Piston engine induction icing
 

In a random moment last night (ok, it wasn't totally random, I was reading up on carb icing) I started wondering why fuel-injected piston engines don't suffer from induction icing.

Air, as I understand it, enters an intake. It passes across a throttle butterfly, just as it would in a carburetted engine. With a low throttle setting, the butterfly will create a considerable venturi effect and cause a depression, and an associated drop in the temperature of the air. Why does the water vapour in the air not then come out of suspension and freeze, as it does in a carburettor?

I can think of two differences. First of all, the fuel is not mixed with the air until a considerable way after the butterfly in the case of the fuel injected system. The latent heat used in fuel vapourisation helps reduce the temperature of the air, and this temperature reduction wouldn't be present with fuel injection. The second difference is the absense of the carburettor venturi in the fuel injected system, the sole venturi effect being provided by the throttle butterfly. But I thought that both of these were relatively minor effects?

I must be missing something somewhere. Can anyone fill in the gaps and tell me what's wrong with my argument?

Thanks,

FFF
---------------

My understanding is that the injectors are positioned downstream of the throttle butterfly valve in the induction manifold.

So as you have said, there is no venturi effect on the fuel.

I also think that because the fuel is atomised by the injectors there is less of a temperature drop as compared to the vapourisation in the carburettor.

I'm sure an engineering type of person will know more about it.

Also I have a related question. Why isn't the fuel injected directly into the piston as in car engines?

FIS

You'll have a pressure drop over the butterfly valve but the pressure (and temperature) will still be lower in the venturi in the carburetor. If it wasn't, we could just drop the venturi... :)

Cheers,
Fred

Fred,

At high throttle settings, I agree completely - the reason the carb has a venturi is specifically to reduce the pressure.

But, at low throttle settings, the venturi in the carb is not significant enough to cause a large pressure drop (at least, not in a constant-venturi carb as used on most aeroplanes). That is why this type of carb is fitted with additional jets in the vicinity of the butterfly - because, at low throttle settings, the pressure drop caused by the butterfly will be far more significant than the pressure drop through the throat of the carb. And we all know that carb ice is predominantly a problem at low power settings - hence my assertion that the carb's venturi is not significant in causing carb ice.

Am I wrong? If so, why?

FFF
--------------

Fuel Injected engines can suffer induction system icing, that is why they all have provisions for either automatic or pilot selected alternate induction air.

Yes. It's not that they're immune from icing, it's that they're less prone to it.

They benefit from:

* no venturi
* no fuel evaporation at/near the venturi
* no fuel evaporation from an idle jet at the throttle butterfly restriction (I hesitate to call it a venturi).

All these things add up to reduce the likelyhood.

I think that FFF's question is resolved by the 3rd item I listed. Whilst both carby & injected intakes have a butterfly valve causing a restriction, only the carby design also has fuel evaporation in the restriction.

Field In Sight,

It is actually a very rare car indeed that, at the moment, injects fuel directly into the cylinders, the Mitsubishi Gdi engine being one example. Car engines started out with single point injection (injected before the manifold splits to the cylinders) and then moved onto 1 injector per cylinder, but this is still generally done on the manifold.

Diesels on the other hand tend to inject directly into the combustion chamber.

The real mystery is why there are so many carburetted engines still about, surely a kit to "injectorise" them would sell?

FFF

BPFE is absolutely correct. And to answer your original question, Fuel injected pistons DO suffer from induction icing. It does not happen as readilly as carb icing but is there all the same. There have so far been less injected aircraft flying around historically than normally aspirated aircraft which is why we may have not seen or heard so much about it.

Icing may occur around the air intake before it occurs around the butterfly valve. The air filter clogs up with ice and blocks the air intake. I have seen several light single pistons land with ice on the filter and the pilot was unaware of it.

If you get a reduction in RPM or performance then select alternate air. It is there to select air from a different source to the air filter and from a warmer source.

Any one who thinks injected engines do not suffer from induction icing is heading for a big fall.

Stay sharp and always expect the unexpected!

MM

Rule of thumb for injected engines. Select alternate air when flying in visible moisture at OAT's below +10 C. I seem to remember that from the Beech Queen Air :yuk: POH.

PIGBOAT

I believe most Queen Air's had Lycoming GSO 480 engines with pressure carburators.

Fuel-injected have a venturi?
 

My understanding (which seems to differ from every other post here - clarification please?) is that a fuel-injection system still has a venturi, but instead of the main jet in the middle of it, the fuel control unit is coupled to it to sense the pressure differential and meter the correct amount of fuel accordingly.

Is this not the case?

And therefoere of the 3 types of icing - impact, fuel and throttle ice, a fuel-injection system is free only from fuel ice (sometimes called vapourisation ice).

So given the 'right' conditions the venturi created by the throttle butterfly can still cause icing, however without the fuel to absorb some of that latent heat it is far less likely.

Thoughts?

BPF, could be. We had one of the first swept tail examples, I can't remember what it had on the wings, but I thought they were injected. Unlike the Continental IO540, they were the absolute sh!ts to start when hot. One of the boys crashed that one landing one night in a snowstorm. The company then bought a couple more, but had them re-engined with 8 cylinder, 400 hp engines which made a honest machine out of it but by that time I'd moved on. It was the only airplane I ever really disliked.

The fuel injected Queen Airs came in two versions.... Lycoming IGSO480's (model 65) and IGSO540's (model 70, 80, 88), both with Simmonds fuel injection.
A superb aircraft!:ok:

NOT hard to start....if you know how:E

I didn't find it quite as superb as the Lockheed 10A it had replaced.

Your are right 411A. I was thinking of the Twin Bonanza. BTW what was the difference between Simmonds fuel injection and Continental or Bendix systems?

The Continental system was designed for the proletariat. The Simmonds system was obviously intended for the bourgeoisie. :E

The Simmonds system (on the QueenAir anyway) does not have mixture controls. Has a switch, fuel on/off.

The mixture is automatically regulated, takeoff, climb, cruise, etc.

It takes a specialist to set this system up correctly, but then is usually trouble-free.

Thanks, everyone - seems that I wasn't quite as far off the mark as I thought I was ;)

ROB-x38 poses a question about how the injection unit measures the rate of airflow, and whether this affects the situation. I don't know the answer to this - does anyone else?

FFF
-----------

FFF

There is a small venturi by the throttle butterfly which adjusts fuel flow automatically. This could theoretically get blocked if there is ice around the butterfly valve. I have never heard of this being a problem in practice though. Appropriate use of Alternate Air will make sure it is not a problem.

It is difficult to explain the fuel metering system in this forum, I cant really explain anything to do with flying without a white board and 4 colour pens! The best way is to make your friendly QFI a coffee and get him to explain it.

MM

NTSB Identification: SEA04FA031
14 CFR Part 91: General Aviation
Accident occurred Thursday, January 01, 2004 in Cave Junction, OR
Aircraft: Piper PA-44-180, registration: N53505
Injuries: 1 Fatal, 3 Serious.

This is preliminary information, subject to change, and may contain errors. Any errors in this report will be corrected when the final report has been completed.

On January 1, 2004, approximately 1835 Pacific standard time, a Piper PA-44-180, N53505, registered to a limited liability corporation, being operated by Auburn Flight Service, Inc., and being flown by two commercial pilots, accompanied by an airline transport rated passenger and a non-rated passenger, sustained substantial damage during an in-flight collision with trees/terrain while making an emergency landing at the Illinois Valley airport, Cave Junction, Oregon (3S4). The pilot-in-command and both passengers suffered serious injuries and the copilot was fatally injured. Instrument meteorological conditions existed at the accident site and an IFR flight plan had been filed and activated. The aircraft departed Oakland International airport, Oakland, California, at 1612 on the afternoon of the accident and was destined for North Bend, Oregon. The flight, which was personal, was operated under 14 CFR 91.

Survivors reported that the aircraft encountered induction icing conditions and Seattle Air Route Traffic Control received a request for a diversion into 3S4. Power was lost and the pilots conducted an emergency descent visually acquiring the 5,200-foot long north-south asphalt runway. The airport caretaker reported hearing radio transmissions from the aircraft as well as the engine(s) cutting in and out, and stepped outside visually observing the aircraft over fly the airport.

The aircraft struck a tree and came to rest inverted on a berm alongside the west edge of Highway 199 and partially within Rough and Ready Creek which was under a flood warning at the time of the accident. The accident site was approximately 1,000 feet southwest of the threshold of runway 36. On site law enforcement personnel reported a strong odor of aviation fuel at the site and heavy snowfall immediately following the accident. There was no post crash fire






:ouch: I guess it does happen...