transient and steady terms of a linear first-order solution

When a linear equation describes something physical settling down over time, its solution naturally splits into two pieces with very different fates: a part that fades and a part that lasts. The transient term is the fading piece — it dies away as time grows. The steady term (or steady-state) is what remains after the transient has decayed, the long-run behaviour the system approaches.

Concretely, take y' + a y = q(t) with a positive constant a and a forcing q(t). The general solution is y = y_p(t) + C e^(-a t). The piece C e^(-a t) is the transient: because a is positive, e^(-a t) shrinks toward zero, so this term forgets the initial condition and vanishes. The remaining y_p(t), which tracks the forcing, is the steady term. If the forcing is constant, the steady term is a constant level; if the forcing is a sinusoid, the steady term is a sinusoid of the same frequency. So the solution starts off bearing the marks of its initial state (through the transient) and then relaxes into a long-term pattern dictated entirely by the forcing.

This decomposition is exactly why a circuit shows a brief surge before settling, or a heated object overshoots and then approaches room temperature: the transient is the memory of the start, the steady state is the eventual equilibrium with the environment. Two honest caveats: the split into 'transient' and 'steady' assumes the homogeneous part actually decays, which needs a positive (more generally, negative real part) so e^(-a t) shrinks — if a were negative the 'transient' would grow and the picture inverts. And mathematically the transient never reaches exactly zero in finite time; it merely becomes negligibly small, characterized by the time constant 1/a.