A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Normal modes are special oscillation patterns where all parts of a coupled system oscillate sinusoidally at the same frequency, with fixed amplitude ratios. Any motion is a superposition of normal modes, each evolving independently.
You know simple harmonic motion (SHM): a single mass on a spring oscillates at a frequency ω = √(k/m), and any motion is a sinusoid at that one frequency. You also know coupled oscillators: when you link two masses through a coupling spring, they can no longer oscillate independently — moving one disturbs the other, and the motion becomes complicated. Normal modes are the key to unlocking that complexity. They are the special initial conditions under which a coupled system *does* behave simply — every part oscillating at the same single frequency — and the crucial theorem is that *any* motion of the system is a superposition of these simple patterns.
The connection to eigenvalues and eigenvectors (your prerequisite) is direct. Write the equations of motion for a two-mass coupled system as a matrix equation: Mẍ = -Kx, where M is the mass matrix and K is the stiffness matrix. Try a solution of the form x(t) = v eiωt: substituting gives Kv = ω²Mv, which is a generalized eigenvalue problem. Each eigenvalue ω² gives a normal mode frequency, and each eigenvector v gives the amplitude ratio — how the masses move relative to each other in that mode. For the symmetric two-mass system, the two eigenvectors correspond to the in-phase mode (both masses moving together, ω₁ = √(k/m)) and the out-of-phase mode (masses moving in opposition, ω₂ = √((k + 2k_c)/m) where k_c is the coupling spring constant). The in-phase mode is lower frequency because the coupling spring is unstretched; the out-of-phase mode is higher because the coupling spring is compressed and stretched.
The superposition principle is what makes normal modes powerful. The normal mode vectors form a complete, orthogonal basis (in the sense defined by the mass matrix) — any initial condition can be decomposed into a sum of normal mode contributions. Each mode evolves independently: mode 1 oscillates at ω₁ with a fixed amplitude, mode 2 at ω₂, and the actual motion is their sum. This decomposition is the mechanical analog of Fourier analysis: just as any periodic function is a sum of sinusoids at different frequencies, any coupled oscillator motion is a sum of normal modes at different frequencies. This is why the technique generalizes far beyond two masses.
The payoff extends throughout physics and engineering. A vibrating string has infinitely many normal modes — the harmonics you know from music. A molecule has normal modes of vibration that determine which infrared frequencies it absorbs (and thus its spectroscopic signature). A crystal lattice's normal modes are phonons, the quantum mechanical quanta of sound. Structural engineers calculate the normal mode frequencies of bridges and buildings to ensure they are not excited by wind or traffic at those frequencies (the Tacoma Narrows Bridge failed partly because wind drove it near a normal mode frequency). The concept unifies a vast range of oscillatory phenomena under a single mathematical framework, and it is one of the most transferable tools in theoretical physics.