CFD Utilities Files

Original TEMPLATE Description:
------------------------------

   Make a FLO107MB connection file template from a PLOT3D grid file.
   The name 'template' is used because most but not all of the block
   face connectivity/boundary condition information can be deduced
   from just the grid, although some educated guesses are applied to
   the faces that seem not to match other block faces.

   The grid file may be single or mgrid, real*4 or real*8, though it
   only makes sense to process multiblock grids.

   Use "template -v" for verbose template.con file (last column will
   indicate what fraction of quads of discovered abutting faces match
   within epsilon distance at all four corners.

   (This version turns on "-v" explicitly.)

   There are nb blocks, the nth block has dimensions nj(n),nk(n),nl(n).
   Each block becomes a network of six patches, one for each side.  The
   sides are numbered 1 to 6 as:  jmin, jmax, kmin, kmax, lmin, lmax.
   Each quad face of each side of each block is stored as a four-tuple
   of integers that point into xyz.  Component numbers are the sides
   1-6 for the first block, then 7-12 for the second block, and so on
   to 6*nb.  An index array iaf into ndf points to the nth component's
   quads which range from iaf(n) to iaf(n+1)-1.  The xtrema array is
   used to store the xyz minmax values for each component (for each
   side of each block).  The component xyz extrema are used to match
   sides within epsilon.  ihand(n) is the handedness of the nth block,
   right=1, left=-1; it is found roughly by adding some tet volumes and
   checking the sign of the result.

   Scott D. Thomas, Sterling Software, Inc.
   Contract NAS 2-13210/6/5, NASA Ames Research Center.
   25-JUN-1998..30-JUN-1998.

   Scott D. Thomas, Raytheon STX Corporation
   Contract NAS 2-98080/53, NASA Ames Research Center.
   1-JUL-1998..9-JUL-1998.

Adaptation within the Space Technology Division at Ames:
--------------------------------------------------------

DPLR Input and Interface Files:

   This version writes a GASP control file and two DPLR control files
   as well, and handles formatted grids as well as the original real*4
   or real*8 unformatted.

Ancillary control file 'template.inp.2' usage:

   No 'template.inp' control file is in the picture here, but the name
   'template.inp.2' is used for this optional control file to be consistent
   with other grid manipulation utilities.

   The leading keyword on the first of each optional pair of input lines is
   converted to upper case here, and only the first 4 characters are used.

   Pairs of control lines are read until EOF, so the order of the pairs
   does not matter.

Optional usage (1) (obsolete with DPLR V4.03, since ntx/y/z are gone now):

   Control the type of sequencing (grid coarsening) specified in the
   'dplr.inputs.2' output.  Originally, 2 2 1 coarsening was hard-coded.
   Now, a pair of lines as shown can produce whatever is really needed.
   The default coarsening remains 2 2 1.  Example:

   Sequencing controls
   4 4 2              ! Coarsening factors in the i, j, k directions

Optional usage (2):

   Force boundary condition changes in the two DPLR control files as needed
   for proper analysis of cavity or plug repair cases.

   Cavity block(s)    ! Any BC 1 or 2 will be changed to BC 26 (wall)
   6                  ! Any reasonable list of block numbers

or

   Plug blocks        ! BC 2 at jmin will be changed to BC 26 (wall)
   10:13              ! Any reasonable list of block numbers

History beyond Scott's original implementation:
-----------------------------------------------

  [Note that Scott uses a pre-Fortran 90 form of dynamic memory allocation
   that appears more trouble than it is worth to replace with the modern
   forms of allocate and deallocate.]

13-MAY-2004  David Saunders  Command line arguments don't work with
             ELORET Corp./   Intel's ifc compiler;
             NASA Ames       the grid may be formatted now.
14-MAY-2004    "             Added output of two DPLR-related files,
                             derived from the saved FLO107MB data.
17-MAY-2004    "             Added output of GASP format as well.
23-OCT-2004    "             Dinesh asked for dplr.inputs.2 with the
                             i and j block dimensions halved.
23-DEC-2004    "             DPLR-related BCs for a grid with more
                             than one layer of blocks were being
                             affected by the one-layer case, but
                             we shouldn't make assumptions about
                             k = 1 or nk faces unless matching faces
                             have not been found.
13-JAN-2005    "             Changed "init1" to "initi" so that the
                             Perl script for updating DPLR control
                             files doesn't misbehave.
26-MAY-2005    "             Option to output complete dplr.inputs:
                             if 'sample.inputs' is present, its
                             header and trailer are transcribed.
14-NOV-2005    "             Distinguish the symmetry planes in the
                             dplr.inputs files (not always y = 0).
18-NOV-2005    "             With DPLR V3.04 or later now the usual
                             choice, use BC 2 where BC 3 was used
                             (specified inflow or supersonic outflow,
                             rather than just supersonic outflow).
03-JAN-2006    "             The symmetry plane test should not use
                             too loose a tolerance. Make it <= 0.001.
11-APR-2006    "             Introduced ancillary control file
                             'template.inp.2' to allow automation of
                             BC changes in 'dplr.inputs' for cavity
                             or plug repair cases.
13-APR-2006    "             Chun Tang asked for a way to control the
                             type of grid sequencing in 'dplr.inputs.2',
                             so that has been added to the options
                             enabled by 'template.inp.2' (which still
                             need not be present).
10-MAY-2006    "             DPLR V3.05's new grid tailoring controls
                             required changing how the header lines
                             from 'sample.inputs' are transcribed.
27-MAY-2009    "             DPLR V4 expects "ibadpt" on the end of the
                             first line of BC info.  DPLR V3 will not be
                             affected by its presence.
27-MAR-2010    "             Long chemistry file names were being
                             truncated beyond column 80.
11-FEB-2014  DAS, ERC Inc./  The dplr.inputs output file now suits
             NASA Ames       DPLR V4.03: the grid block dimensions of
                             V4.02 and earlier are no longer present.
                             dplr.inputs.2 is therefore redundant.