Case Study 2 — Circulation, Stokes' Theorem, and Why Wings Fly

Field: Aerodynamics / fluid dynamics. How Stokes' theorem connects the swirl inside a flow to the circulation around a wing, and how that circulation becomes lift.

The mystery of lift

A 400-ton airliner stays aloft on two thin wings. The popular "longer path on top" story is wrong, but the real explanation is genuinely beautiful and rests squarely on the mathematics of this chapter. The aerodynamicist's quantitative handle on lift is a single number called the circulation around the wing, and circulation is exactly the left-hand side of Stokes' theorem (§37.2):

$$\Gamma = \oint_{C}\mathbf{u}\cdot d\mathbf{r},$$

where $\mathbf{u}$ is the air's velocity field and $C$ is a closed loop encircling the wing. Stokes' theorem ties this boundary circulation to the vorticity — the curl of the velocity — flowing through any surface that $C$ bounds:

$$\oint_C \mathbf{u}\cdot d\mathbf{r} = \iint_S(\nabla\times\mathbf{u})\cdot d\mathbf{S}.$$

The quantity $\boldsymbol{\omega} = \nabla\times\mathbf{u}$ is the flow's vorticity, its local rate of spin (drop a paddle wheel in and watch it turn, §37.2). Stokes' theorem says: the net swirl you feel walking once around the wing equals the total vorticity enclosed. That sentence is the entire bridge between the microscopic spinning of air near the wing and the macroscopic circulation that lifts the plane.

Step 1 — Circulation from vorticity

Consider a simplified two-dimensional flow around a wing cross-section (an airfoil), with velocity field $\mathbf{u} = \langle u(x,y),\, v(x,y),\, 0\rangle$. The vorticity points out of the plane:

$$\nabla\times\mathbf{u} = \left\langle 0,\ 0,\ \frac{\partial v}{\partial x} - \frac{\partial u}{\partial y}\right\rangle.$$

Notice this is the same $Q_x - P_y$ that appeared in Green's theorem (Chapter 35) — and indeed, for a flat 2D region Stokes' theorem is Green's theorem (§37.2). So the circulation around a loop $C$ enclosing the airfoil equals

$$\Gamma = \iint_S\left(\frac{\partial v}{\partial x} - \frac{\partial u}{\partial y}\right)dA,$$

the integral of all the vorticity packed into the surface $S$ spanning $C$. In an ideal (irrotational) flow far from the wing the vorticity is zero, so by Stokes' theorem the circulation around any loop in that region is zero — air far away does no net swirling. All the vorticity is concentrated in a thin boundary layer clinging to the wing's surface and in the wake trailing behind it. The wing is, in effect, a vorticity factory; Stokes' theorem lets us measure its total output by walking one loop around the outside.

Step 2 — A worked circulation

Make this concrete with a model vortex flow. A clean idealization of the air circulating around a wing is the field

$$\mathbf{u} = \frac{\Gamma_0}{2\pi}\,\frac{\langle -y,\, x,\, 0\rangle}{x^2 + y^2},$$

a swirl whose strength falls off with distance — the potential vortex of classical aerodynamics. Compute the circulation around a circle $C$ of radius $a$ centered at the origin. Parametrize $C$ as $\mathbf{r}(t) = \langle a\cos t, a\sin t, 0\rangle$, so $\mathbf{r}'(t) = \langle -a\sin t, a\cos t, 0\rangle$. On the circle $x^2 + y^2 = a^2$, the field is

$$\mathbf{u} = \frac{\Gamma_0}{2\pi a^2}\langle -a\sin t,\ a\cos t,\ 0\rangle.$$

The dot product $\mathbf{u}\cdot\mathbf{r}'$ is

$$\frac{\Gamma_0}{2\pi a^2}\big(a^2\sin^2 t + a^2\cos^2 t\big) = \frac{\Gamma_0}{2\pi a^2}\cdot a^2 = \frac{\Gamma_0}{2\pi}.$$

Integrate over $t\in[0,2\pi]$:

$$\Gamma = \oint_C\mathbf{u}\cdot d\mathbf{r} = \int_0^{2\pi}\frac{\Gamma_0}{2\pi}\,dt = \Gamma_0.$$

The circulation is $\Gamma_0$, independent of the radius $a$. This is the aerodynamic twin of the radius-cancellation we saw in Gauss's law (Case Study 1): the answer depends only on what is enclosed, not on the size of the loop.

Here is the Stokes-theorem punchline, and the §37.5 pitfall in action. Away from the origin this vortex field is irrotational — you can check that $\partial_x v - \partial_y u = 0$ for $(x,y)\neq(0,0)$. So if the vorticity were zero everywhere, Stokes' theorem would force $\Gamma = 0$. But $\Gamma = \Gamma_0 \neq 0$. The resolution: the vorticity is not zero everywhere — it is concentrated in a singularity at the origin (a real wing's tightly wound boundary-layer vorticity, idealized to a point). Just as the point charge produced flux in Case Study 1, the point vortex produces circulation. Stokes' theorem reports the hidden source: $\nabla\times\mathbf{u} = \Gamma_0\,\delta(\mathbf{r})\,\hat{\mathbf{z}}$. The circulation around any loop enclosing the vortex is $\Gamma_0$; around any loop not enclosing it, zero.

Step 3 — From circulation to lift

The payoff is the Kutta–Joukowski theorem, the foundational result of airfoil theory: the lift per unit span on a wing in a flow of density $\rho$ and free-stream speed $U$ is

$$L' = \rho\,U\,\Gamma.$$

Lift is directly proportional to circulation. No circulation, no lift. A wing generates lift precisely by establishing a nonzero $\Gamma$ around itself — the air is nudged into circulating, faster over the top and slower underneath, and Stokes' theorem guarantees this circulation equals the total vorticity the wing has shed into the flow. Increase the angle of attack and you increase the circulation (up to the stall point, where the boundary layer separates and the vorticity bookkeeping breaks down catastrophically). Designers shape airfoils, choose angles of attack, and add flaps all in service of controlling one number: $\Gamma$.

The deepest part is why the circulation takes the specific value it does. An ideal-flow model alone is underdetermined — many circulations satisfy the equations. The Kutta condition (the flow must leave the sharp trailing edge smoothly, not whip around it infinitely fast) selects the physical value. Vorticity of the opposite sign is shed at the trailing edge as a starting vortex when the wing begins to move; by Stokes' theorem applied to a huge loop enclosing both the wing and the starting vortex — where the total enclosed vorticity must stay zero (Kelvin's circulation theorem) — the wing is left with an equal and opposite bound circulation. The starting vortex you can sometimes see shed behind a wing as it accelerates is the visible receipt for the lift the wing is now generating.

Why this matters

Every aircraft design pipeline computes circulation. In low-fidelity tools (vortex-lattice and panel methods), the wing is replaced by a sheet of vortex elements and the circulation distribution is solved for directly — Stokes' theorem is the accounting identity that links each vortex element's strength to the velocity field it induces. In high-fidelity computational fluid dynamics, the Navier–Stokes solver tracks vorticity transport across millions of cells, and the discrete curl-circulation identity (§37.11) keeps the bookkeeping exact. From a hand calculation on the back of an envelope to a supercomputer simulation of a full aircraft, the link between swirl inside and circulation around is Stokes' theorem.

It is a fitting summit. Case Study 1 used the Divergence theorem to explain how charge sources a field; this one uses Stokes' theorem to explain how a wing sources circulation. Two theorems, two halves of vector calculus, both of them the single slogan of §37.1 — boundary integral equals region integral of a derivative — and both of them, quite literally, keeping the modern world powered and aloft.

Discussion Questions

1. Radius independence. We found $\Gamma = \Gamma_0$ regardless of the loop radius $a$. Connect this to the surface-independence of Stokes' theorem (§37.2, §37.9): why does the enclosed vorticity, not the loop's size, determine the circulation?

2. The hidden singularity. The model vortex is irrotational away from the origin yet has nonzero circulation. Explain how this parallels the point charge of Case Study 1, and why neither the Divergence theorem nor Stokes' theorem can be applied across the singular point (§37.5).

3. Sign and orientation. If you traversed the loop $C$ clockwise instead of counterclockwise, what would happen to the computed $\Gamma$, and what would that imply physically about the direction of lift? Tie your answer to the orientation rule (§37.2 Common Pitfall).

4. From integral to differential. Faraday's law (§37.7) and the circulation–vorticity link have the same mathematical shape. Write both as "$\oint(\text{field})\cdot d\mathbf{r} = \iint(\text{curl})\cdot d\mathbf{S}$" and identify what plays the role of the field and its curl in each.

Annotated Reading

• Anderson, Fundamentals of Aerodynamics, Ch. 3–4. The standard undergraduate aerodynamics text; develops circulation, the Kutta–Joukowski theorem, and the Kutta condition with Stokes' theorem as the connective tissue. Anderson is especially good on the starting-vortex argument.
• Stewart, Calculus: Early Transcendentals, §16.8 (Stokes' Theorem). The mathematical foundation here, including the circulation interpretation of the curl. See appendices/appendix-h-stewart-chapter-mapping.md.
• OpenStax, Calculus Volume 3, §6.7 (Stokes' Theorem). Free, with the fluid-circulation reading of $\oint\mathbf{u}\cdot d\mathbf{r}$ made explicit. See appendices/appendix-i-openstax-chapter-mapping.md.
• Acheson, Elementary Fluid Dynamics, Ch. 4–5. A mathematician-friendly treatment of vorticity, Kelvin's circulation theorem, and the irrotational-flow idealization used in Step 2.
• Chapter 38 of this book. Where this circulation result and the electric flux of Case Study 1 become two instances of the single equation $\int_{\partial M}\omega = \int_M d\omega$.