Orbits in an Axisymmetric Potential
A flattened galaxy is well described by the Miyamoto-Nagai potential $\Phi(R,z) = -GM / \sqrt{R^2 + (a + \sqrt{z^2 + b^2})^2}$, with cylindrical radius $R$, height $z$, disk scale length $a$ and thickness $b$. Because this is not an inverse-square point-mass field, the radial and azimuthal oscillation frequencies are incommensurate, so a generic bound orbit does not close: it precesses, filling a rosette bounded by an inner and outer radius. Energy $E$ and the angular momentum component $L_z$ about the symmetry axis are conserved and pin the rosette, while a third (non-classical) integral confines the motion in the meridional plane. The playground integrates an orbit with a symplectic step and shows the rosette plus the conserved quantities holding flat.
WHAT TO TRY
- Lower the azimuthal speed v_phi below the circular speed: the orbit is eccentric and the rosette tightens, sweeping between a near pericenter and a far apocenter.
- Add a vertical kick v_z: the star oscillates above and below the disk plane, and the meridional (R, z) trace fills its allowed box, the region inside the zero-velocity curve Phi_eff = E.
- Drag the 3D view to rotate it and watch the orbit weave a thick torus, never closing on itself, the generic behaviour in a non-Keplerian potential.
- Watch the readout: the energy and L_z hold flat (the symplectic integrator conserves them), which is why the peri and apo radii stay fixed.