Working with carbon fibre

[21TonyK]

Thought I would ask here as it seems pultruded carbon fibre is a popular modelling material...

Basically I'm asking on behalf of my daughter who is looking to produce scale versions of glulam beams reinforced with carbon to strength test again unreinforced and natural timber. Shes also looking to scale model a new type of connector for steel beams in carbon which she intended to bond with an epoxy resin.

So far I've found and exchanged a few emails with http://www.easycomposites.co.uk/#!/

Has anyone one experience with this sort of thing? Suggestions for epoxies and or suppliers? She's already got meetings set up with various companies to look at the lamination process but the CF bit appears to be a complete unknown to them hence looking to play around with scale 1m beams she can make herself.

[GliderRider]

The Light Aircraft Association are the governing body for homebuilt aircraft. A number of those use wood and cabon fibre construction. In addition, there are a number of factory-built aircraft with pultruded carbon spars such as tthe SD-1 SD-1 microlight

Despite its name, The Welding Institute (TWI) provided me with a lot of useful info when I had an aeroengine component which needed metal to composite bonding Surface Finishing TWI

These people should be worth talking to: Carbon Fibre Profiles

Aeromodellers are also a useful source of information: google 'Bonding Carbon Fiber to wood' (note American spelling).

A few pointers from my own experience:

1. Thoroughly abrade all surfaces to be bonded, then clean with acetone or isoproyl alcohol (IPA) Wear clean gloves whilst doing this, as finger prints make an excellent release agent.
2. Plywood, particularly birch ply, has a release agent which inhibits bonding. Abrading and cleaning with IPA is particularly important in this case.
3. Bonding to wood parallel to the grain is likely to result in bond failure by the outer layers of the grain separating from the rest.
4. Bonding to end grain wood which is used as a compression or shear web is generally the most successful.
5. If the materials to which the epoxy is to be bonded is warmer than the epoxy, the epoxy will wet out better with less likelihood of air trapped between the resin and subtrate, This is particularly when bonding to metals.
6. Post cure is very important. If epoxy is heated under load to higher than the temperature at which it was cured, it is likely to soften and either distort or fail. (Imagine the temperatures that can be reached in a house roof on a hot summer day). This is why composite light aircraft are usually finished in white on the top surfaces.
7. Slower setting epoxies generally give a stronger bond.
8. Evenly applied pressure across a bond always helps. Vacuum bagging is good, but clamps onto wooden load spreaders, with a non stick layer against the parts to be bonded also works.
9. Make test coupons with a sample of resin for every mix you do, and expose it to the same post cure as the main test piece. put the date, time and resin batch numbers on each one.
10. Make sure each test is identical with the exception of the test piece.
11. Video each test to destruction, with a card identifying what the test is and the materials used.
12. Perfect your bonds on small cheap parts before spending time on large, costly, time consuming parts.
13. When doing small test pieces, aim for a minimum of five, so you can get a spread of data. A sample size of one doesn't tell you much.

Tell suppliers that it is a college project and ask for any free offcuts, end of lines, material they can't sell for one reason or another also any release film, vac bags, peel ply and time-expired resin (usually ok if you do some tensile test coupons to prove its ok). Tell them you would like to mention their assistance in any publicity, reports etc.

With regard to materials sources, speak to glider & light aircraft repairers, boat yards, especially specialists such as racing yachts & dinghies & olympics-type rowing boats, super car repairers, etc. Expect to contact at least ten before you get a positive response.

Kite shops sell pultruded tubes.

Free Flight Supplies is also a good source of pultrusions in smaller quantities: Free Flight Supplies

Robotbirds are agents for Van Dijk pultrusions: Robotbirds Pultrusions

Edited by GliderRider on Monday 11th June 20:41

Edited by GliderRider on Monday 11th June 20:53

[21TonyK]

Thank you so much for your reply. There is more useful information and advice in your post than a week full of google!

Will get her to follow up on your links and suggestions.

Thanks again.



Owe you a virtual pint.

[anonymous-user]

Sounds like an interesting project. One minor point which may or may not have been considered is scale-ability of timber and the associated grading process.

We do a lot of timber beam testing (Glulam, LVL, solid softwood and hardwood, I joists, metal web joists etc) and for demonstration purposes we made a small-scale bending rig to show a 4 point bending test (just as a bit of fun/educational tool). Having used it to bend and break hundreds of bits of timber, the variability of small sections was quite surprising. Timber is a naturally variable material at the best of times, but the smaller the section the more variable it can become - it is for that reason that the grading rules limit minimum section size to a cross-section of 2000mm2 or greater.

Having said that, you can of course get some interesting and useful comparisons between different materials, but just bear that in mind.

One other minor point, and I am not sure if it is relevant, but if there is any heating involved with the manufacture of the samples (e.g. autoclaving CF), if you heat timber over about 70c for any significant length of time you can end up with an appreciable loss of strength.

If you want any more info on testing etc, PM me. I don't personally do the mechanical testing, but I can talk to the man (well team) who does!

[21TonyK]

Dead_Donkey said:

Sounds like an interesting project. One minor point which may or may not have been considered is scale-ability of timber and the associated grading process.

We do a lot of timber beam testing (Glulam, LVL, solid softwood and hardwood, I joists, metal web joists etc) and for demonstration purposes we made a small-scale bending rig to show a 4 point bending test (just as a bit of fun/educational tool). Having used it to bend and break hundreds of bits of timber, the variability of small sections was quite surprising. Timber is a naturally variable material at the best of times, but the smaller the section the more variable it can become - it is for that reason that the grading rules limit minimum section size to a cross-section of 2000mm2 or greater.

Having said that, you can of course get some interesting and useful comparisons between different materials, but just bear that in mind.

One other minor point, and I am not sure if it is relevant, but if there is any heating involved with the manufacture of the samples (e.g. autoclaving CF), if you heat timber over about 70c for any significant length of time you can end up with an appreciable loss of strength.

If you want any more info on testing etc, PM me. I don't personally do the mechanical testing, but I can talk to the man (well team) who does!

Very useful, and yes, had thought about the fact its only really a "model" (planning on roughly 50mm x 90mm or there abouts) made from C24. Shes not looking to actually test to breaking, just the variation in deflection.

At this stage its really just compare and demonstrate the variation between beams. Next step is using traditional Japanese joinery to connect them!


And thanks for the offer on testing info. When she gets closer to that stage I'll PM.

[AER]

Do the modelling with FEA - it's much cheaper!