This is a test of a 3D Sedov blast wave problem.

Type "make clean" and "make" to re-compile the 3DSPMHD binary.

Run the code using ./3DSPMHD sedov.in
Or with output to a file: ./3DSPMHD sedov.in >& sedov.out &

Plot the results using "nsplash sedov_0*.dat"
(the splash.* files in the directory are already setup
 to plot the exact solution for this problem)

Things to try
--------------
- Run the code at with the default parameters:

 ./3DSPMHD sedov.in >& sedov.out &

  (being 3D, this problem takes a while to run, so best to run it in
   the background: follow the output using "tail -f sedov.out")
   
  (really we are getting to the limits of the code with this problem
   as it starts to need parallelisation, individual timesteps, etc...)

- Plot the density as a function of radius, alongside the exact solution

 nsplash -y 8 -x 1 sedov_0*.dat

- Plot a rendered plot of the column density. First change the coordinate
  system in splash back to cartesian:
  (o menu, option 7), then plot the density rendered 
  by selecting 2, 1 and 8 as the y, x and rendered data.
  
  Or use the splash settings files with these preset:

  nsplash -p coldens -r 8 -dev /xw    sedov_0*.dat

- Plot a slice of the density:

 (type x then choose option 1 in the splash menu to change to a slice view
  or just type "x1" from the main menu)

- make a movie of any of the above
  
  (choose /png instead of /xw at the PGPLOT prompt, this will produce a
   series of images called pgplot.png_1 pgplot.png_2 etc. Run the script
   in ~/splash/scripts/fixpgplotnames.bash to fix the silly filenames,
   then use your favourite movie-making tool to combine the sequence
   of images into a movie)
   
  (just a hint: movies look better with minimal annotation: so remove the
   axes and the colour bar by pressing backspace with the mouse over them
   in splash's interactive mode)

- Try a calculation that does not include artificial conductivity

 ./3DSPMHD sedovnocond.in >& sedovnocond.out &

  (As for our 1D shock tubes, adding thermal conductivity improves the
  treatment of the contact discontinuity. In this case the radial profiles
  of density and thermal energy are much noisier.)

- Try a calculation at higher resolution (up to around 100^3 is manageable
  if you leave it for a few hours), by changing the initial particle separation:

  0.010000000                      ! particle separation

Added by Daniel Price, July 2010
Revised Feb 2014
Checked and updated April 2015
