Workshop on Mathematical Aspects of Celestial Mechanics

Paris, Institut Henri Poincaré (amphi Darboux), December 11-20, 2007

Abstracts

Short course: Green's fiber bundles –a dynamical study

M.-C. Arnaud, Université d'Avignon (web site, pdf file of the slides of the course)

A well-known theorem due to Birkhoff asserts that every continuous graph invariant by a symplectic twist map of the annulus is Lipschitz. I have recently proved that these graphs are in fact more regular than Lipschitz: they are C1 on a residual set with full Lebesgue measure. The goal of my lectures is to present some analogous results for the so-called Tonelli Hamiltonians.

To do that, I will:

The final theorems that I will state concern the cotangent bundles of compact manifolds, but a lot of the notions that I use are valid in the non-compact case. Two of these theorems are:

  1. When the manifold is a surface: if G is a Lipschitz Lagrangian graph invariant under a Tonelli flow whose all the singularities are non degenerate, then G is C1 on a dense residual subset with full Lebesgue measure;
  2. if H is a Tonelli Hamiltonian which is C0 integrable, then there exists a residual subset of invariant Lipschitz Lagrangian graphs which are in fact C1.

First lecture: Basic results on Hamiltonian, Lagrangians, Lagrangian Lipschitz submanifolds, Hamilton-Jacobi equation and minimization properties

1) Hamiltonian and Lagrangian formalism

2) Lipschitz Lagrangian submanifolds, images of such manifolds and minimization properties

Second lecture: 3) Lagrangian subbunbles

An order relation on the Lagrangian subbundles which are transverse to the vertical. A notion of upper and lower semi-continuity for these bundles. In the convex case: links with the image of the vertical and the notion of conjugate points

Third lecture: 4) Green bundles

Construction of the Green bundles. A dynamical criterion. The reduced Green bundles. A characterization of hyperbolicity

Fourth lecture: 5) Regularity of Lipschitz Lagrangian invariant graphs

Generalized tangent vectors and cones. Generalized derivative and generalized tangent planes. Regularity of the invariant graphs for two degrees of freedom. More results of regularity.

Second species solutions of the 3 body problem

S. Bolotin, University of Wisconsin and Moscow Steklov Mathematical Institute
We discuss second species solutions of Poincaré for the plane three-body problem with two of the masses small. Such solutions shadow chains of collision orbits of two uncoupled Kepler problems. Levi-Civita regularization replaces double collisions by a 2-dimensional normally hyperbolic symplectic invariant manifold, and the problem is reduced to describing dynamics of a finite collection of symplectic maps.

Sur l'équation à l'aide de laquelle on détermine les inégalités séculaires des planètes (1766–1874).

F. Brechenmacher, IUFM du Nord-Pas de Calais

Cet exposé questionne l'identité algébrique d'une pratique propre à un corpus de textes publiés sur une période antérieure à l'élaboration de théories algébriques, comme la théorie des matrices, ou de disciplines, comme l'algèbre linéaire, qui donneront à cette pratique l'identité d'une méthode de transformation d'un système linéaire par la décomposition de la forme polynomiale de l'équation caractéristique associée. Dans les années 1760-1775, Lagrange élabore une pratique algébrique spécifique à la mathématisation des problèmes mécaniques des petites oscillations de cordes chargées d'un nombre quelconque de masses ou de planètes sur leurs orbites. La spécificité de cette pratique s'affirme par opposition à la méthode des coefficients indéterminés, elle consiste à exprimer les solutions des systèmes linéaires par des factorisations polynomiales d'une équation algébrique particulière, l'équation à l'aide de laquelle on détermine les inégalités séculaires des planètes. Élaborée en un jeu sur les primes et les indices des coefficients des systèmes linéaires, la pratique de Lagrange est à l'origine d'une caractéristique des systèmes issus de la mécanique, la disposition en miroirs des coefficients de systèmes que nous désignons aujourd'hui comme symétriques. Elle est également à l'origine d'une discussion sur la nature des racines de l'équation qui lui est associée. Nous questionnerons l'identité algébrique de cette discussion qui se développera sur plus d'un siècle en étudiant les héritages, permanences et évolutions de la pratique élaborée par Lagrange au sein de différentes méthodes élaborées dans divers cadres théoriques par des auteurs comme Laplace, Cauchy, Weierstrass, Jordan et Kronecker. Nous verrons que, préalablement à l'élaboration d'une théorie des formes dont la nature, algébrique ou arithmétique, suscitera une vive controverse entre Jordan et Kronecker en 1874, le caractère algébrique de la discussion renvoie davantage à l'identité historique d'un corpus qu'à une identité théorique et s'avère indissociable d'une constante revendication de généralité. Entre 1766 et 1874, la discussion sur l'équation à l'aide de laquelle on détermine les inégalités séculaires permet de mettre en évidence différentes représentations associées à une même pratique algébrique ainsi que des évolutions dans les philosophies internes portées par les auteurs du corpus sur la généralité de l'algèbre.

Link to the full text (pdf)

For a paper in english on this subject see Algebraic generality vs. arithmetic generality in the controversy between C. Jordan and L. Kronecker (1874) at http://arxiv.org/ftp/arxiv/papers/0712/0712.2566.pdf.

Birth control in generalized Hopf bifurcations

M. Chaperon, Université D. Diderot, Paris (web site)
This is about a very general lemma ensuring, at partially elliptic rest points of families of vector fields or transformations, the birth of normally hyperbolic invariant compact manifolds. A few examples and an idea of the proof will follow.

On the action minimizing retrograde solutions of the three-body problem

K.-C. Chen, National Tsing Hua University

In this talk we briefly outline a variational proof for the existence and minimizing properties of retrograde orbits for the planar three-body problem. These solutions begin with collinear configurations and they trace out pure braids in the three-dimensional space-time on some rotating frames. When some of masses are equal, our arguments can be applied to prove the existence of some prograde orbits and part of the Hénon family.

Spin-orbit resonances in the Solar system: periodic and quasi-periodic attractors in the small dissipation limit

L. Chierchia, Università di Roma Tre (web site)

In our Solar system there are more than 20 satellites revolving around the respective main bodies on an (apparently) exact 1:1 spin-orbit resonance (i.e. in one revolution period the satellite revolves one time around its spin-axis, like our Moon). There is only one other object observed in a spin-orbit resonance: Mercury, in a 3:2 resonance. A simple, nearly-integrable, weakly dissipative mathematical model is discussed:

  1. Numerical simulations show that periodic/quasi-periodic attractors coexist and their basins of attraction may be naturally interpreted as capture probabilities;
  2. A theorem concerning the existence of quasi-periodic attractors and their smooth small dissipation limits is presented.

Short course: Genericity of positive entropy for geodesic flows

G. Contreras, Cimat, Guanajuato

We will show the proof that in the C2-topology the geodesic flow of a generic Riemmanian metric on a compact manifold contains a hyperbolic and in particular it has positive entropy.

Lecture notes:

Perturbations of integrable Hamiltonian PDE

S. Kuksin, École polytechnique, Palaiseau

Majority of works on perturbations of integrable Hamiltonian PDE (under periodic boundary conditions) deals with the situation when the unperturbed equation is linear and depends on a certain number of parameters. These results seem to be rather far from the problems in the celestial mechanics. Indeed, there we also often deal with Hamiltonian problems of (almost) infinite dimension, but usually the unperturbed equations are non-linear and do not depend on extra parameters. In my lecture I will discuss results and open problems for perturbations of an integrable PDE which do not depend on extra parameters.

Geometry of Arnold diffusion

M. Levi, Penn State University
In our joint work with Vadim Kaloshin we give what is perhaps the simplest possible geometrical picture of the mechanism of Arnold diffusion. We choose to speak of a specific model –that of geometric rays in a periodic optical medium. This model is equivalent to that of a particle in a periodic potential in Rn with energy prescribed and to the geodesic flow in a Riemannian metric on Rn. Some related examples where the approach works will be mentioned as well.

Short course: Topics on the three-body problem

R. Moeckel, University of Minessota
  1. Reduction, shape space, Hill's regions and the collision manifold
  2. Topological proof of existence of the figure-Eight orbit
  3. Global regularization of double collisions
  4. Symbolic dynamics in the planar three-body problem

Short course: Adiabatic invariants

A. Neishtadt, Space Research Institute, Moscow, and Loughborough University

There are many problems that lead to the analysis of Hamiltonian dynamical systems in which one can distinguish motions of two types: slow motions and fast motions. Adiabatic perturbation theory is a mathematical tool for the asymptotic description of dynamics in such systems. This theory allows to construct adiabatic invariants, which are approximate first integrals of the systems. These quantities change by small amounts on large time intervals, over which the variation of slow variables is not small. Adiabatic invariants usually arise as first integrals of the system after having been averaged over the fast dynamics.

In the lectures it is planed to consider the following topics: exponential accuracy of conservation of adiabatic invariants in analytic one-frequency systems, jumps of adiabatic invariants at separatrices, destruction of adiabatic invariance due to captures into resonances and passages through resonances, examples of adiabatic invariance in problems of celestial mechanics.

Towards a global study of Area Preserving Maps. Applications to the stability around L4,5 in the restricted three-body problem

C. Simó, Universitat de Barcelona

We present a phenomenological study of Area Preserving Maps, with emphasis on simple models. Paradigmatic models like the standard map, the separatrix map and a new model, the biseparatrix map, are useful to undertand the dynamics.

These paradigms can be analysed in a reasonable way. They allow to explain the main features of dynamics. As test example the conservative Hénon map will be used.

Main points of interest are the escape of points from a given region when they are not surrounded by invariant rotational curves and the measure of the set of non-escaping points. Different rates of escape have been found.

The existence of regions where the dynamics is close to a diffusion, separated by cantori which are difficult to cross, having noble numbers as rotation number, are illustrated.

Applications are also made to the stability regions around triangular points in the RTBP as a function of the mass parameter.

The talk will be closed by several open questions.

Singularities and collisions of generalized solutions to the N-body problem

S. Terracini, Università di Milano Bicocca
The validity of Sundman-type asymptotic estimates for collision solutions is established for a wide class of dynamical systems with singular forces, including the classical N-body problems with Newtonian, quasi-homogeneous and logarithmic potentials. The solutions are meant in the generalized sense of Morse (locally -in space and time- trajectories with respect to compactly supported variations) and their uniform limits. The analysis includes the extension of the Von Zeipel's Theorem and the proof of isolatedness of collisions. Estimates on the contribution of collisions to the Morse index will be discussed. Furthermore, such asymptotic analysis is applied to prove the absence of collisions for locally minimal trajectories and, therefore, existence of new periodic and almost-periodic solutions for the N-body problem wich are equivariant under the action of an appropriate symmetry group.

Polymorphisms and adiabatic chaos

D. Treschev, Moscow Steklov Mathematical Institute

In the end of the last century Vershik introduced some dynamical systems called polymorphisms. Systems of this kind are multivalued self-maps of an interval, where (roughly speaking) each branch has some probability. By definition the standard Lebesgue measure should be invariant.

Unexpectedly polymorphisms appeared in the problem of destruction of an adiabatic invariant after a multiple passage through a separatrix.

We plan to discuss ergodic properties of polymorphisms.