Lift Curve — CL vs Angle of Attack
The curve records itself two ways: pause at any angle for about half a second and a point is measured there (points within 2° of each other join into a line), or press “Trace lift curve” to sweep the whole range automatically. Measurements wait for the flow to settle — lift sampled mid-change would be meaningless.
What You're Looking At
This is a two-dimensional wind tunnel running live in your browser. Air flows from left to right past a wing cross-section (an airfoil). The color shows how fast the air is moving — deep blue is slow or stagnant air, green is about free-stream speed, and orange/red is faster than the tunnel speed. The white streaks are individual parcels of air being carried along by the flow.
Notice that at a positive angle of attack the air over the top of the wing speeds up (warm colors) while the air beneath slows down. Switch the field view to Pressure and you'll see the companion effect: fast air over the top is at lower pressure, slower air below is at higher pressure. That pressure difference, summed over the whole surface, is lift — and it's exactly equal to the downward push the wing gives the air (watch the wake get deflected downward). Bernoulli's principle and Newton's third law aren't competing explanations; they're two views of the same flow. (And no — the air over the top doesn't speed up because it “has to meet up” with the air underneath. It actually arrives at the trailing edge earlier. The equal-transit-time story you may have heard is a myth.)
Reading the NACA code
NACA stands for the National Advisory Committee for Aeronautics, the U.S. flight research agency (1915–1958) that was folded into the newly created NASA. It tested hundreds of wing cross-sections in real wind tunnels and published them as numbered families, and airfoils are still named by its catalog today. The four digits describe the shape. For 2412: maximum camber of 2% of the chord, located 40% back from the leading edge, with a maximum thickness of 12% of the chord. The Cessna 172 you may be training in uses the NACA 2412 at the wing root. Try 0012 — zero camber, perfectly symmetric, like most tail surfaces — and notice it makes no lift at 0° angle of attack, while the cambered 2412 still does.
The critical angle of attack
Slowly raise the angle of attack and watch the lift coefficient climb — until the smooth flow over the top can no longer stay attached. It separates: the wake breaks into large shed vortices, the lift starts jumping around violently (watch the CL readout shake and the green band on the graph widen), and the red annunciator lights. That unsteadiness is buffet — the same aerodynamic shaking you feel in the airframe just before and during a real stall, and the reason the amber caution comes on first. Switch on the 🌬 Wind sound and you can hear it too: the airflow noise turns ragged and a low rumble builds before the stall light ever comes on — exactly the cue your instructor wants you to catch in the airplane. In full-scale, three-dimensional air, this separation is also what makes average lift break downward at the critical angle; in a 2D tunnel at low Reynolds number the shed vortices keep generating chaotic lift, so you'll see the buffet band explode more dramatically than the mean line drops.
The checkride point: a wing always stalls at the same critical angle of attack — regardless of airspeed, attitude, or weight. You can stall a wing pointed at the ground with the throttle wide open. Airspeed is only a stand-in; angle of attack is the real variable, and this tunnel lets you fly it directly.
Use Trace lift curve to have the tunnel sweep the angle of attack automatically and draw the whole CL–α curve, with the buffet band showing exactly where attached flow gives way to separation. Compare the linear range with the dashed thin-airfoil theory line (slope 2π per radian): real, viscous air runs shallower than the ideal, then quits flying smoothly altogether.
Honest limits of this model
This is a lattice-Boltzmann fluid simulation at a very low Reynolds number — on the order of a thousand, versus roughly five million for a Cessna 172 wing in cruise. In air this viscous, camber does less, drag numbers run high, and the stall shows up as heavy buffet and vortex shedding rather than the crisp lift break in your textbook (a full-scale wing typically reaches critical angle of attack around 16–18°). The flow patterns, the pressure picture, and the shape of the linear lift range are qualitatively right and genuinely instructive; the exact numbers are not for flight planning. It's also strictly 2D — no wingtip vortices, no induced drag from finite span.
Inspired by crgimenes' desktop wind tunnel; simulation method after the classic lattice-Boltzmann demonstrations of Daniel Schroeder. Built for PilotDECODER.