Meeting Update 1001

Sweet-Parker Problem
AMR issue is kind of fixed, it's due to a typo in the code. However, the error generated on the AMR boundaries depends on resolution. Here is a comparison between a successful 2 level AMR run (top) and a higher resolution 1 level AMR run (bottom). As one can see, the higher resolution run generates error on the inner boundaries.

http://www.pas.rochester.edu/~shuleli/multi_test/amrcompare.png

Here is a movie for a successful 2 level AMR run:
​http://www.pas.rochester.edu/~shuleli/multi_test/spmovie.gif

Magnetic Island Generation
Magnetic field in a sheer pinch configuration can generate many resistive instabilities under perturbation.

http://www.pas.rochester.edu/~shuleli/multi_test/resistive_instab.png

The magnetic island formation is induced by perturbing the sheer pinch by a sinusoidal perturbation. X points and O points can form due to non zero resistivity, and dense regions separated by X and O points are formed, hence "islands". The growth rate and the size of the "islands" depend on resistivity and the strength of the perturbation.

http://www.pas.rochester.edu/~shuleli/multi_test/mihr_0180.png

Full movie:
​http://www.pas.rochester.edu/~shuleli/multi_test/mihr.gif

Triggered Star Formation
I completed a run with the BE sphere defined in the existing module. The density contrast of this BE sphere is 5, the shock is Mach 5, Gamma = 1, 3D.

http://www.pas.rochester.edu/~shuleli/multi_test/tsf1_0040.png

We do see post-shock material bounded by self gravity, but no sink particles formed. The dense material is eventually torn apart by the incoming flow. The full movie is here:
​http://www.pas.rochester.edu/~shuleli/multi_test/tsf1.gif

This one looks promising: with a very low density contrast, the post-shock material is still well bounded by gravity. Next step should be: (1) reduce shock Mach. In the Boss paper, they did Mach 1.5, 2.5 and 5. A slower shock should compress the material to trigger collapse but and the same time not strong enough to fragment stuff. (2) increase density contrast so that more material is available during the compression. (3) cooling and magnetic field can help a lot. Even in the presented setup, a By field will greatly confine the material to prevent the tearing along the x direction. (4) produce Boss' result.

MHD Sink Particle
A paper uses MHD sink particle to produce magnetic field expulsion:
​http://arxiv.org/abs/1105.5739
Their method:
"When the matter from a cell is added to a sink particle, the magnetic flux from the cell cannot be added to the sink as well, on both physical and numerical grounds. Physically, the addition of the magnetic flux to the sink particle would make the stellar field strength much higher than observed (which is, of course, the well known “magnetic flux problem”). Numerically, the sink particle cannot hold a large magnetic flux, which would produce a large, unbalanced magnetic force in the host cell of the sink particle. The needed decoupling of the magnetic field from matter during sink particle mass accretion makes the creation of the DEMS unavoidable in an ideal MHD simulation of the protostellar phase of star formation"
Adam, Eric and I had a short discussion last week via email about their method. This should be easy to do in the code, but is it OK for our application (trigger star formation)?

Resistive Paper
Reynolds number = 10 runs are running now, they are very slow due to the fact that the code does not have efficient subcycling yet for the multiphysics. Maybe I will move to the TeraGrid machines to finish them.

This week
Continue the projects presented above, write up the Qual brief (deadline this Friday), finish the revision for HEDLA proceeding.

Comments

No comments.