minimal surface

Dip a bent wire loop into soapy water and pull it out: the film that spans the loop is a minimal surface. Surface tension pulls the film taut and it settles into the shape of least area among all surfaces with that same boundary. The variational question is: given a fixed boundary curve (the wire), which surface stretched across it has the smallest area? The soap film solves it physically, and the calculus of variations solves it on paper.

Area is itself a functional. For a surface written as a height z = u(x, y) over a region, the area is the double integral of sqrt(1 + u_x^2 + u_y^2) dx dy, where u_x and u_y are partial derivatives. Minimizing this with the boundary held fixed gives an Euler-Lagrange equation — now a partial differential equation, the minimal surface equation — and its geometric meaning is striking: a minimal surface has zero mean curvature everywhere. At every point the surface curves up in one direction exactly as much as it curves down in the perpendicular direction, so the two curvatures cancel. That is why soap films look saddle-shaped, never bulging like a balloon (a balloon has pressure inside and nonzero mean curvature).

The simplest nontrivial example is the catenoid, the surface of revolution you get by spinning a catenary; it is the minimal surface spanning two coaxial rings, and it is the shape a soap film takes between two circular wire hoops. Minimal surfaces matter in architecture (tensioned-fabric roofs use minimum-area membranes), in materials science (interfaces and foams), and in pure geometry (the study of minimal surfaces, the Plateau problem, is a deep and still-active field). An honest caveat: zero mean curvature is the stationary condition, so a minimal surface is a critical point of area; when the boundary frame is too wide, the spanning film can become unstable and is no longer the true area minimizer, even though it still solves the equation.