A natural decomposition of the Jacobi equation for some classes of $N$-body problems
Renato Iturriaga, Ezequiel Maderna
TL;DR
This work develops a natural Hessian-based decomposition of the Jacobi equation for certain $N$-body problems under the mass inner product, enabling a decoupling into invariant subspaces. A central splitting lemma shows that Jacobi fields split along invariant subspaces, leading to a Meyer-Schmidt-type decomposition for planar central configurations and their homographic motions. The paper derives a simple analytic instability criterion for elliptic Lagrange solutions, matching Ou's result via a direct hyperbolicity argument on the essential part of the linearized flow. It further analyzes the orthogonal complement to equilateral configurations in the 3-body problem using a complex mass inner product to obtain explicit Hessian expressions and a signature condition that reproduces the $oxed{ ext{mu}<27/8}$ threshold.
Abstract
We consider several $N$-body problems. The main result is a very simple and natural criterion for decoupling the Jacobi equation for some classes of them. If $E$ is a Euclidean space, and the potential function $U(x)$ for the $N$-body problem is a $C^2$ function defined in an open subset of $E^N$, then the Jacobi equation along a given motion $x(t)$ writes $\ddot J=HU_x(J)$, where the endomorphism $HU_x$ of $E^N$ represents the second derivative of the potential with respect to the mass inner product. Our splitting in particular applies to the case of homographic motions by central configurations. It allows then to deduce the well known Meyer-Schmidt decomposition for the linearization of the Euler-Lagrange flow in the phase space, formulated twenty years ago to study the relative equilibria of the planar $N$-body problem. However, our decomposition principle applies in many other classes of $N$-body problems, for instance to the case of isosceles three body problem, in which Sitnikov proved the existence of oscillatory motions. As a first concrete application, for the classical three-body problem we give a simple and short proof of a theorem of Y. Ou, ensuring that if the masses verify $μ=(m_1+m_2+m_3)^2/(m_1m_2+m_2m_3+m_1m_3)<27/8$ then the elliptic Lagrange solutions are linearly unstable for any value of the excentricity.
