Action-Angle Variables for Axisymmetric Potentials via Birkhoff Normalization
Sam Hadden
arXiv:2404.16941v1 Announce Type: new
Abstract: We describe a method for calculating action-angle variables in axisymmetric galactic potentials using Birkhoff normalization, a technique from Hamiltonian perturbation theory. An advantageous feature of this method is that it yields explicit series expressions for both the forward and inverse transformations between the action-angle variables and position-velocity data. It also provides explicit expressions for the Hamiltonian and dynamical frequencies as functions of the action variables. We test this method by examining orbits in a Miyamoto-Nagai model potential and compare it to the popular St”ackel approximation method. When vertical actions are not too large, the Birkhoff normalization method achieves fractional errors smaller than a part in $10^{3}$ and outperforms the St”ackel approximation. We also show that the range over which Birkhoff normalization provides accurate results can be extended by constructing Pad’e approximants from the perturbative series expressions developed with the method. Numerical routines in Python for carrying out the Birkhoff normalization procedure are made available.arXiv:2404.16941v1 Announce Type: new
Abstract: We describe a method for calculating action-angle variables in axisymmetric galactic potentials using Birkhoff normalization, a technique from Hamiltonian perturbation theory. An advantageous feature of this method is that it yields explicit series expressions for both the forward and inverse transformations between the action-angle variables and position-velocity data. It also provides explicit expressions for the Hamiltonian and dynamical frequencies as functions of the action variables. We test this method by examining orbits in a Miyamoto-Nagai model potential and compare it to the popular St”ackel approximation method. When vertical actions are not too large, the Birkhoff normalization method achieves fractional errors smaller than a part in $10^{3}$ and outperforms the St”ackel approximation. We also show that the range over which Birkhoff normalization provides accurate results can be extended by constructing Pad’e approximants from the perturbative series expressions developed with the method. Numerical routines in Python for carrying out the Birkhoff normalization procedure are made available.