Chaos without Local Spectral Instability in a Kicked Oscillator

arXiv:2611.15525 · nlin.CD, physics.data-an · Submitted 2026-11-11 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Nonlinear Sciences. The day in nonlinear science: chaos, patterns, networks and adaptation.

Kai: We're looking at the paper: "Chaos without Local Spectral Instability in a Kicked Oscillator".

Mira: Here's the gist of the paper. A damped harmonic oscillator kicked by a pulsed periodic potential can be chaotic although every local Jacobian has eigenvalues inside the unit circle.

Kai: The proposed mechanism can be realized with laser-cooled trapped ions in a pulsed optical lattice.

Mira: Chaotic Mechanism and Spectral Stability The paper investigates a non-normal route to chaos in a simple, physically realizable system: a damped harmonic oscillator driven by a pulsed periodic potential.

Kai: Although the eigenvalues of its local Jacobian remain strictly inside the unit circle throughout phase space, the system can sustain chaotic dynamics.

Title and authors: Kai: So we're looking at this paper, "Chaos without Local Spectral Instability in a Kicked Oscillator." It sounds like they are talking about chaos happening even when the local math suggests everything is stable.

Mira: That's right. The title itself sets up a big idea: that you can get chaos without any single local Jacobian eigenvalue ever leaving the unit circle, which is what we usually expect to see in these kinds of systems.

Lev: From an error-correction side, that’s a huge hurdle because standard methods rely on those local eigenvalues being stable for the whole system to be predictable.

Kai: Exactly. And the authors are Citro, Petrillo and Sornette. They're tackling this idea head-on by showing how things get chaotic in a damped harmonic oscillator kicked by a pulsed periodic potential when you look at the sequence of kicks rather than just one step at a time.

Mira: What I find interesting is that they aren't just pointing out this as an anomaly; they are building the math to show *how* it happens, using something called non-normal shears of opposite curvature compounding between kicks.

Lev: That compounds idea is key. If you have two things that individually contract your error, but they do it in opposite ways, the combination can actually push your error out.

The paper's summary: Kai: So what's the core finding of this paper? Basically, they show that a damped harmonic oscillator kicked by a pulsed periodic potential can exhibit chaotic dynamics even though every single local Jacobian has its eigenvalues inside the unit circle.

Mira: It’s showing that local spectral contraction doesn't guarantee global attraction, which is a known issue in dynamical systems where the dynamics are non-normal.

Lev: So when you look at the specifics, they derive an exact joint spectral radius of the Jacobian family and an exact Lyapunov exponent decomposition that shows how this geometric pumping overcomes dissipation.

Kai: That means they’re not just saying it happens; they’s giving us a way to measure it precisely by separating the uniform damping from this cumulative geometric pumping between incompatible contraction metrics.

Mira: The paper highlights that even when the mean mechanical power input doesn't even fully compensate for viscous dissipation, this geometric contribution to the growth rate is what actually sustains chaos.

Lev: That’s a strong point because it suggests that you don't need massive energy input just to get chaos; you need the right kind of timing and structure in the kicks.

The paper's improvements: Kai: They suggest a few ways to improve or clarify this idea. For instance, they look at how ordering matters when you multiply these Jacobians, because the dependence on ordering becomes much clearer when looking at longer lengths.

Mira: That’s important because it means that simply having a set of contracting Jacobians isn't enough; the specific sequence in which you apply them really dictates whether you get contraction or amplification.

Lev: And they show that random permutations of the same Jacobians can exploit this non-normal amplification even more efficiently than the trajectory generated by the actual nonlinear dynamics.

Kai: They have a method for isolating this effect numerically, showing that shuffling those Jacobians reveals the non-normal amplification more clearly than just following a chaotic orbit.

Mira: And they also provide what they call an exact joint spectral radius bound, which sets a sharp boundary between when things contract globally and when geometric amplification can actually happen.

Lev: That sharp boundary is useful for anyone trying to predict if their physical system could actually be chaotic based on its parameters alone.

Conclusion: Kai: So to wrap up, the paper "Chaos without Local Spectral Instability in a Kicked Oscillator" confirms that for chaos to occur, you need the condition where the parameter mu is greater than one minus b.

Mira: That condition essentially tells us that geometric amplification is necessary; if mu is smaller than that threshold, everything contracts globally and there's only one stable fixed point.

Lev: From my view on running this on hardware, it means that for real physical systems, you need to engineer the pulse timing and lattice parameters to push you into the regime where that geometric pumping takes over.

Kai: I think the oscillator realizes only part of this potential because our measured maximal Lyapunov exponent is only about zero point one zero three, which is less than the bound they calculated, which they called ln q(b, mu) <ref:2611.15525#pg2>.

Mira: It's interesting because the oscillator shows us how much of that amplification potential it actually uses; it realizes only about zero point seven four percent of the bound, so we can see that condition mu greater than one minus b is necessary for chaos.

Lev: Yeah, so while the theoretical framework is solid, you still need to get those experimental parameters just right—the pulse interval and intensity—to actually capture all that amplification potential in reality.

V. Citro, G. Petrillo, D. Sornette

Department of Industrial Engineering, University of Salerno · CNR - Nanotec Rome Unit · CNRS ENS de Lyon LPENSL UMR5672 Institute of Risk Analysis Prediction and Management Risks-X Southern University of Science and Technology Shenzhen

nlin.CD, physics.data-an

Submitted: 2026-11-11

Updated: 2026-11-11

Comments: 20 pages, 5 figures

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 82/100

The gist: A damped harmonic oscillator kicked by a pulsed periodic potential can be chaotic although every local Jacobian has eigenvalues inside the unit circle.

Key concepts

Non-normal Jacobian
This is a matrix describing the local linear behavior of a system at a specific point in its state space. A non-normal Jacobian has eigenvalues inside the unit circle, suggesting local contraction, but its structure allows it to stretch perturbations along certain directions over time.
Geometric Amplification
This refers to the stretching of infinitesimal perturbations caused by the geometry of the system's dynamics rather than just linear growth. In this context, it is a mechanism where successive local contractions combine in a way that leads to exponential growth, even when individual steps contract.
Lyapunov Exponent Decomposition
This technique breaks down the overall rate of chaos into two parts: uniform damping (the contraction) and cumulative geometric pumping (the stretching). The paper shows that the geometric pumping term can be larger than the uniform damping term, driving chaos.
Joint Spectral Radius
This is a mathematical tool used to find the maximum possible growth rate of a system when multiple local linear maps are composed. It establishes a sharp boundary: if this radius is less than one, the system contracts globally; if it's greater than one, geometric amplification can occur.

Terminology

Summary

A damped harmonic oscillator kicked by a pulsed periodic potential can be chaotic although every local Jacobian has eigenvalues inside the unit circle. p. 0 The proposed mechanism can be realized with laser-cooled trapped ions in a pulsed optical lattice. p. 0

Chaotic Mechanism and Spectral Stability

The paper investigates a non-normal route to chaos in a simple, physically realizable system: a damped harmonic oscillator driven by a pulsed periodic potential. Although the eigenvalues of its local Jacobian remain strictly inside the unit circle throughout phase space, the system can sustain chaotic dynamics. p. 0 This finding connects to a line of mathematical work showing that local spectral contraction does not ensure global attraction. p. 0 A non-normal Jacobian can stretch a perturbation along one direction despite having all its eigenvalues inside the unit circle. p. 0

Geometric Amplification and Lyapunov Exponent Decomposition

The paper derives an exact joint spectral radius of the Jacobian family and an exact Lyapunov-exponent decomposition showing how this geometric pumping overcomes dissipation. The maximal Lyapunov exponent is given by a decomposition where the first term is uniform damping, and the second term is cumulative geometric pumping between incompatible local contraction metrics. p. 3 This geometric contribution to the growth rate exceeds the uniform contraction 1/2 ln b < 0, even when the mean mechanical power input does no more than compensate for viscous dissipation. p. 5

Exact Amplification Bounds and Dynamical Consequences

The exact joint spectral radius establishes a sharp boundary between global contraction and the possibility of geometric amplification. For µ < 1 − b, one has q < 1, implying that every admissible Jacobian product contracts in the same weighted norm, and the nonlinear map possesses a unique globally attracting fixed point. p. 2 When µ > 1−b, one has q > 1: suitably ordered products of individually contracting Jacobians can amplify perturbations exponentially. p. 2

Experimental Realization and Intermittency

The proposed experimental realization involves a laser-cooled trapped ion driven by a pulsed optical lattice. The bifurcation sequence reveals how geometric amplification develops into chaos through the collision of Floquet multipliers, leading to type-I intermittency. p. 3 Sustained intermittent chaos occurs in the parameter regime specified in Sec. p. 5 S7. The computer-assisted proof establishes the existence of a compact invariant set with positive topological entropy, htop(F60K) ≥ ln 2/60 > 0. p. 5

Ordering and Invariant Measure

The dependence on ordering becomes clearer at longer lengths, as the orbit fixes one particular ordering among all the products allowed by the Jacobian family. Sorting and shuffling the Jacobians provide a way to isolate this effect numerically, showing that random permutations can exploit available non-normal amplification even more efficiently than the trajectory generated by nonlinear dynamics. p. 2 The covering proof guarantees an ergodic invariant measure with a strictly positive Lyapunov exponent and a quantitative lower bound on its asymptotic tangent-space growth. p. 3

Lur'e Representation and Absolute Stability

The kicked oscillator can be written as a discrete-time Lur’e system, which shows that the oscillator is an explicit counterexample to the discrete-time Kalman conjecture. The joint-spectral-radius bounds and geometricpumping decomposition developed in the Letter can thus be viewed as quantitative tools for examining how timedependent feedback gains combine to produce growth that is absent from each constant-gain system considered separately.

Experimental Implementation Details

A possible experimental implementation involves tuning the pulse interval, optical intensity, trap position, and damping to reproduce the dissipative return map. The analysis assumes that successive kicks are separated by exactly a quarter of the damped oscillation period, θ = ωT = π/2. p. 0 The timing window for chaos is determined by the condition 2 cos θ + K sin θ < 2 cosh(γT). p. 0 The principal test is positive long-time Lyapunov growth alongside independent confirmation, from the calibrated return map, that the local Jacobian spectral radius is uniformly below unity throughout phase space. p. 1

Mechanical Energy Balance

The work done by the pulses and the growth of infinitesimal perturbations describe two distinct balances. The term Wn records the change associated with passing between successive local contraction metrics, which is a purely geometric quantity. p. 3 The resulting growth comes from the ordered composition of non-normal Jacobians, not from any creation of mechanical energy by non-normality. p. 2

Discriminating Observables

Sampling the position immediately before each pulse gives Xn = k(xn − x0) and yn = kxn, which can be compared directly with the two-step recurrence (3). Once b, µ, and the optical phase have been independently calibrated, the local curvature and the corresponding one-kick Jacobian follow from tn = µ cos yn. p. 5 The measured sequence can then be used to reconstruct the finite-product rates hL of Eq. p. 5 (6) and to integrate the tangent dynamics to estimate λ1. p. 5

Basins of Attraction and Intermittency

The basin fractions for the fixed point, chaotic attractor, and other periodic attractors were measured on grids covering [−4, 5]2. Sustained intermittent chaos is found for C − Cf ≳ 8 × 10−5. p. 4 The mean laminar residence time scales as (C − Cf)−0. p. 4500±0.002, compatible with type-I intermittency. p. 4 The results are robust to the choice of threshold for the laminar phase. p. 4

Conclusion on Mechanism

The geometric contribution in Eq. (12) is 0.1557±0.0002, which exceeds the magnitude of the fixed spectral contraction, −1/2 ln b ≃ 0. p. 10527, and gives the measured maximal Lyapunov exponent λ1 ≃ 0. p. 2103. This confirms that the geometric contribution in Eq. p. 5 (12) is what sustains chaos. p. 2 The oscillation realizes only part of this potential, as the measured maximal Lyapunov exponent is only 0. p. 274% of the bound ln q(b, µ). p. 2

Final Summary Sentence

The chaotic attractor at b = 0.9 and µ = 1.7076 shows how much of this potential the dynamics actually uses. p. 2 The oscillator therefore realizes only part of the amplification potential permitted by its Jacobians. p. 2 p. 2 p. 2 The condition µ > 1 − b is thus necessary for chaos. p. 2 p. 2 The oscillator therefore realizes only part of the amplification potential permitted by its Jacobians. The condition µ > 1 − b is thus necessary for chaos. The oscillator therefore realizes only part of the amplification potential permitted by its Jacobians.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Develop models for dynamical systems where local Jacobian eigenvalues are strictly inside the unit circle but chaotic dynamics are sustained by non-normal shears of opposite curvature compounding, as described in Chaos without Local Spectral Instability in a Kicked Oscillator. This allows AI to predict chaos even when standard linear stability analysis suggests global contraction.

  2. Implement exact joint spectral radius bounds for families of Jacobians to determine the sharp boundary between global contraction and geometric amplification, specifically using the formula q = mu + p over mu squared + 4b squared. This enables AI to assess whether a system's potential for chaos is realized by the actual nonlinear dynamics.

  3. Create Lyapunov exponent decomposition tools that separate uniform local contraction from cumulative geometric pumping between incompatible local contraction metrics, as shown in Exact decomposition of the maximal Lyapunov exponent. This allows AI to quantify exactly how state-dependent geometric amplification overcomes dissipation along nonlinear trajectories.

  4. Design a symbolic dynamics framework based on covering arguments and interval arithmetic to rigorously establish the existence of chaotic invariant sets with positive topological entropy, as detailed in A rigorous chaotic horseshoe and the onset of intermittency. This ensures that chaos identified numerically is mathematically proven to exist within a compact invariant set.

  5. Build an experimental simulation platform for laser-cooled trapped ions in pulsed optical lattices to test the mechanism, specifically by measuring if the measured maximal Lyapunov exponent aligns with the theoretical geometric pumping budget, as described in S15. This provides a pathway to verify chaotic phenomena without requiring local spectral instability.

Sources

Related papers