Singular KAM theory for convex Hamiltonian systems L. Biasco and L. Chierchia were supported by the grant NRR-M4C2-I1.1-PRIN 2022-PE1-Stability in Hamiltonian dynamics and beyond-F53D23002730006-Financed by E.U.–NextGenerationEU. Santiago Barbieri was supported by the Juan de la Cierva Postdoctoral
S. Barbieri, L. Biasco, L. Chierchia, D. Zaccaria · 2025
Read Singular KAM theory for convex Hamiltonian systems L. Biasco and L. Chierchia were supported by the grant NRR-M4C2-I1.1-PRIN 2022-PE1-Stability in Hamiltonian dynamics and beyond-F53D23002730006-Financed by E.U.–NextGenerationEU. Santiago Barbieri was supported by the Juan de la Cierva Postdoctoral by S. Barbieri, L. Biasco, L. Chierchia, D. Zaccaria (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-2507-12101.txt.
Introduction
Santiago Barbieri and co-authors extend a stability theory for Hamiltonian systems in a brief note from two thousand twenty-five. KAM theory shows that many orderly motions survive small disturbances, and a singular version was built earlier for a simple mechanical case. The note sketches how that theory carries over to convex, analytic, nearly integrable systems of a general form, where the unperturbed part is convex and the disturbance is generic. It builds on the authors' prior work and addresses specialists in Hamiltonian dynamics.
Catalog record
| CREATOR | S. Barbieri, L. Biasco, L. Chierchia, D. Zaccaria |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-12101
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Singular KAM theory for convex Hamiltonian systems L. Biasco and L. Chierchia were supported by the grant NRR-M4C2-I1.1-PRIN 2022-PE1-Stability in Hamiltonian dynamics and beyond-F53D23002730006-Financed by E.U.–NextGenerationEU. Santiago Barbieri was supported by the Juan de la Cierva Postdoctoral Grant JDC2023-052632-I funded by the Spanish National Agency for Research (Agencia Estatal de Investigación) S. Barbieri, L. Biasco, L. Chierchia, D. Zaccaria Abstract In this note, we briefly discuss how the singular KAM Theory of [ 7 ] – which was worked out for the mechanical case 1 2 | y | 2 + ε f ( x ) \frac{1}{2}|y|^{2}+\varepsilon f(x) – can be extended to convex real analytic nearly integrable Hamiltonian systems with Hamiltonian in action–angle variables given by h ( y ) + ε f ( x ) h(y)+\varepsilon f(x) with h h convex and generic f f . MSC 2010: 37J05, 37J35, 37J40, 70H05, 70H08, 70H15 Keywords: Nearly–integrable Hamiltonian systems. Convex Hamiltonians. Measure of invariant tori. Simple resonances. Arnold–Kozlov–Neishtadt conjecture. Singular KAM Theory. In the Springer Encyclopaedia of Mathematical Sciences , Arnold, Kozlov and Neishtadt conjectured that the relative measure of phase space free of invariant tori of a real–analytic, nearly integrable general Hamiltonian System with three or more degrees of freedom is of the same order of the perturbation ; compare [ 2 ] . In [ 7 ] , in the special case of natural systems with a Hamiltonian function given (in action–angle variables) by H=\frac{1}{2}|y|^{2}+\varepsilon f(x) , it has been proven a related result, namely, that, for generic real–analytic f , the relative measure of invariant primary and secondary tori is at least O(1-\varepsilon|\log\varepsilon|^{c}) for some c>0 , in agreement (up to the logarithmic correction) with Arnold, Kozlov and Neishtadt conjecture 1 1 1 For related partial results, see [ 11 ] . . Proofs in [ 7 ] are based on a new “singular KAM theory”, which extends the classical theory of Kolmogorov, Arnold and Moser [ 8 , 1 , 12 ] , so as to deal, in particular, with primary and secondary tori arbitrarily close to action–angle singularities arising near simple resonances, uniformly in phase space . Let us recall that classical KAM theory predicts that the relative measure of primary tory is of order 2 2 2 Compare [ 9 , 13 , 14 , 15 ] . O(1-\sqrt{\varepsilon}) , and that such an estimate is optimal, as simple integrable cases show. In this paper we briefly discuss how to extend [ 7 ] to convex integrable Hamiltonian in place of the purely quadratic term \frac{1}{2}|y|^{2} . The main issue here is geometric: one has to quantitatively describe suitable neighborhoods of simple resonances, so as to be able to put, after averaging, the secular Hamiltonians in “generic standard form” (as defined in [ 6 ] ). More specifically, we introduce action–angle variables so that the secular Hamiltonian, after averaging over the (n-1) fast angles, depends only on one slow angle. Then, we show that the level sets of the derivative of the integrable part with respect to the slow action, are graphs over uniform domains of slow actions; see § 1.2 below. At this point, the arguments to show that the density of all KAM primary and secondary tori is at least of order O(1-\varepsilon|\log\varepsilon|^{c}) for some c>0 , follows, without extra difficulties, the arguments spelled out in [ 7 ] , which lead to the following statement, confirming (up to the logarithmic correction) the Arnold, Kozolv, Neishdatd conjecture, namely: For generic 3 3 3 The generic class we refer to is introduced in [ 6 ] , Sec.1. real–analytic potentials f on \mathbb{T}^{n} , the relative measure of invariant primary and secondary tori for the Hamiltonian H=h(y)+\varepsilon f(x) , with h strictly convex, is at least O(1-\varepsilon|\log\varepsilon|^{c}) for some c>0 . We also mention a technical improvement in the averaging theory discussed here, which may be useful in applications (for example, in celestial mechanics), namely, we allow for different analyticity radii in the angle variables; indeed, in singular KAM theory it is essential to have sharp control of analytic singularities in the angular variables and having a common width of analyticity strip is quite unnatural 4 4 4 For example, in the circular restricted three body problem the angles have different analyticity domains; compare [ 10 ] . . A second extension would be to replace the perturbation f with a generic function depending also on actions, but this is a much more difficult problem 5 5 5 For example, it is not obvious how to generalize the generic real–analytic class introduced in [ 7 ] in view of the zeros of f introduced, in general, by the action dependence. In this respect, as suggested to us by Laurent Niederman, it might be useful to consider the quantitative Morse–Sard theory developed by Yomdin and Comte in [ 16 ] ; compare [ 3 ] . . 1 Resonance analysis Consider a bounded phase space {\mathcal{M}}:=B\times\mathbb{T}^{n} , with B\subseteq\mathbb{R}^{n} a bounded convex open non empty set , and \mathbb{T}^{n} the standard flat n –torus \mathbb{R}^{n}/(2\pi\mathbb{Z}^{n}) , endowed with the standard symplectic two–form dy\wedge dx=\sum_{i=1}^{n}dy_{i}\wedge dx_{i} . Let H(y,x) be a real–analytic, nearly–integrable Hamiltonian on {\mathcal{M}} given by H(y,x)=H_{\varepsilon}(y,x)=h(y)+\varepsilon f(x);\qquad(y,x)\in{\mathcal{M}},\qquad 0\leq\varepsilon\leq 1. (1.1) Analyticity parameters We introduce quantitative analyticity parameters as follows. Let r>0 and {s}=({s}_{1},...,{s}_{n})\in\mathbb{R}^{n}_{+} be a vector with positive components. Denote by 6 6 6 As usual, ‘bar’ denotes complex–conjugated and ‘dot’ inner product. |z|=\sqrt{z\cdot\bar{z}} the standard Euclidean norm on vectors z\in{\mathbb{C}}^{n} , and define the following complex neighborhoods \displaystyle B_{\rho}:=\bigcup_{y\in B}\{z\in\mathbb{C}^{n}:|z-y|<\rho\}; \displaystyle\mathbb{T}_{{s}}^{n}:=\{x\in\mathbb{C}^{n}:|\mbox{Im}\ x_{i}|<{s}_{i}\ \forall\ i=1,...,n\}/(2\pi\mathbb{Z}^{n}). Henceforth, we assume that H is real–analytic and bounded on B_{2r}\times\mathbb{T}_{{s}}^{n} for some r>0 . Remark 1.1 . As mentioned above, the reason for allowing different analyticity widths in the angular variables is motivated by physical examples, such as the three–body problem, where this is the case. Having sharp control on the angle complex singularities is essential in analyzing fine properties of Hamiltonian systems, especially, in the contex of singular KAM Theory or, possibly, of Arnold diffusion. Convexity assumption In this paper we assume that h in ( 1.1 ) is \gamma –convex for some positive \gamma , i.e., we assume that \partial^{2}_{y}h(y)\ \xi\cdot\xi=\sum_{i,j=1}^{n}\partial_{y_{i}y_{j}}h(y)\,\xi_{i}\xi_{j}\geq\gamma|\xi|^{2}\,,\qquad\forall\xi\in\mathbb{R}^{n}\,,\ \forall y\in\,{\rm Re}\,(B_{2r})\,. (1.2) We shall also assume, without loss of generality, that the frequency map y\to\omega(y)=\partial_{y}h(y) is bi–Lipschitz and satisfies for some positive constants \bar{L} , L the inequalities \bar{L}^{-1}\leq\frac{|\omega(y)-\omega(y_{0})|}{|y-y_{0}|}\leq L\,,\ \forall\ y\neq y_{0}\in B_{2r}\,,\qquad\big(\omega(y):=\omega(y)=\partial_{y}h(y)\big)\,; (1.3) We also let M:=\sup_{y\in B_{2r}}|\omega(y)|<+\infty . 1.1 Geometry of resonances and coverings of B In singular KAM Theory [ 7 ] , as well as in Nekhoroshev’s Theory (see, e.g., [ 2 ] ), it is fundamental to consider carefully the geometry of resonances , which allows for high order averaging theory. We recall, that, given an integer vector k\in{\mathbb{Z}}^{n} , a resonance {\mathcal{R}}_{k} for the integrable Hamiltonian h(y) is the set \{y\in B:\omega(y)\cdot k=0\} ; a double resonance {\mathcal{R}}_{k,\ell} is a set given by {\mathcal{R}}_{k,\ell}={\mathcal{R}}_{k}\cap{\mathcal{R}}_{\ell} with k and \ell linearly independent integer vectors. In these definitions, the integer vectors may be assumed to have no common divisors; indeed, it is enough to consider integer vectors in the set {\mathcal{G}}^{n} defined as the subset of integer vectors k\in\mathbb{Z}^{n} with