Rice University logo
Top blue bar image

Archive for March, 2015

LSAP Meeting Notes 3/30, 2015- Simulation Codes – Edison Liang

March 30th, 2015 by toffo

Edison Liang – Codes

My group uses 3 different types of simulation codes to model relativistic plasma phenomena in high-energy astrophysics and intense laser-target interactions:

1. Particle-in-Cell (PIC) codes for plasma kinetic processes
2.  MHD codes for global macroscopic fluid simulations
3.  Monte Carlo (MC) codes for high-energy radiation and nuclear particle physics (NPP) processes
These codes are often linked/coupled to each other to perform multi-physics, multi-scale simulations.  Most codes are imported or licensed.

PIC Codes :

1. EPOCH (U. of Warwick-Ruhl, open-source version of PSC used at LLNL and Germany): 3D explicit, ok but not great for high-g. (Rice & LANL clusters)

2. ZOHAR (LLNL, detailed in classic book by Birdsall & Langdon): 2D explicit, excellent for high-g. (only on LLNL cluster).

3. ipic3d (KUL-Lapenta, still under development): 3D fully implicit, so for only for low-g regimes. (to be imported to Rice).

4. LSP (SAIC, commerical code, high license fee): 3D explicit. Has collisions and MHD options. (under negotiation for license).

5. PICiinGPU (German GPU code): 3D explicit (Rice and Georgia Tech GPU clusters, not sufficient memory for production).

6. NNS code (Noguchi-Nishimura-Sugiyama): 2D explicit includes radiation and rad. damping (Rice & LLNL clusters).

MHD codes

FLASH (U. of Chicago): 3D Eulerian, multi-physics includes radiation transport, nuclear physics, EOS, SR, laser coupling. 3T, kinetic ions … (Rice and LANL clusters).

HARM (U. of Illinois): 2D Eulerian, axi-symmetric polar grid only. multi-physics includes radiation, GR, 1T (Rice clusters).

ATHENA (Princeton): 3D Eulerian, multi-physics includes radiation, 3T, gravity, SR, shear box…(Rice clusters).

Hyades (Commerical code, license fee): 1D Lagrangian, multi-fluid, EOS, laser coupling, radiation, 3T….(Rice cluster).

MC codes:

1. GEANT4 (CERN): 3D. has all high-energy NPP processes. fixed background. lacks lower energy physics. (Rice clusters).

2. Rice-OU Radiation Code: 2D slab, sphere or cylinder grids only. Includes most astrophysics radiation processes. FP solvers for coupled particle distributions (Rice clusters).

3. MK Compton code: 1D slab, sphere or cylinder grid only. Time-independent. Fixed thermal or nonthermal electron distributions. Fast and simple: ideal for training beginning and undergrad students in MC.


MHD simulations of accretion onto black holes -Edison Liang

March 3rd, 2015 by toffo

Panels from final timestep t = 8000 M (1.4×105 s) of HARM data, as labelled for different physical quantities. We show a cut-out in the r-z plane, with the black hole’s horizon shown on left. As detailed in the text, the radial direction is scaled logarithmically, and angular dimension is scaled more finely toward the equator. Therefore, the finest resolution is obtained closest to the horizon and equator. This causes a “spreading” effect in the images, as they are shown evenly spaced by cell number, not physical value. Each quantity is shown on a logarithmic scale, with dark red being the highest values, black being the lowest.A selection of movies demonstrating the advanced numerical modeling and visualization capabilities of the Solar Physics Group. The movies show the predicted evolution of an active region core subject to a large number of small-scale heating events (nanoflares), as it would be observed by the SDO/AIA 94, 131, 171, 193, 211, and 335 Å channels. There are 400 magnetic field lines comprising the active region core, which was subject to a total of 4000 nanoflares with energies drawn from a power-law distribution between 10^23 – 10^25 erg and a spectral index of -2.5. The delay between nanoflares re-energizing plasma along a field line was chosen to be proportional to the energy of the next event. These movies represent ‘synthetic’ observations and they can be analyzed using exactly the same tools and techniques as the real observations provided by our fleet of space-based instrumentation. This allows direct comparisons to be made between model results and observational data, and provides extremely strong constraints on the physics of coronal loops and on the properties of potential coronal heating mechanisms. Figure courtesy of Edison Liang.