Second-order solutions of the equilibrium statistical mechanics for self-gravitating systems

In a previous study, we formulated a framework for the entropy-based equilibrium statistical mechanics for self-gravitating systems. This theory is based on the Boltzmann–Gibbs entropy and includes the generalized virial equations as additional constraints. With the truncated distribution function to the lowest order, we derived a set of second-order equations for the equilibrium states of the system. In this work, the numerical solutions of these equations are investigated. It is found that there are three types of solutions for these equations. Both the isothermal and divergent solutions are thermally unstable and have unconfined density profiles with infinite mass, energy and spatial extent. The convergent solutions, however, seem to be reasonable. Although the results cannot match the simulation data well, because of the truncations of the distribution function and its moment equations, these lowest-order convergent solutions show that the density profiles of the system are confined, the velocity dispersions are variable functions of the radius, and the velocity distributions are also anisotropic in different directions. The convergent solutions also indicate that the statistical equilibrium of self-gravitating systems is by no means the thermodynamic equilibrium. These solutions are just the lowest-order approximation, but they have already manifested the qualitative success of our theory. We expect that higher-order solutions of our statistical-mechanical theory will give much better agreement with the simulation results concerning dark matter haloes.