The pressure drop in the venturi is proportional to air density and the fuel is at ambient pressure in the float chamber, so I would expect the fuel flow to reduce proportionally with density, and that response to preserve the fuel-air ratio over changing altitude.

But in practice that does not seem to be the case. Proper response to altitude requires additional modification that most carburetors don't have, so the pilot usually has to lean out the engine as they climb. What am I missing here?

(Update) More specifically, I would expect that at the same RPM, the volume flow rate will be the same—because the engine pulls in its displacement per revolution. Now velocity in the venturi $v$ is just

$$ v = \frac{\dot V}{A} $$

Where $\dot V$ is the volume flow rate and $A$ is the cross-section of the venturi. So it will also be the same independent of altitude. Since dynamic pressure

$$ p_d = \frac 1 2 \rho v^2 $$

And that is also the pressure that pulls in the fuel (when the float chamber is open to ambient pressure). Substituting mass flow

$$ \dot m = \rho v $$ $$ p_d = \dot m \frac 1 2 v $$

and as long as $v$ is mostly constant,

$$ p_d \sim \dot m $$

This still leaves open the dependence of fuel flow on $p_d$. If the relation is reasonably close to linear, it should mean the venturi mixes properly by mass. I can see a reason why higher pressure should cause less than linear increase in fuel flow, but not much why it should cause more than linear increase in fuel flow—but that is what the actual behaviour would need.

Note: Did you notice the extra $v$ conspicuously missing from the last equation? That suggests the mixture should become richer with throttle opening—which is actually the reason there are extra fuel jets after the throttle that pull in extra fuel when the throttle is at idle and slightly above. It might also explan what is going on here, but would require explaining why the speed should increase when the density is lower.

  • $\begingroup$ The fuel could be going through an engine driven fuel pump. Then the fuel might not be at ambient pressure. $\endgroup$
    – Jan
    Commented Jan 23, 2019 at 7:20
  • 1
    $\begingroup$ @Jan is on right track, fuel orifice is designed to add excess fuel even below sea level. Fuel lean control valve pinches off fuel flow VOLUME, to maintain proper mixture at all altitudes. Air density changes with altitude, fuel density does not. It has to be adjusted. $\endgroup$ Commented Jan 23, 2019 at 14:30
  • $\begingroup$ Possible duplicate question: aviation.stackexchange.com/questions/24097/… $\endgroup$ Commented Jan 23, 2019 at 14:57
  • $\begingroup$ @MikeSowsun, no, that is not a duplicate. The question asks about what the mixture should be. $\endgroup$
    – Jan Hudec
    Commented Jan 23, 2019 at 19:15
  • 1
    $\begingroup$ I figured out the answer and posted it at engineering.stackexchange.com/a/52181/27170. $\endgroup$ Commented Aug 18, 2022 at 15:38

3 Answers 3


The simple reason is that high altitude changes the density of air, but not the density of fuel.

Most simple carburetors use a simple venturi design that mixes air and fuel by VOLUME.

But, the correct air/fuel mixture must be based on MASS, not Volume.

If the air/fuel Volume is correct at sea level, it will be too rich at altitude. This is because a given volume of air will have less Mass at high altitude, while a given volume of fuel will have the same Mass at high altitude.

Aircraft carburetors have a mixture control to "lean" the volume of fuel to the volume of air, and are then able to maintain the correct air/fuel ratio at all altitudes.

  • 4
    $\begingroup$ But why does it mix by volume? As the same volume of air flows by (if rpm stays the same, it should), it generates less pressure difference (because that is proportional to density) and therefore pulls in less fuel—by both volume and mass, because density of that didn't change. But it pulls less, proportionally by density of the air, therefore mixing properly by mass, not by volume. I understand there is an error in that reasoning, because it does not match the reality—but I am asking where specifically. $\endgroup$
    – Jan Hudec
    Commented Jan 23, 2019 at 19:40
  • 1
    $\begingroup$ It is probably simple physics, but not exactly obvious. In either case, the Venturi effect is simply the kinetic term of Bernoulli's equation, so density and square of velocity, but I don't see a reason for the velocity to be higher (piston engines are constant volume, which suggests it should be the same). $\endgroup$
    – Jan Hudec
    Commented Jan 23, 2019 at 21:17
  • 2
    $\begingroup$ This answer is incorrect. Carburetors are more complex than mixing air and fuel by volume. There is a pressure differential from the float bowl to the venturi and this pushes fuel out the jets. As the plane climbs, the pressure in the float bowl reduces, which means less fuel is ejected per stroke. This in turn means less fuel is ejected per volume of air. $\endgroup$ Commented Aug 18, 2022 at 13:22
  • 1
    $\begingroup$ I figured out the answer and posted it at engineering.stackexchange.com/a/52181/27170. $\endgroup$ Commented Aug 18, 2022 at 15:38

Leaning is not done to counter carburetors that do not "preserve the fuel-air ratio over changing altitude".

The [operationally] ideal stoichiometry (air:fuel ratio) is not fixed, it depends on what is being sought. Leaning is done because it provides better fuel consumption when less than full power is needed (see graph below), which coincides with cruise.

enter image description here

Rich mixtures also burn cooler, so leaning becomes practicable and easier on the engine in the colder air of the cruise altitudes.

  • $\begingroup$ But do carburetors preserve the fuel-air ratio over changing altitude? At least the simulator thinks they definitely don't. $\endgroup$
    – Jan Hudec
    Commented Jan 23, 2019 at 21:33
  • $\begingroup$ @JanHudec - My understanding is that main metering circuits "provide a constant fuel–air ratio at any engine speed and condition above the engine’s idle speed". $\endgroup$
    – user14897
    Commented Jan 23, 2019 at 21:51
  • $\begingroup$ It certainly would make sense. Because, after all, the pressure difference does depend on density. $\endgroup$
    – Jan Hudec
    Commented Jan 23, 2019 at 22:02
  • $\begingroup$ @JanHudec - That's the way I see it too. If the differential pressure were to be fixed at high altitude, then the fuel mass would come into play, but I don't see how. $\endgroup$
    – user14897
    Commented Jan 23, 2019 at 22:18
  • $\begingroup$ (i.sstatic.net/7kplh.jpg) That isn't entirely true. When under load? Yes. When there is little/light load an AFR of 16:0 - 17:0 will burn cooler than 12:0 - 15:0 14 - 15 AFR is where the hottest EGTs will be always. $\endgroup$
    – udidwht
    Commented Oct 30, 2019 at 21:09

It's not pressure that forces fuel into the venturi as much as the vacuum pressure sucking it out. Fuel is pumped or gravity fed to the carb, but then it's metered into the bowl by the float. Fuel sits in the bowl before getting pulled into the jets and then into the venturi. There is air in the bowl, sure, and as that air pressure decreases with altitude there is a bit of a decrease in the pressure behind the fuel, however the part you may be missing is that air can mix with a much higher ratio of fuel than what is ideal for an engine. Airplane engine carbs are generally tuned so at sea level they deliver slightly over the ideal ratio, this is done by restricting the fuel flow. If it wasn't restricted the engine would run rich on the ground.

As density altitude increases the amount of air going through the venturi decreases, but the air still can hold more fuel so the mixture richens even though there's less pressure.

  • 1
    $\begingroup$ What amount of fuel the air “can” hold is not really relevant. The amount it pulls in is. The amount it pulls in depends on the pressure drop in the venturi. But this pressure drop decreases linearly with the density of that air. So why does the amount of fuel pulled in not decrease the same way? $\endgroup$
    – Jan Hudec
    Commented Jan 23, 2019 at 19:43

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .