Carbon Fiber Stressed Skins

[Will98Cobra]

We were thinking about trying some composite stressed skins this year. I was wondering if anyone had any info that they could pass my way.

Thanks

Will Austin
ODU Motorsports

[Rob Davies]

We too are thinking of using stressed skins instead of a leap of faith to monocoques (more to do with our lack of ability, as monocoques do seem to be the best solution)

Any sort of information is so hard to come by however.

[leclercjs]

We are also looking at this for next year to save weight on our 05 chassis. It doesn't seems to hard to fill out the Design Equivalency form, just some research.

[Denny Trimble]

Our chassis guys from a couple years ago, when they did it, said it was 75% as much work as making a full monocoque chassis, and it didn't add much to the stiffness of the spaceframe. But, you could probably do it easier than we did, we had molds CNC'd out of foam, which didn't match up with the spaceframe (something about tolerances... http://fsae.com/groupee_common/emoticons/icon_smile.gif ) and it was a pain in the ass to fix. But, I bet you could make sheetmetal molds off the spaceframe itself, and save yourself lots of time.

[Rob Davies]

Denny??

So you have a sheet of carbon fibre to fit over a section of the bodywork. How do you make it stressed and attached to the steel space frame. I was thinking of riveting it but that would weaken the space frame.

Any help ?

Thanks

[Denny Trimble]

Ours was bonded, and eventually some joints failed in a few places (idiots stepping through the floor, etc). I don't think riveting CF is the best idea, but I'm no expert. Also, on some of the larger sections, we would notice the CF panels would "buckle" or oilcan under load. So, that gave the car a decreasing slope on the torque vs. displacement curve.

Aluminum stressed skins are frequently bonded and riveted on "standard" formula cars. Weakening the frame tubes isn't a major concern, most spaceframes are not highly stressed (near yielding), because of stiffness requirements.

[leclercjs]

On that, guys from the solar car suggested me rivets and an epoxy to join tubular and carbon fiber stressed skins. On T.O car, it seems glued with epoxy.

Can you confirm Vihn ?

[James Waltman]

Composites don't work well with many traditional fastening techniques (rivets, bolts, studs, etc). Avoid them!

Bonds can be incredibly strong in shear but they are pretty worthless in peel. Keep this in mind when you are joining things by bonding and you will be in good shape.

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by leclercjs:
It doesn't seems to hard to fill out the Design Equivalency form, just some research. <HR></BLOCKQUOTE>
Can you elaborate on this? I assume that you are researching the strength numbers for an individual ply and then planning on using that as justification for the Safety Structure Equivalency Form (SSEF).

[leclercjs]

On this, the FSAE rule book says:
--------------------------------------------------------------------------------------------------------
3.3.3.2 Alternative Tubing and Material
3.3.3.2.1 General
Alternative tubing geometry and/or materials may be used. However, if a team chooses to use alternative tubing and/or materials:
(A) The material must have equivalent (or greater)
Buckling Modulus EI (where, E = modulus of Elasticity,
and I = area moment of inertia about the weakest axis)
(B) Tubing cannot be of thinner wall thickness than listed
in 3.3.3.2.2 or 3.3.3.2.3.

Note: To maintain EI with a thinner wall thickness thanspecified in 3.3.3.1, the outside diameter MUST be
increased.

(C) A Safety Structure Equivalency Form must be submitted per Section 3.3.2. The teams must submit
calculations for the material they have chosen, demonstrating equivalence to the minimum requirements
found in Section 3.3.3.1 for yield and ultimate strengths in bending, buckling and tension, for buckling modulus and for energy dissipation.


Approved alternatives per Section 3.3.3.2
ITEM or APPLICATION OUTSIDE DIAMETER x WALL THICKNESS
Main & Front Hoops 25.4 mm (1.0 inch) x 2.4 mm (0.095 inch)
Side Impact Protection, Front
Bulkhead, Roll Hoop Bracing &
Safety Harness Attachment
25.4 mm (1.0 inch) x 1.65 mm (0.065 inch)
-----------------------------------------------------------------------------------------------------------

Basically, you need to prove that you stressed skin is able to do the same thing that if you've used steel for side impacts and other parts that are governed by rules. So you need to make equations to achieve the same EI than a 1'' X 0.065'' side impact. My suggestion on this, because I'm also a beginner on this subject, team up with your solar car team and ask them for their equations. I'm fortunate enough that the guys helped me a bit to see through those rules and land me their excel spreadsheet that makes the calculation for the thickness of carbon layer and orientation and the choice of the core material and thickness also. They said to me that they had to do it for a course and it was part of their design process. So bottom line: Go see your solar car team (if you have one, pay them a couple of beer of course, and ask them for their equations!!

Pain in the ass to prove when you already know exactly what it needs to be ok!!!

[James Waltman]

It looks like you are off to a good start. I am familiar with the SSEF because we have been doing it for a while now. We don't have a solar car team. If we did they would be coming to us for questions about composites like the students from our plastics department do.

I have also made a similar spreadsheet to the one you are talking about. I know that mine has a few errors in it. It's really damn complicated. Would the guys on your solar car team be willing to share?

The spreadsheet is probably set up to determine stiffness – mine is. Defects can dramatically impact the actual strength of a composite part and how it fails. There are a lot of ways that defects can appear in a composite part. Most of them are related to manufacturing: poor lay-up, poor vacuum bag, wrong temperature schedule, and on and on. So although you may be able to predict the stiffness with great precision your actual part can fail because of many other variables. How will you predict the strength of the joint where the panel attaches? Will the panels be flat? Will they be perfectly perpendicular to the load of a side impact?

I recommend actual physical testing to verify all of your calculations – that's what we do. I think that teams have been approved for the SSEF without any actual tests to verify what they have – that really bothers me.

[leclercjs]

James,

I would be glad to offer you the spreadsheet, but It isn't mine and I think the Solar Car guys would kill me if they knew some US teams are using their year long project they've build up so, I think you will have to sort it out and do the math. But their spredsheet is a little bit complicated, it only claculates the EI, exactly what we need to find, depending on the orientation of the carbon fibers (i.e: 0/90 or 45/45...), the Poisson Coefficient, the modulus of Young for the fiber and the core material and their thickness. Also, how they build it, you need to have access to the class notes to have the engineering constants.

But on the otherside, what I can help you is how they found their equations where right. They've calculated the strenght of beams and after they submitted it to charges and saw that with good lamination, low porosities and good bounding between the core material and your carbon fiber, you are able to approach the theoritical results. (But not exactly it, like with any material. But, who will check or be able to see if you stress skins has porosities, bad lamination bettween the carbon and the core, etc.. So I think, this is why you need to prove that theoriticaly, your stress skin IS equal or greater that a 1.00'' x 0.065''. That they can see that by simply mesuring the thickness of you side impact and see if you've no tricked your Safety Equivalence Structure.

You can do a search on this forum and you can find what other team used for their side impacts. Plenty of core material are available, from Nomex (that need prepeg) to regular blue insulation foam used in home construction. We are more looking into a core material that can be used in a VARI (Vacuum Assisted Resin Infusion) or RTM (Resin Transfer Molding) molding technique. So Nomex is out of the ball game since with VARI or RTM, the Nomex will stay full of resin and it's a big no no and just there, you gained all the weight you've desperately tried to lose by doing carbon stress skins.

Carbon fiber is a pretty awesome material to work with for it's properties, lightness and ease of molding (ok, depends on the experience of the guy). But you can face also problems if you do not proceed correctly. The solar car guys told me to be aware of surprises when you are going to be de-molding your carbon fiber.
A flat sheet can give you some bumps or curves if you've not placed well your carbon fiber. The reason to this: residual constraints!!

Also, carbon fiber is there to take the shear stress, so you need a core that will be there to take compression stresses. (I've read that, I'm I right? or it's the other way round, because carbon fiber is pretty lame in compression)

[Didier Beaudoin]

Actually, carbon fiber is the best composite material in compression. A lot better than aramid fiber, for example.

I wonder if one or some kevlar layers in the side impacts would effectively improve the resistance to impacts. Kevlar is extremely strong in tension, and is often use in ballistics, to armour buldings and vehicles, for example. Also, a kevlar fabric is used in bulletproof jackets.

Therefore, I think it may have a use if an impact would actually happen between two formula cars.

Any ideas on this?

[leclercjs]

Hey Didier,

don't talk to much in here, they may discover our innovations http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

But one thing is, I don't know how much kevlar runs in the cost report, it may hurt a lot to use it!! http://fsae.com/groupee_common/emoticons/icon_eek.gif

[Charlie]

Kevlar is actually the same cost as fiberglass.

That's our real-world cost report for you. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[leclercjs]

Sweet!!!!

Thanks Charlie for the info, I've look in the rule book and didn't find the price for Kevlar, just the name scared me off and though it was expensive.

[MikeWaggoner at UW]

I wouldn't bother with kevlar because it has poor specific stiffness (relative to carbon), especially in compression. If you make your car stiff enough from glass or carbon (or god forbid, steel or aluminum), it'll be strong enough in side impacts.

[Travis Garrison]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Didier Beaudoin:
Actually, carbon fiber is the best composite material in compression. A lot better than aramid fiber, for example. <HR></BLOCKQUOTE>

Boron should be stiffer in compression...

Travis Garrison

[Didier Beaudoin]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Travis:
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Didier Beaudoin:
Actually, carbon fiber is the best composite material in compression. A lot better than aramid fiber, for example. <HR></BLOCKQUOTE>

Boron should be stiffer in compression...

Travis Garrison <HR></BLOCKQUOTE>

Yeah, but boron is actually a metallic reinforcement... http://fsae.com/groupee_common/emoticons/icon_wink.gif

But you're right. I should have said "among the common composite materials."

[leclercjs]

Ok,

just a question like that, it is sure that Kevlar is maybe not carbon fiber for its mechanical properties but, what about the mix Carbon/Kevlar stuff. Since Charlie says it is the price of fiber glass, is it calculated in the cost report 1/2 price of the carbon fiber.

And we're can we find the price of composite. I remember that Carbon fiber is around 100$ per pound, but what about the rest? http://fsae.com/groupee_common/emoticons/icon_confused.gif (like Nomex?)

[MikeWaggoner at UW]

The carbon/kevlar stuff is made for failure mode. Kevlar fails through fibrillar failure (frays and increases in size, using up energy) so it's good for impacts etc). The problem is that you have materials with very different moduli, so the carbon will be bearing most of the load. Basically it'll be a lot worse than carbon alone, but it'll fail better.

[Matt Gignac]

I'm curious how teams go about bonding the composite panels to the rest of the chassis. Is the core bonded to the tube, or is it the carbon, or both? Before or after powdercoat/paint? Have any teams who have gone the stressed skin route seen seperation between the panel and the rest of the chassis?

Matt Gignac
McGill Racing Team

[Rob Davies]

Is the stressed skin a simple couple of layers of carbon or more complex with honeycomb core etc??

How do the judges view stressed skins if you say we just put it there for some extra stiffness instead of the bodywork, instead of designing in the skin and hence reducing the required stiffness (and weight) of the space frame.

We are only dabling with carbon fibre at the moment and hence we dont want to rely on it too much.

Thanks, Rob

[MikeWaggoner at UW]

"How do the judges view stressed skins if you say we just put it there for some extra stiffness instead of the bodywork, instead of designing in the skin and hence reducing the required stiffness (and weight) of the space frame"

It's all about engineering decisions. Are the skins heavier than the bodywork? Did you need the additional stiffness? If you're just making it stiffer and heavier for no reason they'll dislike it.

"Is the stressed skin a simple couple of layers of carbon or more complex with honeycomb core etc??"

The UW car I've seen was just skins, and they oil-canned (see Denny's post above). I'd recommend coring them or putting in ribs (linear orthotropic FEA will not necessarily show buckling behavior).

[Erik C]

We used prefabricated carbon panels. Not necessarily the best option, but for ease of manufacture they worked very well. The panels act as shear webs. A quick shear buckling calculation (see Roarks or Bruhn for formulas) should convince you that you laminate is stiff enough (shouldn't take much as the loads should be minimal). For areas where you will need to absorb impacts in the out of plane direction, I would do a couple of calculations. First, check to make sure that your shear stresses are low enough at the attachments to the structure that the laminate will not have a "pull thru" failure. Most carbon laminates have an ETW (elevated temp wet) short beam shear allowable of between 4,000 and 5,200 psi, which is the strength of the laminate in the 1,3 and 2,3 directions. Second, I would look at the skin stresses due to bending caused by an impact load in the middle of a simply supported flat plate. This will be heavily dependent on the laminate thickness (obviously). Finally I would check that I don't have any fasteners fail in bearing or pull-thru. Fasteners in bearing MUST have an e/d of 2.5 or greater in composite structures to prevent tear out. We use a bearing strength based on, tests completed on single shear strength at 4% hole elongation (stabilized) of 40,400 psi. For pull thru, use an allowable of 3,800 psi (ETW again). These are all values, which come from test results and are difficult to predict using hand calculations or FEA (yes, you could do a detailed FEM of the area around your fasteners...but why? do you really have that much time on your hands?)
As far as calculating stiffness' and strength allowables for laminates, there are some good examples in "Engineering Mechanics of Composite Materials" by Daniel and Ishai, MIL17 and of course Tsai's "Composites Design"� which should be able to give you a good start on how to calculate your A, B, and D matrices. Have fun!

[Rob Davies]

hey Erik thats some great info you have for us there so thanks !!

we too are looking at purchasing/being given (we hope) pre fabricated flat sheets of carbon fibre. They advised us that 3mm is what would be required for stress skins. Not sure on the fibre orientation and number of layers that would be best for this application? Any help

Some more questions if anyone is feeling kind and would like to help us?

1) when you say bond the carbon fibre to the frame is this bond with a simple epoxy?

2) How do you do a cut and fold method like the Vikings (washington). Looks good but not sure how they pull it off.

Thanks

[Erik C]

For shear webs, the primary load is well, shear. Assuming that the structure is designed as such, the orientation of your plies should be along the axis where shear stress is highest.........hint, good ol Mohr's circle contains the answer you desire