20 replies to this topic

[Sciman]

Hi, I'm studying a Bachelor of Engineering Technology and for my end of year project I have chosen to model up the front end of a Formula 1 car showing the workings of the steering and push-rod suspension.

I was wondering if anyone out there had any detailed information on the steering and suspension set-up, as I am finding it much harder than I thought to find anything useful. There are many pictures out on the web and in books, but I really need some detailed specifications, or technical drawings/illustrations so that I can model it up using SolidWorks.

Any help would be greatly appreciated,

Thanks!

[Roll_Centre]

I am sorry to tell you buddy but you are not going to do well on your project. You should try something else and just come up with a suspension model on your own. Why? If you can get pictures or models of a suspension from a F1 car than you got what every other team outhere would LOVE to know. That is one of THE biggest secrets of those cars.

Also, I hope you realize how intricate a suspension geometry is, especially the front since you make the front for roll and the rear for ride. Just pick the F1 wheel size, which you can get, pick two points inside of it than from there on you can figure out the hard points on the chassis which will give you the curves you want, toe, camber, roll center etc.

Anyhow, you will learn a lot more buy doing your own and it will probably be ten times easier than trying to figure out the exact geometry of a modern F1 car. Good luck with it.

[Lukin]

Have to agree with Roll_Centre here. Any decent design information will be bloody difficult to come across. I dont know a heap about F1 suspension, but I do think that it is extremely different to most motorsport suspension setups and modelling it will be very hard.

The F2004 has no spherical bearings and gets the up and down motion from the flex of the wishbones through flex of titanium blades (source: F1 Racing May 2004). Modelling it will be bloody hard!!

That said, if your still keen there is a book detailing the 2000 Ferrari car. I have never read it but other people on the forum reckon its a great book. The link is included down the bottom of the page.

Just modelling it in Solidworks wont be a huge breakthrough, as it will come down more to 3D modelling ability than understanding of the physics and also simply put, only a few will know if its right or not.

I guess a lot depends on what the scope of the project is. Modelling, static analysis, dynamic analysis etc.

I would suggest maybe modelling up a suspension system and then doing a dynamic analysis with a heap of variables (camber, springs, dampers, ride heights, weight transfers etc etc) under a heap of conditions (lateral acceleration, load, speed) but to do this Solidworks will help the square root of bugger all.

There is a program Susprog the FSAE teams use a bit which has a baseline integration of these variables from what I can tell. As much as it pains me to say it, Matlab or MathCAD might be the best software to do a comprehensive model to input anything and get anything worthwhile out. A complex system will require comprehensive simulations and a lot of effort and patience.

Goodluck in whatver you choose to do.

http://www.amazon.co...=books&n=507846

[Sciman]

Thanks for your interest, and the information you have given has been helpful. That book you recommended Lukin looks pretty good - I'm looking at ordering it now, could be quite helpful.

Your right about Solidworks not being the best place for a decent analysis, although I do have a copy of MATLAB so I mite try having a play in there. Thanks, didn't think of that!

The aim of the project is mainly to just get a working model drawn up in Solidworks, so I am still looking for some basic suspension layout diagrams. Even if they are from a few years back they would still be of a great help, my tutor acknowledges that its bloody hard to get this info so he doesn't mind if its slightly out of date - I just need to have something to draw.

Well thanks for you time and your replies!

[Lukin]

G'Day Sciman,

Have you thought at looking at the suspension setup for a small formula car? Being able to do the full design and analysis of a F1 car would be great, but the principals of suspensions design will still be evident in smaller formula cars (the laws of physics dont seem to change much for some reason...). But I think it would be a great deal easier to gain access to a small formula style car ie formula ford, formula vee, formula holden (or whatever its called now) and the principals will still be the same.

I was thinking about the solidworks modelling. It would help to a degree, in the sense you can contrain axis, shafts etc and model the suspension as rigid links and watch how a displacement input into the wheel will produce related motion in the dampers/springs or how lateral acceleration will cause roll, camber change etc (though that would be very hard to model)

Love the comment about having a play with Matlab. HAHA. No one plays with Matlab, people play with sharp knives and bottles of liquor after Matlab. Its a pain in the arse program that is so unintiutive that it makes oval racing and drag racing look practicle and enjoyable.

Also, for a good overview of car setups and effects, download Lapsim by Bosch. There is a thread somewhere on the BB about Lapsim with where to download it and some of the things people have done.

Good Luck
Jas

[red300zx99]

Lukin just wondering what your beef is with MATLAB

[Lukin]

red300zx99,

We didnt get any formal training with Matlab at uni and got completely thrown in the deep end with it and I have had a lot of trouble picking it up. It does some good stuff but to me it's weird. Though its probably just me not being smart enough. Just not a fan.

[Sciman]

I agree with Lukin on that one!

I've had basic training with it at uni, and the program does have huge potential, it can just be extremely hard to actually make it do what you want it to do!

Also, thanks Lukin, I'm going to follow your advice by modelling up a basic version of the suspension, mainly focussing on the push-rod, rocker, and springs/dampers etc, and will perform my analysis by watching the result of a displacement in one of the wheels.

I'm still looking for some decent detailed drawings tho - would like to see more of the components in relation to each other so I can get a better idea of sizes.

[Greg Locock]

Matlab is not really the weapon of choice for this, in my opinion.

Having said that I wrote a 2D kinematics program in Excel, so abusing mathematical packages is something I am familiar with.

If you can get your hands on ADAMS, Working Model 3D, or Visual Nastran 4D, then they are designed for this sort of thing.

One very fundamental point is - are you interested in forces? Another is are you interested in compliant effects? If you are then a kinematic approach is likely to founder. Susprog3d gets around this by averaging the compliances out (ie, when the mechanism locks up it somehow distributes the errors around to get it going again).

These problems come about because many suspensions these days are not true mechanisms. They are overconstrained and rely on compliances to give particular effects, including full suspension travel.

[red300zx99]

Greg, love your comment about abusing math packages, spent a year programming kinematic anylsis in MATLAB, only to find out that they have a modeling package. Still havent used it....doubt I every will, in the process of redoing all the code in visual .net, wish me luck on my sanity.

Sorry as I have nothing to say about pushrod suspensions, other then the fact that my life is easier right now as the racing series I perticipate in doesn't allow them.

[Sciman]

Sounds like everyone has had fun at some point abusing maths programs!

Yeah would've been a lot easier had I chosen something else to model up beside a push-rod system, didn't realise it was going to be this impossible to find information on it.

Unfortunately, I'm not sure how easy it will be to get my hands on any decent analysis programs, as I'm doing this project through my University and therefore only have access to their programs. Although I'm sure they'll have something capable of doing it otherwise my lecturer wouldn't have suggested the idea. If they don't, then I'm gonna have to have a bitch to him at some point :

Well I'm gonna continue looking for whatever I can find, and if anyone does happen to see anything related then please pass it on

Thanks again!

[Lukin]

Unless your uni has a different policy to mine, you should be able to use any program as long as you get them legaly. When I did my intro talk for my final year thesis the first question I got was whether or not the copy of Nastran I was using is legal! If it wasnt, I would of been in the brown stuff very quickly.

There are a fair few programs out there for free, though you have to be careful to ensure you check what the program gives you is correct in both it's make up and your operation of it. Simulation Software will only give you crap if thats what you feed into it.

Susprog3D is a good one, you can get a free demo, just google it. From speaking to the Suspension guy in our SAE team it is kinetic based and will give you good information on wheel rates and the like, but seems to ignore load transfer/forces as Greg pointed out a few posts ago.

I also found that a lot of companies are happy to give evaluation/educational versions. I downloaded e-drawings to help transfer drawings to other members on the SAE team and a few days later I got a letter from Solidworks offering tutorials and a free personal (LEGAL!) copy of solidworks while Im at uni.

Its a good investement on their behalf and it helps student. It might be a bit (a lot) harder to get your hands on Adams or Nastran unless your involved with SAE however, but if you find a sympathetic rep you never know. Also, check out other departements in the uni and also post-grad students in case they have what your after. I had no idea Solidworks was available in our Industrial Design labs till this year.

Have you checked out Desmo's list of Technical Papers? There are some good articles there written by some smart people but Im not too sure how many pertain to suspension design/analysis. Also there is a few sites around on the topic, though I think a lot of information is generic so you will have to look hard for the good stuff.

[Greg Locock]

MSC runs a campus software license program, the chances are reasonable that your university has at least one license for every MSC package out there, including Nastran and ADAMS and Visual Nastran 4D.

I'm pretty sure you'll find that getting the geometry and compliance data is much more difficult than finding or writing an appropriate analysis package. Of course you could just contact a local racing team or low volume manufacturer...

[Lukin]

Greg,

On the topic of compliance, what does the compliance data give? I dont know a heap about suspension systems, but as much as I understand, compliance refers to the rigidity of the attachement of all the components. Like in our SAE car last year, there was a degree or two of roll before the chassis/suspension provide roll resistance, is that what they mean by compliance? And things like deflection in bushings/rod ends/bearings etc etc.

Ta

[Greg Locock]

Every rubber bush has six degrees of freedom (3 translational and 3 rotational). It is likely to have a non linear force vs deflection curve for every one of those. Everystructrural mounting point also has 6 degrees of freedom. Usually you can assume those compliances are linear, but you'll still need a value for them.

Compliance = 1/stiffness, we use them interchangeably, except when we are talking values.

[Lukin]

Ok, so for any given loading, you can determine the load that goes into deflecting the mounts and therefore the loads transfered into the chassis.

What is ideal for compliance? I guess if there is a high compliance a lot of the load there will be a smaller load transmitted to the chassis, but also I assume the suspension geometry would change a lot and effect the supension performance?

Conversely I guess lowering the compliance would give more consistent suspension geometry but the high stiffness in joints/mounts/bushes will increase the load into the chassis? How far am I off?

I really should know this doing chassis design for the SAE car.

[Greg Locock]

What is ideal for compliance?

Generally, in sports supension NO COMPLIANCE is the ideal, unelss you are trying to be tricky.

I guess if there is a high compliance a lot of the load there will be a smaller load transmitted to the chassis, but also I assume the suspension geometry would change a lot and effect the supension performance?

Mr Lukin, meet Freebody Diagrams. Freebody Diagrams , this is Mr Lukin.

The compliances have no effect on the forces transmitted in the steady state, in a mechanism, to a first approximation. They /do/ affect the deflections and the geometry. The thing is, they almost always conspire to make your curves go weird (technical term) or the wrong way. For instance, the tie rod/rack system is pretty securely mounted to the chassis, and takes fairly small loads. Now introduce a compliance in the lower wishbone. Hallo, now I've got a steer effect due to lateral loads at the contact patch that I didn't have before. Now add a compliance at the upper ball joint. Hot damn, it just got worse! Draw a free body diagram and all is made clear. So now I have to move my hard points just to get back to where I was, in terms of steer response to lateral loads. But now the bump steer curve has gone funny.

So, this is the sort of idiocy I do all day sometimes. Fortunately I have a lot of matrix crunching routines available that simplify these interactions, so when we change a bush stiffness we know how to move each hardpoint to get back to where we thought we wanted to be. These matrices have to be built for each suspension individually, by doing a Design of Experiments on the ADAMS model for every location and bush stiffness of interest. Typically this is about 200-1000 runs of the model, because we actually do a lot of interactions with 5 settings for each variable. As I am very lazy it is fortunate that there is a scripting tool (ADAMS Insight) that generates all these runs automatically.

I really should know this doing chassis design for the SAE car.

Yeah, you do need to get a handle on this stuff. Get very confident with free body diagrams, if you can't work the forces out then you have a good chance of getting lost.

[Artisan]

Go to www.teamserpent.com and see if you can still get a manual for the Serpent V-Spec 1/8 scale Radio Controlled racer ( Don't laugh guys- if you are not familiar with Nitro powered R/C, you have no idea how sophisticated these things are). The V-Spec had a pushrod front and rear suspension, and the geometry and other specs will be fairly well spelled out. Unfortunately, this car has been discontinued as a kit because the unsprung weight savings was not enough of an advantage to offset the increased cost and complexity compared to a conventional control arm setup, but then, most R/C guys don't have a 200-500 mil budget to operate on, or they would be in F1!

[red300zx99]

There is a program made by performance trends, i believe their website is performancetrends.com, makes sense, but anyways in the downloadable demo there is an example of a pushrod suspension, not F1 but it might be something to look at

[Lukin]

Thanks for that Greg. I guess I was sort of looking at it from an energy point of view in the sense that some of the energy produced by the suspension motion will go into deforming the joints/bushes and therefore less energy will be transferred into the chassis. But for compliance your right, it is easier to look at it from the force/moment and yeah, Im not too flash on the suspension side of things.

[Sciman]

Thanks very much guys for your interest - I'm definitely going to try and find those programs to download.

Lukin, that Susprog3D sounds really good, and if I can get a free version, even better!

Its really interesting following your conversations in here, I'm learning a lot!

Thanks