Regularized variational principles for the perturbed Kepler problem (Q2020409)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Regularized variational principles for the perturbed Kepler problem
scientific article

    Statements

    Regularized variational principles for the perturbed Kepler problem (English)
    0 references
    0 references
    0 references
    0 references
    23 April 2021
    0 references
    In this paper the authors propose a new method that combines the use of variational techniques with regularization methods in order to study existence and multiplicity results for the periodic and the Dirichlet problem associated to the perturbed Kepler system \[ \ddot{x} =-\frac{x}{|x|^3} + p(t), \quad x\in\mathbb{R}^d, \tag{1}\] where \(d\geq 1\), and \(p: \mathbb{R}\to \mathbb{R}^d\) is smooth and \(t\)-periodic, \(T>0\). The \textit{action functional}, associated to (1) is \[ \mathcal{A} (x) = \int_0^T \left[\frac{1}{2}|\dot{x}(t)|^2 + \frac{1}{|x(t)|} + \langle p(t),x(t) \rangle \right] \: dt, \] which is (well defined and) finite in \[ \mathfrak{X} := \left\{ x\in H^1(0,T;\mathbb{R}^d): \int_0^T \frac{dt}{|x(t)|} <\infty\right\}. \] It is natural to search for \(T\)-periodic solutions of (1) by looking for critical points of the action functional \(\mathcal{A}\) on \(\mathfrak{X}\). The main problem encountered with this approach is related to the so-called possibility of ``binary collisions'', i.e. \(x=0\). In order to regularize the singularity due to collision, the authors exploit some changes of variables (both for \(t\) and \(x\)) in such a way that the action functional in the new variables is regularized, that is, it does not contain a singular potential term. The basic idea is to regularize the system (1) by a change of variables from \(x =x(t)\) to \(y=y(\tau)\), where \[ \tau = \int_0^t \frac{d\xi}{|x(\xi)|}, \quad x= \Phi (y). \] Here \(\Phi\) is an appropriate homogeneous function of degree 2 (different choices are available, depending on the spatial dimension \(d\)). The different choices for \(\Phi\) are inspired by classical regularization techniques for the Kepler problem (Sundman, Levi-Civita, Kustaanheimo-Stiefel). Accordingly, the functional \(\mathcal{A}\) is transformed into a new functional \[ {\mathcal B} (y) = \int_0^1 \beta \left( \tau, y(\tau),y'(\tau), \int_0^\tau |y(\xi )|^2d\xi, \| y\|_2\right) \: d\tau, \] where \(\|\cdot \|_2\) is the norm in \(L^2((0,1); \mathbb{R}^d)\). Once the variational principles are established, the rest of the paper is devoted to applications, for either Dirichlet or periodic boundary conditions. First, they consider the minimization \({\mathcal B} \) in the Sobolev space \(H^1_0(0, 1; \mathbb{R}^d)\). This lead to the existence of a generalized solution of (1) satisfying the Dirichlet boundary conditions \(x(0) =x(T) =0\). This is done for any dimension \(d \geq 1\). In the last part of the paper, the authors assume that \(p\) is \(C^1\) on \(\mathbb{R}\) and \(T\)-periodic, and they look for periodic generalized solutions to (1) in dimension \(d =2\) where action functional is taken in the space of 1-periodic loops, and adapt the definition of \(\Phi\) to the classical Levi-Civita change of variables. Lastly, an analogous method is used for dimension 3, but the proof is more delicate and the definition of \(\Phi\) is now inspired by the Kustaanheimo-Stiefel change of variables.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    Kepler problem
    0 references
    forced problem
    0 references
    generalized solutions
    0 references
    regularization
    0 references
    constrained critical points
    0 references
    periodic solutions
    0 references
    0 references
    0 references