Much depends on the facilities, skill level and budget you have available. Its all well and good designing the "ultimate" only to find practical or budgetary restrictions make it impossible to manufacture correctly.
Given the relatively small loadings it should be possible to keep the layup simple by using the profile of the the shaft to add strength. Braided sleeve comes in a variety of sizes and can be expanded or contracted easily. If you go up a size, when you pull it to contract it the fibres become more inline getting nearer to uni directional. With some thought you can use this to your advantage.

For me I would increase the number of layers in the area where the head will attach, giving a tapered wall thickness along its length. To try and disperse any point loadings. You could add some E-Glass or similar to improve impact resistance if required.  I would be concentrating more on how to attach the head to the shaft reliably.
By keeping the layup simple you should get much better repeatability, making any testing you do more meaningful and accurate.

Obviously this is just my opinion  Warren.

Carbon Copies Ltd

I thought steel could be designed for infinite load cycles. As long as you keep the load below a given amount it never reaches the fatigue limit. Aluminum on the other hand will always fail after enough cycles, no matter how small the load is. I think titanium's the only other metal that can be designed with infinite load cycles.

One thing I have found from testing is that carbon is extremely tolerant to fatigue.(when designed correctly) If the shaft on your project was made from steel or aluminum tube you would expect the cycle of on/off loads to cause fatigue cracking at the point where the shaft joins to the head, so you would have to design in the number of expected cycles plus a safety factor. With carbon if it is capable of taking the load 10 times it will take the load 1000 times if that makes sense.
On the downside unlike steel or ally a carbon part gives little or no indication if it has been overloaded, it simply breaks.
We actually now use this in the wishbone design, eliminating the rose joints on the inner mounts. The wishbone itself now provides the flex to allow suspension movement.

As for layup. If you are new to composites and planning on using some sort of split mould with an internal bladder I would use braided sleeve. As this will expand and contract easily. You could place the sleeve over the bladder, wet out with resin, place in the mould and expand the bladder. If you wrap conventional cloth around the bladder and don't allow sufficient "slack" then the bladder wont be able to expand inside the mould.

Perhaps now you are beginning to understand the high cost of the ones at the beginning of your post, you are paying for the R&D that has gone into getting the product to market. I would imagine they have also designed in a lot of "what if" scenarios as in the real world people tend to use these things for purposes they were not necessarily designed for.

Sorry the wishbone development like much of my work is covered by a wagon load of NDA's so I cant give too much away on a public forum.

Hope that helps Warren


Carbon Copies Ltd

Awesome, thanks guys!

Looks like you're heading down the right road with you're thinking so far. I was just thinking the right hand load point on the 2nd sketch looked pretty sharp. Carbon doesn't really like point loading. Warren was highlighting this as well I think. 

I can find little fault with your thinking so far. All things being equal failure will occur at the point where there is a change in section, the more you can smooth that transition the better. By layering the reinforcement along the length of the shaft you can transfer the loads from the head reducing the point loads where the two connect. I do very similar with race car wishbones where there are aluminium inserts bonded into carbon tubes.

Perrrrrrfect. Now to make friends with the professor who runs the machine shop....

Most adhesives I have used had a recommended clearance of 0.2-0.4 mm so your not far off at 10 thou. Too tight a fit and most of the adhesive gets pushed out when the parts are assembled.

I had more stuff but I got to an actual computer and didn't want to type on my phone anymore.

I'm sure my machinist buddy will be thrilled to hear that instead of just cutting a channel with constant radius fillets on the sides, I now want a superellipse, haha. I wish they let us actually use the 3D printer at school, then I could print out a superellipe handle and make a mold of it in fiberglass. Oh well.