Escaping orbits are also rare in the almost periodic Fermi-Ulam ping-pong

Henrik Schlie{\ss}auf

REVIEW 2 major objections 2 minor 27 references

Escaping orbits are also rare in the almost periodic Fermi-Ulam ping-pong

T0 review · 2 major / 2 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 For a smooth almost periodic forcing of the Fermi-Ulam ping-pong, initial conditions that lead to unbounded speed form a measure-zero set for almost every phase.

desk verdict A clean, honest generalization of the KO18 result to Bohr almost periodic forcings; one small repairable gap in the oscillation bound, but the main theorem stands. read the letter →

arxiv 1908.02529 v1 pith:PBKEY6JP submitted 2019-08-07 math.DS

Henrik Schlie{ss}auf This is my paper · ORCID

classification math.DS MSC 37C4037A05

keywords Fermi-Ulamping-pongescapingorbitsalmostperiodicforcingmeasurezeroadiabaticinvariantFermiaccelerationPoincarerecurrencequasi-periodic

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 studies the Fermi-Ulam ping-pong, a point particle bouncing elastically between a fixed wall and a moving wall whose position is an almost periodic function of time—that is, a function that approximately repeats itself at arbitrarily large times. It establishes that, for any sufficiently smooth almost periodic forcing, and for almost every phase of the forcing, the set of initial times and velocities that produce escaping orbits (particles whose speed tends to infinity) has Lebesgue measure zero in the plane. This generalizes a known result for quasi-periodic forcings, and it matters because it says that unbounded Fermi acceleration is atypical in this model: typical initial conditions keep the particle's speed bounded despite the time-dependent wall.

What carries the argument

The load-bearing object is the generalized adiabatic invariant $W(\omega,E)=P(\omega)^2 E$, defined on the compact hull $\Omega$ of the almost periodic forcing. The proof requires two inequalities: the derivative of $W$ with respect to $E$ is bounded between positive constants, and along one iteration $W$ grows by at most $k(E)$, where $k(E)\to 0$ as $E\to\infty$. That second inequality is obtained from an estimate, imported from earlier work, which bounds the change of $p(t)^2 E$ over one collision by a quantity $\Delta(E)=E^{-1/2}+\sup\{|\partial_\psi^2P(\omega)-\partial_\psi^2P(\omega')|: \|\omega-\omega'\|\le CE^{-1/2}\}$. The uniformity of $\partial_\psi^2P$ on the compact hull is exactly what makes $\Delta(E)\to 0$. The other essential mechanism is the decomposition of Haar measure on $\Omega$ along the flow into a cross section and time intervals, which lets the recurrence proof be lifted from one forcing to almost every phase.

What would settle it

Compute, for a two-frequency quasi-periodic forcing with an irrational frequency ratio, the fraction of initial conditions in a large box whose speed exceeds a large threshold $V$ after $N$ collisions, for a dense grid of phases; if for a positive-measure set of phases this fraction does not tend to zero as $V,N\to\infty$, Theorem 5.1 is false. More directly, one can check numerically whether the error quantity $\Delta(E_0)=E_0^{-1/2}+\sup\{|\partial_\psi^2P(\omega)-\partial_\psi^2P(\omega')|:\|\omega-\omega'\|\le CE_0^{-1/2}\}$ goes to zero as $E_0\to\infty$; a failure there would invalidate Lemma 5.4 and with it the proof.

Watch

Extended reading notes

Core claim

The central claim is Theorem 5.1: let $P$ belong to $C^2_\psi(\Omega)$ with $0<a\le P(\omega)\le b$ on the compact hull $\Omega$, and consider the family of almost periodic forcings $p_\omega(t)=P(\omega+\psi(t))$. For almost every phase $\omega$, the escaping set $E_\omega=\{(t_0,v_0)\in\mathbb{R}\times(v_*,\infty): \text{the orbit is defined for all } n \text{ and } \lim v_n=\infty\}$ has Lebesgue measure zero in $\mathbb{R}^2$. The proof rewrites the ping-pong map as a measure-preserving embedding of $\Omega\times(E_*,\infty)$, uses $W(\omega,E)=P(\omega)^2 E$ as a generalized adiabatic invariant, and applies a refined Poincaré recurrence theorem to show that almost every unbounded orbit returns to bounded energy infinitely often. The escaping set is thereby contained in a measure-zero exceptional set. A final step converts the statement from energy coordinates back to original time-velocity coordinates, using the fact that the ping-pong map is area-preserving.

Load-bearing premise

The argument depends on the second derivative of the wall motion being uniformly continuous over the whole set of phases, because that uniformity is what makes the one-step error in the conserved quantity shrink to zero as the particle's speed grows.

Editorial extensions

If this is right

For every sufficiently smooth almost periodic forcing, typical initial data do not experience unbounded Fermi acceleration; the escaping set has Lebesgue measure zero.
The theorem extends the quasi-periodic result to the whole class of almost periodic forcings, so the escape is rare not only for periodic or quasi-periodic walls but also for walls whose motion is only almost periodic.
The same framework applies to other systems with almost periodic forcing, in particular the boundedness problem for a superlinear oscillator, where the associated escaping set is also measure zero for typical phases.
Because the ping-pong map is area-preserving, the measure-zero property is invariant under time translations: if a phase $\omega$ has zero escaping measure, then so does every time-translate $\omega\cdot s$.

Reading between the lines

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

The proof suggests a general principle: any forced system whose error in the adiabatic invariant decays with energy and whose forcing hull is a compact minimal flow will have a measure-zero escaping set, regardless of whether the forcing is periodic, quasi-periodic, or almost periodic.
If the forcing's second derivative is not uniformly continuous on the hull, the key error $\Delta(E)$ may fail to vanish, so the natural place to search for exceptions is just below the $C^2_\psi$ regularity imposed here.
A numerical experiment on a two-frequency quasi-periodic forcing could check the finite-time analogue: the fraction of initial conditions whose velocity exceeds a large threshold after many collisions should shrink to zero for a generic phase as the threshold and collision number grow.

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.

Desk Editor's Note

read the letter

Schließauf proves what the title says: for Bohr almost periodic forcing functions with P in C^2_psi, the escaping set has measure zero for almost every phase. This was the open question at the end of KO18, and the generalization is not a formality—it requires replacing the torus by a compact minimal flow and doing real work in that setting. The paper does that cleanly. I think the reader's verdict is right: accept after a small revision.

The main work is adapting KO18's machinery to the almost periodic setup. The measure decomposition along the flow (Lemma 2.4) is proved in detail; the lifted map f on Omega times (E*, infinity) is shown to be injective and measure-preserving; and the adiabatic invariant W = P^2 E is verified using the KO10 estimate. The proof checks the hypotheses of the generalized recurrence theorem honestly. The Littlewood remark is a nice bonus, and the citation pattern is appropriate—the author's own prior work appears only in a remark.

My main concern is Lemma 5.4. The text replaces the KO10 oscillation term, which involves time differences |t-s| <= C E0^{-1/2}, with a sup over metric balls in Omega of radius C E0^{-1/2}. That step silently assumes psi is Lipschitz. It is not in general: for a solenoid hull, d(psi(t),0) ~ C |t| log(1/|t|), so a time window of size delta can leave the metric ball of radius delta. The estimate is still true if you keep the oscillation in time and use uniform continuity of d^2_psi P and of psi itself (any continuous homomorphism from R to a compact group is uniformly continuous). So the gap is repairable with a couple of lines, but as written the proof is incomplete. Second, the final paragraph calls g_omega area-preserving; actually it preserves v dt dv, not dt dv. Since v > v* > 0 on the domain, the measures are equivalent, so the measure-zero transfer still goes through. Also, Lemma 4.3 has a sign typo (>= should be <=), but it does not affect the conclusion.

Bottom line: the result is new, the architecture is sound, and the flaws are minor and local. The paper deserves a serious referee; with a corrected Lemma 5.4 and a 'preserves an equivalent measure' caveat, it should be accepted.

Referee Report

2 major / 2 minor

Summary. The paper proves that for a Bohr almost periodic forcing function P(ω + ψ(t)) of class C^2_ψ(Ω), the escaping set Eω = {(t0,v0) : lim_n v_n = ∞} in the Fermi–Ulam ping-pong model has Lebesgue measure zero for almost every phase ω ∈ Ω. The proof represents almost periodic functions on a compact abelian group, introduces an abstract class of measure-preserving successor maps, proves a generalized recurrence theorem for escaping sets (Theorem 3.1), and then verifies its hypotheses for the ping-pong map using the adiabatic invariant W(ω,r) = P(ω)^2 r and an external analytic estimate (Lemma 5.3) controlling the change of this invariant over one collision.

Significance. If correct, the theorem answers a question raised in [KO18] by extending the quasi-periodic result to Bohr almost periodic forcings. The paper is genuinely non-perturbative in the sense that it uses no fitted parameters and the main mechanism is a recurrence argument rather than invariant curves. The exposition is largely self-contained: the Haar-measure decomposition, the abstract recurrence theorem, and the verification of its hypotheses are given in detail, and the one substantial external input, Lemma 5.3 from [KO10], is cited transparently. The two technical issues described below are local and repairable, but they affect load-bearing steps and therefore require revision.

major comments (2)

[§5, Lemma 5.4 and proof of Theorem 5.1] The bound replacing the time-window quantity Δ(t0,E0) of Lemma 5.3 by the metric-ball supremum in Δ(E0) implicitly uses an estimate of the form d(ψ(t),ψ(s)) ≤ C|t−s|. The manuscript only assumes ψ is a continuous homomorphism with dense image (§2.1), and for the standard solenoid metric this Lipschitz bound fails: d(ψ(t),0) behaves like |t| log(1/|t|) rather than |t|. Since Lemma 5.4 is exactly the input that produces k(E0) = CΔ(E0) → 0 in condition (3.5), the proof of Theorem 5.1 is incomplete as written. The gap is repairable: define Δ(E0) directly as a supremum over the time constraints |t−s| ≤ CE0^{-1/2} and use uniform continuity of ∂²ψP on the compact hull Ω, or choose a compatible metric making ψ Lipschitz.
[§5, final paragraph of the proof of Theorem 5.1] The claim that gω is area-preserving is not correct in (t,v) coordinates. From (5.7)–(5.8), the map preserves dt dE, so the Jacobian of the (t,v) map is v0/v1 and the preserved measure is v dt dv, not dt dv. The transfer from λ2(˜Eω) = 0 to λ2(Eω) = 0 can be made rigorous because v dt dv is equivalent to Lebesgue measure on {v > v*}, but the current argument, which relies literally on area preservation, is not valid as printed and needs a short correction.

minor comments (2)

[§4, proof of Lemma 4.3] In the estimate for λ(A_{j,ω}), the displayed inequality λ(A_{j,ω}) ≥ 2β^{-1}ε_j has the wrong direction; Lipschitz continuity of w^{-1} with constant β^{-1} gives ≤, which is the direction used in the following display. The statement also mixes γ and δ when writing 1/(4γ)W_j versus 1/(4δ)W_j.
[§3.2 and §4] There are small typos: in §3.2 the forward iterates are written as f^n_ω(t0,t0) instead of f^n_ω(t0,r0), and in Step 2 of the proof of Theorem 3.1 'To j ∈ Z' should read 'For j ∈ Z'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main theorem is derived from the external KO10 adiabatic estimate and the KO18 recurrence framework; no fitted prediction is renamed as a result.

full rationale

The derivation chain is not circular. Theorem 5.1 is proved by constructing the map f of the form (3.2), choosing the adiabatic quantity W(ω0,E0)=P(ω0)^2E0, and verifying inequality (3.5) via Lemma 5.4, which is a uniform version of Lemma 5.3 quoted from Kunze-Ortega (KO10, Lemma 5.1). That estimate is an external, parameter-free input controlling |p(t1)^2E1 - p(t0)^2E0| by C(E0^{-1/2}+ oscillation of ∂²ψP); it is not a restatement of the measure-zero conclusion. Theorem 3.1 is attributed to KO18 but is stated and proved in Section 4 using Dolgopyat's recurrence lemma, an independent tool, so the proof does not assume the target result. The function W is not fitted to the escaping set, and no predicted quantity is equal by construction to an input. The only same-author citation, [Sch19], appears in Remark 5.5 as a non-load-bearing remark about an analogous problem. A technical concern that Lemma 5.4's transfer from a time window to a metric ball may require a Lipschitz property of ψ is a possible proof gap or regularity issue, not an instance of circularity; it does not change the finding that the paper's claimed derivation reduces to independent external results rather than to its own conclusion.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

The central claim rests on standard ergodic-theoretic and topological-group facts plus two external analytic lemmas from the Fermi-Ulam literature (generating function and adiabatic estimate). There are no fitted constants, no ad hoc parameters, and no new physical entities; the constants β, δ, C, E* are chosen analytically and are not fitted to data.

assumptions (7)

standard math The hull H_u of a Bohr almost periodic function is a commutative compact topological group under pointwise convergence of translations, and u is represented by U(w)=w(0).
Used in Section 2.2 to reduce almost periodic functions to continuous functions on a compact group; standard theory due to Nemytskii and Stepanov.
standard math Every compact commutative topological group carries a unique translation-invariant Borel probability measure, the Haar measure.
Used in Section 2.3 and Lemma 2.4 to compute measures; cites Pontryagin and Hewitt-Ross.
standard math Poincaré recurrence theorem for finite measure spaces (Lemma 4.1).
Basis for showing almost every unbounded orbit is recurrent; stated and used in Section 4.
standard math First return map to a finite-measure set preserves the induced measure (EW11, Lemma 2.43).
Used inside the proof of Lemma 4.2 to build the induced map S.
domain assumption The ping-pong map in (t,E) coordinates is measure-preserving because it possesses a generating function (KO10, Lemma 3.7).
Imported from KO10; used in the proof of Theorem 5.1 to establish that f and g_ω preserve the product of Haar and Lebesgue measure.
domain assumption Adiabatic invariant estimate: for the ping-pong map, |p(t1)^2 E1 - p(t0)^2 E0| ≤ C(E0^{-1/2} + sup oscillation of p'' on a time window of length C + C E0^{-1/2}), reproduced as Lemma 5.3 from KO10.
Central external analytic input; ensures W = P(ω)^2 E satisfies the Lyapunov condition (3.5) with k(r) → 0.
standard math Implicit function theorem for continuous functions (Biasi-Gutierrez-dos Santos) applies to solve τ = P(ω0+ψ(τ))/√(2E0).
Used near equation (5.9) to define F and G for the embedding f.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Escaping orbits are also rare in the almost periodic Fermi-Ulam ping-pong." pith.science (2026). https://pith.science/paper/PBKEY6JP

@misc{pith2026190802529,
  author       = {Pith},
  title        = {Pith review of: Escaping orbits are also rare in the almost periodic Fermi-Ulam ping-pong},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PBKEY6JP}},
  note         = {Machine review of arXiv:1908.02529}
}

read the original abstract

We study the one-dimensional Fermi-Ulam ping-pong problem with a Bohr almost periodic forcing function and show that the set of initial condition leading to escaping orbits typically has Lebesgue measure zero.

Figures

Figures reproduced from arXiv: 1908.02529 by the authors.

**Figure 1.** On the 2-torus T 2 , intersections of Σ = {0}×T and the orbit of ψ(t) are separated by time intervals of length S = 1/ν1. 2.2 Almost periodic functions The notion of almost periodic functions was introduced by H. Bohr as a generalization of strictly periodic functions [Boh25]. A function u ∈ C(R) is called (Bohr) almost periodic, if for any ǫ > 0 there is a relatively dense set of ǫ-almost-periods of this function. … view at source ↗

**Figure 2.** Let χ(σ, t) = ( ˜σ, s). The map χ ‘divides out’ every complete period of ϕ ◦ ψ, i.e. s = t mod S , while preserving the relation ˜σ · s = ω = σ · t. for every (σ, t) ∈ Σ × R, where ⌊·⌋ indicates the floor function. This representation shows that χ is measure-preserving on every strip Σ × [t, t + S ) of width S , since µΣ and λ are invariant under translations in Σ and R, respectively. Moreover, the equality χ(Φ −1 (… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 27 canonical work pages

[1]

Besicovitch

A.S. Besicovitch. On generalized almost periodic functions. Proceedings of the London Mathematical Society , s2-25(1):495--512, 1926

work page 1926
[2]

Biasi, C

C. Biasi, C. Gutierrez, and E.L. dos Santos. The implicit function theorem for continuous functions. Topological Methods in Nonlinear Analysis , 32(1):177--185, 2008

work page 2008
[3]

S. Bochner. Beiträge zur Theorie der fastperiodischen Funktionen . Mathematische Annalen , 96(1):119--147, 1927

work page 1927
[4]

H. Bohr. Zur Theorie der fast periodischen Funktionen . Acta Mathematica , 45:29--127, 1925

work page 1925
[5]

Campos and M

J. Campos and M. Tarallo. Nonmonotone equations with large almost periodic forcing terms. Journal of Differential Equations , 254(2):686--724, 2013

work page 2013
[6]

Dolgopyat

D. Dolgopyat. Lectures on Bouncing Balls . https://www.math.umd.edu/ dolgop/BBNotes2.pdf. [Online; accessed 6-August-2019]

work page 2019
[7]

Dolgopyat

D. Dolgopyat. Bouncing balls in non-linear potentials. Discrete and Continuous Dynamical Systems , 22(1):165--182, 2008

work page 2008
[8]

Dolgopyat

D. Dolgopyat. Geometric and Probabilistic Structures in Dynamics , chapter Fermi acceleration, pages 149--166. AMS, Providence/RI 2008

work page 2008

Show all 27 references

[9]

de Simoi and D

J. de Simoi and D. Dolgopyat. Dynamics of some piecewise smooth Fermi-Ulam models . Chaos: An Interdisciplinary Journal of Nonlinear Science , 22(2):026124, 2012

2012
[10]

Einsiedler and T

M. Einsiedler and T. Ward. Ergodic Theory . Springer London, 2011

2011
[11]

E. Fermi. On the Origin of the Cosmic Radiation . Phys. Rev. , 75:1169--1174, 1949

1949
[12]

Hewitt and K

E. Hewitt and K. Ross. Abstract Harmonic Analysis I . Springer New York, 1979

1979
[13]

Kunze and R

M. Kunze and R. Ortega. Complete Orbits for Twist Maps on the Plane: Extensions and Applications . Journal of Dynamics and Differential Equations , 23(3):405--423, 2010

2010
[14]

Kunze and R

M. Kunze and R. Ortega. Escaping orbits are rare in the quasi-periodic Fermi-Ulam ping-pong . Ergodic Theory and Dynamical Systems , page 1–17, 2018

2018
[15]

Laederich and M

S. Laederich and M. Levi. Invariant curves and time-dependent potentials . Ergodic Theory and Dynamical Systems , 11(02), 1991

1991
[16]

J. Moser. On invariant curves of area-preserving mappings of an annulus . Nachr. Akad. Wiss. Göttingen, II , pages 1--20, 1962

1962
[17]

Nemytskii and V.V

V.V. Nemytskii and V.V. Stepanov. Qualitative Theory of Differential Equations . Priceton Univ. Press, 1960

1960
[18]

Ortega and M

R. Ortega and M. Tarallo. Almost periodic linear differential equations with non-separated solutions. Journal of Functional Analysis , 237(2):402--426, 2006

2006
[19]

Pontryagin

L.S. Pontryagin. Topological Groups . Gordon & Breach, 1966

1966
[20]

Pustyl'nikov

L.D. Pustyl'nikov. On Ulam's problem . Theoretical and Mathematical Physics , 57(1):1035--1038, 1983

1983
[21]

Schließauf

H. Schließauf. Escaping orbits are rare in the quasi-periodic Littlewood boundedness problem . Nonlinear Differential Equations and Applications NoDEA , 26, 2019

2019
[22]

De Simoi

J. De Simoi. Fermi acceleration in anti-integrable limits of the standard map. Communications in Mathematical Physics , 321(3):703--745, 2013

2013
[23]

Stepanoff (V.V

W. Stepanoff (V.V. Stepanov). Über einige Verallgemeinerungen der fast periodischen Funktionen . Mathematische Annalen , 95(1):473--498, 1926

1926
[24]

S.M. Ulam. On Some Statistical Properties of Dynamical Systems . In Proc. of the Fourth Berkeley Symposium on Math. Statistics and Probability, Volume 3: Contributions to Astronomy, Meteorology, and Physics , pages 315--320, Berkeley, 1961. University of California Press

1961
[25]

H. Weyl. Integralgleichungen und fastperiodische Funktionen . Mathematische Annalen , 97(1):338--356, 1927

1927
[26]

Zharnitsky

V. Zharnitsky. Instability in Fermi-Ulam ping-pong problem . Nonlinearity , 11(6):1481--1487, 1998

1998
[27]

Zharnitsky

V. Zharnitsky. Invariant curve theorem for quasiperiodic twist mappings and stability of motion in Fermi-Ulam problem . Nonlinearity , 13(4):1123--1136, 2000

2000

Pith tools

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

[1] [1]

Besicovitch

A.S. Besicovitch. On generalized almost periodic functions. Proceedings of the London Mathematical Society , s2-25(1):495--512, 1926

work page 1926

[2] [2]

Biasi, C

C. Biasi, C. Gutierrez, and E.L. dos Santos. The implicit function theorem for continuous functions. Topological Methods in Nonlinear Analysis , 32(1):177--185, 2008

work page 2008

[3] [3]

S. Bochner. Beiträge zur Theorie der fastperiodischen Funktionen . Mathematische Annalen , 96(1):119--147, 1927

work page 1927

[4] [4]

H. Bohr. Zur Theorie der fast periodischen Funktionen . Acta Mathematica , 45:29--127, 1925

work page 1925

[5] [5]

Campos and M

J. Campos and M. Tarallo. Nonmonotone equations with large almost periodic forcing terms. Journal of Differential Equations , 254(2):686--724, 2013

work page 2013

[6] [6]

Dolgopyat

D. Dolgopyat. Lectures on Bouncing Balls . https://www.math.umd.edu/ dolgop/BBNotes2.pdf. [Online; accessed 6-August-2019]

work page 2019

[7] [7]

Dolgopyat

D. Dolgopyat. Bouncing balls in non-linear potentials. Discrete and Continuous Dynamical Systems , 22(1):165--182, 2008

work page 2008

[8] [8]

Dolgopyat

D. Dolgopyat. Geometric and Probabilistic Structures in Dynamics , chapter Fermi acceleration, pages 149--166. AMS, Providence/RI 2008

work page 2008