Ethiopean 787 fire at Heathrow

[amicus]

syseng68k,
Thank you for your kind request and I trust that I can make a sensible and accurate response to your good self regarding conduction and insulative properties of both aluminum alloys and CFRP, whilst simultaneously attempting to correct some misunderstanding by others and make some suitable, accurate and helpful remarks regarding relevant differentials of Coefficients of Thermal Expansion (CTE) and other composite properties.
First, please let me address those who claim that CFRP quote "chars'" and that it thus self extinguishing. Sadly this is not the case for epoxies, I again repeat to those that will hear, the fact that all epoxies have been banned from aircraft interiors for around the last 30 ears and for excellent reasons, and it was not merely the flammability and combustability of such epoxies, but primarily the Fire, Smoke and Toxicity (FST) emisions that occur.
And, yes, a char develops. but only as a result of the epoxy having first burned with deadly initial combustion, flaming, smoking and emitting copious FST hazardous products in the process. Incidentally, that is what we used to employ way back when to make carbon/carbon, but that process was a controlled, pressurised and in an enclosed chamber. We used phenolics, primarily, as they have a greater carbon yield than evil epoxies. However, even that process , required over seven re-denisification and burn cycles to come close to the desired theoretical density of carbon/carbon.
The epoxy itself,and I have both performed or witnessed many 100's or 1000's of OSU ( Ohio State University) tests as developed back in the 60's re testing flammability of epoxy interiors, seat cushions, fittings,overhead bins and the like which became a semi-standard interior burn test over the decades to prove this point. However, and it's a very large however, the OSU test was designed around a mere burning cigarette or lighted pipe scenario, with a 50 watts level, far from 28V electrical fires and shorts and fuel fed fires in a survivable crash which requires 200-250 watts per square metre and their ilk. This point is cited and detailed in my paper as some have read on this board.
But the charring of epoxies does not, repeat NOT, prevent burn-through which is why the FAA and Boeing developed a whole new set of high wattage burners to try and adequately simulate and test for burning fuel fires and resulted in the FAA, over Boeing's weeping, wailing and lobbying to stipulate that the lower half, and, most regrettably only the lower half, of the 787 fuselage be expensively and heavily insulated with a special anti-burn through insulation to allow a five minute escape window in case of a wheels-up or off runway survivable crash.
However, and here is yet another huge however, the FAA and Boeing still had to assume and design the burn-through test only for a totally intact, all doors closed, no slides deployed no doors opened and no fuselage fractures, given the FST loaded epoxies employed by Boeing. Clearly, in most survivable crashes, this is not the case and I supplied the FAA, Boeing and Airbus with over 150 such commercial airline survivable crashes since the 1980's as my paper cites. All were survivable crashes with fuselage fractures and/or opened doors. The simple and clear fact is that Boeing and the Northwest FAA Office did not have a prayer of certificating the 787 unless an intact and closed fuselage was assumed as a basis for fire testing in the realm of 200-250 watts as is the case in fuel fed fires. Specifically, EADS also presented totally independent, but similar findings re the A350 two or three years ago.
Now , allow me to wander back, please, to the char that some have cited. All such claims are nonsensical, tendentious and worthless twaddle, to put it mildly. A "CHARRED" CFRP is no longer CFRP, it is merely a residue of strands of free CF, either woven or UD, but it is no longer a composite structure capable of meeting any design loads or criteria at all. So 'char' is a silly diversionary tactic , pure and simple,with the emphasis on simple to my mind and has nothing whatsoever to do with any structural load carrying ability or any chance to sustain any pressurisation loads. I hope that I see no more silly inputs re"char"and would note that the non-structural and useless char would be blown away by the 500 plus mph airflow.
Next, please let me outline briefly the relative CTE's regarding Aluminum and CFRP and epoxies. Let me work in old fashioned degrees F units, please. just to humour me.
Typically an aerospace alloy will have a CTE of 12-13 x10 to minus 6 inches/inch /degree F, and epoxy will have (in its pristine state epoxy only state it will have 22-23 in same units). But not in a CFRP composite whose CTE is controlled by the CF with a modulus of 40 MSI vs epoxy at a mere MSI of less than 1.
MSI matters, just as loads follow stiffness (same degree F units,of course). This differential generates internal stresses in the composite, but they are usually of a second order concern, but some of we composite folks like lower cure temperatures to minimise such internal and residual autoclave induced curing stresses. For CF itself, as a unidirectional material, it has an expansion close to zero (typically 1 or less re CTE units quoted, which is why it is employed for space mirrors encountering wide night/day temperature swings. This is also why INVAR is typically used as Boeing and other aerospace companies as tooling material to minimise problems involving thermal distortions. Boeing and others typically employ a quasi-isotropic layup, with, I note, emphasis on the quasi, as there are some significant differences between a true isotropic material and a quasi-isotropic material. Such CFRP structures typically have CTE's in F units of around 3-3.5. I hope that this clarifies and settles the CTE issues previously discussed in this thread.
And, prior to getting to my final points regarding thermal insulative and conductivity properties of aluminum alloys and CFRP's as I promised that the outset of this overly long input, let me give some relevant metallic vs. CFRP cross plied properties for others to mull over. CF is widely touted, and oft over-touted, regarding its strength, but let us be very careful here. A decent CFRP will have a tensile composite strength, if a UD material at 60 % fiber volume, of around 280- 320 KSI,( note, this is the sigma 1-1 equivalent to metallics).
A quasi-isotropic composite in the same CFRP will have a tensile value of around 90-120 KSI. Now, let us look a couple of weaknesses without boring you all with the hygroscopic nature of that nasty epoxy.
The shear strength of metallics is typically 60% of the tensile, so for a decent steel, for example, say in the 240 ksi range we will have a shear strength of around 145 KSI. Now look at composites and up pops that nasty epoxy again. Whereas Steel will have a SBS or ILSS strength of 144 KSI, a quasi-isotropic composite will have an ILSS of 8-10 KSI static on a good day and, if we allow for fatigue et al, we are down in the 4 KSI area. And finally another key nasty is the short transverse tensile strength ( equivalent to metallic sigma 3-3), which is entirely epoxy dependent and no CF failure is involved, there we find on a good day around 3-4 KSI with fatigue knocking that down to around 1-1.5 KSI which is close enough to zero in this composite engineer's mind and no tensile Sigma3-3 asallowed by an decent and competent comosites engineer.
Achilles only had one heel , lucky for him, but for we in composite design and analysis,there are plenty of other heels to fret about. This is not an anti-composite rant of any ilk, but rather to emphasise why composites need to be analysed far differently and much more closely than most metallics Further, large scale repairs become much more demanding and possibly questionable in contrast to the blithe assertions of some posters on this board. And, as a final point,why has nobody in the composites community,after a full half a century of designing and using CFRP, has yet developed "A'" basis allowables for composites as is so common for metallics? Why and we are still operating with "B" basis allowables with significantly lower probabilityand confidence levels?
Now, belatedly back to the original question that I was requested to respond to and, perforce, my final zinger for today for those seeking to minimise the magnitude of the 787's inherent epoxy based FST et al difficulties.
Yes,CFRP is indeed an thermal insulator compared with aluminum alloys, but also exhibits significant differences depending upon fiber orientation and layup. This, in turn, leads fires, heat and the like to be concentrated in one local area rather than with aluminum's high thermal conductivity, which lowers and significantly lowers and dampens peak temperatures in event of fires from whatever the source.
I trust that this helps the ongoing discussion.
And now I promised you a zinger and here it is; for all proclaiming and advocating the efficacy and safety and non -flammable nature of CFRP, let me issue this very simple challenge.Go out and buy (or borrow from your missus or similar)some aluminum frying pan. Then please cook something requiring 600-650 degrees F on a gas stove. Next, just to humour me, reproduce that same test, but this time first build a CFRP frying pan of similar size and construction and cook the same food with 600-650 degrees F gas cooker. There is only one stipulation, as I do not want you or your loved ones or pets to suffer or die in such a foolish attempt, please do this test of your cooking skills on an outdoor barbicue well away from all humans and pets and let me know the results. In closing we all know, I would hope, why there are no CFRP cooking utensils on the market and,just as a clue,it is zero to do with cost and all the do with flammability and FST.
Any takers, 787 defenders and "it only chars and is better than aluminum" folks, please put your frying pan where your mouth is?
Cheers and apologise for such a long post, but I hope is deemed of some value.