View Full Version : Engine failure due carb icing. Will it restart?

18th Aug 2001, 18:28
I thought about putting this on Instructor Forum due accent on training type aircraft, but feel that it might be more appropriate in Tech Forum. Part of the trouble check after engine failure in light singles (and twins)is to select carb heat on, presumably in case carb ice had caused the failure. The purpose being to get warm air into the carb system to melt any ice.

If the engine has stopped (ie not just rough running), then the warm air from around the exhaust muff will quickly dissipate. Does it become useless in 10 seconds, 30, or 40? I don't know, do you?

Therefore, is it not a waste of time following the time honoured cockpit drill of selecting carb heat to melt the ice?

And how do you know if the ice is melted anyway, unless you try an engine restart via the primer and electrical starter. Assuming the ice that caused the failure had accumulated around the throttle butterfly, then if the throttle is closed (which is standard procedure following engine failure - or is it?), would any remaining residual heat from the engine only reach one side of the butterfly valve and maybe not de-ice the other side?

And for any carb heat to be immediately effective in melting ice prior to restart drills, should you actually open the throttle wide before carb heat is selected, in order to get max residual warm air flow right down the throat of the carby past the wide open butterfly valve? Finally, how long should you wait after selecting carb heat on for it to work and before attempting a re-start?

As I see it, it is all very well to carry out these rote drills following engine failure, but are we fooling ourselves and the students to whom we give sage advice?

More to the point, is it just another myth that carb heat should be selected on as part of the trouble check following engine stoppage due carb ice, when the heat dissipates so quickly once the engine stops?

Bally Heck
19th Aug 2001, 01:36
Hudson, unless you are very lucky, the carb heat is pretty useless for a restart once the engine has stopped. Both for the reason you state (cold muff) (oooer!) and because the airflow into the carb will be negligible.

Use the carb heat when it starts running rough or preferably before. If the engine stops, plan a forced landing and if you still have time, try a restart. Conducted heat from the engine block may have dislodged it after a minute or two.

Exercising the throttle may also help.

Opening the throttle wide before trying a restart wont help the engine due to providing an over weak mixture

I would still select carb heat hot even if the engine has stopped as you need all the help you can get and if it does restart it will melt the residual ice.

Best advice.....follow the flight manual.

19th Aug 2001, 20:11

As you realise, there is probably not much heat remaining but the reason for selecting carb heat in the drills may not be for reasons of heating as you assume.

On many aircraft, selecting carb heat allows an alternative intake path which allows air to bypass the intake filter and then go around the exhaust jacket for heating. If the reason for engine stoppage was because the filter was blocked by ice, then bypassing it may result in a successful start.

I recall some years ago, an RAF Bulldog had a birdstrike where the bird hit the intake filter and damaged it. The engine either ran very rough or failed. Selecting induction air to hot (no carb on a Bulldog, it's got fuel injection) allowed the engine to pick up and recover.

This does mean that if carb / intake heat is selected on whilst on the ground, unfiltered (gritty!) air may be ingested into the engine which won't do any good.

Hope this helps.


20th Aug 2001, 01:42
You should still get plenty of airflow through the carburettor even with icing, as the propeller is still windmilling (unless you're much slower than typical gliding speed).

Even thirty seconds of warm air should melt enough ice to allow rough running, I would imagine.

I agree that closing the throttle altogether might be counterproductive. I have started teaching my students that they should leave the throttle partly open until shutdown checks, mainly so they will be able to detect a restart if it happens! (However this is mainly applicable to practising forced landings - not enough time to fool around in an EFATO.)

O8 :)

20th Aug 2001, 03:23

I wouldn't agree that leaving the throttle partly open as the normal thing to do following engine failure. It may result in a perfectly judged forced landing ending up in the far hedge if the engine briefly picks up again.

Even worse, an attempt might be made to fly away from the only decent landing area, only to have the engine fail again with nowhere to go. A situation like this killed a very good mate of mine. He and his student ended up in the undershoot of a very small field, unfortunately they didn't get away with it.


20th Aug 2001, 05:13

Your comments are valid. Note however that I suggested the throttle be slightly open only until shutdown checks, presumably at or shortly after the low key (1000' agl) area.

Below 1000' agl or so, I quite agree that you are committed and should make sure the engine stays off until landing!

A partial engine failure is a different topic, taught separately.


20th Aug 2001, 18:03
Chances of restart not very good, I suspect. Think how long it takes an ice cube to melt in a room-temperature drink! So from a couple of thousand feet or so there will not be much likelihood of a re-start. Number one priority is 'fly the aircraft', finding a good field and executing a well planned forced landing pattern. NB that if propellor is stationary, rate of descent will be greater than for PFLs with throttle at idle.

Bally Heck
21st Aug 2001, 00:44
It's worth noting that the prop will almost certainly windmill even thought the engine is producing no power. Had this demonstrated to me while doing an FI rating. To stop the engine windmilling it is necessary to raise the nose and bring the airspeed close to the stall. Incidentally, I don't recommend trying it at home. I considered it a rather unwise demo.

A windmilling engine does have more chance of self restarting itself if the ice melts.

Not sure that I agree with Multp. I think a windmilling prop will cause considerably more drag than a stationary one. Open to discussion though.

21st Aug 2001, 01:56
Bally Heck,

If the engine is at idle, as Multpb stated, then the engine is self sustaining and AVGAS provides the energy to drive the engine and its ancillaries. As the prop is being driven round it will still provide some thrust and therefore an idling engine will give a reduced ROD over a stopped prop.

However, with a failed engine and a windmilling prop, the energy to drive the engine round comes from the potential energy of the aircraft i.e. its height. Airflow alone is driving round the engine, pumping its oil, driving its electrical generator, overcoming friction, compressing air etc. This will cause an increase in ROD over a stationary prop.

ShyT :)

cosmo kramer
21st Aug 2001, 03:29
I don't agree to Bally Heck's if the engine stops, plan a forced landing and if you still have time, try a restart (I do, however, agree if below 500 feet).

Most engine failures can be restarted. The most commen reason for engine failures is an empty fuel tank. So switching tanks should be the first priority. Then turn in a generally good direction but do not start looking for specific landing sites (it would be foolish to make a landing in a rough field, with possible risk of damage, because the time was prioritied looking for a landing site, rather that restarting the engine). After that perform the memorized emergency checks. All of the above shouldn't take more than 20-30 seconds and will most like get the engine started again

Then, if nothing works, plan a forced landing.

21st Aug 2001, 14:38
Back to RoD. Thanks, Shy-T.
PFL, engine at idle: some residual thrust, thus RoD less than if engine and prop stopped. Windmilling engine needs energy to turn it. OTTOMH, I think the drag would probably be greater than any thrust from the auto-rotating prop and thus RoD greater thna engine at idle.
Now my curiosity has been stirred, wonder if I can find anything in the books......

23rd Aug 2001, 06:09
On GA aircraft, the ram air entering the ducting and surrounding the exhaust will stay quite warm, even after engine shutdown, If the engine failed in flight, by selecting carb heat, then attempting a restart, the ice will almost certainly melt and allow a restart. The "ice" we are talking about is more of a frost that covers fuel discharge ports in a carburetor, and will readily dissipate when heat is applied. The only likely time after an engine failure the carb heat wouldn't work is if you were in known ice conditions.

23rd Aug 2001, 17:06
I've done windmilling, idling & prop stopped glides in in 150 & 152 aerobats.

Performance was worse for the windmilling condition, better for idling & best with the prop stopped.

I used to have the figures for this but now they're long since lost. :(

[ 23 August 2001: Message edited by: Tinstaafl ]

A and C
23rd Aug 2001, 18:02
I hope this may be of some help to you all on this subject because it makes me go cold when i think about it !.

first the tech details.....aircraft Robin DR400-180..engine Lycoming 0-360-A3A

The met situation.......FL90 temp +1 cloudbase 3000ft tops well above FL100 moderate rain.

I had been using carb heat about 60% of the time at FL100 and was having no trouble clearing the ice ,at top of drop about 27 NM south of the isle of man i was cleared to FL70 i selected carb heat and reduced the power about 250 RPM for a cruse descent after about 500ft it was clear that the ROD was much higher than normal for that power setting so i pushed the throttle all the way up and very little happend the EGT was indicating that the engine was making some power as the temp was well above the bottom stop ,the next thing i tried was setting the mixture full rich ....RPM and EGT droped instantly so that came back to the highest EGT setting that i could get this gave the lowest ROD.
It was now becoming clear that i could not maintain FL70 so a call to ATC got me clear to descend as far as i needed to with the other trafic getting moved clear i was all ready on a HDG that was going to get me ditching as near to land as i could get.

AS we passed about FL50 the engine started to pick up and by 4000Ft was running normaly,engineering investigation found no problems but i changed the carb as the butterfly valve shaft was showing evidence of wear.

The lessons from this ?

1 dont give up on trying to start the engine.
2 stay with the carb heat even if it seems to be having no efect.
3 Full rich on the mixture may be too rich at altitude with most of the carb blocked with ice.
4 Have the proper safety equipment for the flight , if i had ditched i had all the kit to give me the best chance of being alive and found IE life jackets ,liferaft,ELT,dye markers,pyros and the training to use them.

[ 23 August 2001: Message edited by: A and C ]

24th Aug 2001, 18:00
Interesting replies all round and thank you all for your inputs. I usually save Pprune info like these and hand them out to students for "required "reading.

I had a prop stop on me during stall recovery practice in a C150. Decided to aim for a ditching in a nearby river estuary due lack of suitable fields. Fortunately I managed to convince the student to lean back in his seat while I stretched across him to reach the primer pump situated near his left knee. Three hasty pumps and the engine started with first twist of the key. Cancelled Mayday and 7700 and flew home. Snagged aircraft in tech log for low idle rpm.

Occasionally one finds a primer pump that binds badly and is difficult to operate.
This particular Cessna had previously had one of those and I had written it up a few days earlier. Fortunately the techs fixed it, because if it had played up on me when the prop stopped, I guess we would have had wet feet.

28th Aug 2001, 06:03
Many years ago I was ferrying new C-150's from Wichita to Miami to build time, once I got caught in WX and was climbing to get through a sucker hole which promptly closed up on me. Suddenly the engine started sputtering and then died. Needless to say I headed for terra firma, the prop was still turning, throttle was at max and mixture was still where I had it during climb, the engine suddenly came back to life and continued to operate normally thereafter. It did not seem to be bothered by the smell in the cockpit!

Jaun Huw Nose
30th Aug 2001, 01:46
Apparently in days of old they flicked the mags on and of to get the engine to backfire and so clear the ice!? ;)