Detailed comparison of the structures and kinematics of simulated and observed barred galaxies

• Published in: Monthly notices of the Royal Astronomical Society 2003-11, Vol.346 (1), p.251-264
• Main Authors: O'Neill, J. K., Dubinski, John
• Format: Article
• Language: English
• Subjects: galaxies: evolution; galaxies: haloes; galaxies: kinematics and dynamics
• ISSN: 0035-8711; 1365-2966
• DOI: 10.1046/j.1365-2966.2003.07085.x

We examine the observable properties of simulated barred galaxies, including radial mass profiles, edge-on structure and kinematics, bar lengths and pattern speed evolution for detailed comparison to real systems. We have run several simulations in which bars are created through inherent instabilities in self-consistent simulations of a realistic disc+halo galaxy model with a disc-dominated, flat rotation curve. These simulations were run at high (N= 20 million particles) and low (N= 500 000 particles) resolution to test numerical convergence. We determine the pattern speeds in simulations directly from the phase angle of the bar versus time and the Tremaine–Weinberg method. Fundamental dynamics do not change between the high and low resolution, suggesting that convergence has been reached in this case. We find that the higher resolution is needed to simulate structural and kinematic properties accurately. The edge-on view of the higher-resolution system shows the bending instability and formation of a peanut-shaped bulge clearly. We determined bar lengths by different means to determine that the simulated bar is fast, with a corotation to bar length ratio of under 1.5. Simulated bars in these models form with pattern speeds slower than those observed and slow-down during their evolution. Dynamical friction between the bar and dark halo is responsible for this deceleration, as revealed by the transfer of angular momentum between the disc and the halo. However, even though the pattern speed is reduced at later times, the instantaneous scalelength of the disc has grown sufficiently for the bar motion to agree with many observations. By using a different model and simulation technique than other authors, we are able to compare the robustness of these methods. An animation of the face-on and edge-on views of the 20-million-particle simulation is available at.