Fourier Series Solutions for Bounded Domains
You'll learn: How to systematically solve the wave equation on finite intervals by combining separation of variables with Fourier series expansion.
The Core Idea
When solving the wave equation on a bounded domain (like a vibrating string of length L fixed at both ends), separation of variables gives you spatial eigenfunctions and corresponding time-dependent coefficients. The general solution is an infinite sum of these separated solutions—each representing a normal mode of vibration. The Fourier series provides the machinery to match arbitrary initial conditions by expressing them as sums of these eigenfunctions.
The Solution Strategy
For the wave equation u_tt = c²u_xx on 0 < x < L with fixed ends:
- Separate variables: Assume u(x,t) = X(x)T(t), leading to eigenvalue problems
- Solve the spatial problem: X'' + λX = 0 with boundary conditions yields eigenfunctions X_n(x) = sin(nπx/L)
- Solve the temporal problem: T'' + λc²T = 0 gives T_n(t) = A_n cos(ωₙt) + B_n sin(ωₙt), where ωₙ = nπc/L
- Build the general solution: u(x,t) = Σ [A_n cos(ωₙt) + B_n sin(ωₙt)]sin(nπx/L)
- Apply initial conditions: Use Fourier sine series to determine coefficients A_n and B_n from u(x,0) and u_t(x,0)
The Fourier Connection
Think of this like decomposing a complex musical chord into pure tones. Each term sin(nπx/L) is a pure spatial "note" (standing wave pattern), and the Fourier coefficients tell you how much of each note to include to match your initial pluck or displacement.
Key Takeaway: Fourier series solutions transform the PDE problem into finding coefficients through orthogonality, allowing any reasonable initial condition on a bounded domain to be expressed as a superposition of standing wave modes.