Project > Computational Astrophysics

Computational Astrophysics (CompAS, formerly CFD-MHD)

Computational Astrophysics (CompAS) Website

Research Gallery

Click the image for more...

Figure: Example of a nested-grid mesh refinement (NMR) run of a global protoplanetary disk with an embedded circumplanetary disk surrounding a 0.05-Jupiter-mass planet at 5.2 AU from the star at origin. The simulation has 6 levels of refinement, with the finest resolution of 6.25×10–4 AU, resolving the Hill radius of the planet (0.13 AU at 5.2 AU) with ~200 cells.

Formation and Evolution of Protoplanetary and Circumplanetary Disks with Antares Three-Dimensional Global Hydrodynamic Simulations

the plot visualizes the contribution of the specific transition of the molecule from a certain observing angle

Cotribution analysis of molecular line of a disk

In the bottom panels, square sections of surface density simulated with <b>Antares</b> are shown centered at the locations of protoplanets. The surface density is obtained by vertical integration over <i>z</i> < 0.0156 AU (the height of the finest level) for protoplanets with masses of 4, 8, and 16 Earth Masses, respectively. The central star is located on the left and the protoplanets are orbiting the central star counterclockwise. On top of each surface density is the corresponding volume density cutting through the plane defined by

Formation of Isothermal Circumplanetary Disks with Three-Dimensional Global Simulations for Sub-Neptune-Mass Protoplanets

$Distribution of the physical quantities along the jet propagation axis. We perform special relativistic MHD simulations in a cylindrical 1.5-dimensional pproximation. Our model reproduces a paired super/subluminal knot motions, taken by the HST observations (Biretta et al. 1999), as a quad RMHD shock system. By assuming a viewing angle of 14$^{\circ$ (Wang \& Zhou 2009), observed apparent speeds 6 $c$ and 0.84 $c$ correspond to $\sim$ 0.99 $c$ and 0.80 $c$, respectively. In our model, these knots are identified as the forward/reverse fast mode MHD shocks (FF/RF in the figure). By the nature of a slow mode MHD shock, it may not be appropriated to mediate a diffusive shock acceleration of high energy particles.$

A Self-consistent Model for Super/sub-luminal Motions and Particle Acceleration in the M87 Jet

A comparison between the numerical simulation result and the observation of NGC 1097. Left panel: The projected surface density distribution of the simulation result. Middle panel: Optical image of NGC 1097 taken with the Visible Multi-Object Spectrograph (VIMOS) instrument on the 8.2-m Melipal (Unit Telescope 3) of ESO's Very Large Telescope. Right panel: Superposition of the left two figures. North is at the top and east is to the left.

HYDRODYNAMICAL SIMULATIONS OF THE BARRED SPIRAL GALAXY NGC 1097

In simulations of a star-disk system in the closest vicinity of a star, we
obtain very fast, light ejecta of matter. Those micro-ejections are of small
mass and angular momentum flux, and can be launched from a resistive
magnetosphere above the disk gap. Shown is the absolute value of toroidal
current in code units, in logarithmic color grading. In solid lines are
drawn poloidal magnetic field lines, with lines along which magnetic field
vector points towards the star painted in black color, and those where the
field points in the opposite direction, in white color. Reconnection is
ongoing along the boundary sheet between the opposite-directed magnetic
fields. The quasi-stationary state is reached, with matter pushed out of the
magnetosphere by pressure gradient and magnetic forces, helped by
reconnection. Vectors show poloidal velocity, normalized to Keplerian
velocity at R=2.8 (for more details, see Cemeljic et al., 2013, ApJ768, 5).

Micro-ejections in star-disk resistive magnetohydrodynamic simulations

Our simulations indicate that in the study of the runaway (type III) migration, it is important to carry out a fully self consistent treatment of the gravitational interaction between the disk and the embedded planet.

On the Orbital Evolution of a Jovian Planet Embedded in a Self-Gravitating Disk

The numerical solutions of a D2 disk for N=1024, the profiles are surface density (left-top), the x-directional force (right-top), the y-directional force(left-bottom), and radial force(right-bottom).

Self-gravitational force calculation of infinitely thin gaseous disks

The left panel corresponds to the particle image showing the face on and two edge on views of an initially axisymmetric disk. The right panel shows the formation of a bar which produces a peanut shaped morphology.

The Effect of Bars and Transient Spirals on the Vertical Heating in Disk Galaxies

Numerical calculations reveal that the striking features in NGC 6782 can be reproduced provided that the gas flow is governed by the gravitational potential associated with a slowly rotating strong bar.

Hydrodynamical Simulations of the Barred Spiral Galaxy NGC 6782

The efforts of theoretical and computational astrophysics in ASIAA date back to the CFD-MHD initiative, established in 2000, as a Key Project of the Academia Sinica. The CFD-MHD initiative started as a joint research project of the Institute of Astronomy and Astrophysics (ASIAA), the Institute of Mathematics (ASIM) of the Academia Sinica, and the Department of Mathematics of the National Taiwan University. Its main goal was to develop high-performance codes of computational fluid dynamics and magnetohydrodynamics (CFD-MHD) for astrophysical problems. Built upon Prof. Chi Yuan’s vision, the first grass-root solver for astrophysical fluid in Taiwan, Antares, was made with the 2nd-order Godunov scheme with exact/approximate Riemann solvers.

The name of CFD-MHD was changed into CompAS in Feb. 2012 for Computational Astrophysics, for the generalization of scopes into the broad context of computational astrophysics. Applications of the Antares suites of codes include structure and evolution of gas disks in spiral galaxies, planet migration in protoplanetary disks, planet-disk interactions, and the formation of the circumplanetary disks.

The new CompAS (now Computational Astrophysics Code and Algorithm Development) group continues to thrive from the founding CFD-MHD effort in code and algorithm development for theoretical and computational astrophysics. Since the inception of the first Antares code in 2D in the early 2000’, the group has not only been able to advance on algorithm development but also been the key incubator for several new platforms of different designs. The ongoing efforts are staying at the forefront of numerical algorithms and technologies of scientific computation.