Galactic bulges can be built by mergers, by satellite accretion, or through internal evolution of disk galaxies. In this latter case, one of the well-studied mechanisms is the formation of bulges during the evolution of barred galaxies. Barred galaxies can produce both disk-like bulges and boxy/peanut bulges.

Observationally, bars are present in a large fraction of nearby disk galaxies, 72% when looking at H-band images. The fraction of strong bars is constant up to redshift z~1. Bars are long-lived structures with important dynamical effects on the galaxies which host them. Bars are responsible, for example, for the inflow of gas into the inner parts of the galaxy, which may lead to the formation of a nuclear bar and to fuelling of an AGN. Bars also trigger the formation of spiral arms and of stellar and gaseous rings via resonances.

Stellar bars in N-body simulations evolve with time from a flat bar to a boxy/peanut bulge through two buckling instabilities, as shown in the figure. This evolution is mainly determined by the exchange of angular momentum between the bar and the disk, bulge, and dark matter halo. We know now that also the Milky-Way and our neighbour galaxy M31 have such boxy/peanut bulges.

SPIN-UP OF LOW MASS CLASSICAL BULGES IN BARRED GALAXIES

The secular processes driven by the bar may cause dramatic changes in the dynamical structure of a preexisting low-mass classical bulge, such as might be present in galaxies like the Milky Way. Such a bulge absorbs angular momentum emitted by the bar, mostly through resonances, particularly Lagrange point (-1:1) and ILR (2:1) orbits, but also retrograde non-resonant orbits absorb angular momentum while the bar  grows rapidly. Thus an initially non-rotating low-mass classical bulge transforms into a fast rotating, radially anisotropic and triaxial object, embedded in the similarly fast rotating boxy bulge formed from the disk. Towards the  end of the evolution, the classical bulge develops cylindrical rotation.

THE INNER GALACTIC BULGE: EVIDENCE FOR A NUCLEAR BAR?

Ortwin Gerhard & Inma Martinez-Valpuesta, 2012, ApJL, 744, 8 (to the paper)

Starcount observations show evidence for a structural change in the Milky Way bulge inward of |l|~4◦. With an N-body barred galaxy simulation we showed that a boxy bulge formed through the bar and buckling instabilities matches these observations. The change in the slope of the model longitude profiles is caused by a transition from highly elongated to more nearly axisymmetric isodensity contours in the inner boxy bulge. This transition is confined to a few degrees from the Galactic plane.

The same simulation snapshot was earlier used to clarify the apparent boxy bulge—long bar dichotomy. Furthermore, the nuclear star count map derived from this simulation snapshot displays a longitudinal asymmetry similar to that observed in the TwoMicron All Sky Survey (2MASS) data. These combined results

• support the interpretation that the Galactic bulge originated from disk evolution,
  
• and question arguments based on star count data for the existence of a secondary nuclear bar in the Milky Way.

UNIFYING A BOXY BULGE AND PLANAR LONG BAR IN THE MILKY WAY

Inma Martinez-Valpuesta & Ortwin Gerhard, 2011, ApJL, 734, 20 (to the paper)

We have known for sometime that the Milky Way is a barred disk galaxy. But more recently, several studies inferred from star count observations that the Galaxy must contain a separate, new, flat long bar component, twisted relative to the barred bulge. With a simulation we showed that these observations can be reproduced with a single boxy bulge and bar structure. In this simulation, a stellar bar evolved from the disk, and the boxy bulge originated from it through secular evolution and the buckling instability.

We calculated the star count distributions for this model at different longitudes and latitudes, in a similar way as observers have done for resolved star counts. We found good quantitative agreement with the observations for a suitable model snapshot. The long bar signature in this model is partially is due to a volume effect in the star counts, and partially because of choosing a snapshot in which the planar bar has developed leading ends by interacting with the nearby spiral arm heads. We also calculated radial velocity predictions from this model for comparison with upcoming surveys.