What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Early fire detection

Webb

Webb

Well Known Member
Sponsor
I’m going to put a temperature probe in the airstream that exits the cowl. Probably use the TIT input instead of a discrete input on the G3X since it is unused.

With all the discussion about firewall safety and fire/smoke, it makes more sense to me to early detect a large increase in air temp exiting the cowl and shut the gas off ASAP rather than discovering an established fire.

I’m looking for any tips on best place to install the probe if anyone has done this before.
 
I have placed my sensor on the engine mount.
It shows about 55 C or 131F on a normal flight.
If the fuel hoses are insulated and the exhaust system is looked after, there is
not much to worry about.
An early warning is nice to have.

Good luck
 
My understanding is that in flight the heat/fire is (initially) blown down out the cowl outlet so In my 9A I drilled a hole in the floor just aft of the firewall/cowl outlet for the probe.

I also installed a cheap go cart/motorcycle CHT probe in the same location so with 2 gauges I can confirm there is actually a fire and not just a faulty probe.

Fin. 9A
 
fire sensors are very common in the marine world. Can probably find an off the shelf sensor or solution.
 
I found these. Inexpensive type K.

I like the idea of a pair to verify there really is a problem. I assume that since it is set up for a single engine, then only TIT-1 is available and not TIT-2. However, a free discrete would prove the second input.
 

Attachments

  • 1CA042B9-48E6-4625-AF04-2C50D490C0E8.jpg
    1CA042B9-48E6-4625-AF04-2C50D490C0E8.jpg
    176.4 KB · Views: 146
At one time I had a home brew system with two fusible link type sensors wired in parallel at the cowling outlet. When/if both thin solder links mounted on non conductive blocks melted, it activated a "red" light in the cockpit. (Mine was mounted to the engine mount in the lower sections of the engine compartment.) I removed it a few years ago decluttering the engine area. Pros, it's good to know if there's a fire. Cons, will it really work and what would result from a false indication. A commercial system should reduce the reliability factor. There must be more robust systems for cars, boats. But the thermocouples feeding an existing EFIS input sounds pretty simple.
 
Last edited:
Actual 6 fire during run up

You can do a search for a detailed more extensive write up since it’s been a long time ago, but I helped a fellow who broke a primer line during runup and had a blazing fire under the cabin. It melted his shoes to the cabin floor and quickly burned up every hose and wire fwf

The fire was put out within minutes
 
False Indication

I'd be afraid of the risk of a false indication. Fire light goes off, and we react according to that. Now we shut off the fuel and land on a highway, swamp, ditch in a lake or other dangerous maneuver. Turns out it was a false indication and all was good. I have personally had false fire indications and have seen several others happen to co-workers. These were photo cells in a King Air and I'm not sure how the system worked on the other twin turboprop that coworkers were flying.

False indications aren't such a big deal on a turbine twin, the other engine is more than adequate to permit a safe landing. On a single, we turn into a glider with few options.

Some posts mention two sensors to rule out a false indication. This is where I start to feel more comfortable. Engine fire procedures are dangerous in a single, although not as dangerous as a real engine fire. I don't want to see anybody conduct an engine fire procedure for a false indication.
 
I will be installing 2 sensors for the reasons others have pointed out.

I found out earlier today that both turbine inlet temperatures can be used even if there is just one engine which leaves the unused discretes available for later use.

Now I just have to come up with a different name instead of turbine inlet temperature.
 
Perhaps parallel one type of sensor with another. I'm thinking a robust but fusible thread stretched across the cowl exit airflow that holds open a plunger type of normally closed switch. Thread burns through, switch closes, warning light goes on. Cross-check that with the aforementioned spare channel CHT probe. If both are indicating a fire, it's time to run the procedure. If not, then check and fix issue before next flight. Thread would be a scheduled replacement item due to possible deterioration from environmental conditions in the lower cowl - hot, oily, vibration while under constant spring tension...

Could even go triply-redundant, but I'd want two different species of indicator at least.

If this seems silly, I'm only one cuppa Joe in so far this morning...
 
I have hours of video, realistic firewall samples from point of ignition, with a timer and thermocouples. There is merit in fire detection, but with two caveats.

The time advantage will vary a lot. In the worst combinations, it will be almost non-existent...open flame in the cabin in less than 10 seconds. I'm not kidding. You won't need a fire indication system.

In the best configurations, a fire detection alert will be valuable, but only if you believe it, and act on it. Don't sit there wondering if it's true, or how bad. Forget about punching the GPS to find the nearest airport. The immediate response is "fuel valve off" and emulate a rock all the way to the flare, because even with the best configuration, insulation, and warning, the clock is ticking.
 
Afterthought

Playing with the new boroscope I just got (I bet Amazon's AI is scratching its silicon head about the sudden rush of orders for the 5mm models with side cameras), I found myself wondering if a remote video feed from inside the lower cowl would assist in interpreting a fire warning indication.

It would have to be robust enough to allow a peek for a few seconds before the whole thing melted, I guess. "TV snow" on the video monitor would be an additional Very Bad Sign, if your other sensors were blaring the FIRE klaxon at you.
 
The sensors were installed and the G3X displays. i think the sensors are rated to 600f. If one shows hot and the other doesn’t, I’ll ignore but if both show hot, I’ve got a problem. For the curious, I used the TIT discretes.

My way of thinking is an alarm from the G3X will buy me just enough time to turn off the fuel so I can fly a glider instead of a fireball to the ground.
 

Attachments

  • IMG_1565.png
    IMG_1565.png
    12.8 MB · Views: 38
Last edited:
I think I remember the Cozy/LongEZ crowd talking about fire sensors. Imagine a fire in a pusher... a long time will go by before you see or smell smoke.
Worth doing some searching over that way.
 
Would it be possible to use a link made from low melting point metal, and detect when the low voltage circuit is broken by the link having been melted? It seems like this would be pretty foolproof - if it gets hot enough to break the link you know for sure something is wrong. You could maybe mount it on the firewall where wind would push heat towards it?
 
inktomi said:
Would it be possible to use a link made from low melting point metal, and detect when the low voltage circuit is broken by the link having been melted? It seems like this would be pretty foolproof - if it gets hot enough to break the link you know for sure something is wrong. You could maybe mount it on the firewall where wind would push heat towards it?
Click to expand...
You are speaking of a thermal fuse. I used two 350 deg F fuses in parallel on a terminal board mounted on my firewall just behind the fuel pump. When both melt (redundancy) the ground is removed and my Dynon D180 displays "FIRE".
 
Thermal fuse! That's the name. It looks like 350F is very common for clothing dryers. If you wanted something a little smaller, Newark carries this which breaks 360F

https://www.newark.com/thermodisc/g5a01184c/fuse-thermal-184-c-20a-250vac/dp/39T4439 $2.71 each isn't bad at all.

It would be super easy to solder two of these in parallel on a little breadboard and run some leads out from there and then mount the whole thing anywhere.

39T4439-40.jpeg
 
A major issue with the Lycoming powered low wing aircraft is the engine driven fuel pump. Rocket fatal years ago in CA was caused by that. MAY have been a factor in the McSpadden accident. The commonly used fittings require install procedures that many do not understand.
Video of a Beech Duke on fire and then exploding. New fuel flow installation with line that had never been tightened.
Turbocharged Twin Cessnas have what I call "make believe" firewalls. The issue is that a portion of the exhaust goes thru the firewall with questionable protection.
Widely accepted belief that if you have an exhaust failure in a critical area the wing will burn off in 90 seconds. Glad my Twin Cessna days are behind me. Nice flying airplanes but I hated them.
 
inktomi said:
Thermal fuse! That's the name. It looks like 350F is very common for clothing dryers. If you wanted something a little smaller, Newark carries this which breaks 360F

https://www.newark.com/thermodisc/g5a01184c/fuse-thermal-184-c-20a-250vac/dp/39T4439 $2.71 each isn't bad at all.

It would be super easy to solder two of these in parallel on a little breadboard and run some leads out from there and then mount the whole thing anywhere.
Click to expand...
I like your idea, thanks for posting! A few of these will definitely be on my next component order so I can experiment with them! (y)
 
Let us know what you come up with! I've been toying with an idea of also putting a couple LEDs in series with the parallel fuses along with a TEST button that we could mount remote. It would let us preflight the sensor, push and hold TEST and see that both LEDs light up.

Or, even better, maybe the TEST button would break the circuit so that the display would show FIRE as if both had been blown. You could preflight by ensuring that FIRE is not shown (fuses working), TEST to ensure that FIRE is shown, and that FIRE goes away when TEST is no longer pressed.
 
Last edited:
I gave thought to thermal fuses but by the time they pop, you are well on the way to 400F. They are also all or none which is passive and is why I choose the thermocouple route that is actively monitored and will alarm when the threshold temp is reached.

Also once they pop, you have no longer know what the temp is doing.
 
You must log in or register to reply here.
Back
Top