Near infrared images from the COBE satellite presented the first clear evidence that our Milky Way galaxy contains a boxy shaped bulge. Recent years have witnessed a gradual paradigm shift in the formation and evolution of the Galactic bulge. Bulges
were commonly believed to form in the dynamical violence of galaxy mergers. However, it has become increasingly clear that the main body of the Milky Way bulge is not a classical bulge made by previous major mergers, instead it appears to be a bar seen somewhat end-on. The Milky Way bar can form naturally from a precursor disk and thicken vertically by the internal firehose/buckling instability, giving rise to the boxy appearance. This picture is supported by many lines of evidence, including the asymmetric parallelogram shape, the strong cylindrical rotation (i.e., nearly constant rotation regardless of the height above the disk plane), the existence of an intriguing X-shaped structure in the bulge, and perhaps the metallicity gradients. We review the major theoretical models and techniques to understand the Milky Way bulge. Despite the progresses in recent theoretical attempts, a complete bulge formation model that explains the full kinematics and metallicity distribution is still not fully understood. Upcoming large surveys are expected to shed new light on the formation history of the Galactic bulge.
We explore the kinematics (both the radial velocity and the proper motion) of the vertical X-shaped feature in the Milky Way with an N-body bar/bulge model. From the solar perspective, the distance distribution of particles is double-peaked in fields
passing through the X-shape. The separation and amplitude ratio between the two peaks qualitatively match the observed trends towards the Galactic bulge. We confirm clear signatures of cylindrical rotation in the pattern of mean radial velocity across the bar/bulge region. We also find possible imprints of coherent orbital motion inside the bar structure in the radial velocity distribution along l=0 degree, where the near and far sides of the bar/bulge show excesses of approaching and receding particles. The coherent orbital motion is also reflected in the slight displacement of the zero-velocity-line in the mean radial velocity, and the displacement of the maximum/minimum in the mean longitudinal proper motion across the bulge region. We find some degree of anisotropy in the stellar velocity within the X-shape, but the underlying orbital family of the X-shape cannot be clearly distinguished. Two potential applications of the X-shape in previous literature are tested, i.e., bulge rotation and Galactic center measurements. We find that the proper motion difference between the two sides of the X-shape can be used to estimate the mean azimuthal streaming motion of the bulge, but not the pattern speed of the bar. We also demonstrate that the Galactic center can be located with the X-shape, but the accuracy depends on the fitting scheme, the number of fields, and their latitudinal coverage.
Dust lanes, nuclear rings, and nuclear spirals are typical gas structures in the inner region of barred galaxies. Their shapes and properties are linked to the physical parameters of the host galaxy. We use high-resolution hydrodynamical simulations
to study 2D gas flows in simple barred galaxy models. The nuclear rings formed in our simulations can be divided into two groups: one group is nearly round and the other is highly elongated. We find that roundish rings may not form when the bar pattern speed is too high or the bulge central density is too low. We also study the periodic orbits in our galaxy models, and find that the concept of inner Lindblad resonance (ILR) may be generalized by the extent of $x_2$ orbits. All roundish nuclear rings in our simulations settle in the range of $x_2$ orbits (or ILRs). However, knowing the resonances is insufficient to pin down the exact location of these nuclear rings. We suggest that the backbone of round nuclear rings is the $x_2$ orbital family, i.e. round nuclear rings are allowed only in the radial range of $x_2$ orbits. A round nuclear ring forms exactly at the radius where the residual angular momentum of infalling gas balances the centrifugal force, which can be described by a parameter $f_{rm ring}$ measured from the rotation curve. The gravitational torque on gas in high pattern speed models is larger, leading to a smaller ring size than in the low pattern speed models. Our result may have important implications for using nuclear rings to measure the parameters of real barred galaxies with 2D gas kinematics.
About one third of early-type barred galaxies host small-scale secondary bars. The formation and evolution of such double-barred galaxies remain far from being well understood. In order to understand the formation of such systems, we explore a large
parameter space of isolated pure-disk simulations. We show that a dynamically cool inner disk embedded in a hotter outer disk can naturally generate a steady secondary bar while the outer disk forms a large-scale primary bar. The independent bar instabilities of inner and outer disks result in long-lived double-barred structures whose dynamical properties are comparable with observations. This formation scenario indicates that the secondary bar might form from the general bar instability, the same as the primary bar. Under some circumstances, the interaction of the bars and the disk leads to the two bars aligning or single, nuclear, bars only. Simulations that are cool enough of the center to experience clump instabilities may also generate steady double-barred galaxies. In this case, the secondary bars are fast, i.e., the bar length is close to the co-rotation radius. This is the first time that double-barred galaxies containing a fast secondary bar are reported. Previous orbit-based studies had suggested that fast secondary bars are not dynamically possible.
We describe a method to extract resonant orbits from N-body simulations exploiting the fact that they close in a frame rotating with a constant pattern speed. Our method is applied to the N-body simulation of the Milky Way by Shen et al. (2010). This
simulation hosts a massive bar, which drives strong resonances and persistent angular momentum exchange. Resonant orbits are found throughout the disc, both close to the bar itself and out to the very edges of the disc. Using Fourier spectrograms, we demonstrate that the bar is driving kinematic substructure even in the very outer parts of the disc. We identify two major orbit families in the outskirts of the disc that make significant contributions to the kinematic landscape, namely the m:l = 3:-2 and 1:-1 families resonating with the pattern speed of the bar. A mechanism is described that produces bimodal distributions of Galactocentric radial velocities at selected azimuths in the outer disc. It occurs as a result of the temporal coherence of particles on the 3:-2 resonant orbits, which causes them to arrive simultaneously at pericentre or apocentre. This resonant clumping, due to the in-phase motion of the particles through their epicycle, leads to both inward and outward moving groups which belong to the same orbital family and consequently produce bimodal radial velocity distributions. This is a possible explanation of the bimodal velocity distributions observed towards the Galactic anti-Centre by Liu et al. (2012). Another consequence is that transient overdensities appear and dissipate (in a symmetric fashion) on timescales equal to the their epicyclic period resulting in a periodic pulsing of the discs surface density.
A vertical X-shaped structure was recently reported in the Galactic bulge. Here we present evidence of a similar X-shaped structure in the Shen et al. (2010) bar/boxy bulge model that simultaneously matches the stellar kinematics successfully. The X-
shaped structure is found in the central region of our bar/boxy bulge model, and is qualitatively consistent with the observed one in many aspects. End-to-end separations of the X-shaped structure in the radial and vertical directions are roughly 3 kpc and 1.8 kpc, respectively. The X-shaped structure contains about 7% of light in the boxy bulge region, but it is significant enough to be identified in observations. An X-shaped structure naturally arises in the formation of bar/boxy bulges, and is mainly associated with orbits trapped around the vertically-extended x_1 family. Like the bar in our model, the X-shaped structure tilts away from the Sun--Galactic center line by 20 degrees. The X-shaped structure becomes increasingly symmetric about the disk plane, so the observed symmetry may indicate that it formed at least a few billion years ago. The existence of the vertical X-shaped structure suggests that the formation of the Milky Way bulge is shaped mainly by internal disk dynamical instabilities.
Bulges are commonly believed to form in the dynamical violence of galaxy collisions and mergers. Here we model the stellar kinematics of the Bulge Radial Velocity Assay (BRAVA), and find no sign that the Milky Way contains a classical bulge formed by
scrambling pre-existing disks of stars in major mergers. Rather, the bulge appears to be a bar, seen somewhat end-on, as hinted from its asymmetric boxy shape. We construct a simple but realistic N-body model of the Galaxy that self-consistently develops a bar. The bar immediately buckles and thickens in the vertical direction. As seen from the Sun, the result resembles the boxy bulge of our Galaxy. The model fits the BRAVA stellar kinematic data covering the whole bulge strikingly well with no need for a merger-made classical bulge. The bar in our best fit model has a half-length of ~ 4kpc and extends 20 degrees from the Sun-Galactic Center line. We use the new kinematic constraints to show that any classical bulge contribution cannot be larger than ~ 8% of the disk mass. Thus the Galactic bulge is a part of the disk and not a separate component made in a prior merger. Giant, pure-disk galaxies like our own present a major challenge to the standard picture in which galaxy formation is dominated by hierarchical clustering and galaxy mergers.
Many barred galaxies harbor small-scale secondary bars in the center. The evolution of such double-barred galaxies is still not well understood, partly because of a lack of realistic N-body models with which to study them. Here we report the generati
on of such systems in the presence of rotating pseudobulges. We demonstrate with high mass and force resolution collisionless N-body simulations that long-lived secondary bars can form spontaneously without requiring gas, contrary to previous claims. We find that secondary bars rotate faster than primary ones. The rotation is not rigid: the secondary bars pulsate, with their amplitude and pattern speed oscillating as they rotate through the primary bars. This self-consistent study supports previous work based on orbital analysis in the potential of two rigidly rotating bars. We also characterize the density and kinematics of the N-body simulations of the double-barred galaxies, compare with observations to achieve a better understanding of such galaxies. The pulsating nature of secondary bars may have important implications for understanding the central region of double-barred galaxies.
Although at least one quarter of early-type barred galaxies host secondary stellar bars embedded in their large-scale primary counterparts, the dynamics of such double barred galaxies are still not well understood. Recently we reported success at sim
ulating such systems in a repeatable way in collisionless systems. In order to further our understanding of double-barred galaxies, here we characterize the density and kinematics of the N-body simulations of these galaxies. This will facilitate comparison with observations and lead to a better understanding of the observed double-barred galaxies. We find the shape and size of our simulated secondary bars are quite reasonable compared to the observed ones. We demonstrate that an authentic decoupled secondary bar may produce only a weak twist of the kinematic minor axis in the stellar velocity field, due to the relatively large random motion of stars in the central region. We also find that the edge-on nuclear bars are probably not related to boxy peanut-shaped bulges which are most likely to be edge-on primary large-scale bars. Finally we demonstrate that the non-rigid rotation of the secondary bar causes its pattern speed not to be derived with great accuracy using the Tremaine-Weinberg method. We also compare with observations of NGC 2950, a prototypical double-barred early-type galaxy, which suggest that the nuclear bar may be rotating in the opposite sense as the primary.
