Carbon fibre and Aluminum, Thermal expansion

[sheidebr]

I'am building a carbon fibre box made from carbon fibre panels. I am using aluminum angle on the corners to prevent damage to the edges. The aluminum is throughly bonded to the carbon fibre.
What I am worried about is thermal expansion since the box is about 6 feet long. Over this length I predict that the aluminum will srink 0.083 inch when I cool the box from 20C to -30C.
Does anyone have any recommendation on how to avoid the aluminum from de-laminating from the carbon fiber?
I was think of breaking up the alumium into 6 inch pieces to reduce the overall effect of expansion, but I do not know if this makes sense.

 

[Joshmo]

Can you use something other than Aluminum that has similar CTE properties to the carbon panels? What about SMC or pull-truded rails?

What is the aluminum for? If it is only used for locally fastening of the Panels perhaps only bond the aluminum where necessary and match drill...

 

[CompositeGeek]

Have you considered making the box edge a U channel instead of an L angle? Then the carbon will move within the channel.




Composites and Airplanes - what was I thinking?

There are gremlins in the autoclave!

 

[Prabbit]

Making the Aluminum into strips will help decrese the interlaminar shear forces that the Aluminum will exhibit

Aluminum must be properly prepared before bonding. It must be etched and then prime with an epoxy primer. Otherwise the adhesive/epoxy resin will not provide a proper bond.

Alumium and Carbon is not a good combination. Any moisture, condensation, will accelerate the corrision process.

 

[berkshire]

sheidebr (Aerospace)
As several other people have pointed out, aluminum and carbon fiber are not a good idea, (Think battery) If the metal edge is simply to protect the box from damage, have you considered substituting a thinner stainless steel angle for the aluminum? This can be etched and epoxied in the same manner for bonding to the carbon fiber.
B.E.

 

[ewh]

If you have to use aluminum and carbon, use a glass ply between them to isolate the aluminum.

 

[jabonet]

what impact are you protecting the edge from, could you consider other materials less prone to galvanic corrosion?

 

[nbucska]

if you protect the edge from impacts, you could use short segments of "L" of any metal.

Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[RPstress]

A delta T of 50 deg C will probably be ok.

The shear stresses in these sort of joints peak at the ends, typically over the last inch or two. (Depends heavily on relative stiffnesses of the substrates, the glue shear modulus and the glue line thickness.) Thus cutting the aluminium into six inch segements probably won't help. You'd probably need even shorter lengths. To quantify the critical length for peaking you need to do a little stiffness analysis, preferably with FE. There ought to be some little formulas for doing it by hand, but I've never found any.

If you've got equal-ish stiffness-times-areas of ally and carbon then they'll both strain the same amount and the total relative length change will be half the difference in the two CTE*delta-T's. This is the best possible situation.

If one material totally dominates the EA then the total length change will be the difference in CTE*delta T. Thus, if the carbon laminate has a CTE of 3e-6/deg C and the ally is 24e-6/C, the worst case strain (say the panels dominate) is 21e-6*50 = 0.00105 tensile in the aluminum. It won't be quite that bad, because the panels will be shrunk a bit by the aluminum pulling on them. Anyway, a strain of 0.00105 translates to a stress of 0.00105*10.5e6 = 11 ksi.

Now, assume the angles are 0.02" thick and have legs 1" long. That gives a cross-sectional area of 0.04 in^2. With a stress of 11 ksi this will generate a force of 0.04*11000 = 440 lb. This will be reacted with some sort of distribution at the end of the joint. Assume it is reacted over the end one half inch. Shear area is 0.5*(1+1) = 1 inch square. Total shear stress would then be 440/1.0 = 440 psi. A good adhesive will have a shear strength of at least 2000 psi.

If you're just looking to the aluminum to protect the carbon then it's probably a bit OTT. A thermoplastic would probably be cheaper and lighter. However, if you're also using the ally to join the panels together then it's a reasonable choice.

With regard to insulating the metal from the carbon, it's usual to put a layer of glass scrim in the laminate in way of the metal, but it depends on your service environment. If it's indoors then you'll almost certainly be ok with just the adhesive and the primer on the metal. It's not as if you've got any fasteners potentially making a circuit with actual fiber contact.