Tuesday

=Test Problems: Tuesday=

Pages: **Workshop**, Overview, CADAC, Data Access, Work Area Setup, Path Setup, Utilities, Grid Data, SPH Data, Running, Reporting, LOG

Selfgravity and sink particles
Since most modelers have not yet done Problem 2, we have the opportunity to try to actually **//predict//** what will happen when selfgravity is turned on. Thus, the take-home project for today is to analyze the initial snapshots; find the potential sites of collapse, and try to tune criteria so as to predict which sites will actually collapse, and which will not!

To do this, you may either use your own favorite recipe / clumpfind method, or else modify one that we supply. Tools to read the data are already available, and there is an IDL procedure find_cores.pro and a Fortran utility g2c.x, respectively, that both initially do the same thing. However, none of them use any information about velocity or magnetic fields. These quantities are made available, and it's up to you to device criteria to apply.

You may choose to do this on either the HD or the MHD initial snapshots, but obviously the MHD case is the more relevant and interesting one; how does the presence of a magnetic field (often strong compared to the thermal energy) affect the chances for collapse? Even if you are not doing Problem 2 yourself (and even as a non-modeler), you have a chance to come up with predictions, and then check against the simulation results that will be uploaded to CADAC!

Problem parameters
Here's how to figure out what physical parameters to use to be compatible with Test Problem 2: The ratio of kinetic to gravitational energy (alpha if you will) scales as

math {v^2 \over 4\pi G\langle\rho\rangle L^2} math

This dimensionless ratio should be the same when measured in physical units and in model units. In the Test Problem 2 setup, most things are normalized to unity, so

math {4\pi G \over g}=\langle\rho\rangle=L=1 math

where the gravity parameter //g// = 100. If we choose size along the Larson relation btw. velocity and size, which we can approximate as

math L_{pc}\sim ({v / 1.7\,10^{5} \mbox{cm\,s}^{-1}})^{2.5} math

we find, correspondingly, using physical dimensions and working out the combination of constants, the simple approximate relation

math n_3 \approx {g \over L_{pc}^{1.2} \mbox{Mach}^2} math

Thus we may choose to take //L//=1pc and //n=//1e4 cgs; this will correspond approximately to Mach=9 and g=1000 -- there is no point in being very precise with these parameters. Note also that, since the run is decaying, the Mach number is dropping quickly and hence the model alpha is dropping even more quickly, so (spot-wise) gravitational collapse is expected to be accelerating.

General information about the tools may be found in the utilities wiki.

Science questions related to this will be discussed at lunch time.