Almost everything confusing about flight clears up once that sentence lands. This page works through the four forces, what an aerofoil is really doing, why the same wing behaves differently at the same speed, and how six wing shapes solve six different problems.
Straight and level at constant speed means all four are cancelling out. Every manoeuvre an aircraft performs is a deliberate imbalance between them, held only as long as it is wanted.
Generated by the wing turning a mass of air downward. By Newton's third law the air pushes the wing up by exactly as much. It acts perpendicular to the oncoming airflow, not to the ground — which is why a banked turn loses altitude unless you pull.
Acts straight down through the centre of gravity, and unlike the others you cannot switch it off. Where it sits along the fuselage decides how stable the aircraft is, which is why loading is calculated before every flight.
Forward force from the propeller or jet. A propeller is itself a rotating wing, generating lift in the forward direction; a jet throws mass backwards. Different machinery, identical principle.
Two kinds, pulling in opposite directions as speed changes. Parasite drag — friction and shape — grows with speed. Induced drag, the price of making lift, shrinks with speed. Their sum has a minimum, and that speed is where an aircraft is most efficient.
The explanation most people were given at school is wrong, and the correct one is simpler.
The story about air splitting at the leading edge and having to "meet up again" at the back is not true. Nothing requires the two flows to arrive together, and measurements show that the air over the top gets there considerably sooner. A wing built on that reasoning would not fly.
What actually happens is that the wing deflects air downward. The curved upper surface pulls the flow around and past the trailing edge on a downward slant; the lower surface pushes air down more directly. A large mass of air leaves the wing moving downward every second, and the reaction to that is an upward force. That reaction is lift.
How hard the wing turns the air depends chiefly on one number: the angle between the wing's chord line and the air actually coming at it. That is the angle of attack, and it is not the same as the aircraft's nose attitude. A machine can be pointed at the sky and have a small angle of attack, or be pointed at the ground and have a large one.
Camber, the asymmetric curvature of most wing sections, simply means the wing produces useful lift at zero angle of attack. It is a convenience, not a requirement. A flat plate held at an angle generates lift perfectly well, which is why paper aeroplanes work and why aerobatic aircraft use symmetrical sections that fly identically upside down.
Speed appears nowhere in this sequence. A wing can be stalled at any speed at all, in any attitude, and it will unstall the moment the angle is reduced — which is why the recovery is always the same movement.
Cambered sections still make lift here. Flow follows both surfaces cleanly.
Near the efficient range. Lift is high, drag still moderate.
Maximum lift. Beyond this angle the upper flow can no longer stay attached.
Flow breaks away into turbulence. Lift falls sharply, drag rises, the nose drops.
The recovery follows directly: lower the angle of attack. Not add power, not pull up — reduce the angle, let the flow reattach, and the wing works again. Every stall-recovery drill ever taught is that one action with details attached.
All three run through the centre of gravity and all three meet there at right angles. Each has one primary control, and each control works by changing the camber of a small part of a fixed surface.
Nose up and down about the lateral axis, running wingtip to wingtip. The elevator at the tail pushes the tail down or up, levering the nose the opposite way.
Banking about the longitudinal axis, nose to tail. Ailerons move in opposite directions — one wing gains lift, the other loses it, and the aircraft rolls.
Nose left and right about the vertical axis. Contrary to instinct the rudder does not turn the aeroplane — it keeps the nose aligned with the flight path while the wings do the turning.
Seen from above, the outline of a wing tells you most of what the designer was optimising for. None of these is better than the others; each is a different answer to a different question.
Constant chord from root to tip. Cheap to build and beautifully predictable: it stalls at the root first, so the ailerons out at the tips keep working while the centre gives up. That is why trainers use it.
Chord narrows towards the tip. Lighter and structurally more efficient than a rectangle, with less drag, at the cost of a slightly sharper stall. The commonest compromise in aviation.
The theoretical optimum for induced drag, because lift falls off along the span in exactly the right curve. Almost nobody builds them any more — the aerodynamic gain is small and the manufacturing cost is not.
Angling the wing back delays the drag rise that appears as an aircraft approaches the speed of sound. The penalty is poor low-speed behaviour, which is why swept-wing aircraft carry elaborate flaps and slats.
A triangle from root to tip. Enormous internal volume, very strong, and it copes with a wide speed range — but it needs a high nose attitude on approach, which is why delta aircraft land looking like they are climbing.
Very long, very narrow. Induced drag falls as span rises, so a glider chasing the best possible glide angle stretches its wings as far as structure allows. Roll response is slow and the wings visibly flex.
Numbers on a specification sheet are hard to feel. Drawn against one another they are not.
Bars drawn to a common scale. A widebody's wing is longer than the entire first powered flight, which covered roughly thirty-seven metres.
Every flight is the same shape, whether it lasts twenty minutes or twelve hours. Only the cruise segment stretches.
Accelerate until the wing can carry the weight, then raise the nose to set an angle of attack. The aircraft leaves the ground because the wing is ready, not because the pilot pulled.
Excess thrust converted into height. Climb rate is not about lift exceeding weight — in a steady climb it does not — but about having power left over after overcoming drag.
The efficient middle, flown as high as conditions allow because thin air means less drag for the same speed. Everything is balanced and nothing much changes.
Height traded back for distance at a controlled rate, flaps extended to fly slowly, and a final flare that arrests the descent just as the wing runs out of angle to give.