Introduction to PDEs
Ch 18 — Part IV — PDEs & Fourier Series
A partial differential equation (PDE) involves an unknown function of several variables and its partial derivatives. We meet three classical linear second-order PDEs.
The three classics
Heat (parabolic): \(u_t = \alpha u_{xx}\) — diffusion of heat, concentration. Wave (hyperbolic): \(u_{tt} = c^2 u_{xx}\) — waves on a string, sound, light. Laplace (elliptic): \(u_{xx} + u_{yy} = 0\) — steady-state temperature, electrostatic potential.
Boundary vs initial data
PDEs on bounded spatial domains require boundary conditions (Dirichlet \(u = g\), Neumann \(\partial u/\partial n = g\), or mixed) and often initial conditions. These data dramatically shape the solution.
Linear superposition
All three classical PDEs above are linear; solutions add and scale. This is what makes Fourier methods so powerful.
Practice Problems
Hint
Solution
\(u_t = -\sin(x) e^{-t}\); \(u_{xx} = -\sin(x) e^{-t}\). Equal.
Answer: Yes.
Hint
Solution
\(u_{tt} = -c^2 \sin(x-ct)\); \(u_{xx} = -\sin(x-ct)\). Equal after multiplying by \(c^2\).
Answer: Yes.
Hint
Solution
\(u_{tt}\) and \(-u_{xx}\) have opposite signs, so hyperbolic (wave-like).
Answer: Hyperbolic.