Continuation of Hamiltonian dynamics from the plane to constant-curvature surfaces

This paper establishes a geometric framework using Inönü–Wigner contractions and local slice constructions to prove that non-degenerate relative equilibria and periodic orbits in Hamiltonian systems persist when deforming from the Euclidean plane to constant-curvature surfaces, with specific application to the Newtonian $n$-body problem.

The Gist

Imagine you are a physicist playing with a set of toy planets. For centuries, we've studied how these planets dance around each other on a perfectly flat, infinite sheet of paper (the Euclidean plane). We know exactly how they move: they form triangles, they spin in circles, and sometimes they chase each other in a figure-eight pattern.

But what if the paper isn't flat? What if the universe is actually a giant, smooth ball (like a sphere) or a saddle-shaped surface (like a hyperbolic plane)? Do those beautiful dance moves still work?

This paper, by Cristina Stoica, answers that question with a resounding "Yes, but..."

Here is the breakdown of the research using simple analogies.

1. The Big Idea: Bending the Stage

Think of the flat plane as a trampoline that is perfectly taut. Now, imagine slowly inflating a balloon underneath it. The trampoline starts to curve.

• The Goal: The author wants to prove that if you have a stable dance move (a "Relative Equilibrium") on the flat trampoline, you can gently curve the surface, and the dancers will find a new, slightly different way to keep dancing without crashing into each other.
• The Catch: This only works if the curve is gentle. If you try to bend the trampoline into a tiny, tight ball instantly, the dancers might fly off. But if you curve it slowly (mathematically, as a parameter $\epsilon$ gets smaller), the dance persists.

2. The Secret Weapon: The "Magic Lens" (Exponential Coordinates)

To prove this, the author needed a way to compare the flat world and the curved world directly. You can't easily compare a flat map to a globe because the shapes get distorted.

• The Analogy: Imagine you have a magic lens (called Riemannian exponential coordinates). If you look at the North Pole of a sphere through this lens, the curved surface looks almost exactly like a flat piece of paper right in front of you.
• The Trick: The author uses this lens to "flatten" the curved surface temporarily. This allows her to write down the laws of physics for the curved world using the same "language" (mathematical coordinates) as the flat world.
• The Result: She shows that the curved laws of physics are just the flat laws plus a tiny "curvature tax" (a small correction term). As the curvature gets smaller, the tax disappears, and you get back the flat world.

3. The Dance Moves: Equilibria vs. Drifting

The paper focuses on two types of special dances:

• The Perfect Spin (Relative Equilibrium): Imagine three planets forming a triangle and spinning around a center point. On a flat surface, if they spin, they stay in one spot.

  • The Discovery: The author proves that on a curved surface, this spinning triangle still exists! It just spins at a slightly different speed or size to compensate for the curve.
  • A Surprising Detail: On a flat surface, if the whole triangle is spinning, it must be sitting still (not drifting). If it were drifting, it would be a different kind of dance. On a curved surface, "spinning" and "drifting" get a little mixed up, but the author shows how to untangle them.

• The Drifting Dance (Relative Periodic Orbits): Imagine the planets spinning, but the whole group is also slowly sliding across the floor. On a flat floor, they slide in a straight line.

  • The Discovery: On a curved surface, you can't slide in a straight line forever (the floor curves away). Instead, the "slide" turns into a slow, gentle rotation or a "drift" along the curve. The author proves that these drifting dances also survive the transition from flat to curved, provided they are stable to begin with.

4. The "Figure-Eight" Example

One of the most famous dances in physics is the "Figure-Eight" choreography, where three equal-mass planets chase each other in a figure-eight loop.

• The paper confirms that this specific, complex dance is robust. Even if you live on a curved universe (like a sphere), the planets can still perform this figure-eight, provided the universe isn't too curved. They just have to adjust their speed slightly to match the curvature.

5. The Mathematical "Glue" (Lie Algebra Contraction)

How did she connect the math of a sphere to the math of a flat plane?

• The Analogy: Think of the symmetry groups as the "rules of movement."
  • On a flat plane, you can move Left/Right, Up/Down, and Spin.
  • On a sphere, you can't move "straight" forever; you eventually curve back.
• The author uses a mathematical technique called Inönü–Wigner contraction. Imagine a rubber band connecting the rules of the sphere to the rules of the flat plane. As you stretch the rubber band (making the sphere larger and flatter), the curved rules slowly morph into the flat rules. This allows her to use the known stability of the flat dance moves to prove the stability of the curved ones.

Summary

In plain English, this paper says:

"If you have a stable, non-colliding dance of planets on a flat map, you can gently bend the map into a sphere or a saddle shape, and the planets will find a way to keep dancing. They won't crash; they will just adjust their steps slightly to fit the new shape of the universe."

It's a proof of resilience. The beautiful patterns we see in our flat universe aren't fragile accidents; they are robust features that survive even if the geometry of space changes.

Technical Summary

1. Problem Statement

The paper addresses a fundamental question in mathematical physics: Do known solutions of dynamical systems on the Euclidean plane persist when the underlying geometry is deformed to a constant-curvature surface?

Specifically, the author investigates the Newtonian $n$-body problem (and general cotangent bundle Hamiltonian systems) on a surface $M^\sigma_R$ of radius $R = 1/\varepsilon$ and curvature sign $\sigma \in \{+1, -1\}$ (representing the sphere $S^2$ and hyperbolic plane $H^2$, respectively). The goal is to prove that Relative Equilibria (RE), Periodic Orbits, and Relative Periodic Solutions (RPOs) existing in the flat limit ($\varepsilon \to 0$) continue smoothly to the curved space for sufficiently small $\varepsilon > 0$, provided standard non-degeneracy conditions are met.

2. Methodology

The author employs a geometric perturbation approach that reduces the problem of varying geometry to a perturbation problem on a fixed symplectic manifold. The methodology relies on three key ingredients:

A. Riemannian Exponential Coordinates

The author introduces a coordinate system centered at the North pole of the curved surface using the Riemannian exponential map, $\text{Exp}^\varepsilon_N$.

• Chart Properties: This map provides a single chart covering the sphere (minus the antipodal point) and the entire hyperbolic plane.
• Metric Expansion: In these coordinates, the metric and momentum map admit explicit expansions in powers of $\varepsilon$.
• Symplectic Invariance: Crucially, the pull-back of the canonical symplectic form via the exponential chart is independent of $\varepsilon$. This allows the curved problem to be viewed as a family of Hamiltonian systems on a fixed phase space $T^*U^n$ (where $U \subset \mathbb{R}^2$).

B. Lie Algebra Inönü–Wigner Contraction

To handle the changing symmetry groups, the paper utilizes the Inönü–Wigner contraction of Lie algebras.

• Symmetry Groups: The symmetry group of the curved space is $G_\varepsilon \cong SO(3)$ (for $\sigma=+1$) or $SO^+(2,1)$ (for $\sigma=-1$). In the flat limit, this contracts to the Euclidean group $G_0 \cong SE(2)$.
• Contraction Mechanism: By defining an $\varepsilon$-dependent Lie bracket on a fixed vector space, the author interpolates between the curved algebras ($\mathfrak{so}(3)$ or $\mathfrak{so}(2,1)$) and the planar algebra ($\mathfrak{se}(2)$).
• Momentum Map: This framework ensures that the momentum map and group actions vary smoothly with $\varepsilon$, converging to the standard planar momentum map as $\varepsilon \to 0$.

C. Local Slice Construction

The author constructs local slices (transversal submanifolds) within the level sets of the momentum map.

• Reduction: These slices represent the symplectic reduced spaces locally.
• Smooth Dependence: The reduced symplectic forms, reduced Hamiltonians, and reduced vector fields on these slices depend smoothly on $\varepsilon$.
• Implicit Function Theorem: This smooth dependence allows the application of the Implicit Function Theorem to prove the existence of continued solutions.

3. Key Contributions

1. Unified Framework for Full Symmetry Groups:
   Unlike previous works (e.g., Bengochea et al.) that restricted analysis to rotational symmetry ($SO(2)$), this paper treats the full symmetry groups ($SE(2)$ in the plane, $SO(3)$/$SO^+(2,1)$ on curved surfaces). This is a significant generalization that captures translational dynamics and their curvature-induced modifications.

2. Explicit Hamiltonian Expansion:
   The paper derives the explicit expansion of the Newtonian $n$-body Hamiltonian in exponential coordinates:
   $$H_{\varepsilon, \sigma} = H_0 + \sigma \varepsilon^2 H_2 + O(\varepsilon^4)$$
   where $H_0$ is the standard planar Newtonian Hamiltonian and $H_2$ is the first-order curvature correction term involving both kinetic and potential energy modifications.

3. Distinction between REs and RPOs in Planar Limits:
   A novel insight is the clarification of planar Relative Equilibria under full $SE(2)$ symmetry. The author proves that a planar RE with non-zero rotational angular velocity must have zero total linear momentum to be a true RE. If the linear momentum is non-zero, the solution is actually a Relative Periodic Orbit (RPO) (a rotating configuration with a drifting center of mass). This distinction is crucial for correctly applying continuation theorems.

4. Continuation Theorems:
   The paper establishes rigorous theorems for the continuation of:

   • Relative Equilibria (RE): Non-degenerate REs on a slice continue to curved REs.
   • Periodic Orbits: Non-degenerate periodic orbits on regular energy surfaces continue.
   • Relative Periodic Solutions (RPOs): Non-degenerate RPOs continue, with the planar translational drift transforming into a "rotational drift" (small rotation or boost) on the curved surface.

4. Key Results

• Theorem 7 (Continuation of REs): Any non-degenerate planar RE (with zero linear momentum) continues to a unique family of REs on the constant-curvature surface for small $\varepsilon$.
  • Examples: The Lagrange triangular solutions, Euler collinear solutions, and regular $n$-gon configurations continue to the sphere and hyperbolic plane.
• Theorem 9 (Continuation of RPOs): Any non-degenerate planar RPO (including those with non-zero linear momentum) continues to a curved RPO.
  • Geometric Interpretation: The translational drift of the center of mass in the plane becomes a small rotational drift (or boost) on the curved surface, consistent with the Lie algebra contraction.
• Application to the Figure-Eight Choreography: The paper confirms that the famous figure-eight solution of the 3-body problem (which is a periodic orbit with zero angular momentum) continues to the curved surfaces as a nearby RPO (or periodic orbit if the drift is zero).
• Momentum Behavior: For curved REs, the "linear momentum" components are generally $O(\varepsilon^2)$ non-zero, reflecting the tilt of the momentum vector away from the vertical axis in the ambient space, vanishing only in the flat limit.

5. Significance

• Mathematical Rigor: The paper provides a rigorous geometric proof for the persistence of dynamical structures under curvature deformation, moving beyond numerical evidence or specific case studies.
• General Applicability: The framework is not limited to the $n$-body problem; it applies to any smooth family of $G_\varepsilon$-invariant Hamiltonian systems on cotangent bundles.
• Physical Insight: It clarifies how Euclidean symmetries (translations) deform into curved symmetries (rotations/boosts) and how this affects the classification of solutions (RE vs. RPO).
• Foundation for Future Work: The results lay the groundwork for studying collision regularization in curved spaces and the numerical continuation of specific choreographies, bridging the gap between flat-space celestial mechanics and curved-space dynamics.

In summary, Stoica's work successfully unifies the study of Hamiltonian dynamics on flat and constant-curvature surfaces by leveraging Lie algebra contractions and exponential coordinates, proving that the rich structure of planar solutions (including complex choreographies) persists in curved geometries under non-degeneracy conditions.