I’ll take a stab (I thought of using the word “strike” but didn’t want to get flamed too much!) at the problem with carbon vs glass or metal during lightning strikes.
1) The lightning bolt can be thought of as a constant current generator, big, but constant. By this I mean, whether your plane is in the middle or not is immaterial to the current in that stroke – the current is being controlled by much larger factors than your insignificant plane. (I know, from you’re viewpoint, it’s a big deal, but not from the lightning bolt’s!)
2) 50% of the strokes are in the current range of 10KA for some 100us to 400us. The voltage before strike can be estimated in the 1MV to 10MV range. (If you’re interested in more about the subject, a cheap book is “Lightning” by Martin A. Uman and available from Dover, ISBN 9-486-64575-4.
3) So this 50KA stroke now tries to include your plane as a path. Here two different scenarios emerge: Metal and carbon both are more conductive than the air so they can provide a preferential path. The problem now devolves to total power dissipation. We’re interested in total watt*seconds that heats up our plane, and that formula becomes current squared x resistance x time. Since current is essentially constant as is time the important factor is the ratio of resistance of the path through the plane. Unfortunately for equal volumes this is the ratio of Carbon = 1375 to Aluminum = 2.6548 or about 519 times. Add to this the fact that carbon cloth is not solid so the number is larger (worst) and finally add in the fact that aluminum (pure) melts at 660 degrees C and boils at 2467 while the epoxy binder probably starts failing in the 300 degree and boiling at 600 (?). The specific heat of Al =.215 vs C = .17 ratio also means the carbon composite heats more for equal power. So I estimate that this means our carbon plane is about 2000 times or more susceptible to heat build-up from the lightning stroke. As another data point, I recall Hal Woodruff and I measured about 5 ohms on his IV while under construction. (correct me if I’m wrong, Hal.) Using just this number (ignoring many other effects like current crowding, path of least resistance, etc.) we now find that 10K^2 * 5 * 100e-6 Ws = 50KWs. This is enough to thoroughly heat a path in carbon cloth, but conversely divide the number by some 2K and you only have 25 Ws which is probably why the aluminum isn’t fazed.
4) And finally, the glass cloth model. Glass/epoxy is less conductive than air and therefore doesn’t carry any current at all therefore no power dissipation. The lighting bolt can strike through the glass to a conductive surface, such as carbon or aluminum at which point the equations above take over for the path. But typical numbers used to design electrical insulation have values like 1KV/milliinch of thickness so with only a few mills of glass insulation, the air breaks down and the electricity follows the ionized path in air (read lightning bolt!) It just plain doesn’t want to go in the glass/epoxy at all.

So in a rather long winded explanation the problem is that the carbon is not conductive enough for these large currents. Foil or light screen may not be either taking these kind of numbers into account. I honestly don’t know the solution, but I wanted you to have a bit of insight into the nature of the problem.

Comments?

Charles R. Patton
LNC2 N360JM