27 replies to this topic

[briggs]

Hello,
My Name is Briggs and I am a full time student at Southern Polytechnic State University. I am also a member of Formula SAE, where we build a formula car from the ground up using a 600 cc motorcycle engine. Our newest car with a tublar cromally chassie designe weighs in just under 450 lbs. However, we want to go lighter. I am researching carbon and or aluminum monocoques. I found this forum and was wondering if anyone knows anything about carbon monoques that may be able to help answer some questions? Also I was wondering if anyone has any web addresses that may help, or pictures?
Thanks,
Briggs

[mariner]

I hope this does not sound too offputting but building a light and safe carbon monocoque is a large challenge.

I have not done it but I would quote from two people who have.

I once attended a seminar by Patrick Head of Williams on carbon monocoques. His answer to the question " what do you stress it for" was " falling down the side of a mountain" i.e any and every possible load in any dierction. As this was just months before the death of Senna that kind of stuck with me.

Burt Rutan of Voyager and Scaled Composites fame, who has to be one of the world's top composite designers, once published a wonderful guide to building aircraft using his moldless construction techniques. In it he gave a very clear warning on the gap between building in composites ( carbon or otherwise ) and designing them.

I hope that does sound too negative, I think I am just saying that building up carbon monocoque design skill can be a long process with lots of empirical testing needed BEFORE putting a driver in it at speed. If you have the funds and time and facialties good luck but it may be an expensive and slow weight saving route.

One thing you can consider is using pre made flat CF panels with bonded joints. That way you have access to the strength/stiffness data of the supplier and the critical design work is restricted to selecting the required panel skin thickness and panel depth plus the jointing approach.

The only comment I can make from my experience is how you feed the point loads in to the skin is critical as you are trying to distrbute high loads into a weak substrate as quickly as possible. Maybe concentrating the loads into just the bulkheads and then interfacing the whole bulkhead perimeter to the CF skins might be the safest approach.

[McGuire]

Originally posted by mariner
I hope that does sound too negative, I think I am just saying that building up carbon monocoque design skill can be a long process with lots of empirical testing needed BEFORE putting a driver in it at speed.

On the plus side, it is FSAE. If the driver falls out he will just skin his bum.

[Greg Locock]

I wonder how many true monocoque cars there have ever been - by that I mean a car where the skin is largely responsible for carrying the structural loads, rather than box sections, cross members and tubes?

The problem with cars is that almost all the loads are concentrated at a few points. So the structure tends to triangulate to these points. I'm not convinced that a stressed skin would actually be lighter for a given strength/stiffness. If you run an optimising FEA program like Altair Optistruct on the typical loads a car sees, to find the perfect shape for the structure .... well I won't tell you the answer, in fact if you were to investigate that it would make a good presentation. I do know that if you do that on the package volume under the floor you end up with a ladder chassis (partly because I used a lot of torsional input when developing the model).

Alternatively you could start from a hand waving design exercise. A car's skin is roughly the same shape as an egg, which is a monocoque. Now bolt your suspension to it. Where will it break? OK so add a front bulkhead and a rear bulkhead to take those loads etc etc.

As to materials - for the solar cars we've built aluminium chassis and carbon fibre ones. They weighed the same. The CF one was stiffer and stronger, and more durable. It is also a major investment in time and effort compared with the ally one, something like 1 man month instead of 2 days to build. The technique we used was to shape foam to form a male plug, set up the hardpoints (integrating the hardpoints into the carbon fibre is the biggest single problem) If as I suspect you would need a lot of bulkheads and other things to take the point loads you'd build those in prior to skinning the shape. For a monocoque you would then want to get rid of the foam, which would be messy and annoying but not difficult.

As a design judge my first question would be how are you going to mass produce a carbon fibre bodyshell for less than $20000, or is it true that USAn teams always ignore that rule?

[J. Edlund]

There was an article series in Race Tech some time ago where Peter Elleray went through some basics.

He also mentioned the first CFRP F1 chassi he had worked on, and how they had managed to get a carbon fibre chassi only a few kg lighter than an aluminum one, while at the same time having reduced the torsional stiffness considerably.

Designing an aluminum chassi is in other words much simpler as the material itself is isotropic, carbon fibre composits being anisotropic. For best result different grades of carbon fibres should also be used, with thecorrect layup and fibre directions. High stength and high modulus grades tend to be quite costly, and getting the lay up and fibre directions correct require a bit of extra work.

[Nathan]

Originally posted by McGuire


On the plus side, it is FSAE. If the driver falls out he will just skin his bum.

Originally posted by Greg Locock
As a design judge my first question would be how are you going to mass produce a carbon fibre bodyshell for less than $20000, or is it true that USAn teams always ignore that rule?

...followed by another judge asking how it would be quickly fixed incase of a crash. One thing I noticed from my FSAE days is quite a few schools lose focus of the whole point of the competition.

[briggs]

Thankyou everyone for your response. I must agree that this will not be an easy task and I am positive that I will be learning many things as I go. I do have a new question. I know that many monocoque designes use a honey comb aluminum or carbon to fill it with and stiffen up the chassie. My question is how, when laying the carbon and resin, do you keep the resin from filling up the holes in the honneycomb? Should I use a prepreg lay up? Also, looking at nose cones on F1 and Lemans cars, I notice that the weaves appear to be at a 45 degree angle. Where my stresses seem to be the highest should I be laying the carbon weave at that angle? If not, how should I go about laying the cabon?

[Greg Locock]

Yes, you'll need to use prepreg with a hex core.

The design of the directions of each ply is either very complex or very simple. If you use a two ply layup with each ply at 90 degrees to the other then you've covered the two dimensional loads. You have also doubled your use of carbon compared with the optimum for a given load. That's pretty much what boat builders do for example - since they don't have a really good handle on the exact loads in the first place. Then add tow (uniaxial) to handle known large point loads.

Designing plies properly in response to known loads is possible, it is a very advanced feature of FEA programs for example. It can be done by hand. Sorry I haven't got any references, it is outside my experience.

[Fat Boy]

Originally posted by Nathan

One thing I noticed from my FSAE days is quite a few schools lose focus of the whole point of the competition.

Every once in a while you find a team that doesn't lose sight of what the point is. They are definitely a minority. Pretty much everyone else just makes me shake my head.

From what I understand, the European teams are the ones putting massive money into their cars. I heard that one team in Detroit had put something like 150,000 Euro into their car/team. Governmental funding at it's finest.

For reference, our first year car way back was something like $4000. A 'big' sponsor for us was $500. I think it's fair to say that FSAE has spun out of control for most schools.

[NTSOS]

Originally posted by Fat Boy


For reference, our first year car way back was something like $4000. A 'big' sponsor for us was $500. I think it's fair to say that FSAE has spun out of control for most schools.

As long as more money is being spent, cannot they "adjust" the rules to make the cars look at least slightly more race car like in appearance!

John

[mariner]

In all the excitment of discussion I falied to read the question properly ( as they say in education), you wanted to know if there are any books!

Try these

1) fiberglass and composite materials - Forbes Aird ISBN 1-55788-239-8

2) sucessful composite techniques Keith Noakes ISBN 1-85532-261-7

These cover all glass fibre stuff but much of it is relevent and 2) has a large section on composite design with detail material properties and sandwich design failure mode calculations

and

3) Formula 1 technology Peter Wright ( a good source I think) ISBN 076800234-6

This has an 11 page appendix specifically on CF. If you are on limited time and budget the description of the Lotus process on page 316 may help even if it old by current F1 standards.

[briggs]

Well, as far as FSAE goes... yes there is lots of money going into it. However, the car that we build at my school has been in the top 20th place the past 2 years out of 140 teams. The catch is, the majority of teams have over $50k to spend on their car, we have a total of $8k. Just this year we may be getting better sponcerships which is why we are looking into the monocoque design. Also, we are trying to keep up with technology and learn as much as we can. I personally will take and use this to my advantage.
Thank you for the books to look up. This I'm sure will be a great help. Also thank you for asking me questions. There will be design judges, and I'm sure I will hear those questions being asked over again. This will help me to deside what I need to kinds of tests to run. So thank you again everyone.
I do have another question. Is there any way in particular to keep the resin out of the holes in the honeycomb, or is the pre-preg resin viscus enough to not really worry about it filling the holes?

[Greg Locock]

The prepreg is only just wet, I'd have thought the problem is more likely to be a dry joint than excessive epoxy dripping into the hex.

Now at the risk of sounding like a boring pedantic twit:

"Pyour is the adjusted cost of your car (with penalties),
and Pmin is the adjusted cost of the lowest cost car. If
Pyour is greater than $25,000 US, the car will be
disqualified from the Cost Event"

"4.3.9 Costing Tables
To assist in your process the following tables must be used in costing:
COMMON MATERIALS AND COST MINIMUMS TABLE
Mild steel, e.g. 1010, 1025 $0.30/pound
Alloy steel, e.g. 4130, Chrome Moly $0.60/pound
Aluminum $0.75/pound
Magnesium $2.25/pound
Non-graphite composites $88.18/kg
($40/pound)
Graphite-based composites $220.50/kg
($100/pound)
This table represents the lowest cost of these materials allowed. Regardless of what
price a team actually paid for the material and/or what the receipts show the cost as,
this table represents the MINIMUM cost.
Other materials such as plastics span such a vast range of uses and costs that a
common price standard is impractical. Cost for composites and structural
construction similar to fiberglass should be cost separately with a clear
identification of the costs of all materials and processes. Obviously, process costs
are in addition to the above material cost minimums."


So, think long and hard about how much of your budget you want to squander on aerospace materials and manufacturing techniques.

[briggs]

I understand. This is something the team has decided on. That is why it is my job to research it. I realize that the price of carbon in the cost report is astronomical, wich is the reason we have made our previous bodies out of fiberglass. We also have to cost out the labor on the welding of the crome-moly chassie. Yes it is still cheaper to run the tubular chassie, however, we also know that it is not impossible to run a carbon monocoque. This is why I am researching it.
And don't get me wrong. I apperetiate the criticism and will use that to my advantage. I am at the point right now where I am more worried about making a structurally sound monocoque more so than I am about the cost aspect.

now about making the carbon pre-preg bond together, I know I will need to heat the resin up, but is there any way to make sure that their will be resin in those places? Could I go about mixing up a little more Epoxi and pasting it on there on should I not worry about it?

[Fat Boy]

Originally posted by briggs
However, the car that we build at my school has been in the top 20th place the past 2 years out of 140 teams.

And in oh-so-typical FSAE fashion, you take a car design that is competitive, throw it out the window and start from a blank sheet.

While you're at it, don't forget to do a completely different engine/fuel/oiling/engine management package. If you don't have a turbo, put one on. If you do, get rid of it. Change the wheel/tire size. Change the differential. Put completely different suspension geometry on both ends of the car. I'd also like to see a massive science experiment in there someplace, like automatic shifting, big wings, or a self-built data system.

And God forbid, don't justify any of it with numbers. That should really move you up in the rankings.

[Greg Locock]

You have to use some vac bag or pressurising system when moulding the carbon skins onto the nomex.
You don't need extra epoxy slopping around if it is done properly.

Now, back to the more interesting question - designing a true monocoque, out of whatever material.

Is it possible for an open cockpit car? how much weight will you save compared with a spaceframe and a non structural skin, or a unibody?

[Fat Boy]

Originally posted by NTSOS


As long as more money is being spent, cannot they "adjust" the rules to make the cars look at least slightly more race car like in appearance!

John

It's kind of the nature of the beast. Look at other cars of similar size, like the SCCA's F500 or various SCCA Mod solo cars. You have to run small wheels and wheelbases. If you stretch the car out so it's visually 'right', then you get a car that is way too big for what it's meant to do. It'd probably be much better on an actual racetrack, but that's not the object of the deal.

[Ross Stonefeld]

Do the students still drive or do they have a 'test driver' that runs all the cars?


150k in a FSAE car?! What's an Atlantic roller cost?

[NRoshier]

Briggs sounds like your team has been reading the FSAE website too much.
How about a simple ali honeycomb panel chassis with bonded on carbon/foam crush/torsion boxes. Then do what no fase team has done enough of yet: work out the ergonomics - make it adjustable and test and test the damn thing. So many teams have the car fall apart because so much time is spent designing and then redesigning a Ti dash bracket that the driver often never sits in the car before the event...then a wheel falls off in the slalom!

[Engineguy]

Originally posted by Fat Boy
And in oh-so-typical FSAE fashion, you take a car design that is competitive, throw it out the window and start from a blank sheet.

While you're at it, don't forget to do a completely different engine/fuel/oiling/engine management package. If you don't have a turbo, put one on. If you do, get rid of it. Change the wheel/tire size. Change the differential. Put completely different suspension geometry on both ends of the car. I'd also like to see a massive science experiment in there someplace, like automatic shifting, big wings, or a self-built data system.

And God forbid, don't justify any of it with numbers. That should really move you up in the rankings.

If I'm not mistaken, you're ridiculing starting from a blank sheet. In an educational environment, shouldn't that be encouraged, as a greater learning experience, versus putting a fresh coat of paint and slight mods on a design someone long gone engineered? If the students aren't going to experience the challenge of conceptualizing and building a car from the ground up, why not just have all the schools buy Miatas and race 'em?

[Fat Boy]

Originally posted by Engineguy


If I'm not mistaken, you're ridiculing starting from a blank sheet. In an educational environment, shouldn't that be encouraged, as a greater learning experience, versus putting a fresh coat of paint and slight mods on a design someone long gone engineered? If the students aren't going to experience the challenge of conceptualizing and building a car from the ground up, why not just have all the schools buy Miatas and race 'em?

Absolutely, I was ridiculing the habit of FSAE teams to start from a blank sheet. For good reasons, though.

There are 2 things going on. One is an engineering project and the other is racing. What happens over and over is that FSAE teams don't understand the design that they inherited, they have no idea what compromises were made or why, and they only have a vague idea of it's short-comings. Often, because of inter-team rivalries, egos, and politics, they throw proverbial baby out with the bath water.

I've been on a FSAE team in the dark ages and know what kind of stupid **** goes on. I've also been a design judge. I can't this year because of schedule conflicts, but I think I'm qualified to talk on the matter.

I'd much rather see a team bring a car that is an evolution of a previous car **and be able to quantitatively justify every bit on it**. More often than not, they produce a completely new design and when I ask them 'Why this' or 'Why that' they tell me 'Well, it seemed to work' or 'Packaging constraints'. In this case, they may have learned manufacturing skills, but not engineering design.

A couple of good race mechanics working on nights and weekends could put together a car nicer than 95% of what you find at FSAE. They wouldn't have a lot of justification for things, but they could use general rules of thumb and accumulated knowledge to put a car together. They'd build it in 1/4 of the time most FSAE teams spend and they'd go drive the **** out of it, break it, fix it, repeat. They'd do well in the dynamic events. They'd hurt in design, but they could manage a top 20 overall.

What I look for teams to do it understand their car. Almost none of them do. Some years that number really is zero. Building a car from the ground up and having a justification for everything is _tough_. Major constructors don't do it every year. Lola, Dallara, Panoz, etc. all build a car and then refine the design over a period of time before making the step to the next design. I don't think there is any question as to whether these cars are 'engineered' or not. They are. The consistently competitive FSAE teams take a similar approach.

You don't have to do a complete redesign a suspension or chassis to understand the engineering justification for it. I would encourage the teams to make evolutionary changes. By playing with a FEA program they might be able to change tubing thicknesses and change the stiffness of the frame. It would be a completely reasonable engineering project to change the handling properties of the car by leaving the suspension alone and evolving the chassis design. They could justify what they did with an ADAMS model and then show through simulation and real world testing how that produced a better car. I've never seen it done or heard of it considered.

-------------------------------------------------------------------

Here's the life of a design judge:



What they tend to do is say "Our front roll center last year was 1/2" above ground last year. The car had understeer, so we put it 2" below ground this year. We also raised the rear roll center and inch, have different motion ratios, dampers, and we went from a space frame to a monocoque with a stressed engine."

Then you ask them, "Evaluate your roll center change and justify on paper why that particular change was necessary and what it has done."

"Ummmmm, Carroll Smith said that if you lower the front roll center and raise the rear that the car would push less. There was a guy that did some work on the suspension, but he graduated. We don't have any of his paperwork."

"OK......how does your car handle?"

"It still pushes."

"Why?"

Blank stares........

-------------------------------------------------------------------------

The young lad here has a car that can get in the top 20. That's no easy feat. In all actuality, what they need is to refine the design. Be better prepared for cost/presentation/design. Going fast in the dynamic events are about 2 things, seat time and tuning. Good drivers....of which there are precious few....can put you 20 places up the rankings against a mediocre one. A bad driver can drop you 20 spots. A reasonably well handling car that has been adequately track tuned is also a rare find. Put them together with a car that doesn't break and you're in the top 10 and maybe top 5. It's just not something you can do if you completely redesign the car every year.

So there's the long-post edition of my previously condensed rant. Hope that clears it up.

[Ross Stonefeld]

Has anyone rocked up and said "we have the same car as last year, but we hired a driving coach and found more laptime in the nut holding the wheel" ?

[Fat Boy]

Originally posted by Ross Stonefeld
Do the students still drive or do they have a 'test driver' that runs all the cars?


150k in a FSAE car?! What's an Atlantic roller cost?

Students still 'should' be driving it at the competition. I wonder if they don't get ringers some times, though. Most of them aren't open enough to allow a test driver to help them out away from the competition. A couple years ago I got the chance to drive 2 different teams' cars. Both of them complained of push. Both cars were crazy loose. One team listened to what I had to say a little. The other team completely ignored me. On the first car I was +/- 0.1 second between their 2 best drivers (who had actually taken a 3-day schools). On the second car I was quite a bit faster than the next guy, like 1 1/2 seconds or so. Keep in mind, I'm not a good driver. I can feel what the car is doing, though, and that put me ahead of the young-uns.

As far as money goes, the desparity there can be pretty massive. It doesn't mean they spend the money wisely, just that they spend it.

[Fat Boy]

Originally posted by Ross Stonefeld
Has anyone rocked up and said "we have the same car as last year, but we hired a driving coach and found more laptime in the nut holding the wheel" ?

Never that I know of. Having said that, that approach would not fulfill the engineering side of the competition.

[Ross Stonefeld]

But isn't the variance in driving skill at that level going to be massive? I'd have though you'd want to track test the cars with a baseline semi-pro.

[Fat Boy]

Originally posted by Ross Stonefeld
But isn't the variance in driving skill at that level going to be massive? I'd have though you'd want to track test the cars with a baseline semi-pro.

It would be a big help to nearly every team. They just won't do it. In all honesty, most probably don't have access to anyone that fills the bill.

We would drive around in our parking lot in 5 or 10 minute intervals. The trick was seeing how many times you got to drive before it broke. There were maybe 10 guys waiting to drive. If you drove 3 times for 15 minutes total during a day it was a lot of driving. There's no way to get any tuning on the car or real drive time doing what we did.

The good teams (UTA was always fast when I did it) would pick drivers at the beginning of the year and they would drive the old cars for hours at a time 2-3 times a week. They were very good at evolving their design. The car from 2 years ago was damn near the same as the car they were building. If you could drive one, you could drive the other. What a concept.

Cornell had a guy one year that was a shoe. I don't know what his actual background was, but the rumor was F3. He made everyone else looks silly. Again, though, it was a team that would produce an evolutionary car.

We were never that smart....and we didn't win.

[Ross Stonefeld]

Sorry I meant at the competition. Are they graded very much on what sort of laptime they do? I meant in the sense that the relative inabilities of students as racing drivers might make bad cars look good or vice versa.

[Fat Boy]

Originally posted by Ross Stonefeld
Sorry I meant at the competition. Are they graded very much on what sort of laptime they do? I meant in the sense that the relative inabilities of students as racing drivers might make bad cars look good or vice versa.

The drivers at the competition are meant to have been involved with the work. What that means if you happen to have a really good driver going to your school is open to interpretation. One of the guys on our team had driven midgets. I don't really know if that was an advantage or not! I heard that a couple years ago one of the top drivers was a way up in the Rotax Max US Championship, like a podium finisher. That would have to be a massive advantage. All the dynamic events are timed and the points you receive is based on your position relative to the leader. Lap times are everything.

I really think they should get a proper driver at each competition to drive the top 5 or so cars prior to the design awards. It's damn near impossible to evaluate these cars based on looking and talking. It won't happen, though, so that probably means some cars win design when a better car places 3rd or 4th. It probably all gets shaken out in the dynamic events.