Bruce,
It’s about 10 years since Tim had his windshield frame repaired with carbon fiber. the repaired area still looks perfect.
I agree that one has to be carful with the finished product. It is not designed to be used as a hand hold to get in and out of the car.
As to the repair, I did not see the repair layup until it was finished, but if it was me, I would follow good practice and grind out both the front and back side in the affected area, plus some, to a tapered shape sufficient to ensure perhaps 3 layers of cloth can be accomodated accros the crack area and still arrive at the correct finished geometry. As in any structural shape, highest loads will be at the surface, so internal layers will provide progressively less benefit. I don’t know how difficult it will be to wet out the repair layers of carbon fibre, but we’re talking about relatively small areas here, so it seems with care, a proper job could be done. Use of epoxy, like perhaps West Systems epoxy, used for boad building, might be advantageous due to it’s superior mechanical properties relative to polyester resin. Orientation of the fibers, primarilly in the length wise direction, will also add additional strength. I sometimes do this with coarse fiberglass cloth where I strip out some individual yarn, slightly separate the individual strands, and orient them in the desired direction, in this case axially or lenght wise.
Use of a steel rod for the core, is also a good idea. For this application, I would also prefer to use stainless, especially if polyester resin were used to bond the cloth or matt overlay. Polyester resin is not impervious to moisture, epoxy is, if the overlaying thickness is at least 5-10 mills.
Different coefficients of expansion due to temperature are something to keep in mind. A steel core probably should be uncoupled from the overlaying layup by not going out of your way to ensure perfect bonding with the steel. If one looks closely at how steel rods are glassed into the Elan body, it appears some axial movement is allowed as the joint is not completly filled, just lightly wrapped. Polyester resin will likely expand and contract more with temperature change than will a glass/restin matrix, especially if the lenght is significant. Carbon fibre expands and contracts an almost negligible amount due to temperature change, that’s why they used it on the Hubble space telescope framework. Still, over a limited length of say 4-5 of the repair the relative movement of the material is small and the underlying glass matrix should accommodate the movement due to creep at the repair interface.
A proper repair should be at least as strong as a new frame and be visually undetectable.
Bill