Meeting Update 0924

Implemented multiphysics: resistivity, conductivity, viscosity.
The switches are in physics.data:
http://www.pas.rochester.edu/~shuleli/multi_test/screenshot.png

Gaussian profile heat conduction test Initial condition: gaussian temperature profile, magnetic field is on the diagonal direction. The mhd heat conduction is turned on.
In the following figure (density), top left is single processor fixed grid, bottom left is 8 processors fixed grid, top right is 8 processors AMR.
http://www.pas.rochester.edu/~shuleli/multi_test/compare.png

Petschek reconnection test The following figure shows the Petschek shock from a reconnection spot at the center. Colored variable is the kinetic energy in log scale, magnetic field is illustrated by white lines.
http://www.pas.rochester.edu/~shuleli/multi_test/sp1_0020.png
See a movie here:
​http://www.pas.rochester.edu/~shuleli/multi_test/sp1.gif

There is a bug in AMR case. See the following figure, colored is kinetic energy. You can see instabilities at the corner of inner boundaries.
http://www.pas.rochester.edu/~shuleli/multi_test/resist_amr.png

KH instability with Braginskii viscocity. Please refer to a previous post of mine for the images.

Multiphysics is interlaced. Depending on total number of steps(maxlevel) you have taken, the code shuffles the order of calls to resistivity, conductivity and viscosity. Multiple component: tested conduction + resistivity and conduction + viscosity. Both ran well.

Potential tests for multiphysics.
Conduction: MTI, Parker instability
Resistivity: Hartman flow, Petschek reconnection
Viscosity: Hartman flow, MRI

New generic data transfer interface. This will be used for anything involving subcycling AMR. Explicit solvers mentioned above and their implicit versions will be using this new interface to speed up multiphysics calculation. Currently the implementation is near finished, but I haven't got time to debug or test it. Will be the emphasis for the next few weeks and the next release.

Papers. Clump Paper: Submitted to ApJ. Resistive paper: waiting for the last two runs to finish.

This week. Fix bugs, test more on multiple components, merge the code (this version is my own version of 1020, derived from 1010 main repo), start writing the resistivity paper draft, start writing the qual brief.

Comments

No comments.