Sturm-Liouville problem
/ STURM lee-oo-VEEL /
Many of the most useful expansions in physics — Fourier series, Legendre series, Bessel series — look like miracles: an arbitrary function gets written as a sum of special functions that happen to be orthogonal. The Sturm-Liouville problem is the single theory that explains all of them at once. It asks: for a given second-order differential operator on an interval, which special values of a parameter (call it lambda) allow a nontrivial solution that also satisfies the boundary conditions? Those special values are the eigenvalues, and the solutions are the eigenfunctions.
Concretely, a Sturm-Liouville problem is the eigenvalue equation (d/dx)(p(x) dy/dx) + q(x) y + lambda w(x) y = 0 on an interval [a, b], paired with boundary conditions at the two endpoints. Here p(x) > 0 and w(x) > 0 are given functions, w being the weight. We seek the pairs (lambda_n, y_n) for which a nonzero y satisfies both the equation and the boundary conditions. Just as a matrix has only certain eigenvalues, this operator admits only a discrete set lambda_1 < lambda_2 < ... , each with its own eigenfunction y_n. The whole machinery — orthogonality, completeness, generalized Fourier series — follows from the special, self-adjoint shape of this equation.
This matters because almost every boundary-value problem solved by separation of variables produces a Sturm-Liouville problem in disguise. The vibrating string gives sin(n pi x / L); the circular drumhead gives Bessel functions; the sphere gives Legendre polynomials. Recognizing the common Sturm-Liouville skeleton lets you reuse one set of guarantees — real eigenvalues, orthogonal eigenfunctions, complete expansions — instead of reproving them for every new geometry.
Take y'' + lambda y = 0 on [0, L] with y(0) = y(L) = 0. This is Sturm-Liouville with p = 1, q = 0, w = 1. Nontrivial solutions exist only for lambda_n = (n pi / L)^2, giving eigenfunctions y_n = sin(n pi x / L) — exactly the sine series of a vibrating string fixed at both ends.
The plain Fourier sine series is the simplest Sturm-Liouville problem; every richer expansion changes only p, q, and w.
Not every second-order ODE is a Sturm-Liouville problem out of the box, but any linear one can be put into self-adjoint form by multiplying through by an integrating factor — that step is what unlocks all the spectral guarantees.