I’ve finished reading an annual review on galactic winds as well as one on rings and shells / ripples.  Galactic winds alone are a fairly complicated topic, and very important in galactic evolution.  Evidently the publicly released version of the GADGET-2 code does not incorporate the processes necessary to simulate the wind.  The publicly released version is actually missing a lot of the more interesting components which allow feedback.  I’m guessing that the public version is simplified so that only preferred users can easily publish major new results.  The full code may also still be buggy and difficult to run–especially since it uses MPI rather than OpenMP.

I’ve started reading the description of the GADGET-2 code.  The most significant things to note are: (1) The multipole moments used in the TreeSPH gravitational potential calculations are just monopoles rather than quadrupoles or higher. (2) The code now includes the option of using a hybrid Tree-particle mesh (TreePM) technique which uses a mesh and Fourier techniques in the far field while using TreeSPH for local calculations. (3) The SPH formulation is based on entropy, so it explicitly conserves entropy with fully-adaptive smoothing length.  Anomolous entropy changes were evidently an issue with the original GADGET code. (4) Cosmological expansion is included explicitly. (5) As far as I can tell, the gravitational model is essentially Newtonian with infinite propagation speed.  Implementing a consistent gravitational model would probably greatly complicate the code and make it run much more slowly.

My goal for the next week is to complete the first program described in my first post. It will be written in C++.  I would also like to be able to create a simple plot of 3D data with IFrIT.

This entry was posted on Friday, July 31st, 2009 at 11:15 pm and is filed under regular update. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.