|The Physics Behind the Applet|
A fully self-consistent treatment of tidal stripping and dynamical friction in satellite galaxies is well beyond the current capabilities of web-based java. To calculate these models we have taken advantage of a number of analytic approximations to evolve the system with time.
Rigid Potential for the Galaxy and Companion: The parent galaxy is assumed to be a fixed gravitational potential comprised of a Hernquist (1993) halo model with mass M=5.5x1011 Msun and scale radius a=10 kpc and a Miyamoto-Nagai disk model with mass M=6x1010 Msun, radial scale factor a=6.5 kpc, and vertical scale factor b=1.3 kpc. The satellite is also modeled as a Hernquist (1993) model, with varying mass and scale radius. The mass of the satellite drops with time due to tidal stripping.
Dynamical Friction: We use a Chandrasekhar-like frictional formula, where the braking due to dynamical friction is determined by the local galaxy density, and the mass and velocity of the companion. We use a factor for the Coulomb logarithm of Lamda=3.
Tidal Stripping: To calculate the tidal radius, we simply calculate the radius inside the satellite at which the satellite density is equal to the local mass density of the satellite. The mass profile of the satellite is then truncated at this radius, and we leave the mass distribution inside this radius unchanged.
Orbital Integration: When only the satellite is being evolved (ie stars are shut off), we use a 5th order Runga-Kutta integrator with variable timesteps to ensure accuracy. If stars are turned on, the variable timestep algorithm is shut off, and we use a 4th order RK integrator with fixed timestep. Note that in this situation, the orbits of the stars is only crudely calculated (ie the errors are large...).