"Self-extinguishing" material stops burning, because it's already burnt part doesn't shed or become porous, blocking oxygen from the non-burnt part underneath it, so the fire is smothered. Think of a fire in your fireplace. When it's about to go out, you take the poker, and knock the burnt part of the wood off, exposing the unburnt part to oxygen, and the fire increases. In self-extinguishing material, the burnt material tends to stay on and block the oxygen, and, there's no one with a poker, knocking the burnt part off.

That is one explanation and the other, closer to the sponsored research, is that, like a candle, with heat applied the material gives of vapours which can burn. Remove the heat source and everything cools to being inert again.

The issue in the 787 case is to detremine what was the initial heat source. If it was the ELT itself then it would be a first. If it was something in that area which then caused the ELT to generate enough heat, maybe. But what if the heat source managed to bring e.g. CFRP material up to the temperature where it and/or other local materials gave off combustible gasses which then sustained the burning and then caused the disruption to the ELT?

FAA to issue AD ?
 

Boeing 787 Dreamliner: FAA to issue 'airworthiness directive' in wake of fire | Business | guardian.co.uk

Boeing 787 Dreamliner: FAA to issue 'airworthiness directive' in wake of fire

Regulator to order mandatory inspections of 787's emergency beacons just two months after lithium-ion battery problems

..
The Federal Aviation Authority was due to order mandatory inspections of the emergency beacons aboard Boeing's ill-fated 787 Dreamliner as early as Monday after a fire broke out aboard one of the planes in London.
The FAA "airworthiness directive" comes just two months after the 787 was certified to fly again following a global grounding triggered by problems with its lithium-ion battery system.
A fire aboard an Ethiopian Airways jet at London's Heathrow airport earlier this month forced the closure of both runways for more than an hour.
The UK's Air Accidents Investigation Branch (AAIB) concluded that the fire started near the aircraft's emergency locator transmitter – a distress beacon used to help rescuers find a plane if, for example, it is forced to land on water or in polar regions. The focus of attention appears to be whether a pinched wire under the battery cover triggered a short circuit.
The FAA spent the weekend telling other aviation safety regulators around the world of its concerns. It is expected to formally issue the directive early this week, possibly as early as Monday.
The beacons, made by Honeywell, are powered by a lithium manganese battery, which could have suffered a short circuit. Last week the FAA said its inspections would call for operators to check "proper wire routing and any signs of wire damage or pinching, as well as inspect the battery compartment for unusual signs of heating or moisture."
Robert Mann, an aviation expert at consultant RW Mann, said it was unclear whether the latest issue was caused by installation, quality control issues or design issues specifically related to the 787. Honeywell had issues with its emergency locators in 2009 that led to the beacons failing to send out signals. But Mann said the issue could also be specific to the Dreamliner.
The 787 is the world's most technologically advanced passenger jet. Its use of lightweight materials means its uses 20% less fuel than its peers. Aluminum wiring and Teflon coating are used to save weight. "We got rid of aluminum wiring in the US in the 1980s because of the tendency of aluminum oxide to cause problems," said Mann. "And there have been reports of fragility with Teflon insulation." ...

The battery box was to meet the new RTCA standard for LiIon batteries, the new Cessna Citation has a similar box and had to have similar tests.

Could be. I think it's a great solution to the problem.... Here's the scenario:

1) Plane flies high in the sky.... Airframe gets very cold....
2) Airframe made of composites has much lower thermal conductivity than aluminium. Nett result... airframe takes longer to get cold on the way up.... once cold... takes longer to warm up on the way down...
3) Plane lands, doors open, passengers get off. Cabin fills with air far more humid that the 15% talked about on this thread.
4) The moist air, inside the aircraft, getting close to the airframe, forms condensation.... Since the airframe is cooler than we see in aluminium airframes.... we see more condensation....
5) Water and electrical circuits don't go too well together...
6) Issues that previously were unlikely to occur, become more likely, in the more moist environment.

So by putting the battery in a sealed box.... nicely ensures that the condensation doesn't get to the battery... It wasn't poorly made batteries, or overcharging.... it was a short caused by the battery being covered in condensation.... Oh and meant to say.... the technology behind the battery wasn't the problem and this is why it didn't make sense to change to conventional batteries...

I could be wrong of course, just a theory.

Not necessarily. Airframe made of composites has much lower thermal capacity as well, so less "coldness" stored in the airframe.
GA experience: Aluminium airframes parked outside have water accumulation inside in the morning, GFRP airframes parked outside do not (or significantly less). Might be an issue of the sandwich construction as well, so for a monolithic dreamliner it may be different again.

Here's a thought: The new battery-box in the 787 is vented to atmosphere, so at altitude it's going to be at low pressure, and maybe quite cold (I don't know if the box is insulated, or how much of the aircraft's warmth will penetrate). On descent the box will "breathe in" the outside air, and if it's humid, will the cold battery have a tendency for condensation to form on it? I also wonder if they had to strengthen the battery's case to withstand the low pressure around it?

An interesting theory...
 

@ simple-simon post #666

but could you condense it down to a simple version?


Looking at the pattern of the charred area it appears that the longitudinal struts (also a composite structure?) shielded the skin or acted as a heat-sink to pull heat out of the surface.

i wonder if the coeff of thermal expansion (CTE) of the composite is much greater than aluminum and may allow more aircraft growth and contraction such that wire chaffing might occur?

You are presuming no overpressure valve or rupture disk in the vent? I don't know either way but would have assumed the opposite. I cant believe those cells would like to spend most of their flying lives at unpressurised Ps & Ts?

It's not.

It is only vented to atmosphere once the pressure inside the box ruptures a disc in the vent. This protection is set to rupture at a much higher pressure than the differential at normal altitude.

Maybe twice as much as aluminium, I am not sure about type of resin:

Aluminium 22.2 . 10-6 m/mK
Epoxy, castings resins & compounds, unfilled 55 . 10-6 m/mK

That means ca 1 mm per 1°C for hole long fuselage. Fuselage is getting shorter at FL so wires are not tighten but some movement surely exists.

2 simple-simon: Interesting hypothesis...

I think you are all reading way too much into this. If they shorted one of the battery leads under the cover, it would only take time to make an ELT immolate.

CFRP thermal expansion?
 

CFRP typically has a very low thermal expansion, at most about 6 times less than Aluminum at 300K. If designed right, it can have almost zero. Many sophisticated structures (that are not limited by weight) use it, despite its cost, to take advantage of this property. If there's any chafing it would probably be because the metal wiring changes size while the composite structure does not.

Your data is wrong.
Thermal coefficient of expansion of CFRP is about 5 times smaller than aluminium. (2.3 e-5 per degC for Al and 0.5 e-5 per degC for CFRP).

Is 'unfilled resin' the right number to use? I thought carbon composite is much better than that - because of the fiber, optical tables made of it have a COE typically 1/10th that of steel for instance, so I thought the number is lower than Al, not higher.

The hugely dominant effect is flexure in flight, have you seen the wings on takeoff?

If condensation is the cause of battery fires, and Boeing "knew it", it can be easily tested in test scenario. Are you suggesting conspiracy that Boeing is trying to hide 787's condensation issue as it effects not just the battery but the entire electrical systems on the 787s (which is a lot), and may even force Boeing to redesign the entire plane?

There are a lot of "ifs" in your statement. I could bet my home that condensation did not play any role here. But even if it did you wouldn't have to redesign electricity - you simply turn down humidity to much lower levels. There is nothing in the design of A/C-ventilation in 787 that says you MUST have higher humidity.

In light of the discussion re: composites and aluminium alloy in the
presence of fire, I wonder if someone (amicus ?) could comment:

In a composite structure, serious fire at one side of the sheet could
cause serious delamination and strength reduction before it burnt through,
due to the insulating properties of the composite materials. Aluminium
alloy, on the other hand, even though thinner, conducts heat very well
and has a high melting temperature.

Seems to me that, in the same way that an aluminium pan doesn't melt when
exposed to a gas flame, so long as it doesn't boil dry, aluminium should
be safer, especially at 30k feet, lots of air cooling the panel and air
temperature of say, -50 C ?...

Wasn't that a selling point of this aircraft to the passengers?

Yes, it was ONE OF selling points. It doesn't mean that higher humidity is mandatory.