canard down by fire

[Cozy Girrrl]

Dear Jack,

In the interest that your tragic loss does not pass without possibly helping to prevent loss and injury to others can you give us any information on your control system and what you think may have given way when it failed?

What were your firewall materials?

Also from your description it was not until you opened the canopy and added fresh oxygen to the cockpit did the fire advance forward, this is contrary to what Shirl said about his fire with Jeanna Yeager on board at Jackpot he was talking of the virtues of a forward hinged canopy in such a condition and I am reading what he said and dissagreeing, thinking that keeping it closed is the only thing that kept you from becoming a serious burn victum.

No oxygen, no fire, but a few moments later you may not have been able to remain concious either. Can't put a price on your health and safety; you made all the right instinctive decisions and good fortune prevailed.

Don't know who you pray to but you owe big time. So glad you are still with us.

Regards, Chrissi & Randi

[Jack Morrison]

Dear Chrissi and Randi

 

This will be a long post but necessary. My E Racer was originally a mark I, with the engine in the center behind the pilot in front of the firewall. My opinion a bad design and am suprised it was ever approved. The firewall in this configuration was a standard 1/4 in 5 ply birch plywood. A .040 fiberfrax layer secured with the high temp red silicon, and a .032 6061 t6 aluminum cover over that. The mark1 main spar had a deep v in the center which allowed room for a drive shaft to run through the spar and connect to the reduction drive unit. When I changed from the mark 1 to the mark 11 configuration, it was necessary to widen the firewall to accomidate the 540 lyc. engine. I removed the fiberfrax and aluminum cover, resined the second 1/4 birch plywood firewall over the original firewall, added the fiberfrax and aluminum cover. I also left that driveshaft hole in the second firewall, a perfect place to run all the wireing harness's for the electronic fuel injection system. BIG MISTAKE ON MY PART. I'll explain later. Today I and a good friend with a cozy 4 took a good close look at the firewall in this configuration and actually was amazed at how well it survived the tremendous fire and temp. When I installed the fiberfrax on the firewall to secure the fiberfrax, I only smeared it in globs to hold it on. After the fire, all the plywood under the high temp blobs showed no burnt wood. Half of the aluminum cover melted away and the fiberfrax turned brown and crumbled where the aluminum was gone. The fire still did not penetrate the plywood, burnt the heck out of it though. (Note #1) When you glue your fiberfrax (get the thicker fiberfrax) to the firewall, spred the silicon on the ENTIRE wood firewall. (Note #2) Although I believe the aluminum firewall is sufficent, I would rather see everyone useing a SS material, just a little more protection. The fire started in the engine compartment somewhere on the main fuel line to the two fuel rails. Because of the extreme heat it could not be determined the exact failure because most components were melted. Back to my big mistake. After the fire started in the engine compartment and the engine went silent, the flames blasted through the driveshaft hole, even though I had an aluminum cover plate in the engine compartment side siliconed with hi temp silicon. There was a rubber gromet about an 1 1/2 in size to secure the wireing harneses to the aluminum plate. I estimate a few seconds after the engine went silent, that 1 1/2 in hole was like a blow torch blasting flames and smoke into the cock-pit. Inside the cock-pit and mounted right above that hole was the high pressure fuel regulater mounted on an aluminum bracket and the fuel return lines to the main tanks. The bracket and return lines melted in a very short time and fuel ran in the cock-pit from both fuel tanks. This is why I was running out of time for survival very quickly. The inside of the original firewall was burnt as bad as the second one in the engine compartment.

Aileron control loss was just before touch down. Most likely my senerio would not happen in a cozy 4 AC because of the location of the bellcranks. Both my right and left bellcranks failed just before I was about 15 ft agl and pulled a hard left aileron for runway alignment, no response. Had they melted 20 seconds sooner, I would have never been able to make the turn to final and you would not be reading this long message, and yes I did have a great co-pilot and thanked him accordingly.

I new I could not open the canopy until the last second even though I could barely breath. The cock-pit is a low pressure area and as soon as you introduce oxygen into the flamses, whosh, and that is exactly what happened. The head rest and headset deflected and protected most of my head when the flames shot forward. You are correct Chrissi. Seal any hole through the firewall as tight as possible, you cannot be careful enough.

 

Jack

[Cozy Girrrl]

Thanks Jack, I want to thank you for revisiting this and it must be difficult.

So I guess the E-Racer does not have the aluminum bellcranks out inside the wing roots?

Does it have the aluminum or stainless wing root covers?

Were all the pushrods behind the firewall steel?

Were there any aluminum fittings or aluminum fuel lines aft of the firewall?

You say the failure was probably one of the feed lines to the fuel rails, what were the feed line materials?

Were they firesleeved?

What is the pressure of the pump(s) for the feed lines?

 

---Fire is my greatest concern---

 

Great observation about the silicone RTV. At this point we may go:

-plywood firewall

-silicone RTV

-stainless steel

-intumescent paint

..yeah I know; belt and suspenders.

 

Thanks so much for sharing with us all.

Regards, Chrissi & Randi

[brainfart]

Ever considered using self-extinguishing epoxy in that area?

I know of several suppliers which sell brominated laminating resins to the general public. They aren't prohibitively expensive either. Such resins are being used extensively in commercial airliners and trains. Might be worth using around the firewall.

 

Quote:

"Low-viscosity, modified, bromine-containing epoxy resin, ideal for manufacturing self-extinguishing fibre composite components exhibiting good mechanical properties.

The burning behaviour of these components complies with the requirements und FAR 25.853, even for small wall thicknesses below 1mm."

 

Current prices:

2.5kg resin €36 + VAT

5.0kg resin €69 + VAT

10kg resin €115 + VAT

[brainfart]

FAR 25.853 requirements (burn tests, smoke density) briefly explained:

 

http://www.kydex.com/briefs/131.pdf

[Cozy Girrrl]

Thanks for the information on the self extinguishing resins. Too late for us but good to add to the knowledge base.

...Chrissi

[chasingmars]

Jack, condolences on the loss of your aircraft, I too am happy to hear you're ok. I followed what you did with the blended winglets closely and with interest, and it's a great shame such a unique aircraft is no longer flying.

 

Regarding brominated resins, I came across this interesting link that talks about some of the restrictions coming in force on brominated additives (it looks like the epoxy systems aren't the focus of the toxicity concern from my brief read of it.

http://www.newmoa.org/prevention/chemicalspolicy/Harriman.ppt

 

One thing that I wonder about is whether a brominated resin would still meet strength requirements, no sense in trading one rare failure mechanism for a more likely one, So I think doing a resin switch to brominated epoxy in the engine area would need a bit of an engineering look first to make sure the matrix performance still meets the requirements. Anyone have any familiarity with strength and elasticity of the brominated resins?

[Jack Morrison]

Chrissi and Randi:

1. The E racer plans control system was the same as the long ez and the cozy. There was an E racer retro fit with push pull cables from the stick torque tube bellcrank to the aileron bellcrank. I had this system in the AC for 325 hrs. I never liked it because there was 3/4 inch side slop in the aileron control caused from the backlash in the cable, tried several different cables with the same result, the auto pilot did not like either, would rock slightly side to side no matter what setting changes you would make to the pilot. I decided to go to the pushrod system but because of limited room, I had to mount the bellcranks just outside the engine mount on the firewall. This location is why they failed on my AC and this would not happen on a long or cozy.

2. The plans never called for a shield of any type but I did have a rolled piece of fiberfrax in the gap between the wing and spar and the spar was not compromized in the fire.

3. All pushrods behind and forward of the firewall were steel.

4. All fuel lines in the engine compartment were braided SS , teflon lined with steel fittings and firesleeved.

5. Here is where I believe the fire started. The fitting going thru the firewall was aluminum and appears to have failed on the engine compartment side. When the engine started to loose power in my climb, as though you pulled the power back, very smooth, I switched on the second fuel pump and the engine picked up about 200 rpm and again started to loose power until stopping. I believe the thru hull fitting failed when the engine first lost power and sprayed high pressure fuel on the exhaust and ignited the fire. By adding the second fuel pump, only sprayed more fuel on the fire. The smoke and fire inside the cockpit started almost immediatly. The fuel pressure on my fuel regulator is set at 45 lbs.

I was at the airport today and noted a few other items which should be taken into consideration.

I had top cowl naca inlets on the AC before I had side inlets. That made it necessary to cut the firewall with 2 inch by 11inch cutouts for the naca inlets to rest on. When I went to the side inlets, I filled the 2/11 cutouts with proper plywood and covered the engine side with aluminum, when I opened the canopy just before I stopped he AC and when the flames blew forward, those two pieces of plywood blew forward and I cannot find them in the burnt rubble. Point being, the one piece firewall withstood the heat and flames remarkably well but the seamed pieces of the firewall completely seperated.

I believe I am the only one that has a one piece top cowl on their AC. Although I really like this arrangement because of the access to the complete AC for service, I would not recommend it. The top cowl was made with 5 layers of cabon and it appears that from the extreme heat the cowl softened and raised off of the firewall which also helped in the fire coming forward. You need a solid seal between the cabin and the engine, like the cozy4 and long ez have, just in case of fire.

[Cozy Girrrl]

Dear Jack,

Thanks again for the details and clarification. So many of these things we learn only in the most unfortunate of circumstances.

Regards, Chrissi & Randi

[brainfart]

> One thing that I wonder about is whether a brominated resin would still meet strength requirements,

 

I don't know if they are inherently weaker, except that they have a higher specific gravity due to the heavy bromine and the inorganic additives they contain, which would affect some of the numbers [strength/mass]. They also have a higher viscosity which might lead to a greater resin content in the laminate. None of these are serious problems, just wanted to point out the little I know.

 

In mass transportation only flame retardant plastics are used for the interiors for safety reasons.

[Edge 513]

Jack,

Thank you for sorting through this here in the forum. Thanks for being here to BE ABLE to sort this through here on the forum. While my own wife was not seriously affected adversely to my explanation of your events and loss. She was concerned, and did ask if that event was easily possible in a MarkIV.

I am sure that has been talked about between other builders and partners. It is helpful to have the differences explained by yourself in black and white so that some feathers possibly might be de-ruffled.

On behalf of those significant others, thanks a million.

[Steve Innova]

This month's Kitplanes has an artical on Halon fire extinguishers, made by "SafeCraft". http://www.safecraft.com/.

 

A lot can be learned from how NASCAR protects its crew.

[Edge 513]

Without sounding like a nit-picker.

Its just more weight in the airplane. There are billions and billions of Rutan derivitive canard flight hours flown without fire incidents.

Granted, if it was YOU that had the incident, it could ruin your whole day. FWIW.

[WileEZ]

My thinking here is that brominated resins, while self-extinguishing, would not have done Jack one bit of good.

 

This fire was caused and fueled by gasoline, so even if the resin is self-extinguishing, it would still melt, puddle up and drip off.

 

I would pay more attention to what Jack said he would have done in hindsight.

 

Actually, I get the impression that if Jack had used brominated resins, he would be dead.

 

From http://www.compositesworld.com/ct/issues/2005/December/1108/2:

 

"Traditionally, fire-retardant additives in organic resins contain halogens. That is, the additive compounds contain either fluorine, chlorine, bromine, iodine or astatine. Brominated resins, which are the most common, are strong oxidizers. When exposed to heat, the weak bonds between the bromine and the rest of the resin's atomic structure are replaced with covalent bonds, which means the resin cannot be further reduced by flame. While it is this reaction that gives brominated resins their fire retardant properties, the bromine that is removed by this reaction then reacts with hydrogen to form hydrogen bromide (HBr). If the brominated resin is exposed to more and/or hotter flames, HBr density can increase. The resulting smoke is considerably more toxic than smoke from nonhalogenated resins. In mass transit applications, therefore, where smoke toxicity is of paramount concern, halogen use is declining."

 

There's more at the above site.

 

I suspect, based on a quick Google search, that brominated resins are being phased out for most use. There's a lot of information about "non-brominated" resins which should give us a clue not to use it.

 

Thank you Jack, for sharing this with us and I'm very glad you are still among us!

[Steve Innova]

Without sounding like a nit-picker.

Its just more weight in the airplane. There are billions and billions of Rutan derivitive canard flight hours flown without fire incidents.

Granted, if it was YOU that had the incident, it could ruin your whole day. FWIW.

So are seatbelts, unless you plan on crashing or flying inverted.

 

There have not been all that many fatal canard crashes, but I can think of two recent ones where the occupant(s) burned to death after otherwise survivable crashes.

 

Many of those Halon systems weigh less than 5 lbs. Stay off the bacon for a couple weeks and you'll buy that weight back and reduce your chances of a really horrible death.

[Big Steve]

Steve I. Said:

 

 

So are seatbelts, unless you plan on crashing or flying inverted.

You will be really greatful for those shoulder harness's with out crashing or flying upside down. I flew across the badlands of Colorado in the summer time once. The turbulence was so bad that without the seat belts I would have been in the canopy continually.

 

Fire extinguishers are not the answer here, better through points in the firewall and a different way of getting fuel to the engine I have never been a fan of hard lines with fuel in them. They crack and leak unsuspected.

That said I will probably use the plans method of getting fuel to the gascolator then braided lines wherever fuel connects to that rattly Lycoming motor. Build on STeve

[Edge 513]

So are seatbelts, unless you plan on crashing or flying inverted.

 

There have not been all that many fatal canard crashes, but I can think of two recent ones where the occupant(s) burned to death after otherwise survivable crashes.

 

Many of those Halon systems weigh less than 5 lbs. Stay off the bacon for a couple weeks and you'll buy that weight back and reduce your chances of a really horrible death.

Okayyyy...so you have a fuel line fire in the rear real estate section...your halo system shoots its shot....and the fire goes out. the fuel line keeps spraying the cyl head and the halo system is exhausted...

wha, am I missing here during a reoccurring event? If the engine bay was in front one could shut down the fuel line...but alas, it isn't, its where you can't see it. You're buggered because odds are you are not scanning your panel** to see the nifty warning light alerting you that the halon system has gone off. [presuming one would wire it that way- it would make sense]...so unhappily you motate on down the mile high skyway trailing smoke and fire because the firewall is doing its effective thing, being a firewall, until what...

 

**heck many guys forget to put their gear down on final- so whats up with the discipline of continually checking to see if that little "Halon expended" light is EVER lit up?

So thats/this is my point..and I usually cover it with my hat.

[chasingmars]

Lets make a separate thread for the brominated stuff if there's any continued interest in it perhaps? I'm not trying to be the post police or anything, but the on-topic discussion (and the lessons that might be learned from the unfortunate experience gained) is I think important enough to be kept so?

 

Didn't mean to veer off topic by ending my previous post with a question. My bad.

[Lifessamsara]

Fair call Mars.

So back on topic........... let's look at what we know about the root cause so far. A fuel line was compromised leading to fuel under the cowls with hot exhaust... Further discussion about the best possible ways to avoid fatigue in the rigid fuel lines please? Good mounting, use of appropriate fuel line material, appropriate placement of flexible lines etc.

Next point; if there is a fire under the cowls, early positive detection methods. I've heard of fire detectors being devised of twin twisted pairs of wire (if there is a fire, the insulation is burned away and the 2 wire short making an alarm signal) and someone using the bi-metalic thermastat from a tumble drier to indicate elevated temp's (what is the margin of difference between a heat soaked engine and a fire?). An aerospace (aviation) fire wire system would be nice, but quite expensive I expect? This works by a fine metal tube having a fine metal wire running down the center insulated by a dielectric. Once a predetermined temp is reached, the dielectric melts and conducts, thus shorting and raising an alarm signal. This 'firewire' typically does a loop around the engine bay.

From Jack's accounts, the standard Cozy and LongEz firewalls are of a suitable design so that we would not encounter Jack's experience with fire in the cabin. Thicker fiberfrax an/or use of SS or titanium in place of alumiminium for the heat shield is a consideration.

Smoke is still a possible issue, so close consideration needs to be made to all points of entry through the firewall and how well they are sealed, and with what. Anyone know what the pressure delta is between the engine bay and the cabin in flight? Should consideration be given to a suitable overboard vent to produce the function of a smoke door (to evacuate smoke from the cabin)?

One last point, fire needs 3 elements (remember your fire training sessions?), oxygen, heat, and fuel. If all 3 are present even after dumping the fire bottle, there is a chance it will re-ignite. Any solution to extinguish a fire must effectively remove one of those 3 elements. There will still be high airflows through the cowls and even halon will be dispersed at that rate after a minute or so.

I'd say the initial focus should be on resolving the root cause, and then suitable fire detection (no false alarms due to non-fire elevated temp's alone). Nobody wants to shut down a functioning engine and take immediate 'fire onboard'' emergency procedures due to an over temp' engine bay.

We are blessed to still have Jack with us, and that he has been so clear minded and open to share his assessments. Let's make the most of this learning opportunity for the best of the whole Canard Community. Separate threads as required on specific topics?

 

Cheers,

 

Bruce.

[brainfart]

"Traditionally, fire-retardant additives in organic resins contain halogens. That is, the additive compounds contain either fluorine, chlorine, bromine, iodine or astatine. Brominated resins, which are the most common, are strong oxidizers. When exposed to heat, the weak bonds between the bromine and the rest of the resin's atomic structure are replaced with covalent bonds, which means the resin cannot be further reduced by flame.

The number of errors in that paragraph alone is enormous.

Astatine... yeah totally dude! And organic bromides are oxidisers?!?

 

BTW, halon is just another brominated gaseous alkane... and that means it will disappear quickly, in less than a second under the conditions found in an average airplane.

 

> There's a lot of information about "non-brominated" resins which should give us a clue not to use it.

 

Sure, I was merely giving an example from my supplier, trying to point out that self extinguishing resins exist. If you find other resins which fulfill the requirements, use them. Other flame retardants are known and in use. Add a bunch of hydrated alumina to the resin, or throw in some red phosphorous (yes, contrary to common sense that stuff ist is a flame retardant too).

I don't think that brominated resins are a big problem here when they are used in the engine compartment. Poisonous gasses aren't an issue as long as the fire is behind the firewall and everything is airtight, or you can somehow let air into the cockpit.

[WileEZ]

I started a new thread to cover brominated resins, etc at: http://www.canardzone.com/forum/showthread.php?p=24108#post24108.

 

Jack did give an excellent "after action" report, including listing what he, in hindsight, would have done differently. I appreciate Jack telling us about this in detail and I will pay attention to what he recommend doing differently from what he did originally.

 

If I was in Jack's place, I'm not sure what I would have done differently. But this incident did give me plenty to think about and while I have not come up with anything concrete as yet, I'm giving this a lot of thought on preventing fires and how to deal with a situation like this if it ever does happen.

 

This is a worthwhile subject to discuss.

[Jack Morrison]

OK, here is the final report on what caused the fire. I was finally able to get to the thru firewall fitting, removed it with the braided lines to the injectors. I discovered that the fitting I had installed was in deed the proper one and a steel fitting. The aluminum firewall had melted onto the fitting and it appeared the fitting had melted, not the case. aai was able to pry off the melted aluminum from the fitting. When I was removing the fitting and lines, the inside nut on the fuel line connected to the fitting on the cock-pit side of the firewall was a full turn + unscrewed. Also the fuel line on the firewall side was unscrewed the same amount. This fuel system was installed by me in 1998 and has survived many many fuel leak pressure inspections and has never failed in 10 years and 360 hrs of flight. What made these two nuts come loose on this fitting I can only speculate was caused by some type of vibration. I have worked in the auto industry all my life and used these same type fittings in many race vehicles and have never had or heard of a failure of this type. I'll post some photos of the fitting and lines as soon as I can.

 

Jack

[TMann]

Jack, were you running 100LL at the time of this incident?

 

T Mann

[Jack Morrison]

Tom:

Thats all I can use with the setup I have. I was talking to an engineer today as to what would cause the two hose nuts to come loose. He summized that the steel nut on the firewall side had come lose after all this time and as soon as the steel on steel nut came loose and started the fire in the engine compartment, the fitting heated up and the fuel line nut in the cockpit side which was aluminum, expanded and came loose also, spraying fuel into the cockpit. After the engine lost all power, it was less than 5 -6 seconds when the fire started in the cockpit , and not just a small flame.

 

Jack

[TMann]

Jack..... everytime you post the conditions you were operating under just seem worse.

You must be living right!