Wind Tunnel: Lift Without Solving the Navier-Stokes Equations

Potential flow: a cheap approximation that works remarkably well

The full equations governing air flow around an aerofoil, the Navier-Stokes equations, are nonlinear and expensive to solve. Far from the surface, though, viscosity barely matters, and the flow can be treated as inviscid, incompressible and irrotational — a potential flow. That single simplification turns a system of nonlinear PDEs into one linear equation, Laplace's equation for a velocity potential, which can be solved by superimposing simple building-block flows: a uniform stream, sources, sinks and vortices.

Streamlines, stagnation points, and the pressure coefficient

Streamlines trace the path a massless tracer particle would follow through the steady flow field, and they never cross. Where the oncoming stream splits to pass over and under the body, the local velocity drops to zero at a stagnation point. Bernoulli's equation, valid along any streamline in this inviscid flow, converts local speed directly into local pressure through the pressure coefficient:

Cp = 1 − (v / U)²

U = free-stream velocity far upstream
v = local velocity at a point on/near the surface
Cp = 1   at a stagnation point   (v = 0, maximum pressure)
Cp → −∞  where v is largest (though real flow limits this)

Where the flow speeds up — over the curved upper surface of a cambered aerofoil, or around a cylinder's shoulders — Cp goes negative, meaning pressure below the free-stream value. That pressure difference between the fast, low-pressure upper surface and the slower, higher-pressure lower surface, integrated over the whole body, is the source of lift.

Circulation and the Kutta-Joukowski theorem

A symmetric cylinder in a uniform stream produces zero net lift by symmetry — unless a circulation Γ is superimposed on the flow, a rotational component that speeds up flow on one side and slows it on the other. The Kutta-Joukowski theorem gives the resulting lift per unit span exactly:

L' = ρ · U · Γ

L' = lift per unit span
ρ  = fluid density
U  = free-stream speed
Γ  = circulation, set by the KUTTA CONDITION:
     the flow must leave the trailing edge smoothly, with
     no infinite velocity around the sharp edge

The circulation Γ is not a free parameter — it is fixed uniquely by the requirement that flow leaves the sharp trailing edge smoothly rather than wrapping impossibly around it (the Kutta condition). This is the physically real explanation of lift, and it exposes the popular "equal transit time" myth as false: nothing requires air splitting at the leading edge to reunite simultaneously at the trailing edge, and in real flow it generally does not.

Angle of attack, and where potential flow runs out

Tilt the aerofoil into the stream (increase the angle of attack) and the Kutta condition demands more circulation to keep the trailing-edge flow smooth, so lift rises — roughly linearly, in thin-aerofoil theory, at about 2π per radian of angle of attack for a thin symmetric section. But potential flow assumes the boundary layer stays attached, and real viscous flow eventually cannot keep up: past a critical angle the adverse pressure gradient on the upper surface overwhelms the boundary layer, it separates, and the neat circulation pattern collapses. That is stall — a fundamentally viscous phenomenon that inviscid potential-flow theory, on its own, cannot predict; it can only tell you the (wrong) answer keeps climbing forever.

Frequently asked questions

Try it live

Everything above runs in your browser — open Wind Tunnel and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.