1. The Enzo Project
  2. Untitled project
  3. enzo-dev
  4. Pull requests

Pull requests

#305 Merged at 2fe9203

New 2d test problems for hydro and MHD in 2D

  1. Tom Abel
  • Added initial conditions and parameter file for the problem studied by Lecoanet et al (2015) http://arxiv.org/abs/1509.03630 It also works in parallel.

  • Make it work with AMR and RefineAtStart in parallel.

  • Added AMR parameter file. This uses basic slope refinement on colur field only. This is not the best strategy, os is just useful to test that RefineAtStart and AMR is workin in general.

  • Correct bugs in Colourfield advection for HydroMethod=3 which were already correct in HydroMethod=4.

This test has now been tested with HydroMethod 0, 3, 4.

Comments (15)

  1. dcollins4096

    Would it be possible to put the 2D-LongKH-AMR.enzo into its own directory? There's a general one-to-one between tests and directories, I'm going to totally forget about this one where it is.

  2. dcollins4096

    I ran the test, and I think I did something wrong, so maybe usage could be explained. I ran 2D-LongKH.enzo on 8 cores, with changeset 9ec0bc7176a6. I do yt.ProjectionPlot(yt.load('DD0101/data0101'),2,'density').save() and I get this http://imgur.com/WcmdM2K and for snapshot 18, I get this: http://imgur.com/Qdb637Q neither of which look like what I expect from the linked paper. Did I mess up? If so, could you clarify? Or are my expectations that it should look like one of the figures from the paper off?


    1. dcollins4096

      Ah, ok. That looks more reasonable. It doesn't look like it gets as chaotic as in the paper, is that a parameter space issue? It just kind of looks like a wind-up. Here's what I get: http://imgur.com/a/e8OQk

      Its a useful starting place for people who want to do that test and looks good. I'll approve it.

      1. Tom Abel author

        Yes they did much higher resolution in the paper, but I wanted to run fast for testing and so stuck with lower resolutions default parameters. It does work in parallel however...

  3. Greg Bryan

    This looks good to me. I compiled and did a few quick runs with various parameters to make sure it ran, but didn't have the patience to see it through to useful output. I'm approving but a merge should wait until Dave confirms the run produces correct output on his machine...