Carb Heat

[barbershopquartets]

I was once told by a fairly experienced Pilot / Engineer that Carb Heat should only ever be used to melt accumulated ice and never as a preventative measure to stop it's formation in the first place.

The theory being that under certain operating conditions, it is actually possible for the application of CH to riase the air temp into the range optimal for ice formation. At which point you're left with no possible way to clear it.

Have never flown anything with a Carb Heat Temp guage before - anyone out there who has and could confirm / refute this one?Sounded like bull**** to me but you never know.

PS Please no replies along the lines of "The correct answer is in you flight manual"......

[Paranoid Parrot]

The correct answer is not in your flight manual it is in the knowledge of your engine and your local atmospheric conditions. You are correct - there is no set answer.

[mazzy1026]

Well if you think about it, the only way to prevent the formation of ice, is to leave it on all the time so in theory, if you turn it on say, after 15 mins of flight, then you may be getting rid of tiny ice particles :confused:

[strafer]

Carb Heat should only ever be used to melt accumulated ice and never as a preventative measure to stop it's formation in the first place Keep up the PFLs bsq, you may be needing the experience sooner rather than later.

[Lister Noble]

I'm a new boy, but the instructors at our club really empasise how important carb heat is.
My feelin is that carb icing is the number one killer for light aircraft.
We are taught to put on the carb heat whenever the engine is running below cruise power and to always check from time to time in flight.
I suppose you could run with it on all the time but it would reduce engine power?
Also I think it should either be on or off,never in between.
Last weekend I did a touch and go and forgot to put the carb heat off before going round,but my instructor said OK so you've lost a bit of power but the engines still running!
Recently a locally owned Stearman was diving with idle engine and the engine stopped due to ice.He was lucky and managed to land on an old airstrip nearby without harm.
I wonder why a carb ice warning system isn't fitted to all light aircraft?
Lister:D

[DubTrub]

A very quick search finds this thread (http://www.pprune.org/forums/showthread.php?threadid=148655&highlight=carburettor+ice)

[Whirlybird]

The theory being that under certain operating conditions, it is actually possible for the application of CH to riase the air temp into the range optimal for ice formation. At which point you're left with no possible way to clear it.

Logically, it's probably possible. In which case, you apply carb heat for long enough to raise the temperature even more. And you apply it frequently, just to be sure.

This argument will go on and on, and has, on numerous threads. When all engines are either fuel injected or fitted with carb ice detectors, they can finally stop. Till then, I would suggest applying carb heat frequently, for long enough to melt ice if you have it OR if you've inadvertantly caused it to form by applying the heat in the first place. In misty, damp conditions, leave it on all/most of the time, especially if over mountains or water. Get to know your engine, and if the RPM is dropping, don't keep increasing it; ask why, as carb ice is the most likely answer. And don't listen to people who don't really know what they're talking about, how ever many hours or qualifications they've got; check to see if they're right first.

At least, all of the above is what I do.

[S-Works]

I Agree with Whirly here 100%. I have done the same as her for the odd thousand or so hours and it works for me and her and I bet many others on here!

[mad_jock]

And if you do turn it on and the engine starts burping and farting don't turn it off again. Which is what your instinct will tell you to do.

Pick a field and stay close to it and try and maintain as much height as possible without going IMC. Cover your bum and tell ATC if you like and run your normal engine drills as if it was a PFL. Then wait. After a while it will stop burping and farting and you can continue. If you have told ATC remember to tell them that you are OK now. Then continue on your way being a bit more careful how often you use it. If it doesn't sort itself out put it in the field and wait for the police to turn up.


MJ

[Dop]

Don't Chipmunks have the carb heat permanently on?

I understand one of the reasons for not having it on all the time is the power reduction caused by the air being less dense when it's hot. Then the other reason is that the carb heat intake isn't filtered, so on or near the ground there's the chance of airborne dirt getting into the engine.

[IO540]

I vaguely recall reading an explanation (based on physics of water vapour/liquid/ice; the old "steam tables") that merely warming the airflow (e.g. with partial carb heat) is highly unlikely to create ice, in a temp/DP scenario where ice would not have been created in the absence of carb heat.

I don't have the reference for this however. Bookworm, where are you?

If correct, this would explain why planes which have their air intake warmed up slightly by the exhaust (various PA28s, I think) don't get carb icing.

Anyway, I think carburettors should be banned :O

[gasax]

Chipmunks (well ex military Chipmunks) do have their warm air arrangements wire locked on.

You'll note warm air - not hot air - the Gypsy draws air off the cylinders and so it is nothin glike as warm as coming from an exhaust muff. Warm air is recommended on all Gypsys in dusty conditions as the incoming air takes a 90 degree bend and so a lot of dust is removed from the incoming air when in 'warm air'.

I cannot remember the exact specification but hot air systems are supposed to be capable of at least a 30 deg C temperature increase. So to cause icing to occur when hot air is selected it is going to have to be bl**dy cold for icing to occur hot air is selected. So the next time you're flying on a crystal clear -20 Deg C day you might need to be careful with carb heat. All the rest of the time it is a theoretical blind alley.

[strafer]

IO540 - I can assure you that PA28s can get carb icing, I've had it with an OAT of 25C! I think as a general spamcan rule, Lycomings are less prone to ice than Continentals, but as a useful rule of thumb, if you've got a carb heat lever in the plane, your carb can ice up.

Lister Noble - carb icing is not the number one pilot killer, that's pilots doing stupid things - like not knowing exactly how and when the plane they are flying can generate carb ice. Also being 'fairly experienced' at anything does not give you an immunity from spouting bullsh1t. Look at Sven Goran.

[Chippik]

The carb heat isnt wired on in all chipmunks (like ours).We select the carb heat on the climb-out and deselect when vacated the runway (and a check during the 1800 power check pre takeoff)

[dublinpilot]

I've heard this one before too. And likewise it was from an engineer. But I don't believe it.

Lets say you're nice and high on a cold day, and the outside air temp is -20C. You use Carb heat and this brings the air temp up 30C to +10.

Then surely there is such a large gap between the +10C and the dew point of the air, that there is no possibility of Carb Ice forming?

At -20C I can't imagine the air holding very much water. Just because you raise it's temperature to 10C very quickly, isn't going to make much difference. Where is the water going to come from?

Of course this is just my amateur physics, and but I can't see the logic.

dp

[robin]

If you wait for ice to form on a VW or Continental engine there is a possibility that by the time you have recognised it, the engine won't be able to generate enough heat to melt the ice - then you are really stuffed.

I've had the engine stop on the ground when taxying. As a result I tend to use it in a precautionary way before it is needed, but always to leave it on for some time, not a quick blast

[B Fraser]

Manual carb heat control always amazes me. Car manufacturers solved this problem in the 1970s before almost universally ditching this technology in favour of fuel injection. I don't recall ever having a fault develop on a car where the induction heating system failed and the engine developed a problem as a result.

In some aspects, aviation is still in the dark ages and it's a wonder that we have managed to progress to electric starters.

Perhaps that's part of the charm :confused:

[A and C]

The ice formation in a carb is not only down to the atmospheric conditions I have noticed that the Robin DR400 aircraft are much more prone to carb ice than Pipers with the same engine/carb the only differance is the intake/filter.

When shedding ice it may take some time to shift and if you are at altitude with the mixture lean and the engine starts to loose power or run rough then persist with the carb heat and dont what ever you do put the mixture to rich This will only send the mixture over rich reducing not only the power from the engine but also reducing the heat avalable to the carb heat heat exchanger.
If you can moniter the EGT and set the mixture to the hottest setting that you can get if the carb has a lot of ice then the EGT will be well below the normal cruise setting but keeping the heat in the engine is the only way to melt the ice. If you have no EGT indication then set the mixture for max RPM (the throttle should already be fully open).

[bookworm]

Then surely there is such a large gap between the +10C and the dew point of the air, that there is no possibility of Carb Ice forming?

At -20C I can't imagine the air holding very much water. Just because you raise it's temperature to 10C very quickly, isn't going to make much difference. Where is the water going to come from?

Exactly. The only place it can come from is solid water that would otherwise have not presented a hazard, but melted because of the heat. Thus you might want to avoid carb heat application in glaciated (e.g. cirrus) cloud and in snow. I don't see how carb heat application can make icing worse in clear air.

[3 Point]

Don't forget that the temperature in the carb venturi will be significantly lower than the OAT; this is due to a combination of adiabatic cooling as the pressure drops in the venturi and to the latent heat taken from the air and the carburettor body by the evaporation of the fuel. This is why you can get icing in the carb with the OAT above zero.

So, in dublinpilot's scenario with an OAT of -20 and a further drop in the carb temperature of 10deg due to these two effects, to -30 deg, any water will be ice crystals which will go straight through the carb. Now apply heat and raise the temperature 30 degrees to...... around zero. Bingo, ice forming in the carburettor because you applied heat!

3 Point

[Mike Cross]

My feelin is that carb icing is the number one killer for light aircraft.
I think you'll find the subsequent unconventional landing is more likely to be what does the killing.

[Audio]

iv only had carb icing once, and am always flying over water, and i NEVER keep it on whole time, just a quick 30 second balst from time to time to ensure that its clear of ice and when needed. remember if you keep it on, and ICE can form at these times how are you going to melt it then?

its a complex thing that not many understand, but in some conditions carb heat on can cause ice to form.

but its more about local conditions, and finding what is best, but under my training, its best to keep it off, onless you are bringing the rpm right down.

[ShyTorque]

My (RAF) training was not to ever use partial carb heat, to avoid putting the intake air into a condition where ice might form. Either use it fully ON, or fully OFF, as a preventative rather than a cure where possible.

Hot air will cause a reduction in power to some extent due to its lower density. Detonation might also be made more likely by the inappropriate use of hot intake air.

I recall that the CAA did a safety bulletin on this some time ago.

[A and C]

If you think that a quick "30 second blast" of the carb heat will clear the ice then it is time that you started practicing your forced landings and ditching drills.


Some times the ice seems to have cleared but if the conditions have not changed the ice will start to build up the instant that the carb heat is selected cold.

If you have to use the carb heat in IMC I would think that "blast" of 2 min would be a good place to start with keen observation of the RPM, MP and EGT to check that the ice has gone.

This usualy works only to show that as soon as the carb heat is selected cold the ice slowly starts to form once more, I have had trips in IMC when the carb heat has been left selected hot for most of the cruise because as soon as it was put to cold the ice formed rapidly.

As to this thing about carb heat making ice form I think it is just another bit of aviation folk lore with very little practical use, if the air is so cold that it requires the carb heat to melt the ice crystals in the air so that they can reform as ice in the carb then the volume of water to air is so low as to make the chances of any ice that forms insignifficant and lightly to subliminate as qiuckly as it forms.

[IO540]

Rather than repeating the standard teaching, can someone demonstrate, using physics, how carb ice can form through the partial use of carb heat?

The air will be warmed up, sure, but its dew point is still going to be what it was before.

[DBo]

Manual carb heat control always amazes me. Car manufacturers solved this problem in the 1970s before almost universally ditching this technology in favour of fuel injection. I don't recall ever having a fault develop on a car where the induction heating system failed and the engine developed a problem as a result.

I've had several experiences of carb ice causing engine failure - fortunately in a Ford Escort in the early 80s. The automatic carb heat selector failed and after a 30 mile run up the motorway from Bristol the engine would quit in the contraflow near Gloucester. A couple of minutes of cursing & the car would start & run normally. After the 2nd or 3rd time it dawned on me that Ice might be the problem. It was a classic for carb ice, damp autumn mornings, no trouble at high power (70-80 mph), but running slowly (about 40-50 mph) through the roadworks & the engine dies.

It's amazing how many car mechanics have never heard of carb ice.. "you'll never get it cold enough for petrol to freeze" they told me. I jammed the selector flap on hot and had no problems for two years until the bit of tat I stuffed in the mechanism fell out...

Dave

[ShyTorque]

IO540, I see where you are coming from. A complex equation of maths and physics though, probably somewhat different for each design of aircraft, hence the advice in the manuals not always being the same.

As far as I was taught, you must keep the surface temperature of the intake venturi / throttle plate material (usually metal) out of the range where moisture in the air can condense and rapidly freeze on it.

It's there that critical ice forms, not so much in the air itself. If you can get the H2O in the air past the throttle plate and through the venturi without the ice particles precipitating onto the metal, you have succeeded. If the moisture has already precipitated out into tiny ice particles and they remain so, they will go straight through. If you keep the intake area warm enough, the moisture doesn't precipitate, so it also goes straight through. It's the sticky bit in the middle that causes the problem. Partial carb heat might produce that condition.

You will already be aware that two things making the metal cool down - the pressure drop in the intake/around the edge of the throttle plate and the latent heat of evaporation of AVGAS. You must apply sufficiently heated intake air to keep the metal out of the critical range.

By design, it is possible to heat the body of the carburettor but not so easy to heat the throttle plate, hence the need for intake air heat.

I brought a car back from warmer climes. It too used to conk out from carb ice as it had no provision for intake hot air. Taught me a lot about the conditions condusive to carb ice (not too cold but lots of airborne moisture). Two ways round it. Either wait at the side of the carriageway for engine heat to percolate up and clear it or use full throttle till it got sucked into the engine - lots of coughing and banging while it did so and obviously not always possible in traffic.

[IO540]

ST

I see your reasons but the basic issue with partial carb heat is

The air will be warmed up but its dew point is still going to be what it was before

and that is not dependent on the aircraft or the carb design.

If the OAT is say 0C (and the DP is say -5C) and you heat it up to say +20C, it's DP will still be -5C so at +20C its RH will be very low. It is going to be very hard to get ice out of it.

Is the above wrong?

The reason, AIUI, for not having permanent carb heat is that it reduces the engine power. (I am not sure whether it reduces it's efficiency in terms of HP per fuel flow though; I think it doesn't).

[wbryce]

I was trained with much emphasis on the carb heat....but I have a quick question that came into my mind when seeing this thread:

If you have carb icing at say 3000ft, can you slightly lean the engine to make it run hotter with carb heat on to help your case? Or is leaning the mixture slightly a no no in this situation?

[Mariner9]

(I am not sure whether it reduces it's efficiency in terms of HP per fuel flow though; I think it doesn't).

Not sure that can be the case IO. If it were, surely aircraft engines would be designed to have carb heat on at all times except when climbing, rather than the current off except when decending?

wbryce Carbice obviously reduces (and can eventually stop) fuel flow into the engine. I'm not sure restricting fuel flow even more (by leaning the mixture) would therefore be advisable. Although I have to confess I don't actually know how the mixture control works - does it restrict the fuel flow or increase the air flow?

[FlyingForFun]

There's been a couple of mentions of Chipmunks.... but none of Tiger Moths, where the carb heat is arranged by the manufacturer to be permanently on, unless full power is applied. Is there any reason why this arrangement shouldn't work in every/any other aircraft?

And another question: I was always taught not to apply partial carb heat. I'm currently teaching a PPL(H) for his PPL(A). He tells me (and maybe some of the other heli pilots here can confirm?) that the use of partial carb heat is normal in the R22 and many other helicoptors. Why the difference?

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

[Whirlybird]

I'm currently teaching a PPL(H) for his PPL(A). He tells me (and maybe some of the other heli pilots here can confirm?) that the use of partial carb heat is normal in the R22 and many other helicoptors. Why the difference?

As far as I'm aware, it's because the R22 has a carb heat temperature gauge. You are supposed to always apply enough carb heat to keep it at 10 degrees, unless you lower the lever (ie low power) when you pull full carb heat. Presumably since you know the temperature of your carb, you're not ever in the position of having applied just enough heat to make ice more likely to form. At least, that's how I've always understood it.

[High Wing Drifter]

I am not sure whether it reduces it's efficiency in terms of HP per fuel flow though; I think it doesn't
Wouldn't heated air have the same properties as high alt air - a lower density? Heated air would suck up the same amount of fuel because the same volume of air is being drawn through, but the mass is less so the mixture is rich(er).

Is there any reason why this arrangement shouldn't work in every/any other aircraft?
My understanding is that heated air from the exhaust manifold is unfiltered, therefore it is a bad idea to leave it on in most cases. However, our Vagabond requires CH on during taxing, especially on the potentially moisture laden grass fields it usually frequents.

[Piltdown Man]

And now for my two-penneth worth. A few years back doing a CPL in another place, there was a question about the most likely temperature at which you could expect carb-ice. From memory, the answer was +16C. The reason was that the evapouration of fuel and the venturi effect cooled the incoming air by 20C or so, thus enabling ice to form. This didn't preclude ice formation at other temperatures, just that this was optimal. It is also important to realise that cooler air is less able to contain water vapour than warm air so when flying in sub-zero temperatures carb-icing is less of an issue. The time to be really switched on with surface temperatures of 18-25C and high humidity.

[bookworm]

I am not sure whether it reduces it's efficiency in terms of HP per fuel flow though; I think it doesn't.
You're right. See also Deakin's sidebar (http://www.avweb.com/news/columns/182583-1.html).

[A and C]

Most carbs control the mixture by controling the pressure difference between the carb venturi and the float bowl so at full rich the float bowl has atmospheric pressure above the fuel and as the mixture is leaned the presure is reduced towards ventui pressure ....... fuel flow is proporsional to the pressure difference.

Mariner 9

The problem is that it is the airflow that is restricted by the ice and so the mixture goes towards rich this reduces the temp in the engine and so reduces power. The EGT will go down and so the heat in the carb heat exchanger will be reduced.

In short if you have carb icing the mixture should be set to the hottest EGT setting that you can get.

Lycoming say that as long as you are below 75% power you can lean to peak EGT so I see no problem with leaning to peak EGT as the ice in the carb at cruise power settings will be taking the engine power well below 75%.

[Mariner9]

In short if you have carb icing the mixture should be set to the hottest EGT setting that you can get.

You may be right A&C, but I would have thought the EGT of an over-rich mixture would still be more than sufficient to melt carb ice. My instinct would still to be put the mixture fully rich and carb heat fully on if the engine started running rough. Leaning would be counter-intuitive in my view.

I no longer fly an ac with a mixture control though, so I don't have to face that particular quandry at the moment.

[wbryce]

A & C,

Does that mean what I mentioned in a couple of posts above this one is no problem?


If you have carb icing at say 3000ft, can you slightly lean the engine to make it run hotter with carb heat on to help your case? Or is leaning the mixture slightly a no no in this situation?

[Piltdown Man]

Don't mess about. If you have carb ice, get rid of it as quickly as you can. Farting about with mixture, EGT and other stuff will end in tears. Pull the carb heat and and if ice keeps returning, change your level.

[A and C]

So you think that I'm "farting about" well at least I'm not swimming !.

A few years back I was going to the Isle of Man in the cruise at FL100 between WAL ant the IOM just on top all seemed fine but it all started to go wrong.
I was asked to take a lower level so ATC could get a Turbo prop through my level so down I went to FL80 as I started down I put the carb heat on and mixture slightly richer. After I leveled at FL80 it was clear that ice was slowly building up so on went the carb heat and It seemed to do the trick and clear the ice but as the carb heat was returned to cold the RPM instantly started to drop so the carb heat went back on and remained in that positon.

Very soon it was time to descend so I was cleard to FL60 with the carb heat in the hot position and the throttle about 3/4 shut the aircraft started a cruise descent approaching FL60 I opend the throttle and NOTHING happend to the RPM I told ATC thatI had no option but to continue the descent.

I had always been told to go full rich if the engine won't respond to throttle inputs but when I did this the RPM dropped further and the EGT fell to zero.

I then leaned the mixture and the RPM and EGT increased and the ROD decreased untill approaching FL40 the engine started to produce enough power to maintain FL40.

Pitdown man I ask you that if I had not leaned the mixture and increased the EGT How would the carb ice melted with out the heat from the engine ?.

As you say the carb heat should be used as soon as ice is suspected and if ice persists then change the altitude but why oh why do you think that not getting as much heat in the engine is a bad idea ?

wbryce

If you are in the cruise at 3000ft the mixture should be leaned as a matter of normal engine management.

[ShyTorque]

A carburettor cannot sense density of the air - it senses airflow velocity only. If the intake air is less dense, the mixture will tend towards rich, because the same fuel flow (mass) will be drawn out from the jet for less mass of air.

That's why we need to "lean" an engine with increased altitude - we are not really making the engine run "leaner", rather we are keeping the air / fuel ratio where it should be, with a lower outside air density prevailing. The power produced will decrease but the still air fuel consumption will be better. (There will be a more beneficial IAS / TAS ratio which helps in that respect, also pumping losses are decreased due to lower exhaust back pressure).

Ice in a carburettor system can begin to block the filter, the intake pipe, the venturi or the throttle plate aperture. This is likely to cause the mixture to become much more rich as the jet experiences a stronger depression across it (think of the starting choke on a car - those who can remember manual chokes on cars will know, it's a bit like putting your hand over the end of the carb, causing the engine cylinder "suction" to pull much harder on the jet for a constant throttle setting), greatly increasing the amount of fuel drawn out.

I don't think going "mixture fully rich" will alleviate that symptom, it's going to make it worse - possibly to the extent that combustion doesn't take place and fuel goes unburnt out of the exhaust on a windmilling engine. This sounds like what probably occurred on A&C's flight!

This link is useful:

www.gasco.org.uk/upload/docs/GASCo%20paper%20final.doc

[IO540]

A carburettor cannot sense density of the air - it senses airflow velocity only

Not really correct. The assertion that the carb measures velocity only is easily disproved by considering two cases: sea level and say 100,000ft. There is about a 100:1 difference in air density. Clearly the latter case (roughly the sea level pressure on Mars) will not cause the same amount of fuel to be sucked up into the airflow.

The reality is that a carb measures something part-way between mass flow and plain velocity. Same for fuel injection incidentally.

The carb does however measure the FUEL by mass flow - it's easy to do with a liquid.

This disparity between the way the air is measured and the way the fuel is measured gives rise to the need to lean for increasing altitude.

There are altitude compensated carbs which implement a bodge to approximate true mass flow air measurement.

Direct air mass flow measurement is standard on cars but planes are some decades behind...

I don't really see a useful connection between carb icing and the mixture setting though. A lean mixture is likely to increase the EGT but the carb doesn't see any of that obviously. Higher EGT would mean a hotter exhaust pipe and if the carb is getting some exhaust-heated air then than a higher EGT would help but the effect is bound to be marginal. The change in EGT from fully rich to peak-EGT is only about 30%.

[ShyTorque]

IO,

You seem to like to split hairs. I don't think you would be flying your piston engined aircraft in the extremes of the Martian atmosphere or at 100,00 feet; not a good example. The point is that changes in air density caused by increased altitude effects makes little measurable difference to the fuel draw in the working environment of a piston engine carburettor. If it did then aircraft wouldn't need a mixture control. Aircraft carburettor designs became hugely complicated about fifty years ago, for just that reason.

"I don't really see a useful connection between carb icing and the mixture setting though."

This is basic stuff. A carburettor depends on the smooth passage of air through the venturi. Anything that alters or disrupts the flow has a marked effect on absolute mixture strength or on the formation of a uniform fuel/air emulsion in the manifold.

As a hobby I have worked on and modified carburettors and intakes for ground-based vehicles for many years; I used to be amazed how a seemingly minor modification to the intake, or even the filter type or position, can make a large difference in mixture strength - now I'm not at all surprised to hear how a small build-up of ice can stop an engine altogether.
........................

"ST, I see your reasons but the basic issue with partial carb heat is: The air will be warmed up but its dew point is still going to be what it was before".

A point you have perhaps missed is that there might well be liquid water (droplets) present in the ambient / intake air, especially so under a cu-nim cloud. Once the intake air becomes warmed it will contain more water than ambient. This is where the use of partial carb heat may put the air in the band for icing to occur (a small increase in the amount of water will absorb much more heat in the carb) and so the design of an effective carb heat system must allow a surprisingly large increase in temperature, to overcome the powerful cooling effects in the intake / carb combination.

As we know, the trick is to keep the intake temperature above zero degrees C so that ice cannot precipitate. The designer of an effective carb heat system will have to over-compensate for safety reasons, especially if a carb temperature gauging system is not fitted. If you look at a typical carburettor equipped engine hot air intake for a car, you might be surprised how closely the hot air shroud fits round the exhaust manifold. It looks like an over-compensation but it's like that for good reason.

The information is all out there; here is something else for you to read. This article mentions partial carburettor heat being possibly ineffective and also gives advice to lean mixture (rather than richen it) if icing is encountered:

www.casa.gov.au/fsa/2001/jul/28-31.pdf

My input now complete, I'll go back to tidying up my garden as I only fly turbines these days. ;)

[IO540]

Unfortunately the PDF leaflet, written in the standard UK CAA-style patronising manner, contains the same old assertions

Under certain conditions, partial carburettor heat may be worse than none at all.

I am still looking for a proper physical proof of this, rather than a repeat of the same statement that's dished out everywhere.

As for my 100,000ft example, I was merely using the standard way of proving or disproving something easily, by considering obvious extreme cases :O

A point you have perhaps missed is that there might well be liquid water (droplets) present in the ambient / intake air, especially so under a cu-nim cloud. Once the intake air becomes warmed it will contain more water than ambient

Only if the droplets are able to vapourise in the time it takes for them to pass from the part of the induction system where the air gets heated, to the part where ice would form which is the venturi. Can this happen? I don't know.

However, if the temperature of the carb internal surfaces was below 0C and there were liquid droplets in the air, one would expect to get extremely rapid carb icing - just like flying through freezing rain causes a very rapid clear ice buildup.

[High Wing Drifter]

The reality is that a carb measures something part-way between mass flow and plain velocity. Same for fuel injection incidentally.
I beg to differ! The amount of fuel metered (for a given carb setup) relates to the pressure differential between the venturi and the ambient pressure in the fuel chamber, density doesn't affect this relationship. At 100,000' the pressure differential at 27mb between the venturi and the fuel chamber will be very small, that is why less will be drawn. It doesn't matter of you pass a hot gas or a gold bar through the carb, the same amount of fuel will be drawn; the pressure differential is related to the volume of air drawn through the carb.

[ShyTorque]

"Only if the droplets are able to vapourise in the time it takes for them to pass from the part of the induction system where the air gets heated, to the part where ice would form which is the venturi."

IO,

As long as water is present, liquid OR vapour (and don't forget that water can exist as supercooled droplets), it can form ice if the carb body temperature is low enough. Partial carb heat may not overcome the "refrigeration cooling" effect in the carb caused by the evaporation of fuel issuing from the jet, which is the main cause of the temperature drop therein, not the Bernoulli effect.
.................

"Can this happen? I don't know."

I'm certain it can. I've watched it happen in an SU carburettor with the air filter removed and the engine sucking in warm engine bay air while I tried to tune it. The engine misfired like hell after a couple of minutes at idle, so tuning it was impossible. I shone a torch in the venturi and noticed ice in there. The ambient conditions were cold and very wet, it was raining but I was inside the garage with the door open. I eventually gave up tuning for the night. It was fine next day once the ambient air was less humid and the engine cover was closed. Increased engine bay heat, coupled with lower humidity, was sufficient to prevent the icing problem.

[bookworm]

A point you have perhaps missed is that there might well be liquid water (droplets) present in the ambient / intake air, especially so under a cu-nim cloud. Once the intake air becomes warmed it will contain more water than ambient.

Where does this "more water" come from then? If the water is there, as liquid or vapour, it will form ice in the carb due to the cooling. And it will do so whether you apply partial carb heat or no carb heat at all. How can partial carb heat be worse than no carb heat (in the absence of ice particles)?

[High Wing Drifter]

Where does this "more water" come from then? If the water is there, as liquid or vapour, it will form ice in the carb due to the cooling. And it will do so whether you apply partial carb heat or no carb heat at all. How can partial carb heat be worse than no carb heat (in the absence of ice particles)?
There isn't any more water. Merely that you have say +5deg ambient, in the carb it is now -25deg. At -25deg, ice will be much less likely to stick. My understanding is that ice sticks because when it impacts, the impact pressure causes the ice particle to quickly melt and then it re-freezes. If the temp in the carb -25, the impact is not sufficient for the particle to melt so it just carries on through and into the cylinder. If you use partial heat in that case, then you may raise the carb venturi to say -5 which is within the temprature range where ice forms most rapidly. Hence the recommendation for a carb temp guage if partial use of carb is required.

[ShyTorque]

Unless the air is cooled sufficiently to cause ice precipitate to pass straight through the engine. It appears that with very cold air, the ice particles go straight by the throttle plate without sticking.

Aah, High Wing Drifter beat me to it.

[bookworm]

There isn't any more water. Merely that you have say +5deg ambient, in the carb it is now -25deg. At -25deg, ice will be much less likely to stick. My understanding is that ice sticks because when it impacts, the impact pressure causes the ice particle to quickly melt and then it re-freezes.

OK, I understand the proposed mechanism, I think, but I don't believe it. The kinetic energy of a 100 kt ice particle is about 1 J/g which is worth about 0.25 degC. Even if the ice particle impacts a surface and just stops (which is highly unlikely), the "impact pressure" will be effective only in marginal cases.

Ice doesn't stick. Supercooled liquid water sticks.

Unless the air is cooled sufficiently to cause ice precipitate to pass straight through the engine. It appears that with very cold air, the ice particles go straight by the throttle plate without sticking.

But the ice particles need to be there in the first place for this to be a problem, right? Again, I can see how heating up a volume of glaciated cloud could cause icing where none would otherwise be observed, but I can't see how heating up a volume of unsaturated air could do so.

[jonnoboy]

With reference to Bookworm's last statement, surely it doesn't matter that the clear air being flown through is unsaturated. It still contains water, however little.

So if it is -25 deg and applying carb heat raises the temperature to -5 deg, then the carb itself will be at about the right temp for water in the air (however little) to freeze on contact with parts in the carb and slowly accumulate. Again if this was the case, then it would be hard to melt the ice (without descending) , as the carb heat is already on full.

Now am I thinking right, or does the air have to be saturated, ie water droplets be present for the ice to actually buiild up?

This thread certainly makes you think!

[slim_slag]

If the statement

"Under certain conditions, partial carburettor heat may be worse than none at all."

is correct, which for reasons given above I don't think to be the case, then surely

"Under certain conditions, full carburettor heat may be worse than none at all."

is equally true (or more likely false).

[Piltdown Man]

Maybe I haven't expressed myself properly. From your post I am given to understand that you were suffering from Carb Ice so you immediately applied full carb heat. That is not farting about! Trying to prevent carb ice by trying other methods first is!

[bookworm]

Now am I thinking right, or does the air have to be saturated, ie water droplets be present for the ice to actually buiild up?

Water droplets have to be present for ice to build up. Because of the cooling in the carb, unsaturated ambient air can become saturated and cause icing.

However, the cases we're comparing are:

1) Take air at a particular temperature and humidity and cool it by 30 degC in the carb. (carb heat off)

2) Take air at the same temperature and humidity, heat it by (say) 20 degC and then cool it by 30 degC in the carb. (partial carb heat)

The amount of water available in the carb as supercooled droplets to cause icing is the same in each case (or in fact, rather more in case 1), unless there are ice crystals in the starting sample of air.

[High Wing Drifter]

Water droplets have to be present for ice to build up. Because of the cooling in the carb, unsaturated ambient air can become saturated and cause icing.
I know where you are comming from, we are taught that if you can't see the water, you won't get ice (except for hoar frost).

However, if the air humdity is high enough, then it is possible for the pressure drop in the venturi to be sufficient to condense the moisture in the air and so form droplets?

[bookworm]

However, if the air humdity is high enough, then it is possible for the pressure drop in the venturi to be sufficient to condense the moisture in the air and so form droplets?

Yes absolutely. But heating the air first (with partial carb heat) can only reduce the amount of water that condenses as (supercooled) droplets.

[High Wing Drifter]

Yes absolutely. But heating the air first (with partial carb heat) can only reduce the amount of water that condenses as (supercooled) droplets.
I think it is simply a matter of degrees of influence that temprature, pressure and humidity have in the process. Partial heat may well increase the carb air temp from what would have been -25 deg unheated to -5deg, but if the OAT is +5deg then the air in the carb will have a higher relative humidity, coupled with the pressure drop in the venturi all the ingredients are perfect for a special serving of Carb Sorbet.

[bookworm]

perfect for a special serving of Carb Sorbet

It begs the question of why lower temperatures are less perfect.

I think the bottom end of the carb icing envelope is determined by the amount of water vapour in the air.

1) If you start at an OAT of 25 degC (saturated or near) and cool to -5 degC in the carb, there's 20 g/m^3 of water that condenses out into supercooled droplets ready to form ice.

2) If you start at an OAT of 5 degC (saturated or near) and cool to -25 degC in the carb, there's something like 5 g/m^3 of water that condenses out. That's much less of a problem from an icing point of view, which is why the -25 degC in the carb is less troublesome than -5 degC in the carb.

3) If you start at an OAT of 5 degC (saturated or near), heat to 25 degC and then cool to -5 degC in the carb, there's only about 4 g/m^3 of water that condenses out, even less than in the previous case.

You might argue that a greater proportion of the water in case 2 would spontaneously form ice than in case 3, but I think the processes that go with glaciation are much slower than the time in which the air is cooled in the venturi and then slams into surfaces that can collect ice. You'd expect all the condensed water to be capable of causing ice.

[High Wing Drifter]

Bookworm,

I don't have the knowledge to argue about how much water condenses out for a given temp drop. Suffice to say that I'm pretty sure that it doesn't matter what order you heat and cool, the amount of water in the air doesn't change (unless it rains which is another somewhat mysterious process). For a given pressure and RH at 5 deg you will get RH' at -5deg and RH'' at -25deg regardless of what route you followed to finally end up at the temprature.

I'm stepping into dangerous territory of guesswork from 1st principles here!!!: Your pointing out of supercooled droplets triggered a memory and got me looking up me old notes.
Something to bear in mind, is that when water condenses it releases latent heat into the atmosphere. The same is true when supercooled water droplets (SCWD) impact and then move to the stable ice state. This process of releasing latent heat (80 cal for deposition?) is the problem for ice formation, because if there is enough released heat to partially melt the now iced droplet, it will re-freeze and this it is the re-freezing that forms the dangerous clear ice. The portion of the SCWD that was instantly frozen is fluffy and aerated (rime ice) and probably gets blown down the tube almost instantly. At -25deg, significantly less of the SCWD is melted and re-frozen preventing serious ice formation. At -5deg the opposite is true and the majority of the SCWD ends up as clear ice.

Obviously, that last paragraph assumes that wing icing and carb icing are two sides of the same coin.

With that in mind, your point about how much time this takes is taken. But I guess if ice forms on the wing of a jet liner at 450kts TAS then 150kts inside the carb presents no real barrier. A quick and rough calculation of carb body at 150kts and airliner wing at 450kts (TAS of course) suggests that they both form ice in about 0.01 seconds!

[bookworm]

With that in mind, your point about how much time this takes is taken. But I guess if ice forms on the wing of a jet liner at 450kts TAS then 150kts inside the carb presents no real barrier. A quick and rough calculation of carb body at 150kts and airliner wing at 450kts (TAS of course) suggests that they both form ice in about 0.01 seconds!

Your point is slightly different, HWD.

I was thinking that one might claim that the condensed water might have already frozen at -25 degC, but not at -5 degC, since we normally observe glaciated (pure ice) cloud at -25 degC, but a high proportion of SCWD at -5 degC, which is what makes clouds much more dangerous to the airframe at -5 degC than -25 degC. That's because the SCWD soon (seconds or minutes) finds an ice nucleus and sticks to it. It has already done that by the time your airliner wing hits it.

But the time over which the cooling occurs in a carb venturi is of the order of 0.1 milliseconds (length of a carb venturi divided by the speed of sound) and I don't think this is long enough for the newly condensed SCWD to find an ice nucleus. So in the carb, I don't think one can take much consolation from the -25 degC.

The latent heat issue is a different one, and a good point. I agree you would expect to deposit clear ice in a carb at -5 degC and rime ice in a carb at -25 degC. While clear ice is more dangerous for the wing because of runback, I don't know how much difference it would make in a carb, nor whether rime ice "probably gets blown down the tube almost instantly" -- you may have a case. My vague recollection of carb ice photos from experiments is that it looks milky like rime.

[Lister Noble]

This is becoming so complicated it hurts my head!.
I reckon I will listen to and follow my instuctor who has just celebrated 50 years of professional flying,he must know what he is talking about.
Be aware of weather that can cause icing.
Regular carb heat checks to make sure there is no icing, and carb heat on when using lower throttle settings.
Lister:D

[High Wing Drifter]

Bookworm,

Interesting stuff and has blown the cobwebs of the whence dusty ATPL manuals!!

(length of a carb venturi divided by the speed of sound)
I don't understand that bit. Why speed of sound or am I missing something obvious?

Lister,
This is becoming so complicated it hurts my head!.
Its a damn sight more interesting than the other subject of the day: Data quality managagaammnt...zzzzzzzzzZZZZ

[bookworm]

I don't understand that bit. Why speed of sound...?

In the broadest terms, when you have a pressure drop in a pipe or orifice, then it's either viscosity or compressibility that causes the pressure difference. In the case of compressibility the flow rate is limited by the speed of sound at the narrowest point. So the speed can't be faster than that (unless you do the clever stuff they do to make supersonic wind tunnels), and if the pressure drop is substantial, it's probably close to the speed of sound. I don't know what the pressure distribution looks like on either side of the venturi, but that's the upper limit.

Perhaps more pertinently, the temperature drop and the flow velocity are related by thermodynamics. For a drop from say 25 degC to -5 degC, the flow velocity has to have a Mach number of about 0.8. So the speed of sound is close enough for the sort of approximation I was making.

[IO540]

my instuctor who has just celebrated 50 years of professional flying,he must know what he is talking about

As Bookworm shows, the above assumption is dangerous in GA :O Old wives tales have a half life of at least 50 years.

[High Wing Drifter]

Bookworm,

I must admit you've lost me now. Poor old me was simply thinking how long the moist air would be in the carb; in simple terms if it isn't relatively very long then there isn't enough time for the moisture in the air to cool to the now brrrrr cold air temp in the carb, and visa-vers of course.

[bookworm]

Poor old me was simply thinking how long the moist air would be in the carb; in simple terms if it isn't relatively very long then there isn't enough time for the moisture in the air to cool to the now brrrrr cold air temp in the carb, and visa-vers of course.

No, I think the cooling can happen very quickly. What would take much longer is the spontaneous formation of ice crystals in the air. Thus the propensity for the flow to form ice in a nasty way is no different whether the temperature is -25 degC or -5 degC, assuming it had the same moisture content.

[ShyTorque]

Back from a few days away working - glad to see the discussion still going! :ok:

Bookworm said: "No, I think the cooling can happen very quickly. What would take much longer is the spontaneous formation of ice crystals in the air. Thus the propensity for the flow to form ice in a nasty way is no different whether the temperature is -25 degC or -5 degC, assuming it had the same moisture content."

I'm not convinced that ice crystals DO form in the air itself inside the carb, any airborne ice crystals before the throttle plate would be cooled even further by the refrigeration effect of the evaporating fuel and would be unlikely to stick but go past it into the manifold without causing an icing problem.

However, supercooled water droplets in cold air (remaining as water below the normal freezing point of 0 degrees C) need a nucleus to freeze onto. Water droplets entering the intake, or precipitating out due to cooling saturation once the air has entered the intake, is more likely to remain as supercooled water droplets that freeze instantly on contact with the metal of the carb throttle plate, spindle, or any other cold projection within the carburettor.

That's why it occurs so rapidly if the conditions are right (or wrong, as far as the pilot is concerned).

As I said earlier, I've watched this happen inside a carburettor, the ice formed directly on the throttle plate edge at the points where it was closest to the wall of the venturi.

[Audio]

A and C,

the 30 second Blast is to check for ICE ONLY, of course if you do find ice youd keep it on.

[High Wing Drifter]

Shy and Bookworm,
However, supercooled water droplets in cold air (remaining as water below the normal freezing point of 0 degrees C) need a nucleus to freeze onto. Water droplets entering the intake, or precipitating out due to cooling saturation once the air has entered the intake, is more likely to remain as supercooled water droplets that freeze instantly on contact with the metal of the carb throttle plate, spindle, or any other cold projection within the carburettor.
I was under the impression that we tacitly agreed this bit earlier on. Bookworm brought up the issue of SCWD, but ice in the air does seem to have reapeared.

[sstaurus]

Hmmm... I have read pretty much every reply on this thread and I still don't get the concept.

I was under the impression, that for water to condense it needs to reach its DP. In order to do this, either more water has to be added to the air or the air has to be cooled. So how does adding carb heat risk ice build up? :confused:

I guess it must have to do with the supercooled water droplets. By heating them up, they are more likely to freeze? Supercooled water droplets only occur at very high levels no?


(forgive me I'm not a pilot, just an interested wannabe)

Cheers

[bookworm]

I guess it must have to do with the supercooled water droplets. By heating them up, they are more likely to freeze?

Well that's, in effect, what we've been debating. We've been discussing a number of mechanisms why that might or might not be the case. That's the only mechanism we can come up with for partial carb heat to make icing worse than none at all. I don't believe it, but I'm not sure if I've convinced HWD.

Supercooled water droplets only occur at very high levels no?

No, they occur anywhere with ambient temperature below freezing where the water in the air has not yet formed ice particles. That includes a carburettor venturi where the air is suddenly cooled, and also a lot of (lowish) level cloud with a temperature between about -10 and freezing. It's less likely at cold temperatures (and therefore at high levels) because the droplets will have formed ice particles.

(forgive me I'm not a pilot, just an interested wannabe)

I think you have to be a wannabe physicist to get excited about this. :8

[High Wing Drifter]

I don't believe it, but I'm not sure if I've convinced HWD.
The only bit I'm unsure of is, at very cold tempratures, does the ice form in the air or does it remain in SCWD form and so deposition only occurs upon contact with the carb body? Otherwise, I fink we's talking the same fing :8

[ShyTorque]

Supercooled droplets can remain liquid down to minus 40C, they need a nucleus to freeze around or onto and will do so instantly (read up Bergeron Process). The cold metal of a carb can provide just that.

That was my earlier point. The idea of correct use of full carb heat is to ensure that the carb body is heated above zero. There is a struggle between the cooling affect in the carb and the heating effect provided by carb heat. Improper partial use can cause the carb body to be just at or below freezing, so it becomes a nucleus and gets rapidly affected by ice.

If the ambient air is very cold (below freezing), it will probably contain only a relatively small amount of water because it will have "dropped out" / precipitated as rain or snow before it enters the carb intake. Warmer air may contain much more water and may provide a much greater RATE of ice build-up once cooled in the carb.

The trick is to not allow the carb body to be operating at the zero degree mark with supercooled water droplets passing through so FULL carb heat is the only safe option.

Carb heat systems are designed to give a temperature rise of 50 degrees C. Compare that with the supercooled droplets able to exist down to minus 40 degrees C (!) and you might get the idea why use of partial heat might just put the carb body at the critical temperature range.

BTW, the scientists are still arguing about the finer points of this process so it's not surprising that we are, too! ;)

[bookworm]

If the ambient air is very cold (below freezing), it will probably contain only a relatively small amount of water because it will have "dropped out" / precipitated as rain or snow before it enters the carb intake. Warmer air may contain much more water and may provide a much greater RATE of ice build-up once cooled in the carb.

But by using partial carb heat, you're not taking in warmer ambient air containing more water. You're taking in the same air that would have gone through if you'd left the carb heat cold, but you're heating it before allowing it to flow through the carb. If there are water droplets in that air, they will become supercooled droplets and ice up the carb whether you heat them first or not.

Only ice particles, in other words, glaciated cloud, can be harmful if heated (and melted), relatively harmless if not.

you might get the idea why use of partial heat might just put the carb body at the critical temperature range

There is no "critical temperature range" for the carb body. The "range" has no bottom end. It cannot be too cold within the carb to make ice. It can only be too cold in the ambient air for it to contain enough water to cause significant icing.

After all this debate, I still can't think of a good reason for applying partial carb heat! If full carb heat enrichens the mixture too much, use the red knob.

[ShyTorque]

"There is no "critical temperature range" for the carb body. The "range" has no bottom end. It cannot be too cold within the carb to make ice. It can only be too cold in the ambient air for it to contain enough water to cause significant icing."

I think this is the nub of the debate. I was taught that if the temperature drop is large (no carb heat), in certain conditions the water vapour does not condense out into liquid, but precipitates straight into ice crystals which go straight past the throttle plate without sticking, despite the low temperature of the metal. If the temperature is kept high (full carb heat) there is no formation of ice. It's the critical phase in between that hurts.

Looks like we're going to continue to disagree on that point.

Here's another link that backs up what I was taught: http://www.tc.gc.ca/civilaviation/general/flttrain/planes/Pubs/TP975/PartII/Ex3.htm

This one is from Canada - they also know a bit about ice over there.

So far the UK CAA, the NZ CAA and the Canadians as well as the RAF hold the same opinion - so I'll stick with it too.

[bookworm]

I think this is the nub of the debate. I was taught that if the temperature drop is large (no carb heat), in certain conditions the water vapour does not condense out into liquid, but precipitates straight into ice crystals which go straight past the throttle plate without sticking, despite the low temperature of the metal.

Yes, that's the nub. Some of the debate above was about analysing whether there is time for those ice crystals to form. I believe the timescales are orders of magnitude too short.

The TC reference indeed suggests that there are conditions in which carb heat (note they don't say, partial) can induce icing. I don't think we disagree about that, only about whether it can do so in clear (but humid) air.

I would add that it would not be the first piece of received wisdom quoted by various aviation authorities that turned out to be bad physics without a trace of empirical support! Just ask the FAA how a wing produces lift...

[High Wing Drifter]

Bookworm,
Just ask the FAA how a wing produces lift...
Ah! The old Bernoulli vs Newton debate perchance? (I've read Stick and Rudder too).

[bookworm]

Well HWD, I prefer to think of it as the flawed "equal transit time" assumption. But that's one for another rainy Sunday. ;)

Just to add to the last, Lycoming's article on carb ice (http://www.lycoming.textron.com/main.jsp?bodyPage=/support/publications/keyReprints/operation/inductionIcing.html) says:

Unless the aircraft is equipped with a carburetor air temperature (CAT) gage, and very few general aviation aircraft are, use of full carburetor heat is recommended. An unknown amount of partial heat can actually cause induction ice in the float type carburetor. This may occur when moisture in crystal form in the incoming air that would ordinarily pass through the induction system without any problem is melted by the partial heat. This moisture then freezes when it comes in contact with the cold metal of the throttle plate.