and quite complicated. That complication can lead to a very in depth analysis though, if you have time.

[Ericyu]

Thinks guys, I really hope I can have you guys around in real world but not just Internet, so I can learn every thing I want from you guys! FSAE is new thing here in China, only few universities have FSAE team, I have been look around my town to find someone has the knowledge of racecar design or some one who want to work with me, but sadly I haven't find one, maybe there are some people there, just didn't find them. so I have to do everything myself, I read some papers about FSAE car design sequence, they suggest to begin with the suspension geometry, I will try your suggestions and give the progress update.
thanks for all your help again, appreciated.
Regareds!

[scholar]

Design Competition could be considered a misnomer. It is really an evaluation of the engineering knowledge of the team.

What would impress the 'Design' judges more, that a team used the tire data to engineer a theoretically ideal suspension system to maximize the performance of the tire contact patch or that a team designed a very adjustable car but could not get it completed in time to do the huge amount of testing necessary to find the ideal suspension system?

Further, there are several publications that suggest the best starting point for selecting wheel rate is to select an arbitrary ride frequency. As with the above, is this good engineering? Wouldn't that tire contact patch, again, be a better starting point with ride rate being a result rather than a design criterion?

[John Grego]

Well suspension design is a bit of a game of cat and mouse. Sometimes you have to pick something, like a ride frequency, design around it and then come back and verify that "arbitrary" number to make sure that it is what you want. It all revolves around assumptions that you make and verifying that these assumptions were the correct ones.

Basically, you have to start some where and usually ride frequency is a good place to get the ball rolling.

[Peter]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by scholar:
What would impress the 'Design' judges more, that a team used the tire data to engineer a theoretically ideal suspension system to maximize the performance of the tire contact patch or that a team designed a very adjustable car but could not get it completed in time to do the huge amount of testing necessary to find the ideal suspension system?
</div></BLOCKQUOTE>

In my opinion a good design judge would be impressed by the team that found the optimal combination between both; If, in the end it is all about laptimes, those will be found by a combination of design and testing.

Like everything in this competition, it's about resource management; where is time spent most sensibly? And I think for a few exceptions most of the teams can improve more by optimizing areas other than the tire contact patch...

For a new, resource limited team I would suggest to get something on wheels and start driving it as soon as possible. I am sure this can be done without expensive (for them) software, like many teams / designers did in the past. Excel, 2D CAD is sufficient.

Good luck,

Peter
Delft FSUK04, FSG06