Originally Posted by Jonny B
The actual build process (laying up) isn't really that hard, but designing the frame (and deciding where to put what type of carbon cloth, and in what direction) and making a foam core or mould are very difficult and time-consuming. Also, you'll have no precident for how strong it will be (whereas a steel builder will now that, as long as his or her brazing is up to scratch, the lugs and tubes will make a good bike), so some form of testing will need to be undertaken (possibly a destructive test, like the guy on Sheldon's site). And bare in mind that it's gonna cost a bit too, for what Sheldon's pal spent you could have a top-spec 'normal' bike (about $2000 I think).
The trick here is the experience & ability of the laminator. Quality control & technique important. Without it, the laminate part's physically properties can run all over the board. A poor lay-up or resin cure problem will never meet the needed physical properties. But, to the experienced fabricator, these are second nature.
If one goes with traditional straight tube geometry, the molds are easy, assuming that you are willing to go a male plug, wet contact lay-up route. Simply use fluorescent light tubes as the mould. Put your mould release agent on the light surface, do your lay-up, once the part has cured, give the part a sharp tad to break the light tube, & presto, finished hollow part. If desired, one could use this with other curved light tubes if their shape & size fits your needed radius & size.
ON EDIT: As for the fiber orientation, my first instinct w/o any reasearch would be: A triaxial layup of 0, +45, -45 deg. relitive to the frame C/L. (strength along the length & tortionally). I doubt that much hoop strength is needed. For a bi-axial, I think +/-45 deg. I would be worried about not having the 0 deg. layup. I thing I will start looking @ the carbon frames to see what is commonly used.
Thanks for the link on the kit bike. Bob