LeSving wrote:
Being prone to ice up at lower throttle setting is function of the the butterfly valve position (part closed) and the icing air is moving slower across it, more time to cool it down and hit it with droplets.
So you mean that the temperature drop due to pressure drop over the valve has nothing at all to do with it?
On starting our O200, it often runs rough, and stops. Ice can be seen on the carb outside. After a few starts, it goes O.K. In these conditions, the DR1050 manual approved by the French DGAC says to start with carb air hot, but initially the exhaust jacket will have no heat.
I seem to get carb ice when opening the throttle for take-off at temperatures too low for this to happen, when the runway is Konsin treated. I now use carb heat for the take-off run in these conditions,
I suggest anyone answering Airborne_again’s question, especially LeSving, read up on the gas laws, especially Gay-Lussac, and reflect on why the temperature at FL180 is not quite the same as at sea level.
Cobalt wrote:
I suggest anyone answering Airborne_again’s question, especially LeSving, read up on the gas laws, especially Gay-Lussac, and reflect on why the temperature at FL180 is not quite the same as at sea level.
and the relevance to the fluid dynamics and thermodynamics of a carb is?
Airborne_Again wrote:
So you mean that the temperature drop due to pressure drop over the valve has nothing at all to do with it?
It’s minimal, if at all compared to the evaporation. Depending on the actual venturi, the shape, efficiency, the temperature could in principle also increase. The pressure drop in a perfect venturi is an isentropic process. This means, for the temperature to decrease (due to lower pressure), the condition after the venturi has to be equal to the condition before the venturi (the energy has to be recovered). Just injecting fuel in there is more than enough to offset this recovery, not to mention the less than perfect shape of such a venture on a typical Lycoming or Continental engine. Things become different at M >= 1, but this is not very relevant for a carb in a SEP.
Evaporation is an endoterm process. It “stores” energy in the form of heat. To release that energy, the evaporated fuel has to condensate again. That will not happen unless the temperature is decreased way below. This means the temperature will remain low, until the mix is heated by the surroundings, in effect it is a refrigerator. This is the very thermodynamic principle for a refrigerator (the cooling part). Evaporation takes heat from the surroundings, condensation releases heat to the surroundings.
The temperature in the throat of a venturi is given as:
T2 = T1 + (V1^2 – V2^2)/(2Cp)
Cp for air is about 1000 J/kgK. Then if you are able to fully recover the energy after the throat (fully isentropic venturi), the temperature difference is abut 1 degree C when V1 = 25 m/s and V2 (the throat) = 50 m/s
What happens to pressure after the butterfly, or whatever is used to restrict the airflow? And what is the temperature of that air?
Why is icing more severe at low power settings, although the amount of fuel per unit of air is constant if the mixture is constant, and the pressure and temperature in the venturi are lower at higher power settings?
How do you think a refrigerator work? The venturi principle? It’s rather amusing and surprising how the most elementary of things of thermodynamics, like the workings of a fridge, is completely lost on people, even as a physical principle.
Look at this sketch:
This is a typical icing situation in a Lycoming. The ice is on the butterfly valve and upstream the butterfly valve. That is the only place ice can form. Downstream the valve, the manifold is heated by oil, all ice will melt. At low power, the air flow through the nozzle is low, the “venturi effect” is minimal, yet this is the most dangerous condition for icing, why? Clearly it has nothing to do with the venturi effect, or this would be more severe, the more power is used. Therefore, some other effect is the cause. That effect is the evaporation of fuel.
Evaporation starts the very instant fuel and air meet, and continues until the air cannot hold any more vapor, or all the vapor is evaporated. At high power, the air moves fast, and most of the evaporation may even happen after the butterfly valve (were there is no danger due to oil heating). When passing the butterfly valve, the bulk of the mix may not have cooled down enough to cause icing. At low power, the air moves much slower, and therefore the evaporation and cooling take place earlier. The butterfly valve is also almost closed, and is hit by small droplets that evaporates directly on the valve itself. It becomes a freezing element of a fridge. Soaked in fuel, and a constant supply of fresh air for evaporation. It cools down, and condensed air droplet from the cooling air freezes on it. Also, because the valve is almost closed, it doesn’t take much to freeze it shut. Maybe more dangerous is ice in the venturi itself. This can stop the supply of fuel, or it can reduce the throat area so the mixture get’s too rich for the engine to run.
That’s how I see.it.
It would be be interesting to enrichen mixture on a given engine/carb type after encountering carb ice in cruise flight, watch to see if it affects rpm (assuming a FP prop) and in that way check whether the carb ice is blocking the fuel jet or just the air flow.
LeSving wrote:
Clearly it has nothing to do with the venturi effect
No, clearly it does not, because a carburetor is not a venturi. It is a pressure reducing valve. The engine creates suction which will cause a marked pressure drop over the carburetor whenever you have partial throttle. With a pressure drop, the temperature will also drop. This pressure drop is the common explanation to carb icing. I have still to see a convincing argument that this is not the main cause of carb icing.
I don’t believe in your theory of fuel droplets striking the butterfly valve because the smaller the valve opening, the fewer droplets will strike the valve as the flow will be lower.
Leaning the mixture reduces rough running when using carb heat to combat carb ice. Carb heat leads to a richer mixture due to warmer air.
Airborne_Again wrote:
I don’t believe in your theory of fuel droplets striking the butterfly valve because the smaller the valve opening, the fewer droplets will strike the valve as the flow will be lower.
It’s not “my” theory (look at the picture, and the youtube movie in this thread). Anyway, you have to include mass balance and heat transfer also. The fuel itself could be anywhere from -40 to +60 degrees (at least). Lets say it is +40, then too much on the butterfly valve would warm it sufficient to not make it subzero temp even with icing droplets impinging it. While just the right amount will cool it just right to make it freeze.
Airborne_Again wrote:
This pressure drop is the common explanation to carb icing. I have still to see a convincing argument that this is not the main cause of carb icing.
Do the math then ! Yes, after the butterfly valve the temperature reduction could theoretically be substantial due to expansion of the air, maybe 40-50 deg C at idle, IF we are talking about dry air, ie ideal isentropic expansion. But we do not have ideal, dry air, very far from it. We have air that is saturated with fuel and water (when at idle). The moment the pressure falls, the temperature decreases, and water and fuel starts to condensate. Water even has to go from vapor to liquid to solid. These are exotherm reactions giving of heat, the same amount that is “absorbed” during evaporation. The moment the temperature decreases, condensation of fuel will stop further reduction of temperature. Maybe there could form some fancy hydrocarbon “ice” due to this?, who knows. IMO it’s on higher power settings that ice could theoretically form in the manifold after the butterfly valve, due to the fact that most of the evaporation happens there. One thing is that the conditions has to be right for ice to form, another thing is the ice particles has to stick on something.
The combined physics here is very complex, so it’s better too look at what works. Lycoming use heated manifold combined with carb heat. Rotax use a Bing carb that by design (or luck?) is very resistant to carb ice, but is often combined with heating elements and/or slightly heated air. Limbach and Sauer use Bing + heated manifold. Sonex (Aerovee) and Revmaster use sliding carb (almost 100% carb ice resistant by design) combined with heated (by air) manifold, and carb heat if you feel for it. In the old days, on cars, we used ant-freeze liquid in the tank. Depending on formula, it worked by lowering the freezing point of the water, or wetting the insides of the carb with a “anti stick” coating.
By far the best solution is fuel injection, even a basic mechanical one, and here comes the kicker A standard fuel injection (Lycoming/Continental) is completely free of icing problems. The funny thing though, they all have a butterfly valve and the same manifold pressure for various power settings. The have the same venturi for metering of fuel. It’s only on diesel engines and very modern, direct (into the combustion chamber) digital EFI systems, like that Evinrude system, that you can also take out the butterfly valve. How is this possible without icing problems ? Obviously, the fuel sprayed out of the carb is a key element Get rid of that fuel evaporating all over the place, and the problem is solved.
