Michael Solway's Graduate Research
I ran collisionless N-body simulations of disk galaxies to study the effect called radial migration. During this effect, a star corotating with the spiral pattern (angular velocity of the star equals the spiral's pattern speed) gains or looses angular momentum and ends up on a nearly-circular orbit with a different radius without an increase in its orbit's eccentricity (Sellwood & Binney 2002). The spirals are transient and keep recurring in simulations of typical disk galaxies. During a duration of about 10 billion years, a simulated disk undergoes about 30 transient spirals. Each of these can have a corotation resonance at a different radius. Thus, the cumulative effect of radial migration causes stars to essentially random walk in radius with average step sizes of 1kpc (3.1×1019m). The region affected by a single spiral is typically a 4kpc-wide annulus centered at the corotation resonance. Hence, radial migration is a significant dynamical effect in disk galaxies. Over time, it mixes different formation environments by moving stars from their birth radii to other radii. In the solar neighborhood, the radial gradient of the abundances of elements heavier than helium is flatter for older stars (e.g. Yu et al. 2012), which supports this mixing, because older stars had more time to undergo radial migration.
Radial migration could also be responsible for the formation of thick disk populations in galaxies. As stars migrate outwards from the inner region of a galaxy to the outer regions, they experience a weaker vertical restoring force due to the radially declining surface density. Therefore, their vertical amplitudes increase as they migrate outwards, and they form a thick disk. This has been shown to occur in galactic chemical evolution models (Schonrich & Binney 2009) and hydrodynamic N-body simulations (e.g. Loebman et al. 2011). However, there are a few other competing thick disk formation models, which likely also play a role.
In Solway et al. 2012, we extended the work of Sellwood & Binney 2002 to 3D disks and showed that radial migration weakens with increasing disk thickness, but remains significant for disks as thick as the Milky Way's thick disk. We also showed that it is weaker in disks with greater velocity dispersions and for spirals with greater number of arms. The net effect is stronger for multiple transient spirals compared to a single one, and is enhanced a little more in the outer regions if a bar is present. Finally, we showed that the particles conserve their vertical actions on average when undergoing radial migration. This can be used to prescribe radial migration in galactic evolution models.
The thick disk of the Milky Way is primarily composed of old stars in the solar neighborhood. The mechanism of thick disk formation through outwards radial migration cannot by itself explain why it does not contain young stars. Were stars allowed to migrate outwards from the inner region till present, young stars forming in the galactic center would have continued to populate the thick disk in the outer region where the sun lies. Hence, something must have shut off this outwards migration a long time ago corresponding to the youngest age of thick disk stars.
In a second project, we showed that bar formation is such an event that suppresses outwards radial migration of stars from within the bar region, which explains why the thick disk does not have young stars at present. The Milky Way contains a bar, so this scenario is reasonable. We also tested whether thick disks form through outwards radial migration with only spirals in our simulations in the first place.