Lightning test on composite aircraft and groundwork for finite element modeling

My interest in this subject is a result of curiosity arising from the conversation with Prof. Rani-el-hajjar at University of Wisconsin at Milwaukee over several aircraft being hit by lightning and yet survived. Although aircraft are not lucky enough to survive such lightning bolts. Recently a Colombian plane crashed after being struck by lightning, on the island of San Andres in the Caribbean Sea(My sincere condolences for those who lost their lives).
   I had known that aircraft is made of aluminum or specifically some sort of aluminum alloy: namely Duralumin.  Since aircraft body consists of a conducting alloy, the lightning-sparks, which hit the aircraft, are instantly conducted and the passengers inside are safe from any natural macroshock electrocution. New Boeing 787 aircraft has replaced several of aluminum parts with composite parts owing to less weight and higher fuel efficiency. Since composite is 1000 times less conductive to lightning compared to aluminum, obvious concerns of passenger safety during thunderstorms arise.
   Boeing came up with the idea of embedding a thin metal mesh in the outer layers of the composite fuselage and wings so as not to cause a potential fuel-tank explosion due to a spark reaching and igniting fuel. The Seattle Times reported that a gap in fastener with the hole could cause sparking as current jumps over the gap. I have observed that many sockets create sparks at home, if the electric plug is loose. A snug fit it therefore desired.
   In this blog, I wish to talk about lightning and a possible test-setup where a destructive testing over composite part could be done. There are tests conducted by Lightning Laboratory for multiple strokes and multiple bursts. I wish to relate the test with some sort of numerical computation technique as well.
   First, a tad about fulminology.

"They all look beautiful. Isn't it? Only till they don't strike you."

    I have loved them; every thunderstorm and every monsoon. My early calculations were trivial. I would use a stop-watch to calculate the difference between the thunder-sound and thunder-strike on the sky. What causes thunder-sound? It is produced by the expansion and contraction of air in millionth of second as electricity passes through air. Hence thunder-strike is precursor to thunder-sound. Any experimental test-setup would therefore produce thunder-strike and not thunder-sound because thunder-sound does not affect the aircraft composite part.

    Wikipedia says that lightning is an atmospheric discharge of electricity accompanied by thunder, which typically occurs during thunderstorms, and sometimes during volcanic eruptions or dust storms.
Amongst various types of lightning, intra-and inter-cloud lightning strikes the aircraft most. Hence the mechanism by which inter-and intra-cloud lightning occurs would also be the mechanism for generating thunder-strike for composite-part testing.

    How much electricity do we need for such an experiment? Enormous. A single flash of lightning contains about one billion joules of electricity which is equal to the energy to light a 100 W light bulb for three months. The temperature of lightning's return stroke can reach up to 50,000 degrees Fahrenheit. Hence such an experiment would require not only enormous electricity but also some sort of thermal insulation for researcher.
 
   Tesla coil could be a good idea to generate arc-lightning. There are other methods for example using a Van De Graff generator. Aircraft part can be subjected to probabilistic lightning criteria. The composite part can be subjected to varying intensity of arc-strikes at various locations. A Monte Carlo technique could be used to create arc-strike at different locations of composite parts, thereby confirming that part has been tested for lightning effects for 99% locations.

   A numerical way of analyzing the effect of lightning would be to :

1. Generate a model of aircraft part in CAD modeling software.
2. Assign the material properties especially thermal and electrical properties to model composite.
3. Using a multi-physics approach, allowing the high voltage to pass the current through such a part. I personally know of COMSOL Multi-physics where this could be done.
4. The current, depending upon the electrical conductivity, would try to flow as much as possible. The carbon-fiber in composite try to pass as much current as possible but owing to less electrical conductivity, also offers resistance to current. Same goes with resin. This resistance would result in loss of electrical energy by means of energy conversion. 
5. I would focus on loss by means of heating of composite. Essentially, the current would cause Joule's heating in the fiber owing to poor electrical conductance. Joule's law of heating from current is the bridge between the two domains of Physics: Electricity and Heat transfer.Joule's heating causes the conversion of electrical energy to heat energy.
6. Now lets say that temperature in each of the elements is increasing as a result of production of thermal energy. When the melting temperature(Tmelt) of composite is reached, the element in the mesh would degrade. The degrade would again be gradual asking for a non-linear analysis. As the element degrades, it looses its electrical and thermal properties. Again this degradation would also result in breaking of the current circuit. Eventually, the element would die as the result of heating.
7. Such an analysis would be conducted on millisecond timescale. Hence some sort of time scaling would also be required for carrying out electro-thermal analysis of composite. 
8. As the elements in mesh dies in milliseconds, the current would quickly change its path. This would to a certain extent also model the random nature of blitzkrieg.