Context Navigation

← Previous Post
Next Post →

Meeting Update 0724

I've been porting the resistive solvers to the newest revision. Also have been working on the scaling and set up for the new project: the resistive clump simulations. The set up of our "standard" case is as follows:

ambient density = 1 mg/cc
atomic mass = 12 (carbon)
ambient temperature = 5 ev
density contrast = 100
clump radii = 0.1
tube radii = 0.3
shock speed = 40 km/s (mach 5)
magnetic beta = 4
magnetic reynolds number R_{em}=10
simulation time = 4 crushing times
resolution = 64 zones per clump radii
adiabatic

\beta=4$, $R_{em}=100	\beta=4$, $R_{em}=10	\beta=4$, $R_{em}=1
\beta=10$, $R_{em}=100	\beta=10$, $R_{em}=10	\beta=10$, $R_{em}=1

with two additional runs:
\beta=100$, $R_{em}=10
\beta=4$, $R_{em}=10 with cooling

The magnetic reynolds number of 10 implies that:
t_{mhd}/t_{res}\approx 4.34
i.e. each hydro step requires 4 resistive steps. Therefore the R_{em} = 100 simulations should not require cycling resistive steps, but R_{em} = 1 will be demanding.

I'm currently working on the low resolution version of the runs (with 24 zones per clump radii) to check things. But the production runs should be doable on bluegene. (the previous paper's runs of 192 zones grid takes 4 days with cooling)

Posted: 12 years ago
Author: Shule Li
Categories: (none)

Comments

No comments.

Download in other formats:

RSS Feed