After many hours of filling and shaping and sanding, the two halves were split (using a very thin kerf Japanese saw) into a right and left side.  These PICS show the middle portion of each plug half and the 1/64 inch plywood that was glued to the right and left sides.  In detail you can see the headset area where I extended past the plywood.  This occurred because when forming a shape, one’s eye (mine in this case) and fingers tells you where a smooth shape must be filled (or reduced) to create a smooth molded and faired shape.

A word on the carbon that was used here!  I wanted an uninterrupted shell (therefore an uninterrupted piece of carbon on the plug) so that the surface would be smooth with no undulations of any sort.  Undulations and distortions would be caused by cutting the carbon in pieces and overlapping it on the male plug.   I choose a 4X4 Twill Carbon which was 8.3 oz and dresses out to approximately .016 inches.  I used two layers which resulted in an overall thickness of .032 inches.  I chose that material because it is very drapable and conforms to curves more easily than other weaves.  I also used a very wide piece (like 10 inches) to insure that the same piece (actually two) could be made to fit over the part that was to be formed.  This made for more waste than normal.  I could have made the shell with 3 narrower pieces of the carbon, but then the overlapping areas would cause interruptions on the surface of the frame.

I cut two 10 inch wide X 50 inches long of carbon, placed them on a flat surface and applied a generous amount of resin.  The carbon was put on wax paper.  I then vacuum bagged the carbon with peel ply on one side of the carbon, and generous blotting material on the other.  With 20 inches of vacuum I removed most of the resin over a period of approximately 15 minutes.  I used a slow cure epoxy resin to ensure that gelling would not occur at this time.  I then put the material (carefully) over the male plug and formed it to the shape.  At this time an inordinate amount of time is spent insuring that the carbon adheres to all surfaces of the male plug.  The peel ply faced outward and the raw carbon is against the waxed mold.

These PICS show the frame half after all trimming has been done.  The shell has consistently thickness throughout the piece, it is extremely smooth and IMHO the results are spectacular.    There are no highs or lows on the surface.  After some minor wet sanding and a light coat of clear, this will look as perfectly smooth as any factory part.  I cannot praise this method enough and consider it very acceptable as a homebuilder method (but not for production).  Allow me to evaluate the two methods at this point.  The first was laying carbon in a female mold and injecting expandable foam.  The second is this method which is laying carbon on a male part.  The disadvantages of the female mold/foam are:

1. All layers must be applied before the foam is injected.  This means the outer layers of carbon as well as the strengthening layers.  You will see this strengthening process in a bit.  In order to accommodate this, you must work like crazy to complete before the resin begins to cure.  You don’t answer the phone, nor do you take coffee breaks, nor do you go to the bathroom!  In the male mold method, you are only applying a single layer (actually two, a double layer) and have more time to ensure that the carbon is forming to the mold.  The strengthening will be done after curing, so you are making it a two part process instead of one.
2. The expanding foam creates so much pressure that it infiltrates the actual carbon.  There is a lot of experimenting that must be done to prevent this.  While this does not seem to be a problem with my BC (going strong after 8,000 miles), it does not look good, and that is why I applied paint and filler (to make the bike look good) to the bike.  This added weight!  I estimate more than 1 pound on my BC.
3. While it is difficult to conceive, foam is not light.  A two pound (per cubic foot) expanding foam interior adds a great deal of weight.  You are better off strength wise to add extra carbon in lieu of foam, with added blocks of balsa wood where penetrations occur.  This latter point will be shown later.

These PICS show the main frame in daylight.  The color variation in the carbon, especially at the headset area, is not a thickness factor, but is caused by the peel ply.  This will go away when I sand the outside.  The weight at this time is 350 grams for both halves.

This is the final step before the two halves are to come together.  Balsa blocks are epoxied in the shells so that penetrations can be inserted.  Looking at the photo, the two blocks on the left are for the rear bolts which will hold the rear of the bike to the main frame.  This would be the rear stays and the rear seat support.  Forward of that is for the front seat brace.  The block at the headset is for the headset tube as well as for the power idler (remember this is a FWD).  Just forward of that is a block which will reduce torsion bending in the main boom, and the extreme right is for the bottom bracket.  Weight wise I am now at a little over 1000 grams for both halves.  This is lower than I was with the BC at the same time.  I predict that if the weight penalty for the wheel drive does not become excessive, the total bike will be in the lower 20 pound range.