Geometry of Almost-Conserved Quantities in Symplectic Maps Part II. Recovery of approximate invariant

T. Zolkin, S. Nagaitsev, I. Morozov, S. Kladov · 2025

Read Geometry of Almost-Conserved Quantities in Symplectic Maps Part II. Recovery of approximate invariant by T. Zolkin, S. Nagaitsev, I. Morozov, S. Kladov (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-07224.txt.

Introduction

Tim Zolkin and co-authors show how near-conserved quantities in symplectic maps can be recovered from discrete symmetry, in a 2025 paper. Noether's theorem links smooth symmetries to exact conservation laws. Reversible symplectic maps hold a discrete symmetry instead, and it can seed approximate invariants. Building functions that respect the symmetry order by order uncovers those hidden quantities. This second part of the work applies the method to wide-ranging motion.

Once averaged, the approximate invariants trace resonances and the edges of stable regions with care. In fully integrable cases, the method recovers the exact invariants. One unified function yields phase portraits, rotation numbers, and tune maps that match direct tracking across wide settings. Against a rival matrix method, the new technique proved more accurate in resonant, strongly nonlinear regimes. Uses in beam physics are highlighted.

Catalog record

CREATORT. Zolkin, S. Nagaitsev, I. Morozov, S. Kladov
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE47 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-07224

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Geometry of Almost-Conserved Quantities in Symplectic Maps Part II. Recovery of approximate invariant T. Zolkin, S. Nagaitsev, I. Morozov, S. Kladov Abstract Noether’s theorem, which connects continuous symmetries to exact conservation laws, remains one of the most fundamental principles in physics and dynamical systems. In this work, we draw a conceptual parallel between two paradigms: the emergence of exact invariants from continuous symmetries, and the appearance of approximate invariants from discrete symmetries associated with reversibility in symplectic maps. We demonstrate that by constructing approximating functions that preserve these discrete symmetries order by order, one can systematically uncover hidden structures, closely echoing Noether’s framework. The resulting functions serve not only as diagnostic tools but also as compact representations of near-integrable behavior. The second article applies the method to global dynamics, with a focus on large-amplitude motion and chaotic systems. We demonstrate that the approximate invariants, once averaged, accurately capture the structure of resonances and the boundaries of stability regions. We also explore the recovery of exact invariants in integrable cases, showing that the method reproduces the correct behavior when such structure is present. A single unified function, derived from the map coefficients, yields phase portraits, rotation numbers, and tune footprints that closely match numerical tracking across wide parameter ranges. Comparisons with the Square Matrix method reveal that while both approaches satisfy local constraints, our technique provides greater accuracy and robustness in resonant and strongly nonlinear regimes. These results highlight the method’s practical power and broad relevance, offering a compact, analytic framework for organizing nonlinear dynamics in symplectic maps with direct applications to beam physics and beyond. I Introduction The first part of this study developed a systematic perturbative method for constructing approximate invariants in symplectic maps — discrete systems closely related to Hamiltonian dynamics and widely applicable across mathematical and physical problems [ 1 , 2 , 3 ] . These invariants are not only conserved up to a given order but also respect two families of discrete symmetries induced by reversibility [ 4 ] . In continuous-time dynamics, Noether’s theorem [ 5 ] establishes a correspondence between symmetries of the action and exact integrals of motion; here, we uncover a similar correspondence in the discrete setting. By benchmarking our method against results from Lie algebra technique and by probing the structure of low-order resonances, we demonstrated both the mathematical consistency and practical accuracy of the approach in the vicinity of a fixed point. In this second part, we shift the focus from foundational formulation to broader applications. Much like the classical Arnold scheme in KAM theory [ 6 , 7 , 8 ] , as the order of approximation increases, the simply connected domain where the approximate invariant holds shrinks toward a point. In this limit, the system approaches an integrable twist map — the same map that lies at the heart of Lie algebra benchmarking. However, our interest lies beyond this infinitesimal regime: we aim to extend the method into the realm of large-amplitude dynamics. On one end, we investigate near-integrable systems exhibiting chaos beyond the fixed point neighborhood. In such cases, the method uncovers coherent structures — such as chains of islands — that are otherwise obscured by standard twist expansions of action-angle variables. On the other end, we revisit known integrable maps and test whether the method can recover their exact global invariants. This is a critical test: failure in the integrable case would cast doubt on capturing global features in chaotic systems, where no exact invariant exists. We begin where the first part left off, by briefly restating the method for general planar maps. The remainder of this article is organized as follows. In Section III , we test the method on all known integrable maps with smooth invariants for transformation in McMillan form [ 9 ] , systematically described by Suris [ 10 ] . For maps with finite polynomial invariants, our results fully recover the expected structure; for maps whose invariants are infinite power series, we observe almost-everywhere convergence. In Section IV , we examine the method’s limitations by applying it to chaotic systems, focusing on the quadratic and cubic Hénon maps [ 11 ] . They are relatively well understood [ 12 , 13 , 14 ] , yet exhibit rich nonlinear dynamics, making them ideal for benchmarking new methods against known results. These maps are directly relevant to accelerator physics, modeling horizontal dynamics in lattices with thin sextupole and octupole lenses [ 15 , 16 ] . We use several case studies where the phase space contains large island chains, analyzing both the approximate invariants and the nonlinear rotation number derived from them, providing an alternative to traditional twist expansion. In Subsection IV.2 , we extend this analysis to the joint space of dynamical variables and map parameters, constructing approximations of the simply connected domain of validity. Building on recent work [Fractal], where integrable systems were shown to approximate the geometry of resulting fractal structures to first nonlinear order, we demonstrate that the perturbative approach enables further refinement — albeit with limitations — beyond the fixed points/ n -cycles framework. One crucial component of this method is the averaging procedure, which resolves under-determinacy in the expansion coefficients and ensures resonance compatibility (i.e., applicability at rational rotation numbers). As a case study, we compare this to the naive alternative of setting all undetermined coefficients to zero. Along the way, we draw comparisons with other analytical tools, particularly the Square Matrix (SM) method, and show how even low-order approximations can capture essential features of strongly nonlinear systems. In the following and final part of this series, we demonstrate its effectiveness by applying it to realistic dynamical models corresponding to accelerator lattice configurations from the FermiLab complex. As we show, this construction naturally extends the classic Courant-Snyder [ 17 ] formalism — a foundational framework in accelerator physics now spanning seven decades and historically focused on linear phenomena. This limitation stemmed from several factors: weak nonlinearities at small amplitudes, design goals favoring linearity, and the limited analytical tools available at the time (1953), preceding the full development of KAM theory and later progress in chaotic dynamics. While this is not the first attempt to describe near-integrable motion using normal forms, we believe the present approach holds promise for broader applications due to its conceptual simplicity. II Perturbation method Consider a map of the plane \mathrm{T} : \mathbb{R}^{2}\mapsto\mathbb{R}^{2} , which is symplectic (i.e., it preserves area and orientation) and admits a power series expansion: \begin{array}[]{cc}\displaystyle q^{\prime}=A_{1,0}\,q+A_{0,1}\,p+A_{2,0}\,q^{2}+A_{1,1}\,q\,p+A_{0,2}\,p^{2}+\ldots,\\[7.11317pt] \displaystyle p^{\prime}=B_{1,0}\,q+B_{0,1}\,p+B_{2,0}\,q^{2}+B_{1,1}\,q\,p+B_{0,2}\,p^{2}+\ldots.\end{array} \bullet We begin by introducing a general polynomial ansatz for the approximate invariant of order n : \mathcal{K}^{(n)}[p,q]=\mathcal{K}_{0}+\epsilon\,\mathcal{K}_{1}+\ldots+\epsilon^{n}\,\mathcal{K}_{n}, where \mathcal{K}_{m} is a homogeneous polynomial of degree (m+2) : \mathcal{K}_{m}[p,q]=\sum_{\begin{subarray}{c}i,j\geq 0\\ i+j=m+2\end{subarray}}C_{i,j}\,p^{i}q^{j}. To determine the coefficients C_{i,j} , we require that the approximate invariant be conserved up to \mathcal{O}(\epsilon^{n+1}) : \mathcal{R}_{n}=\mathcal{K}^{(n)}[p^{\prime},q^{\prime}]-\mathcal{K}^{(n)}[p,q]=\overline{\mathcal{R}_{n}}\,\epsilon^{n+1}+\mathcal{O}(\epsilon^{n+2}). The lowest-order contribution, which can also be derived via standard linearization, is defined only up to a constant multiplier \mathcal{K}_{0}[p,q]=C_{0}\left[A_{0,1}\,p^{2}+(A_{1,0}-B_{0,1})\,p\,q-B_{1,0}\,q^{2}\right]. (1) The seed coefficient C_{0} can be set to unity or chosen specifically to eliminate resonant denominators. \bullet Due to under-determinacy, the invariant (after all two-indexed coefficients C_{i,j} have been fixed) is known only up to a series: C_{0}\,\mathcal{K}_{0}+C_{1}\,\mathcal{K}_{0}^{2}+\ldots. Nevertheless, the approximate invariance condition holds to the required

Read the complete text