<span class="Apple-style-span" style="font-family: Times; font-size: medium; "><div><font class="Apple-style-span" size="7"><span class="Apple-style-span" style="font-size: 26px;"><b><span class="Apple-style-span" style="font-size: medium; font-weight: normal; "><pre>
Hope you had a good Xmas and New Year Eve break.</pre><pre>Attached is the minor changes I made in the SurfaceAdaptor files. The surface is</pre><pre>supposed to be defined on a regular square grid.</pre><pre>To use it in the xml input file you need to proceed like before except that you must add</pre>
<pre>the number of points in x,z directions. </pre><pre><struct name="surfaceAdaptor">
<param name="Type">SurfaceAdaptor</param>
<param name="mesh">mesh-linear</param>
<param name="sourceGenerator">linearMesh-generator</param>
<param name="surfaceType">topo_data</param>
<param name="surfaceName">test.topo</param>
<param name="nx">32</param>
<param name="ny">12</param>
</struct> </pre><pre>I've also attached the topography file to show the format (simply x,z,y).</pre><pre>I was thinking the PolygonShape was only for 2D simulation. The SurfaceShape is made for </pre><pre>defining 3D shape. It consists in giving for each shape 2 surfaces that represent its envelop.</pre>
<pre>It helps me to import easily our simulated sedimentary layers in Gale. I have attached as well as the files</pre><pre>2 snapshots to give you a better idea. To make it work in Gale few changes need to be done in the following files:</pre>
<pre>Init.c, SConscript, Shape.h, types.h </pre><pre>mainly to add the new shape type and the headers of the new file.</pre><pre>Considering the particle tracers it seems that it is exactly what I'm looking for. </pre>
<pre>I had a try with the ManualParticleLayout but there are some problems when the number of particle is 'large'.</pre><pre>Here is the kind of problem I have:</pre><pre>In func Materials_Register_AssignParticleProperties(): for swarm "passiveTracerSwarm"
Assigning initial particle properties to the 101 global particles
done 10% (11 particles)...
done 20% (21 particles)...
done 30% (31 particles)...
done 40% (41 particles)...
done 50% (51 particles)...
done 60% (61 particles)...
done 70% (71 particles)...
done 80% (81 particles)...
done 90% (91 particles)...
[0]PETSC ERROR: ------------------------------------------------------------------------
[0]PETSC ERROR: Caught signal number 11 SEGV: Segmentation Violation, probably memory access out of range
[0]PETSC ERROR: Try option -start_in_debugger or -on_error_attach_debugger
[0]PETSC ERROR: or see <a href="http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html#Signal[0]PETSC">http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html#Signal[0]PETSC</a> ERROR: or try <a href="http://valgrind.org">http://valgrind.org</a> on linux or man libgmalloc on Apple to find memory corruption errors
[0]PETSC ERROR: configure using --with-debugging=yes, recompile, link, and run
[0]PETSC ERROR: to get more information on the crash.
[0]PETSC ERROR: --------------------- Error Message ------------------------------------
[0]PETSC ERROR: Signal received!
[0]PETSC ERROR: ------------------------------------------------------------------------
[0]PETSC ERROR: Petsc Release Version 2.3.3, Patch 11, Thu Mar 27 09:16:38 CDT 2008 HG revision: 0a79e1469a317b0468b3b3310524bf13867c66f5
[0]PETSC ERROR: See docs/changes/index.html for recent updates.
[0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.
[0]PETSC ERROR: See docs/index.html for manual pages.
[0]PETSC ERROR: ------------------------------------------------------------------------
[0]PETSC ERROR: /Users/sal116/workspace/Gale/bin/Gale on a underworl named <a href="http://matris-kf.per.dpr.csiro.au">matris-kf.per.dpr.csiro.au</a> by sal116 Mon Jan 4 09:40:42 2010
[0]PETSC ERROR: Libraries linked from /usr/local/lib/underworld
[0]PETSC ERROR: Configure run at Fri Oct 23 13:33:43 2009
[0]PETSC ERROR: Configure options CFLAGS=-Wl,-flat_namespace --with-shared=1 --with-debugging=0 --with-fortran=0 --with-blas-lapack-dir=/usr/ --with-mpi-dir=/usr/local/mpich2-optimised --prefix=/usr/local
[0]PETSC ERROR: ------------------------------------------------------------------------
[0]PETSC ERROR: User provided function() line 0 in unknown directory unknown file
application called MPI_Abort(MPI_COMM_WORLD, 59) - process 0
[unset]: aborting job:
application called MPI_Abort(MPI_COMM_WORLD, 59) - process 0 </pre><pre><br></pre><pre>I don't know if you have any idea on how to solve this problem of memory access.</pre><pre><br></pre><pre>Cheers</pre><pre><br></pre>
<pre>Tristan</pre><pre> </pre></span></b></span></font></div><hr><pre>Tristan Salles-Taing <<a href="http://geodynamics.org/cgi-bin/mailman/listinfo/cig-long">sallestaing at gmail.com</a>> wrote:
><i> For the moment Gale is not able to restart from a considered time step such
</i>><i> as Underworld so the material characteristics are lost between two starts
</i>><i> but Guillaume Duclaux (CSIRO such as me)
</i>><i> said that it will maybe be available in the next release.
</i>
Restarting from a checkpoint will definitely be available in the next release.
><i> Then I will have
</i>><i> to see how I can restart a Gale simulation with additional materials
</i>><i> (consisting of Sedsim stacked sedimentary layers)...anyway
</i>><i> that's not the purpose of my mail.
</i>><i>
</i>><i> First of all, if you are interested in the modifications I made how can I
</i>><i> send them to you (I suppose you want to check them so is it ok if I use
</i>><i> subversion) ?
</i>
I am very interested in the modifications to SurfaceAdaptor. I always
had such a fix in the back of my mind. As for SurfaceShape, I do not
quite see how it differs from a regular polygon shape. If you just
send me the files, then I would probably have a better idea.
><i> Secondly, I still have some issues when I try to export the modified
</i>><i> sedimentary layers from Gale. I already lost a lot of precision when I
</i>><i> import the SurfaceShape from Sedsim to Gale due to the fact that
</i>><i> Gale is distributing the swarm particle randomly inside a cell. So there
</i>><i> isn't a perfect match between my SurfaceShape coordinates and the swarm
</i>><i> particles. Then at the end of Gale run, I am computing
</i>><i> new surfaces which delimitate the different particles layers after
</i>><i> deformation. These surfaces are used to find the new sedimentary layers
</i>><i> thicknesses. However the surface tracking is really poor mainly
</i>><i> because it relies on randomly located particles.
</i>><i>
</i>><i> Is there a way of forcing a particle two be at a specific location (X,Y,Z)
</i>><i> and can I track this particle during Gale run ?
</i>
You can add tracers at particular points that are then advected along
with the other particles. I am still not 100% confident about how
well they work, so I have not documented them. There is an example of
how to use them in
input/examples/tibet.xml
Cheers,
Walter Landry
<a href="http://geodynamics.org/cgi-bin/mailman/listinfo/cig-long">walter at geodynamics.org</a>
</pre><div><font class="Apple-style-span" face="monospace"><span class="Apple-style-span" style="white-space: pre;"><br></span></font></div></span>