Why isn't the carb heated all the time?

[moggiee]

All this stuff about "heated air being less dense" is a bit of a red herring.

If the appropriately sized carb is fitted and the mixture is set correctly then there would be no appreciable loss of power but there would be a significant reduction in the risk of suffering a loss of thrust due to icing. Fitting a slightly wider bore carb to permit greater airflow and matching mixture to density would give as much power from the engine as you presently get with no loss of economy.

Unfortunately, very few people properly understand the physics involved and FIs trot out the same old inaccuracies to each new generation of pilots. It's just an internal combustion engine, after all.

[aviate1138]

My SkyArrow650 Rotax 912UL was fitted with a device that heated the carb body. Similar to below......

Carb heat, carburator heater for 912 Rotax, carb heat for Bing carburetors.

From the moment the device was fitted [and the original carb heat system via lever retained] was the day any carb icing conditions seemed to vanish.

Using the original system [with the Nigel Beale device attached] on really bad carb ice days engine output was smooth/no misfiring etc whereas before there would be a hiccup as the slivers of ice were consumed.

It was the best Rotax engine add-on device I ever bought.

[SNS3Guppy]

All this stuff about "heated air being less dense" is a bit of a red herring.

If the appropriately sized carb is fitted and the mixture is set correctly then there would be no appreciable loss of power but there would be a significant reduction in the risk of suffering a loss of thrust due to icing. Fitting a slightly wider bore carb to permit greater airflow and matching mixture to density would give as much power from the engine as you presently get with no loss of economy.

Unfortunately, very few people properly understand the physics involved and FIs trot out the same old inaccuracies to each new generation of pilots. It's just an internal combustion engine, after all.

The R3350's I used to operate, or the R2600's, were also internal combustion, carburetted engines, but were very different in operation than say, an O-320. All of them use carburetor heat, but are different animals.

Yes, one does lose significant power with the application of carburetor heat, depending on the specific installation and the phase of flight in which is it applied. This is not a red herring, and can be a crucial issue, especially when one is flying a heavily loaded light airplane at high density altitudes. Limited performance demands all the power one can muster under some circumstances, and the application of carburetor heat may be enough to tip the scales such that insufficient power is available.

One doesn't have the option of rejetting the carburetors, or selecting a bigger carburetor, or one with more barrels.

You can rest assured that the carburetors I used on large radial engines were substantially larger than what you see on small airplane piston engines. Never the less, carburetor icing was still a consideration.

Put a bigger carburetor on the engine and apply carburetor heat, and one still has a decrease in air density. Given the low RPM's and low relative power output from aircraft piston engines (due to limitations by the propeller RPM in fixed-drive installations), a significantly bigger carburetor isn't going to benefit the engine much, but increases in induction air density (temperature and altitude) will always have an impact.

There is no inaccuracy in explaining, nor in demonstrating that the use of carburetor heat causes a loss of power, because it's indeed the case.

Do you intend to swap carburetors during the approach to get the different results you describe? Do you intend that the student install a different carburetor? Have you successfully managed to STC a new carburetor and thus eliminate the need for carburetor heat in certified, piston engine airplanes? Do you in fact, then, have a basis for what you're attempting to suggest here?

Large volume or small, induction air density as a function of both temperature and altitude impacts engine performance. It's just an internal combustion engine, after all.

[IanPZ]

Gotta agree with sns3guppy. from my days as an engineer...

Power is related to the amount of fuel burned, and that has an ideal ratio of fuel to air. Therefore, to burn more fuel (and so to increase power, you need more oxygen. To get more oxygen into the cylinder, there are only two options (three, if you include nitrous oxide injection!) either increase the density of the air by cooling it, or increase the density by pressurising it. (denser air has more oxygen per unit volume)

Hence, subaru impreza turbo intercooler! Cools and pressurising, so getting both benefits.

By the same logic, heating air reduces the density, which makes less oxygen available to burn fuel, which means less power.

Now, where did I leave my propeller hat! (sorry, bit of a geek :-)

[FlyingStone]

Originally Posted by moggiee

If the appropriately sized carb is fitted and the mixture is set correctly then there would be no appreciable loss of power but there would be a significant reduction in the risk of suffering a loss of thrust due to icing. Fitting a slightly wider bore carb to permit greater airflow and matching mixture to density would give as much power from the engine as you presently get with no loss of economy.

Carburetor heat DOES, believe it or not, reduce maximum available engine power. It's - as you say - simple physics. By heating the air you reduce the air's density (the equation is written in my previous post) and by not changing any other parameter, such as dynamic (velocity) and static pressure, you effectively decrease mass flow of the air through the carburetor, the effect being the previous set mixture too rich, so one needs to adjust it to a new setting. But even with correctly (let's say stochiometric) set mixture without and with carburetor heat there will be approx. 5% drop in available power, since with reducing mass air flow due to increased temperature, you also reduce mass flow of the fuel through the jet into the carburetor. Again - simple physics: less fuel, thus less power (with maintaining stochiometric ratio of course).

As you say fitting the wider bore carburetor would reduce chance of carb ice, but the real way to go are injected engines, which eliminate many problems (except ram/intake ice).

[FL575]

Thinking about Funfliers post, the RAF Chipmunks did have the carbheat 'wire-locked' in hot in 1969.

That is when I learned to use 'wire-locking' pliers whilst taxying in. Should any other 'dual' Chipmunk be spotted during a dual sortie, a quick 'I have control' from the rear, followed by 'Carbheat to cold'. Then followed the usual 5 to 10 mins of 'combat'. As I say, it was the students job to 'tidy up' whilst taxying in. Is that correct D H-H?

[Miroku]

I used to fly an AA5 which had a device to monitor whether carb ice was forming (it didn't work mind). I'm surprised that nobody has come up with a device along these lines which could be easily (?) retrofitted due to the number of problems caused by carb icing.

[Genghis the Engineer]

Originally Posted by Miroku

I used to fly an AA5 which had a device to monitor whether carb ice was forming (it didn't work mind). I'm surprised that nobody has come up with a device along these lines which could be easily (?) retrofitted due to the number of problems caused by carb icing.

Generally a pain to retrofit such things to old engines, and if going for a new engine - have something fuel injected, then you don't have a carb ice problem anyhow.

G

[englishal]

We have a carb temp meter....but then we're very unlikely to get carb ice as we have a turbo charger.

[Mark1234]

Chippies are perhaps a little bit of a red herring. They have a warm air system, not a hot air system (hence less impact to performance if warm is continuously selected), and according to my handling notes are 'highly likely' (in red) to quit at low throttle settings w/o the warm air applied. I presume this means they make ice rather readily, I've never felt the need to test that. Certainly application of heat has a much less dramatic effect than in other a/c I've flown.

If the only difference carb heat made was a richer mixture, you could recover the power by using that little red lever that so many appear afraid of moving - no need to re-jet. However, as others have said, hot air is less dense, and cylinder volume is not variable, so hot air = less power. That's one of the reasons why a high ambient cripples your takeoff performance.

[IanPZ]

So, I know I said stuff about air temp and density, cos I've had to deal with that from car mechanics, but I wonder if someone can clarify the whole carb-icing thing.

I understand why it happens, but surely whether you have a carb or injection, ice will still form around the trumpet of the air intake. Is the problem ice there, or actually on the metering needle of a carb.

If its not the metering needle, then why don't injection systems suffer from carb icing? Presumably if ice forms around the venturi (which are still there in an injection system) then the cross sectional area gets smaller, the air flows faster, pressure drops, more ice, etc etc.

Or am I missing something (which I am sure is true)!

[BackPacker]

We have a carb temp meter....but then we're very unlikely to get carb ice as we have a turbo charger.

I've had carb ice twice now in the R2160 (O-360 or something like that; not injected) when taxiing with about 1000 RPM in high humidity conditions (low T/D spread, in both cases just ahead of a front passage).

At those RPMs I would assume that the turbo would not be spinning at all, or only give a negligable rise in MAP/temperature, so you would be able to get carb ice anyway.

(But how many non-injected turbo installations are there anyway?)

[Jan Olieslagers]

(But how many non-injected turbo installations are there anyway?)

Rotax 914 - not a very common engine, indeed.

[englishal]

(But how many non-injected turbo installations are there anyway?)

Quite a few (Rockwell Commander). It is funny because although a carb, we don't have carb heat, just "alternate air".

[Mark1234]

So, I know I said stuff about air temp and density, cos I've had to deal with that from car mechanics, but I wonder if someone can clarify the whole carb-icing thing.

I understand why it happens, but surely whether you have a carb or injection, ice will still form around the trumpet of the air intake. Is the problem ice there, or actually on the metering needle of a carb.

If its not the metering needle, then why don't injection systems suffer from carb icing? Presumably if ice forms around the venturi (which are still there in an injection system) then the cross sectional area gets smaller, the air flows faster, pressure drops, more ice, etc etc.

Or am I missing something (which I am sure is true)!

There are (broadly) two places where ice can form:

External impact icing on the intake trumpet/air filter stopping the supply of air. That's why you have alternate air, and is equally likely for injected and non-injected, but less common than 'carb icing'

Typically what we refer to as 'carb ice', which forms in the venturi/throttle plate etc. There are two contributing factors to the temperature drop here - the pressure drop in the venturi, and also the cooling effect of evaporating fuel (pour fuel or another volatile liquid on your hand, you'll feel a significant chill as it evaporates, so long as you're not close to a flame!)

In an injected engine, the fuel is typically added further downstream, nearer the inlet valves, thereby removing at least one of those.

As for venturis, I'm not familiar enough to comment on the specifics of aviation application, but typically fuel injected automotive engines do away venturi, using other mechanisms such as a heated wire for air mass sensing - I doubt aero engines are that advanced (they'd not need a mixture lever if they were); I rather suspect fuel metering is a simple mechanical linkage to how far open the throttle plate is. Long and short, I wouldn't expect to find a venturi in an injected aero engine - but I may be wrong.

[edit to add]
Turbocharging usually involves a significant rise in inlet air temperature - both through compression, and heat transfer of the turbine being rather warm, having several hundred degrees of exhaust gas flying through one end. Obviously the extra air more than compensates for this, but a turbo will raise inlet temp far more than the normal carb heat control anyways - I'd rather suspect that is why the rockwell has alternate air. I'm also assuming the carby is pressurised, and downstream of the turbo - I'm not sure if it's possible to do it the other way round and draw the fuel through the turbo.

[IanPZ]

Thanks, Mark1234

So, is carb icing really about the fuel cooling (or at least, thickening), and thus less vapourised fuel reaching the engine, rather than ice causing blockage of metering valves etc? I think that's what you're saying.

Certainly, most old-style fuel injection systems on cars (like old Mercedes) just had the equivalent of a distributor for fuel, and then an injector straight into the chamber, rather than into the neck of a throttle body (like modern cars).

I can see that doing that, the fuel mixture wouldn't be much affected by the air temp, especially as the cylinder itself would be hot from exiting exhaust gas....

Hey, why don't they just use 2-stroke engines instead? (Or do they?)

I'm slowly understanding the differences, and the light is starting to dawn. I suppose design has to be based around the premise of "nothing must cause the engine to stop, if possible" rather than improving fuel efficiency and trusting in the AA!

Thanks for the help.

[Mark1234]

Nope, that's not what I'm saying

Carb icing is ice, and occurs when moisture (humidity) in the air is turned into water ice in the carb - most likely the venturi, and/or throttle plate, which restricts the breathing.

That happens because the temperature in the throat is a lot lower than ambient, THAT is driven by the venturi pressure drop, and also evaporative cooling of the fuel - it requires energy to transform the liquid into a vapour (think water spray in high performance automotive applications).

Aero engines are generally pretty crude, and use large capacities to achieve low power outputs. They have to operate over a far wider range of conditions, and are they're deliberately very low stressed; but probably the main reason is it costs a lot to certify, and there are liability concerns doing anything radical. Therefore we're mostly stuck with technology from the 60's.

[IanPZ]

Ooooohhhhh

So carb icing IS about restricting airflow, but its the combination of the pressure drop AND the evaporative cooling that pushes the humidity over the edge and causes freezing. And by taking the fuel away, or squirting it in further down the intake side, you only get one effect (pressure drop) and so icing is less likely.

Is that right?

[n5296s]

I've heard of a mod for Rotax engines that takes warm fluid from the cooling system and makes it circulate around the carburettor body, so that the body gets the heat (and so preventing any ice accretion), instead of the air/fuel mixture.
Now that's a smart solution, always-on and costing no power.

In the Cessna TR182 (I think all the R182s) the carb is at the back of the engine. Mine has a carb temp gauge and it is never below 20 degrees. So in effect this is what happens. Don't know why other designs don't do it, it seems so common-sensical to me.

[Mark1234]

Ian, that's kinda it. As I said, I don't think there's a venturi in an injected (aero) engine either, but I don't take them apart, only sit behind them..