Table of Contents
Fetching ...

The time-dependent harmonic oscillator revisited

Gaetano Fiore

TL;DR

This work presents a novel reduction of the time-dependent harmonic oscillator to action-angle variables, where the angle $\psi$ obeys a decoupled first-order equation and the action $\mathcal I$ is obtained by quadrature. By recasting the $\psi$-dynamics as a Volterra integral equation, the authors prove rapid convergence of a Picard-iteration sequence $\{\psi^{(h)}\}$ on compact intervals, with explicit error bounds tied to the cumulative frequency variation $g(t)=\int |\dot\omega/\omega|$. They further develop a Riccati-based reduction to study zeros of $q$ and $\dot q$, provide adiabatic invariance and slow-time asymptotics, and derive upper/lower bounds for solutions, including parametric resonance analysis for periodic $\omega(t)$ via Floquet theory. The framework accommodates frequency discontinuities through matching conditions and yields practical insights into stability, damping, and beating, with potential applications to Hill-type systems and quantum settings. Overall, the approach offers a unified, efficient route to long-time behavior, rigorous bounds, and resonance phenomena for non-autonomous linear oscillators.

Abstract

We point out a rather effective approach for solving the time-dependent harmonic oscillator $\ddot q=-ω^2 q$ under various regularity assumptions. Where $ω(t )$ is $C^1$ this is reduced to Hamilton equation for the angle variable $ψ$ {\it alone} (the action variable ${\cal I}$ is obtained \it by quadrature}). The fixed point theorem for the integral equation equivalent to the generic Cauchy problem for $ψ(t )$ yields a sequence $\{ψ^{(h)}\}_{h\in\mathbb{N}_0}$ converging to $ψ$ rather fast; if $ω$ varies slowly or little, already $ψ^{(0)}$ approximates $ψ$ well for rather long time lapses. The discontinuities of $ω$, if any, determine those of $ψ,{\cal I}$. The zeros of $q,\dot q$ are investigated via Riccati equations. Our approach may simplify the study of: upper and lower bounds on the solutions; the stability of the trivial one; parametric resonance when $ω(t )$ is periodic; the adiabatic invariance of ${\cal I}$; asymptotic expansions in a slow time parameter $\varepsilon$; time-dependent driven and damped parametric oscillators; etc.

The time-dependent harmonic oscillator revisited

TL;DR

This work presents a novel reduction of the time-dependent harmonic oscillator to action-angle variables, where the angle obeys a decoupled first-order equation and the action is obtained by quadrature. By recasting the -dynamics as a Volterra integral equation, the authors prove rapid convergence of a Picard-iteration sequence on compact intervals, with explicit error bounds tied to the cumulative frequency variation . They further develop a Riccati-based reduction to study zeros of and , provide adiabatic invariance and slow-time asymptotics, and derive upper/lower bounds for solutions, including parametric resonance analysis for periodic via Floquet theory. The framework accommodates frequency discontinuities through matching conditions and yields practical insights into stability, damping, and beating, with potential applications to Hill-type systems and quantum settings. Overall, the approach offers a unified, efficient route to long-time behavior, rigorous bounds, and resonance phenomena for non-autonomous linear oscillators.

Abstract

We point out a rather effective approach for solving the time-dependent harmonic oscillator under various regularity assumptions. Where is this is reduced to Hamilton equation for the angle variable {\it alone} (the action variable is obtained \it by quadrature}). The fixed point theorem for the integral equation equivalent to the generic Cauchy problem for yields a sequence converging to rather fast; if varies slowly or little, already approximates well for rather long time lapses. The discontinuities of , if any, determine those of . The zeros of are investigated via Riccati equations. Our approach may simplify the study of: upper and lower bounds on the solutions; the stability of the trivial one; parametric resonance when is periodic; the adiabatic invariance of ; asymptotic expansions in a slow time parameter ; time-dependent driven and damped parametric oscillators; etc.
Paper Structure (17 sections, 5 theorems, 87 equations, 3 figures)

This paper contains 17 sections, 5 theorems, 87 equations, 3 figures.

Key Result

Proposition 1

If $\omega_l:= \inf_{t\in\mathbb{R}}\{\omega(t)\}>0$, then every nontrivial solution $t\in\mathbb{R}\mapsto q(t)$ of (eq1) admits a strictly increasing sequence $\{t_h\}_{h\in\mathbb{Z}}\subset\mathbb{R}$ such that for all $j\in\mathbb{Z}$: More generally, if $\omega(t)\ge\bar{\omega}_l$, with some $\bar{\omega}_l>0$, holds for all $t$ belonging to an interval $J\subseteq \mathbb{R}$, then ther

Figures (3)

  • Figure 1: The angle $\psi$. We order the quadrants clockwise, i.e. in the direction of the motion.
  • Figure 2: Mathieu equation with $\omega(t)=\bar{\omega}\sqrt{1+ 0.5\, \sin(0.2 t)}$ (this is a non-resonant case). Up: graph of $\omega$; center: graphs of the two solutions $v_a$ ($a=1,2$) of (\ref{['eq1']}) fulfilling $v_1(0)=\dot v_2(0)=1$, $v_2(0)=\dot v_1(0)=0$ (see section \ref{['preli']}) and of their approximations $\hat{v}_a,\tilde{v}_a$, with $a=1$ on the left, $a=2$ on the right; down: graphs of the corresponding action variable ${\cal I}_a$ and of their approximations $\hat{{\cal I}}_a,\tilde{{\cal I}}_a$. As we can see, $\hat{v}_a\simeq v_a$, $\hat{{\cal I}}_a\simeq {\cal I}_a$, namely the approximation (\ref{['def-hat-q']}) is rather good, and much better than the one (\ref{['tilde-approx']}).
  • Figure 3: Mathieu equation with $\omega(t)=\bar{\omega}\sqrt{1+ 0.2\, \sin(2 t)}$ (this is a resonant case). Up: graph of $\omega$; center: graphs of the two solutions $v_a$ ($a=1,2$) of (\ref{['eq1']}) fulfilling $v_1(0)=\dot v_2(0)=1$, $v_2(0)=\dot v_1(0)=0$ (see section \ref{['preli']}) and of their approximations $\hat{v}_a,\tilde{v}_a$, with $a=1$ on the left, $a=2$ on the right; down: graphs of the corresponding action variable ${\cal I}_a$ and of their approximations $\hat{{\cal I}}_a,\tilde{{\cal I}}_a$. Again, as we can see, $\hat{v}_a\simeq v_a$, $\hat{{\cal I}}_a\simeq {\cal I}_a$, namely the approximation (\ref{['def-hat-q']}) is rather good, and much better than the one (\ref{['tilde-approx']}).

Theorems & Definitions (5)

  • Proposition 1
  • Proposition 2
  • Proposition 3
  • Corollary 1
  • Lemma 1