Gravothermal collapse of self-interacting dark-matter halos with anisotropic velocity distributions

Abstract

Self-gravitating galactic halos composed of self-interacting dark matter exhibit the formation of a highly dense core at the galactic center--a gravothermal collapse. Analytic models to describe this evolution have been developed and calibrated to numerical simulations initialized with isotropic particle velocity distributions, an assumption not necessarily warranted by the theory of halo formation. Here we study the dependence of the timescale for gravothermal collapse on the velocity distribution. To do so, we consider self-consistent initial conditions for halos with the same density distribution but with different velocity distributions. We consider models with constant anisotropy and with an anisotropy that increases with radius. The velocity distributions that we explore have collapse times that differ from that assuming isotropic distributions by more than a factor of two. We argue that these variations may depend on the global changes in velocity-dispersion profiles in these anisotropic models and not just on the degree of anisotropy.

Figures (4)

• Figure 1: The initial distributions of particles in the radius and radial velocity. The radial velocities become increasingly spread out as we increase from $\beta=-1/2$ (more tangential orbits) to $\beta=1/2$ (more radial orbits).
• Figure 2: Top: The evolution of the number of particles with radii $r<0.02$ kpc as a function of time for the five constant-$\beta$ simulations showing collapse time variation between $\sim$3 Gyr and $\sim$6 Gyr. Middle: The evolution of the velocity dispersion at radii $r<0.2$ kpc. Bottom: Evolution of the anisotropy parameter $\beta$ (averaged over the innermost 0.2 kpc).
• Figure 3: Top: The evolution of the velocity dispersion at radii $r<0.2$ kpc for OM models. Bottom: The collapse times for simulations with different values of initial anisotropy $\beta_a$ at the scale radius $r_s$.
• Figure 4: The initial velocity dispersion as a function of radius for 5 models with constant $\beta$ and 5 OM models.