Carbon Fibre Shell Tube Impact Attenuator

[Scuderia]

Was thinking, there seems to be a lot of research out there about hollow carbon fibre tubes being used for crash structures (particularly frames), also theres a lot of compression tests on cricular or conical carbon fibre tubes too. They work pretty well, you can arrange plies/geometry so that you can keep the initial peak load down and keep a good average load as the absorption process progresses. You can get good energy absorption figures too.

Was wondering whether any team out there has ever done anything like using a big carbon fibre shell tube as an impact attenuator? If not, well any obvious reason why it cannot be used (in the regulations)? Any advice?

[Drew Price]

There are many teams who do this, particularly with the monocoque cars because the attachment lends itself quite nicely, rather than trying to attach it to a tube frame.

Disadvantages could be complex analysis and high testing cost, advantages are as you stated, being able to tailor the geometry and properties to suite provided you are capable of doing the development.

They are light too, and don't require the full bulkhead mounting plate like a honeycomb structure does, a big advantage on it's own in my oppinion.

There is nothing specifically prohibiting this type of structure, just be very sure your mounting is designed and reinforced adequately, you will get reamed with the new stipulations for the IA report and for design judging if it's not up to snuff, there are much easier ways to do it out there.

Best,
Drew

[Jersey Tom]

We asked a [very smart, mad scientist] professor about this a few years ago.

The analysis is ridiculous. It's also very nonlinear, so building and testing a scale model probably doesn't apply. Way to do it is full scale physical testing.

Tubes can work, so can a carbon nose cone shell filled with your favorite density foam (this is what we used). Just have to definitely have a number of configurations ready to be tested.

[Scuderia]

Thanks for the feedback guys.

@Drew, yes your right about the mounting methods, Ive suggested having a flange at the end of this composite tube so it can be bolted onto an aluminum plate which in turn is bolted onto the front bulkhead. Im aware this plate isnt required but I think for stability purposes its a safer bet.

@Jersey you say tube(s) plural, I was thinking just one big one. I did consider having a "staged pyrimadal" configuration, where at each stage theres a number of mini carbon fibre tubes, but this is very difficult to manufacture......

As for the analysis......yea its a pain! But does anyone reccomend any programmes??? Im using ABAQUS at the moment but im aware that its not much use for crash simulation (so Ive been told)

[Drew Price]

Like Tom said, the simulation is highly non-linear, impact loading analysis is plenty hard enough even with materials that can claim isotropic properties.

I do know of at least one team however that used a carbon nosecone attenuator with no filling element, the geometry and the orientation and weave of the plys was all that they proved analysis of. I know that they cited a paper developed at their school which claimed that energy dissipation curve of impact loading of a carbon composite was very close to the energy curve of the same structure subjected to a slow crush though, like you could do on a tensile-test machine. That was enough justification at least last year.

Unless your tube is tapered or something such that when the structure starts to buckle there is more material underneath so support the crush, I think you will find thay your energy dissipation falls off sharply, I would strongly suggest some sort of pyramidal or other geometry rather than just a single tube. Especially a carbon one. Once the first fracture starts there will be little to hold it together.

Best,
Drew

[Steve Yao]

Regardless of what you do for analysis, SAE likes to see full scale, full velocity impact testing.

Just looking at the "grades" they gave to IA reports submitted at FSAE West last year, teams with dynamic testing were given better grades, period. Even if their analysis appeared lacking compared to another team(who did not do dynamic testing).

As mentioned, the analysis is difficult. So difficult that one of the University of Washington professors that researches composite crash structures says that there only a handful of people in the world who can honestly say they can do it well. So IMO your focus should be physical testing; a bunch of coupons to get your material props, stepping up to full scale tests.

Jersey Tom's approach of carbon over foam is a good and easy one. I believe a team of students from Cranfield received high praise in a impact attenuator design competition a few years back for such a solution.

[TG]

The prof who teaches composites at ASU was the lead engineer for Boeing in getting the 787 structure crashworthiness approved by the FAA and JAA. He's showed his classes the rigs they used to test the structures...pretty cool stuff. They started off with small beams and tested them and worked their way up to larger structures. We also saw some high speed video of some failed beam tested buckling under load.

As far as you should be concerned with the design, you want to have the only mode of failure to be in fracture where energy absorption in composites is very high. You want to prevent and bending, buckling, and crippling within the structure.

Additionally, you need to consider delamination in composites. Again, you want to prevent this from occurring. If you are using UD plies, you will want to keep the ply orientation at or below 45 degree from it's adjacent plies (ie [0 45 90 -45] rather than [0 90 45 -45]) to keep the chances of delamination at a minimal.

Gary Savage (Honda F1's composite guy) has written a few white papers on this subject, you should check those out as well.

[Scuderia]

Once again thanks to everyone for their suggestions. I'll crack on with it then by focussing on physical testing. Generally Ive found the computer simulation very difficult to interpret or carry out accurately.

[impact@cmu]

@TG where can I get Savage's papers? Any suggestions?

[Thrainer]

Check out the latest video: http://www.amzracing.ch/amz/video.ort

I think I spotted some delamination http://fsae.com/groupee_common/emoticons/icon_wink.gif