Skip to main content

Browse the archive

Revision fe53297356da5f02478fe9cafab5d9914a36d2be authored by Thorsten Becker on 14 August 2007, 03:33:21 UTC, committed by Thorsten Becker on 14 August 2007, 03:33:21 UTC 
- added coor=2 option (for full citcom) to assign finer radial node
 
  spacing to top and lower layers of shell. The
  coor_refine=0.1,0.15,0.1,0.2 parameters specify the radius fraction
  of the bottom layer [0], the fraction of the nodes in this layer
  [1], the top layer fraction [2], and the top layer node fraction
  [3]. I.e. the defaults will put 15% of all nz nodes into the 10%
  lower layer, 20% in the top 10% upper layer, and the rest in
  between.
  

- renamed gzipped output version with sub-directory storage ascii-gz

- built in restart facilities for temperature and tracers when using
  ascii-gz I/O with vtkio != 2


- added a composition viscosity function, CDEPV, based on two tracer
  flavors

  - for this to work, I had to move viscosity_input() *behind*
    tic_input() and tracer_input() in instructions


- added tracer_enriched option for internal heating. If tracer = on
  and tracer_enriched = on, will reader Q0_enriched and vary the element heat production 
  between Q0 for C = 0 and Q0_enriched for C = 1. I.e. this only works
  if C varies between 0 and 1.


- added an option to write from all processros to a single VTK file,
  if ascii-gz is activated, and vtkio = 2. The VTK output is of the
  "legacy", serial, single-file type, and requires that all processors see the same 
   filesystem. 

   This will lead to a bottleneck for large # of CPU computations as
   each processor has to wait til the previous is done. 

   More efficient I/O should be possible by using the distributed
   storage version of VTK, but I have no clue how this works. Anyone?
1 parent d6e512c
Raw File
Citcom_init.c 
/*
 *~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 *
 *<LicenseText>
 *
 * CitcomS by Louis Moresi, Shijie Zhong, Lijie Han, Eh Tan,
 * Clint Conrad, Michael Gurnis, and Eun-seo Choi.
 * Copyright (C) 1994-2005, California Institute of Technology.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 *
 *</LicenseText>
 *
 *~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 */

#include "global_defs.h"
#include "citcom_init.h"

struct All_variables* citcom_init(MPI_Comm *world)
{
  int get_process_identifier();

  struct All_variables *E;
  int rank, nproc;

  E = (struct All_variables*) malloc(sizeof(struct All_variables));

  MPI_Comm_rank(*world, &rank);
  MPI_Comm_size(*world, &nproc);

  E->control.PID = get_process_identifier();
  E->parallel.world = *world;
  E->parallel.nproc = nproc;
  E->parallel.me = rank;

  /* fprintf(stderr,"%d in %d processpors, E at %p pid=%d\n",
          rank, nproc, E, E->control.PID); */

  E->monitor.solution_cycles=0;
  E->control.keep_going=1;

  E->control.total_iteration_cycles=0;
  E->control.total_v_solver_calls=0;

  return(E);
}
back to top