Pith. sign in

REVIEW 2 major objections 4 minor 8 references

Unbounded sequences of stable limit cycles in the delayed Duffing equation: an exact analysis

T0 review · 2 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Delayed Duffing equation yields infinite stable limit cycles

desk verdict The exact lift construction is sound and genuinely new, but the stability half of the main theorem is imported from a companion paper and the one quantitative formula printed here has a sign problem, so the paper needs a fix before it fully stands alone. read the letter →

arxiv 1908.06533 v1 pith:66Q5NOD3 submitted 2019-08-18 math.DS nlin.CD

Si Mohamed Sah , Bernold Fiedler , B. Shayak , Richard H. Rand This is my paper · ORCID
classification math.DSnlin.CD MSC 34K1334K2034K1834C25
keywords delayedDuffingequationexactperiodicsolutionsstablelimitcyclesJacobiellipticfunctionstimedelayFloquetstabilityunboundedamplitudesequence
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper establishes that the delayed Duffing equation $x''(t)+x(t-T)+x^3(t)=0$ possesses an infinite and unbounded sequence of exact periodic solutions for fixed delay $T$ with $T^2 < \tfrac{3}{2}\pi^2$. In this family, the $n$-th solution has minimal period $2T/n$ and amplitude $A_n$ that grows without bound as $n$ grows; for every sufficiently large odd $n$ the solution is a locally asymptotically stable limit cycle, while for large even $n$ it is unstable. Earlier studies of the same system found such rapidly oscillating cycles only by approximate averaging or harmonic balance, and mostly at small amplitude. Here the solutions are exact: each is a Jacobi elliptic function lifted from the ordinary, undelayed Duffing oscillator by the symmetry $x(t-T)=(-1)^n x(t)$. The result matters because it gives one of the few exact models in which a delay system demonstrably harbours infinitely many large stable oscillators at once.

What carries the argument

The load-bearing object is the lift identity $x(t-T)=(-1)^n x(t)$, which states that the delayed value is, up to sign, the current value for every member of the constructed family. Inserting this identity into the undelayed Duffing equation $x''+(-1)^n x+x^3=0$ reproduces exactly the delayed Duffing equation $x''(t)+x(t-T)+x^3(t)=0$. The explicit solutions are Jacobi elliptic functions $x_n(t)=A_n\,\mathrm{cn}(\omega_n t, m_n)$ with $\omega_n=\sqrt{A_n^2+(-1)^n}$ and $m_n=A_n^2/[2(A_n^2+(-1)^n)]$. The amplitude $A_n$ is fixed by the period condition $p_n=4K(m_n)/\omega_n=2T/n$, where $K$ is the complete elliptic integral of the first kind; as $n\to\infty$ this equation drives $p_n$ to zero and $A_n$ to infinity. This identity does the argument's work: it turns a countable subset of the continuum of ordinary Duffing periodic orbits into exact periodic orbits of the delay system.

What would settle it

Compute the dominant nontrivial Floquet exponent of the exact lifted cycle for one large odd $n$, say $n=101$ with $T=0.5$, using a high-precision DDE Floquet method; local asymptotic stability requires the real part to be negative. The conflict between the printed sign in equation (5.2) and the asserted stability makes this computation decisive: whichever sign the computation supports determines whether the theorem or the printed formula needs correction.

Watch

Extended reading notes

Core claim

The core claim is a symmetry-based exact correspondence between the ordinary Duffing oscillator and the delayed Duffing equation. Any positive-energy periodic solution $x_n$ of $x''+(-1)^n x+x^3=0$ with minimal period $p_n=2T/n$ satisfies $x_n(t-T)=(-1)^n x_n(t)$, and substituting this identity converts the undelayed equation into $x''+x(t-T)+x^3=0$. Thus every such ODE solution is an exact periodic solution of the DDE with delay $T$. The minimal periods $2T/n$ decrease to zero as $n$ increases, and since the Duffing period decreases monotonically with amplitude, the amplitudes $A_n$ increase to infinity; the paper supplies the exact equation $2T/n=4K(m_n)/\omega_n$ determining $A_n$, plus convergent series expansions for large $n$. The stability part, whose proof is deferred to the companion paper, is that for $T^2 < \tfrac{3}{2}\pi^2$ the odd-$n$ cycles are locally asymptotically stable for all sufficiently large $n$, while the even-$n$ cycles are linearly and nonlinearly unstable.

Load-bearing premise

The stability claim rests on the companion paper's Floquet analysis, which is cited but not carried out here; if that analysis is wrong, the assertion that large odd cycles are stable and large even cycles are unstable does not follow.

Editorial extensions

If this is right

  • For every fixed delay satisfying $T^2 < \tfrac{3}{2}\pi^2$, the delayed Duffing equation has infinitely many coexisting stable limit cycles, one for each sufficiently large odd $n$, with amplitudes tending to infinity.
  • The even-indexed members of the same exact family are unstable and appear to act as repelling boundaries between the basins of the stable odd cycles.
  • The exact amplitude equation $2T/n=4K(m_n)/\omega_n$ makes each amplitude $A_n$ computable to arbitrary precision, and the convergent series expansions cover parameter ranges where direct numerics struggle.
  • At the boundary $T^2=\tfrac{3}{2}\pi^2$, the odd cycles lose stability through a Neimark-Sacker torus bifurcation, producing additional periodic solutions that are not lifts of the undelayed Duffing orbits.
  • The result confirms that the infinite-stable-cycle scenario previously obtained by averaging and harmonic balance is not an artifact of approximation: it is present in the exact equations.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The lift mechanism is generic: any odd restoring force $g(x)$ should yield the same exact construction, so the delayed Duffing equation is a representative example of a whole class of delay systems with unbounded families of stable periodic orbits.
  • Reading the paper's amplitude expansion at leading order predicts $A_n \sim \frac{\Gamma(1/4)^2}{2\sqrt{\pi}\,T}\,n$, so the cycle amplitudes grow linearly with the index $n$; numerical continuation for moderate $n$ could test this scaling directly.
  • The stability theorem is stated for sufficiently large odd $n$, but the numerics in Section 6 show attraction to the $n=1$ cycle from a wide basin; an open question suggested by the paper is whether local stability actually extends to all odd $n$, not only large ones.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. The manuscript studies the delayed Duffing equation x''(t)+x(t-T)+x^3(t)=0. Its central claim is that for any fixed delay T with 0<T^2<3\pi^2/2, the equation possesses an infinite, unbounded sequence of exact periodic solutions x_n with minimal period 2T/n and amplitudes A_n going to infinity; for all sufficiently large odd n these solutions are locally asymptotically stable, while for large even n they are linearly and nonlinearly unstable. Section 3 constructs these solutions by lifting periodic solutions of the non-delayed Duffing ODE (3.1) via the symmetry x_n(t-T)=(-1)^n x_n(t). Section 4 provides numerical and convergent series methods for computing the amplitudes A_n. Section 5 quotes stability theorems from the companion paper [Fie&al19], and Section 6 illustrates the results numerically for n=1,2.

Significance. The exact lift construction in Section 3 is a valuable and largely self-contained contribution: for even n the delay T is an integer multiple of the minimal period, and for odd n it is an odd multiple of the half-period, so the oddness symmetry (3.18) converts the ODE (3.16) into the DDE (1.1). The amplitude formulas are parameter-free, the elliptic-integral equation (4.2) is explicit, and the series expansions in Section 4 are claimed to be convergent, giving checkable quantitative predictions. If the stability claims are correct, the paper offers a clean exact counterpart to the earlier approximate analyses of Wahi-Chatterjee, Mitra et al., and Davidow et al. The main weakness is that the stability half of the abstract's claim is not proved in this manuscript, and the one quantitative stability formula printed, Eq. (5.2), has a sign that contradicts the stated theorems.

major comments (2)
  1. [Section 5, Eq. (5.2)] The printed formula \eta = (2/3)(-1)^{n+1}T^2 + \cdots gives a positive leading Floquet exponent for odd n and a negative one for even n. Since \eta is defined as the nontrivial exponent with real part closest to zero, this directly contradicts Theorem 5.1 (asymptotic stability for odd n) and Theorem 5.2 (instability for even n). The intended formula is presumably \eta = (2/3)(-1)^n T^2 + \cdots; as printed, the sign must be corrected or the sign convention clearly explained. This is load-bearing because Eq. (5.2) is the only quantitative stability asymptotics supplied in the paper.
  2. [Section 5] Theorems 5.1 and 5.2, which constitute the stability half of the central claim, are quoted from the companion paper [Fie&al19] and no proofs or Floquet-setup details are given in this manuscript; the text explicitly says 'For detailed mathematical proofs we have to refer to [Fie&al19]'. Since the abstract's assertion of asymptotically stable periodic solutions rests entirely on these imported results, the paper is not self-contained on its main advertised contribution. The authors should either include the stability proof (or at least a precise statement of all relevant Floquet exponents and their signs), or clearly mark the stability results as imported from the companion paper and verify that Eq. (5.2) is consistent with those results.
minor comments (4)
  1. [Section 5; Section 6] Theorem 5.1 and several later passages refer to 'the delayed Duffing equation (3.1)', but equation (3.1) is the non-delayed ODE; the DDE is equation (1.1). Please correct these cross-references.
  2. [Section 3.2 and 3.3] The monotonicity of the period-amplitude relation p(A) is asserted without proof. Existence of A_n follows from continuity and the endpoint limits, but the uniqueness implied by 'the amplitude A_n' and the use of a threshold n_0(T) depend on monotonicity; please add a proof or a precise reference.
  3. [Section 4] The notation m(A_n) and \omega(A_n) in Eq. (4.2) suppresses the parity dependence given in Eq. (3.10); a sentence making this explicit would improve readability.
  4. [Section 4, Eqs. (4.3)-(4.6)] The claim that the Taylor expansions are convergent would benefit from a brief justification, for example analyticity of K(m) at m=1/2 and the nonzero derivative of p(A) at A=\infty.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: the exact lift and amplitude equation are self-contained; the stability theorem is imported from the authors' companion paper, which is a self-citation but not a circular input, and eq. (5.2) has a sign issue that is a correctness concern, not circularity.

full rationale

The derivation chain in Sections 3 and 4 does not close on itself. The lifted periodic solutions are constructed explicitly via the symmetry x_n(t-T)=(-1)^n x_n(t), using standard Jacobi-elliptic solutions of the non-delayed Duffing ODE, and the amplitudes A_n are defined by the implicit elliptic-integral equation (4.2), p_n=2T/n=4K(m(A_n))/omega(A_n). No parameter is fitted to the target property; the amplitude asymptotics are convergent series expansions rather than calibrated predictions. The stability half is the only part not proved in this text: Theorems 5.1 and 5.2 are quoted from [Fie&al19] with the sentence 'For detailed mathematical proofs we have to refer to [Fie&al19].' The authors overlap, so this is a genuine self-citation, but it is a separate parameter-free theorem with stated assumptions that do not include the target conclusion, so it functions as external proof rather than as a renamed input; I therefore treat it as a minor non-circular self-citation rather than as a circular step. I additionally note that the printed formula eta=2/3(-1)^{n+1}T^2+... appears to have the wrong sign for the claimed stability (odd n would give a positive exponent), which is a correctness/typo issue that should be corrected but is not an instance of circularity. No self-definitional, fitted-input, ansatz-smuggled, or renaming circularity was found.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No free parameters are fitted to data; amplitudes are determined by solving p_n = 4K(m)/omega. The central construction is self-contained once the standard elliptic-function representation of the Duffing ODE is accepted. The imported stability theorems from the companion paper are the main external load-bearing input. The paper introduces no new physical entities.

assumptions (5)
  • standard math Exact periodic solutions of the non-delayed Duffing ODE x'' +/- x + x^3 = 0 are x_n(t)=A_n cn(omega_n t, m_n), with m_n and omega_n given by (3.10) and period p_n=4K(m_n)/omega_n.
    Used in Section 3.1, equations (3.9)-(3.11), as the starting point of the lift.
  • standard math The Hamiltonian H = (1/2)x_dot^2 + (1/2)(-1)^n x^2 + (1/4)x^4 is constant along solutions, and the selected orbits have H > 0.
    Conservation law (3.2) justifies separation of variables and the period integral (3.6)-(3.8).
  • domain assumption For odd n, positive-energy solutions of the double-well Duffing ODE satisfy the half-period anti-symmetry x(t) = -x(t-p/2).
    Invoked in Section 3.3, equation (3.18), to lift to the DDE with x(t-T) = -x(t). It follows from time reversibility for odd force laws.
  • ad hoc to paper The periodicity or anti-periodicity condition x_n(t-T) = (-1)^n x_n(t) with p_n = 2T/n maps the non-delayed ODE (3.1) to the delayed DDE (1.1).
    This is the central construction in Section 3, equations (3.12), (3.15), and (3.20); it is proved by substitution, not taken from prior literature.
  • domain assumption The Floquet stability theorems 5.1 and 5.2 from [Fie&al19] are correct, including the threshold T^2 < 3π^2/2 and the existence of n0(T).
    Quoted without proof in Section 5; the paper explicitly refers to the companion paper for detailed mathematical proofs.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Unbounded sequences of stable limit cycles in the delayed Duffing equation: an exact analysis." pith.science (2026). https://pith.science/paper/66Q5NOD3

@misc{pith2026190806533,
  author       = {Pith},
  title        = {Pith review of: Unbounded sequences of stable limit cycles in the delayed Duffing equation: an exact analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/66Q5NOD3}},
  note         = {Machine review of arXiv:1908.06533}
}
abstract

The delayed Duffing equation $\ddot{x}(t)+x(t-T)+x^3(t)=0$ is shown to possess an infinite and unbounded sequence of rapidly oscillating, asymptotically stable periodic solutions, for fixed delays such that $T^2<\tfrac{3}{2}\pi^2$. In contrast to several previous works which involved approximate solutions, the treatment here is exact.

Figures

Figures reproduced from arXiv: 1908.06533 by the authors.

Figure 2.1
Figure 2.1. (a) Time histories of some periodic solutions [PITH_FULL_IMAGE:figures/full_fig_p003_2_1.png] view at source ↗
Figure 3.1
Figure 3.1. Three dimensional plots of Hamiltonian level sets ( [PITH_FULL_IMAGE:figures/full_fig_p004_3_1.png] view at source ↗
Figure 3.2
Figure 3.2. The relation between amplitude A and frequency ω of the periodic solutions in the non￾delayed Duffing ODE (3.1) obtained from the second equation of (3.10). Upper curve for n odd and lower curve for n even. Only the marked points on these two curves correspond to periodic solutions xn(t) of the delayed Duffing DDE (1.1). The time delay for this plot is T = 3. asymptotically stable, for T 2 < 3 2 π 2 and large n, whi… view at source ↗
Figures from the paper (5 more)
Figure 3.3
Figure 3.3. Figure 3.3: Solutions of the single well Duffing ODE ( [PITH_FULL_IMAGE:figures/full_fig_p007_3_3.png]
Figure 3.4
Figure 3.4. Figure 3.4: Solutions of the double-well Duffing ODE ( [PITH_FULL_IMAGE:figures/full_fig_p009_3_4.png]
Figure 3.5
Figure 3.5. Figure 3.5: Schematic illustration of the lifts from the non-delayed Duffing ODEs ( [PITH_FULL_IMAGE:figures/full_fig_p010_3_5.png]
Figure 6.1
Figure 6.1. Figure 6.1: Time histories (a) and phase plane plots (b) for delay [PITH_FULL_IMAGE:figures/full_fig_p013_6_1.png]
Figure 6.2
Figure 6.2. Figure 6.2: Time histories of x(t), top (a), and of x˙(t), bottom (b), for delay T = 0.5. Exact solutions xn(t) for n = 1 (red) and n = 2 (teal); see (3.9). Their amplitudes are A1 = 7.5139958 . . . and A2 = 14.7834172 . . . , respectively. The numerical solution of the delayed …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

8 extracted references · 8 canonical work pages

  1. [1]

    Bogacki, L

    P. Bogacki, L. F. Shampine. A 3(2) pair of Runge - Kutta formulas. Applied Mathematics Letters 2, 4, 321 ISSN 0893-9659, (1989)

    work page 1989

  2. [2]

    Davidow, B

    M. Davidow, B. Shayak, R. H. Rand. Analysis of a remarkable singularity in a nonlinear DDE. Nonlinear Dynamics , (2017) 90:317-323

    work page 2017

  3. [3]

    Coexistence of infinitely many large, stable, rapidly oscillating periodic solutions in time-delayed Duffing oscillators

    B. Fiedler, A. López Nieto, R.H. Rand, S.M. Sah, I. Schneider, B. de Wolff. Coexistence of infinitely many large, stable, rapidly oscillating periodic solutions in time-delayed Duffing oscillators. arXiv:1906.06602 (2019)

    work page Pith review arXiv 2019

  4. [4]

    Flunkert

    V. Flunkert. Pydelay: A Simulation Package. In: Delay-Coupled Complex Systems. Springer Theses. Springer, Berlin, Heidelberg (2011)

    work page 2011

  5. [5]

    Kovacic and M.J

    I. Kovacic and M.J. Brennan (eds.). The Duffing Equation: Nonlinear Oscillators and their Behaviour. John Wiley & Sons, Chichester (2011)

    work page 2011

  6. [6]

    Mitra, S

    R.K. Mitra, S. Chatterjee, A.K. Banok. Limit cycle oscillation and multiple entrainment phenomena in a Duffing oscillator under time-delayed displacement feedback. J. Vibration and Control , (2017) 23:2742-2756

    work page 2017

  7. [7]

    R.H. Rand. Topics in Nonlinear Dynamics with Computer Algebra, Computation in Education: Mathematics, Science and Engineering. Vol. 1, Gordon and Breach, Langhorne, PA (1994)

    work page 1994

  8. [8]

    P. Wahi, A. Chatterjee. Averaging oscillations with small fractional damping and delayed terms. Nonlinear Dynamics , (2004) 38: 3–22

    work page 2004

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.