Survey Parameters

The simulations analyzed here are grouped into 11 ensembles, each containing 8 realizations. One ensemble serves as a reference; the other 10 ensembles are derived by varying parameters including the pericentric separation, disk orientations, and disk velocity dispersion.

The galaxy models used in these experiments have three components:

• Bulge: a Hernquist (1990) model with mass M_b = 0.0625 and scale radius a_b = 0.04168, smoothly truncated beyond a radius r_t,b = 4.0.
• Disk: an exponential/isothermal model with mass M_d = 0.1875 and scale radius R_d = 1/12. The radial velocity dispersion is twice the vertical velocity dispersion, and the latter is set to make the scale height constant.
• Halo: a Navarro, Frenk, & White (1996) halo with mass M_h = 1.0 and scale radius a_h = 0.2, smoothly truncated beyond a radius r_t,h = 0.4474 using the expression in Springel & White (1999).

All simulations have these parameters in common:

• Orbital energy: the galaxies are launched with relative positions and velocities approximately consistent with a zero-energy orbit.
• Particle number: all simulations were run with a total of N = 65536 bodies.
• Smoothing scale: when calculating forces, the mass distributions were smoothed by convolution with a Plummer model with radial scale 0.01 length units.
• Time-step: integrations were performed using a leap-frog with a step of 1/64 time units.

The reference ensemble has these parameter values:

• Pericentric separation: two point masses launched on the same initial orbit would reach a minimum separation r_p = 0.2 length units.
• Disk orientation: the disks have inclinations i₁ = i₂ = 0 and pericentric arguments w₁ = w₂ = 0.
• Disk velocity dispersions: the disks have constant scale height z_d = 0.005, implying Q = 1.35 over most radii.

Parameter values for all 11 ensembles are: