All,

I've run the head on NS collision with K-driving, using BAM
in stencil_width = 2 octant mode.  I've done the same run in
grid = "full" mode.  I've attached a gif showing trK along
the axis connecting the two NS.  I did this to check that
BAM, stencil=2, octant is working as expected.  

The pannel on the left is octant mode, the pannel on the
right is full mode.  Note that there is a small asymmetry for the
full run, since there are a different # of points to the left
of the origin than from the right, therefore, the boundaries
are not implemented exactly symmetrically.  (this is due to
running with odd number of points with staggered off the origin).

However, things look pretty good, considering this is K-driving
at every step, which should strongly test the octant/stencil=2
elliptic solver.  This certainly makes me feel more comfortable
that BAM/octant/stencil=2 is working properly.  Thanks to Bernd
for implementing this!

Mark


------------------------------------------------
Mark Miller
Department of Physics
Washington University
St. Louis, MO  63130-4899
E-mail: mamiller@wugrav.wustl.edu
office phone: (314) 935-4617,  fax: (314) 935-6219
------------------------------------------------

Filename: trK.gif
Size :59616

Created for the WUGRAV CoCoBoard Page Projects Page.

Created by The CoCoBoard.