Simulations - AstroBEAR

 

AstroBEAR is a hydrodynamic & magnetohydrodynamic code environment designed for a variety of astrophysical applications. It uses the
BEARCLAW package, a multidimensional, Eulerian AMR-capable computational code written in Fortran to solve hyperbolic systems for
astrophysical applications.

AstroBEAR allows simulations in 2, 2.5 (i.e., cylindrical), and 3 dimensions, in either cartesian or curvilinear coordinates.

 

BEARCLAW is developed at the University of North Carolina by Sorin Mitran ().

 

AstroBEAR is developed at the University of Rochester by the Computational Astrophysics Group in the Department of Physics & Astronomy.

 

Physics AstroBEAR supports hydrodynamic (HD) and magnetohydrodynamic (MHD) applications using a variety of spatial and temporal methods as discussed below. MHD simulations are kept divergence-free via the constrained transport (CT) methods of Balsara & Spicer [1].

 

Three different equation of state environments are available: ideal gas, gas with differing isentropic gamma, and the analytic Thomas-Fermi formulation of A.R. Bell [2].

Current work is being done to develop a more advanced real gas equation of state.

Source Terms

There are several currently implemented source terms available for AstroBEAR. Strang-splitting [3] is optional.

* Gravity
* Radiative Cooling[4],[5]
* Cylindrical Symmetry

 

Additionally, using the "self-consistent multifluid advection" model of Plewa and Müller [6], detailed multiphysics is available for H2, H I, H II, He I, and He II. Current work is being done to implement an elliptic solver for applications including heat conduction and self-gravity.

Spatial Reconstruction

The following are the avialable methods for spatial reconstruction :

* Godunov : 1st order (flat, non-interpolating) reconstruction
* Linear : Piecewise linear reconstruction on either primitive or characteristic variables, employing minmod or superbee slope-limiters
* Hyperbolic [7]

Temporal Evolution

Additionally, there are several methods which may be employed for temporal evolution :

* Eulerian: 1st order total-variation-diminishing (TVD)
* 2nd order Runge-Kutta TVD
* 2nd order Hancock two-step MUSCL [8]

Flux Calculation

Finally, the flux functions can be calculated in a variety of manners :

* Roe average [8]
* The Essentially Non-Oscillatory (ENO) flux of Shu & Osher [9]
* The modified ENO flux of Donat & Marquina [10]

For further details please consult the full documentation.

References :

[1] Balsara, D.S. & Spicer, D.S. 1999, J. Comp. Phys, 149, 270
[2] A. R. Bell, "New Equations of State for Medusa", Rutherford-Appleton Laboratories. Internal Report RL-80-091 (December 1980)
[3] Strang, G. 1968, Siam. J. of Num. Anal., 5, 505
[4] Dalgarno, A. \& McCray, R.A. 1972, ARA&A, 10, 375
[5] See writeup, http://web1.pas.rochester.edu/~lijoimc/FuOri/Writeups/coolingroutine.html
[6] Plewa, T., \& Müller, E. 1999, A\&A, 342, 179
[7] Marquina 1994, SIAM J.Sci.Comp., 15, 892
[8] Toro, E.F. 1999, Riemann solver and Numerical Methods for Fluid Dynamics (Berlin: Springer), Ch.'s 13-14
[9] Shu, C.W. \& Osher, S. 1989, J. Comp. Phys., 83, 32
[10] Donat, R. \& Marquina, A. 1996, J. Comp. Phys., 125, 42

 

Tests : http://www.pas.rochester.edu/~bearclaw/tests.htm

Publication : http://www.pas.rochester.edu/~bearclaw/publications.htm