Re: [Gmsh] patching CMakeLists.txt for TAUCS
Dear Joseph,
The Taucs timing code uses ftime, which indeed is obsolete. I've changed our
Taucs config to disable timings - we don't use them anyway. The fix is in SVN.
Christophe
> On 01 Sep 2016, at 16:09, Joseph Mingrone wrote:
>
> Hello,
>
> There is a linking issue with the FreeBSD gmsh package when TAUCS is enabled.
> Here is a build snippet.
>
>
> CMakeFiles/gmsh.dir/contrib/taucs/G/taucs_timer.c.o: In function
> `taucs_wtime':
> /wrkdirs/usr/ports/cad/gmsh/work/gmsh-2.13.2-source/contrib/taucs/G/taucs_timer.c:(.text+0xd):
> undefined reference to `ftime'
>
>
> I believe adding '-lcompat' when linking gmsh solves the problem. How can
> cmake be instructed to do this?
>
> I tried patching CMakeLists.txt as below, but without success.
>
> Regards,
>
> Joseph
>
> --- CMakeLists.txt.orig 2016-08-31 20:17:30 UTC
> +++ CMakeLists.txt
> @@ -887,6 +887,10 @@ if(HAVE_SOLVER)
> else(WIN32)
> add_definitions(-DOSTYPE_linux)
> endif(WIN32 OR CYGWIN)
> + find_library(COMPAT_LIB compat)
> + if(COMPAT_LIB)
> +list(APPEND LINK_LIBRARIES ${COMPAT_LIB})
> + endif(COMPAT_LIB)
> file(GLOB_RECURSE TAUCS_D
> ${CMAKE_CURRENT_BINARY_DIR}/contrib/taucs/D/*.c)
> file(GLOB_RECURSE TAUCS_S
> ${CMAKE_CURRENT_BINARY_DIR}/contrib/taucs/S/*.c)
> file(GLOB_RECURSE TAUCS_Z
> ${CMAKE_CURRENT_BINARY_DIR}/contrib/taucs/Z/*.c)
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] How to find if a variable exists
> On 07 Sep 2016, at 22:30, Jim Monte wrote: > > Hi All, > > I have written macros returning arrays of various items, such as point IDs. > Depending on the arguments to the function, some arrays can be of different > sizes. To ensure that there is no "junk" at the end of the arrays left from a > previous call, I would like to Delete the arrays before building them. > However, the first time the macro is called, the array does not exist, and > that causes an error. One workaround is to force the array to exist by adding > something like x[0] = 0; before the Delete command, but I was wondering if > there is a more direct way to query if a specific variable exists. Thanks for > any information. > Exists(name) will do the trick > Jim Monte > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Question about early post increment
Jim - Indeed, good catch: our post-increment var++ was actually behaving like
++var. This is now fixed in SVN.
> On 10 Sep 2016, at 00:39, Jim Monte wrote:
>
> Hi All,
>
> I was trying to build arrays of point IDs and related things and used the
> post increment essentially as shown in the simple example below. In this
> example, the array element pid[0] is unassigned while pid[1] has a value,
> although I would have expected pid[0] to be the only assigned value. Am I
> misunderstanding how the increment operator is supposed to work in Gmsh?
>
> Jim Monte
>
> /*** start of example */
> i = 0; /* index to current point ID */
> j = newp; /* get a new point ID */
> pid[i++] = j; /* save point ID */
> Point(j) = {1, 2, 3}; /* assign value to point */
> xyz[] = Point{pid[0]}; /* retrieve coordinates of point */
> Printf("x=%g; y=%g; z=%g", xyz[0], xyz[1], xyz[2]); /* output */
> xyz[] = Point{pid[1]}; /* retrieve coordinates */
> Printf("x=%g; y=%g; z=%g", xyz[0], xyz[1], xyz[2]); /* output */
> /*** end of example */
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Questions related to Extrusions
> On 14 Sep 2016, at 20:44, Jim Monte wrote:
>
> Hi,
>
> I would like to know how to find the order of the "side" elements (lines or
> surfaces) that are generated during an extrusion. Thanks for any information.
When you extrude a line, the surface is created from a line loop defined by the
original line, the line extruded from the end point, the (reversed) "top" line
and (reversed) lines extruded from the starting point.
When you extrude a surface, the surface loop is created with the original
surface, the "top" surface, then all the surfaces extruded from the sides of
the original surface (in the order they appear in the surface definitions).
Note that these resulting entities can be changed if they are found to
duplicate existing entities, through the explicit or implicit call to
"Coherence;".
>
> Also, while doing some related testing, I saw some results that I could not
> explain from the script below.
>
> /*** start ***/
> Printf("next line ID = %g; next surface ID = %g", newl, news);
> Point(1) = {0, 0, 0};
> Point(2) = {0, 1, 0};
> Line(1) = {1, 2};
> Printf("next line ID = %g; next surface ID = %g", newl, news);
>
> out[] = Extrude {0, 0, 1} {Line{1}; };
>
> n = #out[];
> Printf("Extrude has returned %g elements", n);
> n -= 1;
> For i In {0 : n}
> Printf("Extrusion value[%g] = %g.", i, out[i]);
> EndFor
>
> Printf("next line ID = %g; next surface ID = %g", newl, news);
> /*** end ***/
>
> The output is
>
> 1 next line ID = 1; next surface ID = 1
> 2 next line ID = 2; next surface ID = 2
> 3 Extrude has returned 4 elements
> 4 Extrusion value[0] = 2.
> 5 Extrusion value[1] = 5.
> 6 Extrusion value[2] = 4.
> 7 Extrusion value[3] = -3.
> 8 next line ID = 6; next surface ID = 6
>
> Here are the things that did not seem right:
> • The value returned by news in line 2 is too high. No surface was
> created yet, so it should remain as 1.
> • The next free surface has a value of 1, so line 5 should give the
> extruded surface a value of 1 instead of 5.
This is for backward compatibility with old Gmsh versions: if you prefer a more
natural numbering, you can set "Geometry.OldNewReg=0;" at te beginning of your
script.
> • I thought the IDs should all be positive, so line 7 seems wrong.
Curves actually exist with both orientations in Gmsh; use "FAbs()" to get the
ID.
Christophe
> I would appreciate some explanations of what Gmsh is doing in these instances.
>
> Jim Monte
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] boundary layer disconnects at the corner of geometry(stl)
Dear Praveen,
The Boundary Layer field was an experiment: the 3D version was not functional
and has been removed in recent versions. The 2D version is fairly usable. New
implementation of boundary layers are forthcoming. In 3D, I currently advise
simply using the Extrude command: it will not create fans, but for simple
geometries like yours it should be sufficient.
Here's an example:
// create a square
lc = 0.3;
Point(1) = {0.0,0.0,0.0,lc};
Point(2) = {1,0.0,0.0,lc};
Point(3) = {1,1,0.0,lc};
Point(4) = {0,1,0.0,lc};
Line(1) = {4,3};
Line(2) = {3,2};
Line(3) = {2,1};
Line(4) = {1,4};
Line Loop(1) = {2,3,4,1};
Plane Surface(1) = {1};
// create a cube
Extrude {0,0.0,1} { Surface{1}; }
// remove the volume
Delete { Volume{1}; }
// extrude the boundary of the cube inwards by 0.05, with 5 layers of elements
Extrude { Surface{26, -1, 13, 17, 21, 25}; Layers{5, 0.05}; }
// create the inner volume
Surface Loop(159) = {92, 158, 70, 48, 136, 114};
Volume(160) = {159};
General.ExpertMode = 1; // disable warning about mixing Delaunay and extrusion
> On 14 Sep 2016, at 10:52, Praveen Kumar wrote:
>
> Hi,
>
> I am trying to generate boundary layers on all the surfaces of a simple
> rectangular duct geometry (imported stl). The prism layers are not getting
> connected on the edges/corners between each surfaces.
>
>
>
>
> Can someone help me understand what is missing here? Please Note, I am using
> 2.11.0 version of gmsh as this is not working at all in the latest version
>
> --
> Merge "duct.stl";
> CreateTopology;
>
> ll[] = Line "*";
> l = newl;
> For j In {0 : #ll[]-1}
> Compound Line(l+j) = ll[j];
> EndFor
>
> ss[] = Surface "*";
> s = news;
> For i In {0 : #ss[]-1}
> Compound Surface(s+i) = ss[i];
> EndFor
>
> bl = newl;
> bs = news;
>
> Field[1] = MathEval;
> Field[1].F = "0.1";
>
> Background Field = 1;
>
> Mesh.CharacteristicLengthExtendFromBoundary = 0;
>
> Physical Surface("Side-01") = {s};
> Physical Surface("Top") = {s+1};
> Physical Surface("Side-02") = {s+2};
> Physical Surface("Outlet") = {s+3};
> Physical Surface("Bottom") = {s+4};
> Physical Surface("Inlet") = {s+5};
>
> sl = newsl; Surface Loop(sl) = {s:s+5};
> vo = newv; Volume(vo) = {sl};
>
> //Physical Volume("V_Fluid") = {vo};
>
> pp[] = Point "*";
>
> Field[2] = BoundaryLayer;
> //Field[2].AnisoMax = 60;
> Field[2].FacesList = {s:bs-1};
> Field[2].EdgesList = {l:bl-1};
> Field[2].NodesList = {pp[0]:pp[7]};
> //Field[2].FanNodesList = {pp[0]:pp[7]};
> //Field[2].IntersectMetrics = 0;
> //Field[2].Quads = 0;
> //Field[2].hfar = 0.1;
> Field[2].hwall_n = 0.025;
> //Field[2].hwall_t = 0.1;
> Field[2].ratio = 1.1;
> Field[2].thickness = 0.05;
> BoundaryLayer Field = 2;
> //Background Field = 2;
>
> --
>
> --
> Praveen Kumar R | Technical Manager
> Centre for Computational Technologies - CCTech
> 1 Akshay Residancy | 50 Anand Park | Aundh | Pune | 411007 | India
> Ph: +91 20 40098381/82 | M: +91-9923406816 | www.cctech.co.in
> CCTech – "Transforming human life by democratization of technology"
>
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Questions related to Extrusions
> On 15 Sep 2016, at 20:47, Jim Monte wrote:
>
>
> On Thu, Sep 15, 2016 at 3:23 AM, Christophe Geuzaine
> wrote:
>
> > On 14 Sep 2016, at 20:44, Jim Monte wrote:
> >
> > Hi,
> >
> > I would like to know how to find the order of the "side" elements (lines or
> > surfaces) that are generated during an extrusion. Thanks for any
> > information.
>
> When you extrude a line, the surface is created from a line loop defined by
> the original line, the line extruded from the end point, the (reversed) "top"
> line and (reversed) lines extruded from the starting point.
>
> When you extrude a surface, the surface loop is created with the original
> surface, the "top" surface, then all the surfaces extruded from the sides of
> the original surface (in the order they appear in the surface definitions).
>
> Note that these resulting entities can be changed if they are found to
> duplicate existing entities, through the explicit or implicit call to
> "Coherence;".
>
> >
> > Also, while doing some related testing, I saw some results that I could not
> > explain from the script below.
> >
> > /*** start ***/
> > Printf("next line ID = %g; next surface ID = %g", newl, news);
> > Point(1) = {0, 0, 0};
> > Point(2) = {0, 1, 0};
> > Line(1) = {1, 2};
> > Printf("next line ID = %g; next surface ID = %g", newl, news);
> >
> > out[] = Extrude {0, 0, 1} {Line{1}; };
> >
> > n = #out[];
> > Printf("Extrude has returned %g elements", n);
> > n -= 1;
> > For i In {0 : n}
> > Printf("Extrusion value[%g] = %g.", i, out[i]);
> > EndFor
> >
> > Printf("next line ID = %g; next surface ID = %g", newl, news);
> > /*** end ***/
> >
> > The output is
> >
> > 1 next line ID = 1; next surface ID = 1
> > 2 next line ID = 2; next surface ID = 2
> > 3 Extrude has returned 4 elements
> > 4 Extrusion value[0] = 2.
> > 5 Extrusion value[1] = 5.
> > 6 Extrusion value[2] = 4.
> > 7 Extrusion value[3] = -3.
> > 8 next line ID = 6; next surface ID = 6
> >
> > Here are the things that did not seem right:
> > • The value returned by news in line 2 is too high. No surface was
> > created yet, so it should remain as 1.
> > • The next free surface has a value of 1, so line 5 should give the
> > extruded surface a value of 1 instead of 5.
>
> This is for backward compatibility with old Gmsh versions: if you prefer a
> more natural numbering, you can set "Geometry.OldNewReg=0;" at te beginning
> of your script.
>
> > • I thought the IDs should all be positive, so line 7 seems wrong.
>
> Curves actually exist with both orientations in Gmsh; use "FAbs()" to get the
> ID.
>
> Christophe
>
>
> > I would appreciate some explanations of what Gmsh is doing in these
> > instances.
> >
> > Jim Monte
> >
> > ___
> > gmsh mailing list
> > gmsh@onelab.info
> > http://onelab.info/mailman/listinfo/gmsh
>
> Thanks for the details on extrusion. But I am still not seeing how to
> determine which index will receive the ID for a particular surface. Perhaps I
> am trying to do something the hard way. My original interest in this question
> is that I want to know where specific surfaces will be put in the returned
> array so I can associate a physical surface with specific plane surfaces that
> get formed as part of the extrusion. Is there another approach to do this
> programatically (without looking at the GUI for information).
>
> I did more testing with a "brick" shape, made by two extrusions,
> line->square->brick. The Geometry.OldNewReg option did help to make the IDs
> closer to the expected values, but it is still confusing what is being done
> by the program. At a high level, the brick has 8 points, 12 lines, and 6
> plane surfaces. But there are
>
> 49-1-8 = 40 unused points,
> 21-1-12 = 8 unused lines (9, 12, 13, 15, 16, 18, 19, 20), and
> 8-1-6 = 1 unused surface (2)
> generated as part of the extrusion. Is there some way to query what will be
> made by the extrusion?
>
The extrusion works by recursively extruding entities of lower dimension, so
when you extrude a surface, you actually extrude the curves first; each curve
extrusion creates a surface but also new curves and new points. Some of these
are duplicates, and they are merged. Hence all the "unused" points, lines, etc.
You can retrieve the IDs programmatically: you ju
Re: [Gmsh] Instabilities of mesh generation in interactive and batch mode of gmsh
> On 20 Sep 2016, at 17:07, andreas.schus...@dlr.de wrote: > > Hi, > > > I’m using gmsh for creating a finite element model of a full aircraft model > for structural analysis. > For that purpose, I have written a python process which writes the input > geometry of the aircraft as .geo file and pass it to gmsh via a simple batch > process in Python. My problem is that when I run that process in batch mode, > gmsh is running and produces some files called “before.msh” or “after.msh” > again and again and seems not to converge the mesh generation. > If I start the GUI of gmsh with the input geometry and start the mesh > generation inside, gmsh finishes after around 4 seconds with a nice mesh! > The aircraft is mainly defined by regular 4 point areas which I recombine to > get quadrilateral elements. At some places due to the local aircraft > structure definition, there are some areas with 5 lines which produces some > small degenerated elements. > > I already tested different meshing algorithms and options directly written in > the geometry input file to overcome to problem, but the batch process of gmsh > doesn’t finish, though. > > Can anybody help me with this problem? Weird... The "before.msh"/"after.msh" files are created for debugging purposes in the triangle->quad recombination algorithm, when the "-saveall" option is used. (This will be removed in a future version.) What happens if you run Gmsh (without Python), directly from the command line, e.g. gmsh myfile.geo -2 > Thanks in advance! > > Best regards > Andreas > > —— > Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR) > Institut für Faserverbundleichtbau und Adaptronik | > > > _______ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] redundant nodes when using transfinite
> On 20 Sep 2016, at 14:53, Thomas Unfer wrote: > > Hi there, > > I am new to this list, so first how all greetings to everybody. > I am trying to mesh two adjacent boxes with hexaedra using transfinite (.geo > attached) plus a few surfaces to apply different boundary conditions in my > solver. > The meshing goes alright and the cells are (visually) conforming. > However in the .msh file the physical surfaces and volumes rely on different > point ids which makes it impossible to retrieve proper connectivity. I tryed > Coherence but it did no good. Does any one has any clue how to fix this? > Your additional surfaces (e.g. surface 26) are not topologically connected to the volumes: the quadrangles on the surfaces will thus be completely independent of the hexahedra in the volumes. With recent nightly builds you could use "Coherence Mesh;" to glue the mesh back together after the fact, i.e., end your script with Mesh 3; Coherence Mesh; Save "mymesh.msh"; Not super elegant, but that should do the trick... > Thanks for your help, > > Thomas > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Problem with periodic mesh around cylinder with extrusion
> On 19 Sep 2016, at 18:41, ashish bhole wrote: > > Dear All, > > I am facing problem with 'periodic surface' in gmsh. I have created 3D mesh > around cylinder using extrusion (.geo attached here). > When I use Periodic Surface feature of gmsh it do not write complete periodic > data in .mhs file. > Indeed, with "Layers" the mesh is generated in a structured way, and no periodic information is saved - the periodic mesh generator is completely bypassed. > Am I using this feature correctly, specially when I using along with > extrusion feature? > > I can generate periodic mesh inside cube without any problem. > > -- With Regards > Ashish > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Interpolation of polynomial solutions
> On 19 Sep 2016, at 18:53, GARNAUD Xavier (SAFRAN) > wrote: > > Dear gmsh developpers, > > I’d like to know if it is possible to interpolate a polynomial solution at > given locations through the python interface. Only the adapted (h-refined) piecewise linear approximation I'm afraid, using the plugins. > Best regards, > > Xavier > # > " Ce courriel et les documents qui lui sont joints peuvent contenir des > informations confidentielles, être soumis aux règlementations relatives au > contrôle des exportations ou ayant un caractère privé. S'ils ne vous sont pas > destinés, nous vous signalons qu'il est strictement interdit de les > divulguer, de les reproduire ou d'en utiliser de quelque manière que ce soit > le contenu. Toute exportation ou réexportation non autorisée est interdite.Si > ce message vous a été transmis par erreur, merci d'en informer l'expéditeur > et de supprimer immédiatement de votre système informatique ce courriel ainsi > que tous les documents qui y sont attachés." > ** > " This e-mail and any attached documents may contain confidential or > proprietary information and may be subject to export control laws and > regulations. If you are not the intended recipient, you are notified that any > dissemination, copying of this e-mail and any attachments thereto or use of > their contents by any means whatsoever is strictly prohibited. Unauthorized > export or re-export is prohibited. If you have received this e-mail in error, > please advise the sender immediately and delete this e-mail and all attached > documents from your computer system." > # > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Query
> On 13 Sep 2016, at 18:18, UMESH KUMAR SETH > wrote: > > Hello, > > I am a new user to GMSH. > I downloaded the current stable release (version 2.13.2) 64 bit Linux > version of GMSH, last week, from the gmsh website (http://gmsh.info/). > When I checked the About Gmsh option in Help menu, I found it says Version > 2.8.3. You probably have a path-related issue: if Help->About says "2.8.3" you are clearly runnning the old version. > The file I downloaded from the website is named "gmsh-2.13.2-Linux64.tgz". > Because of this I am not sure if I am using the latest version 2.13.2 . > > The problems I face : > > 1.) As soon as I launch gmsh from the command prompt I find these two lines, > always. > > libGL error: failed to open drm device: Permission denied > libGL error: failed to load driver: i965 > > My system is Linux 3.13.0-77(Ubuntu version 14.04. 4LTS) > > 2) gmsh window hangs too much, when I go for generating the mesh (2D or 3D) > even with simple cube of 3 nodes per line. > 3) Sometimes, When I change the number of nodes on lines in my script and > then Reload and try to mesh with 3D from GUI window of gmsh, the window > CRASHES abruptly. > 4) I see the error messages : > > gmsh: ../../src/XlibInt.c:245: _XPrivSyncFunction: Assertion > `dpy->synchandler == _XPrivSyncFunction' failed. > [h2-efd-seth:03831] *** Process received signal *** > [h2-efd-seth:03831] Signal: Aborted (6) > [h2-efd-seth:03831] Signal code: (-6) > *** Error in `gmsh': double free or corruption (fasttop): 0x7fdefc000960 > *** > Abandon (core dumped) > > > gmsh: ../../src/xcb_io.c:529: _XAllocID: Assertion `ret != inval_id' failed. > [h2-efd-seth:12728] *** Process received signal *** > [h2-efd-seth:12728] Signal: Aborted (6) > [h2-efd-seth:12728] Signal code: (-6) > gmsh: ../../src/xcb_io.c:529: _XAllocID: Assertion `ret != inval_id' failed. > Abandon (core dumped) > > > Could someone help me with this, please ? > > Many Thanks, > Regards, > > > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] reading stl and meshing
> On 19 Sep 2016, at 17:39, Martin Genet wrote: > > Dear all: > > Attaching a gmsh script + stl that used to work fine (loading the stl, > remeshing the surface and then meshing the volume), at least until 2.11 (see > test.2.11.0.log), but not anymore (see test-2.13.2.log). I tried with 2.12.0, > 2.13.0, 2.13.1, 2.13.2, but it just hangs after "Re-orient all 0 face normals > coherently". Any idea on what might be going on? Thanks a lot! > Indeed, that's a regression. We are currently rewriting the STL remesher from scratch - this will soon be ready in "beta" and should be much more robust than the old code. > Martin > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Possible bug when extruding relatively small surfaces
Dear Kate, This is "normal" : by default the tolerance to distinguish separate points/vertices is set to 1e-6 times the size of the model bounding box. Set e.g. Geometry.Tolerance = 1e-12; at the beginning of your script. Christophe > On 21 Sep 2016, at 14:26, look wrote: > > Hello all, > > I think I found a bug related to the extrusion (and recombination) of > relatively small surfaces. > > The problem occurs when you try to extrude very small 2-D cells into the > third dimension while having a huge domain, like it is often necessary for a > resolved boundary layer mesh for a 2-D airfoil simulation. > Gmsh then fails to recombine the extruded mesh into quadrangles. > > You can test this with the attached minimal example. It creates the mesh well > until you uncomment the last line. You can then see in the GUI > statistics-panel, that gmsh created triangles, which is obviously not > intended. > > Please see also the related discussion at cfd-online.com: > http://www.cfd-online.com/Forums/openfoam-solving/164600-gmshtofoam-fatal-error.html > > Best regards, > > Kate > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Interactive gmsh command line?
Hi Todd, Just put all the commands you want in a text file, and run Gmsh on it: either "gmsh file" to show the GUI, or "gmsh file -" to don't show the GUI. If you really want an interactive command line like gnuplot, have a look at gmsh/utils/solvers/c++/interactive.cpp in the Gmsh source code: it's an example on how you can build such an interactive system by communicating with a Gmsh server process. Christophe > On 3 Oct 2016, at 14:51, Todd Pierce wrote: > > Hi Crew, > > I am completely new to gmsh. I may have posted this question before, but > this time I actually have gmsh installed so this time I am prepared. > > I was hoping to use gmsh as a "back end" to a program I am writing, similar > to the way I utilize gnuplot. What I do there is simply start gnuplot and > hurl commands at it through a pipe and the graphic window magically opens up > with results. > > Now, I know gmsh is a totally different animal. However, if I could have a > command line to work with and throw commands at it to define the 3D structure > of something and perform meshing (among other things) this would open up many > doors. Of critical importance would be to see the output in a graphical > window, since my program has no graphical capabilities. > > Does this make sense to anybody? Has anybody done this? > > Thanks, > > -Todd > _______ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] aaaaaaaaaaaaaaargh
Hi Jeremy - I've removed both debug statements from SVN; this is the cost of
living on the bleeding edge ;-)
> On 3 Oct 2016, at 15:37, German Theler wrote:
>
> Hi all
>
> Following up with (still unresolved) hard-coded printfs,
> http://onelab.info/pipermail/gmsh/2016/010837.html
> I am now getting the message 'aaargh':
>
> gtheler@tom:~/codigos/3ros/gmsh$ svn blame Mesh/meshGRegion.cpp | grep
> -C3 aaargh
> 12845 geuzaine
> 12808remacle for(int j = 0; j < 3; j++){
> 22971ajohnen if (out.trifacelist[i * 3 + j] - 1 >=
> initialSize){
> 22971ajohnen printf("aaargh\n");
> 22971ajohnen //if(v[j]->onWhat()->dim() == 3){
> 22971ajohnen v[j]->onWhat()->mesh_vertices.erase
> 22971ajohnen
> (std::find(v[j]->onWhat()->mesh_vertices.begin(),
> gtheler@tom:~/codigos/3ros/gmsh$
>
>
> Any comment or solution is appreciated.
> Thanks!
>
> --
> Jeremy Theler
> www.seamplex.com
>
>
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Adding items to physical entities
Hi Alessandro - you can use e.g. "Physical Volume(1) += {1,2,3};" (or simply
use lists to construct what you need, and affect the list at the end).
Christophe
> On 30 Sep 2016, at 08:48, Alessandro Vicini
> wrote:
>
>
> Hullo everybody,
>
> is there a way to add elementary entities to existing physical groups? Thank
> you.
>
> A.
>
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] How to generate 'plugin crack' for post processing
Hi Xiaoxin,
Plugin(Crack) works on meshes: you thus need to have a mesh before you can run
it. For example, add "Mesh 3;" before calling the plugin.
Christophe
> On 29 Sep 2016, at 17:44, Xiaoxin Lu wrote:
>
> Hello,
>
> I am a PhD student using GMSH to generate crack faces. However, I could not
> find any example for this kind of post processing.
> My .geo file is attached below, I put 'plugin crack' command inside but it
> did not work, I was wondering how to generate a post process file for this
> command.
>
> Thank you for your help.
>
> best regards,
> Xiaoxin
> //start of the file
> lc1 = 3;
> lc2 = 3;
>
> // We proceed by defining some elementary entities describing a cube:
>
> Point(1) = {0,0,0,lc1};
> Point(2) = {40,0,0,lc1};
> Point(3) = {40,40,0,lc1};
> Point(4) = {0,40,0,lc1};
> Point(5) = {0,0,40,lc1};
> Point(6) = {40,0,40,lc1};
> Point(7) = {40,40,40,lc1};
> Point(8) = {0,40,40,lc1};
>
> Line(30)={1,2};
> Line(31)={2,3};
> Line(32)={3,4};
> Line(33)={4,1};
> Line(34)={5,6};
> Line(35)={6,7};
> Line(36)={7,8};
> Line(37)={8,5};
> Line(38)={1,5};
> Line(39)={2,6};
> Line(40)={3,7};
> Line(41)={4,8};
> Line Loop(42)={30,31,32,33};
> Plane Surface(14)={42};
> Line Loop(43)={-37,-36,-35,-34};
> Plane Surface(16)={43};
> Line Loop(44)={38,34,-39,-30};
> Plane Surface(18)={44};
> Line Loop(45)={-32,40,36,-41};
> Plane Surface(20)={45};
> Line Loop(46)={-33,41,37,-38};
> Plane Surface(22)={46};
> Line Loop(47)={39,35,-40,-31};
> Plane Surface(24)={47};
> vol1[]=Extrude{0,0,40}{Surface{14};};
> Delete {Volume{vol1[1]};}
> vol2[]=Extrude{0,40,0}{Surface{18};};
> Delete {Volume{vol2[1]};}
> vol3[]=Extrude{40,0,0}{Surface{22};};
> Delete {Volume{vol3[1]};}
> Physical Surface(1)={14};
> Physical Surface(2)={vol1[0]};
> Physical Surface(3)={18};
> Physical Surface(4)={vol2[0]};
> Physical Surface(5)={22};
> Physical Surface(6)={vol3[0]};
> Surface Loop(1)={14,vol2[0],vol3[0],18,22,vol1[0]};
> Volume(1)={1};
> Physical Volume(1)={1};
>
> //n=1; add a crack surface.
> Point(5000)={1.645000e+01,7.466800e+00,5.890600e+00,lc2};
> Point(5001)={1.728000e+01,2.112300e+01,1.204100e+01,lc2};
> Point(5002)={2.823700e+01,1.636900e+01,2.111600e+01,lc2};
> Point(5003)={2.740700e+01,2.713500e+00,1.496600e+01,lc2};
> Line(5004)={5000,5001};
> Line(5005)={5001,5002};
> Line(5006)={5002,5003};
> Line(5007)={5003,5000};
> Line Loop(5008)={5004,5005,5006,5007};
> Plane Surface(5009)={5008};
> Physical Surface(11)={5009};
> Surface{5009} In Volume{1};
>
> Plugin(Crack).Dimension = 2 ;
> Plugin(Crack).PhysicalGroup = 11 ;
> Plugin(Crack).OpenBoundaryPhysicalGroup = 0 ;
> Plugin(Crack).Run ;
> //end of the file
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Repeated ids with Duplicata
Alexander - Some of your translations are by (0,0,0), hence creating a copy of
the surface at the same location as the original. One of the two copies gets
removed automatically (with Geometry.AutoCoherence=1).
Christophe
> On 13 Oct 2016, at 21:23, Alexander Lindsay wrote:
>
> I'm getting repeated IDs using Duplicata. Here's the relevant excerpt from my
> script:
>
> Plane Surface(11) = {9, 10};
> Plane Surface(12) = {10};
> fuel_surfaces[] = {};
> moder_surfaces[] = {};
> For xtrans In {0:108:12}
> For ytrans In {0:108:12}
> new_f_surface = Translate {xtrans, ytrans, 0} {
> Duplicata { Surface{11}; }
> };
> fuel_surfaces += new_f_surface;
> new_m_surface = Translate {xtrans, ytrans, 0} {
> Duplicata { Surface{12}; }
> };
> moder_surfaces += new_m_surface;
> Printf("New f surface '%g'", new_f_surface);
> Printf("New m surface '%g'", new_m_surface);
> EndFor
> EndFor
>
> And here's the beginning of the output:
>
> New f surface '13'
> New m surface '13'
> New f surface '13'
> New m surface '22'
> New f surface '23'
> New m surface '32'
> New f surface '33'
> New m surface '42'
>
> How do I ensure that everything has a unique ID?
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Problems with version 2.14.0 - Part 2
It would help if you would send the actual files, and maybe describe what you
want to do? Do you want to mix .brep and .geo entities in the same model?
> On 12 Oct 2016, at 11:43, Alessandro Vicini
> wrote:
>
>
> This is another problem I have with version 2.14.0:
>
> 1) I open an existing .geo file and it looks fine, see attached geometry.jpg
> picture;
> 2) I interactively delete a few surfaces, obtaining geometry_2.jpg. If I
> look at the geo file, I have these lines at the end:
>
>Delete {
> Surface{122, 124, 126, 120};
>}
>
> 3) The problem is that, when I issue the “delete” command, I get these
> (apparent) error messages in the console:
>
> Error : Unknown surface 515812064
> Error : Unknown surface 515812632
> Error : Unknown surface 515826792
> Error : Unknown surface 515813200
> Error : Unknown surface 515812064
> Error : Unknown surface 515812632
> Error : Unknown surface 515826792
> Error : Unknown surface 515813200
>
> 4) I see the same kind of (apparent) errors also if I reload the .geo file
> (I say apparent because in the graphics window the geometry looks ok).
>
> Where is the problem? Thank you.
>
> Alessandro
>
>
>
>
> _______
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Command line options
> On 12 Oct 2016, at 15:10, Alessandro Vicini > wrote: > > > Is it possible from command line to load a .geo file and save it in geo > unrolled format? Thanks. > gmsh file.geo -0 > Alessandro > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Recherche information sur l'installation de GMSH (Linux & windows)...
Dear Gregory, There are no dependencies: you can simple copy the binary file (gmsh.exe) on the machines. Christophe > On 11 Oct 2016, at 16:44, GACE, Gregory (SOGETI) > wrote: > > Bonjour, > > Je suis en charge de l’installation des applications d’analyse au sein du > bureau d’études d’Airbus Defence & Space. > Je souhaite installer gmsh sous Linux et Windows et souhaite donc connaitre > les prérequis et la méthode d’installation. > Existe-t-il une documentation notamment pour réaliser des installations > scriptées sous Windows ? > En d’autres terme, est-il possible d’installer gmsh sous windows via des > lignes de commandes de sorte à pouvoir les automatiser ? > > Cordialement, > > Grégory Gace | Product Engineering Support Team | Sogeti High Tech > IM – TOIS – Information Management > Phone +33 (0)5 62 19 59 26 > gregory.gace.exter...@airbus.com > > > > > *** > Ce courriel (incluant ses eventuelles pieces jointes) peut contenir des > informations confidentielles et/ou protegees ou dont la diffusion est > restreinte. Si vous avez recu ce courriel par erreur, vous ne devez ni le > copier, ni l'utiliser, ni en divulguer le contenu a quiconque. Merci d'en > avertir immediatement l'expediteur et d'effacer ce courriel de votre systeme. > Airbus Defence and Space et les sociétés Airbus Group declinent toute > responsabilite en cas de corruption par virus, d'alteration ou de > falsification de ce courriel lors de sa transmission par voie electronique. > This email (including any attachments) may contain confidential and/or > privileged information or information otherwise protected from disclosure. If > you are not the intended recipient, please notify the sender immediately, do > not copy this message or any attachments and do not use it for any purpose or > disclose its content to any person, but delete this message and any > attachments from your system. Airbus Defence and Space and Airbus Group > companies disclaim any and all liability if this email transmission was virus > corrupted, altered or falsified. > - > Airbus Defence and Space SAS (393 341 516 RCS Versailles) - Capital: > 16.587.728 EUR - Siege social: 51-61 Route de Verneuil, 78130 Les Mureaux, > France > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Orient gui camera along axis
> On 14 Oct 2016, at 01:17, Alexander Lindsay wrote: > > Is there a way to orient the gui camera along an arbitrary axis, kind of like > with paraview? E.g. if I've rotated the geometry around, but now my clumsy > mouse skills won't allow me to perfectly orient along the z-axis? > Sure: cf. the "X", "Y", "Z" and "rotate 90 degrees" buttons on the bottom of the GUI. > Alex > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Problems with version 2.14.0 - Part 2
> On 14 Oct 2016, at 08:59, Alessandro Vicini
> wrote:
>
>
> Hullo Christophe,
>
> here are the .geo and step files I am using.
> In this case yes, I am starting from a .geo kind of geometry, and I want to
> merge a step file into it.
>
Here's what's happening:
1) Delete { Surface{122, 124, 126, 120}; } : you cannot delete these surfaces
- they are embedded in the volume. The error message is bad - this is now fixed
in SVN.
2) Disappearing entities when you merge the .step file in the .geo file : this
is "normal" in the current version, you need to explicitly call "SyncModel;"
before the "Merge", so that the .geo model gets imported right away in the
model database. Otherwise the .step gets imported with entity tags that can
clash with those used in the .geo mode. This explicit sync step is indeed a bit
annoying - the SVN version now does it automatically.
Christophe
> Regards,
>
> Alessandro
>
>
> -Messaggio originale-
> Da: Christophe Geuzaine [mailto:cgeuza...@ulg.ac.be]
> Inviato: giovedì 13 ottobre 2016 21.58
> A: Alessandro Vicini
> Cc: gmsh@onelab.info
> Oggetto: Re: [Gmsh] Problems with version 2.14.0 - Part 2
>
>
> It would help if you would send the actual files, and maybe describe what you
> want to do? Do you want to mix .brep and .geo entities in the same model?
>
>> On 12 Oct 2016, at 11:43, Alessandro Vicini
>> wrote:
>>
>>
>> This is another problem I have with version 2.14.0:
>>
>> 1) I open an existing .geo file and it looks fine, see attached
>> geometry.jpg picture;
>> 2) I interactively delete a few surfaces, obtaining geometry_2.jpg. If I
>> look at the geo file, I have these lines at the end:
>>
>> Delete {
>> Surface{122, 124, 126, 120};
>> }
>>
>> 3) The problem is that, when I issue the “delete” command, I get these
>> (apparent) error messages in the console:
>>
>> Error : Unknown surface 515812064
>> Error : Unknown surface 515812632
>> Error : Unknown surface 515826792
>> Error : Unknown surface 515813200
>> Error : Unknown surface 515812064
>> Error : Unknown surface 515812632
>> Error : Unknown surface 515826792
>> Error : Unknown surface 515813200
>>
>> 4) I see the same kind of (apparent) errors also if I reload the .geo file
>> (I say apparent because in the graphics window the geometry looks ok).
>>
>> Where is the problem? Thank you.
>>
>> Alessandro
>>
>>
>>
>>
>> ____
>> ___
>> gmsh mailing list
>> gmsh@onelab.info
>> http://onelab.info/mailman/listinfo/gmsh
>
> --
> Prof. Christophe Geuzaine
> University of Liege, Electrical Engineering and Computer Science
> http://www.montefiore.ulg.ac.be/~geuzaine
>
> Free software: http://gmsh.info | http://getdp.info | http://onelab.info
>
>
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Problems with structured meshing
> On 15 Oct 2016, at 19:02, Kushagra Vidyarthi
> wrote:
>
> Dear Sir/Madam,
>
> I am working on a CFD analysis around an ellipse. For this purpose, I have
> created a circular domain around the geometry, and I am trying to create a
> structured mesh around it, so that I can have some control over the boundary
> layer. My .geo file is below:
>
> // Gmsh project created on Fri Oct 14 10:40:40 2016
>
> //geometry creation
> //ellipse: major axis 0.075, minor axis 0.009375
> // 1st arc
> Point(1) = {-0.0375, 0, 0, 0.01};
> Point(2) = {0, 0, 0, 0.01}; //global center
> Point(3) = {-0.02, 0, 0, 0.1};
> Point(4) = {-0.026516504, 0.003314563, 0, 0.01};
> Ellipse(1) = {1, 2, 3, 4};
>
> //arc 2
> Point(5) = {0, 0.0046875, 0, 0.01};
> Point(6) = {0.02, 0, 0, 0.01};
> Ellipse(2) = {4, 2, 3, 5};
>
> //arc3
> Point(7) = {0.026516504, 0.003314563, 0, 0.01};
> Ellipse(3) = {5, 2, 6, 7};
>
> //arc4
> Point(8) = {0.0375, 0, 0, 0.01};
> Ellipse(4) = {7, 2, 6, 8};
>
> //arc5
> Point(9) = {0.026516504, -0.003314563, 0, 0.01};
> Ellipse(5) = {8, 2, 6, 9};
>
> //arc6
> Point(10) = {0, -0.0046875, 0, 0.01};
> Ellipse(6) = {9, 2, 6, 10};
>
> //arc7
> Point(11) = {-0.026516504, -0.003314563, 0, 0.01};
> Ellipse(7) = {10, 2, 3, 11};
>
> //arc8
> Ellipse(8) = {11, 2, 3, 1};
>
> //circle
> Point(12) = {-1.7677669, -1.7677669, 0, 0.01};
> Point(13) = {-1.7677669, 1.7677669, 0, 0.01};
> Point(14) = {1.7677669, 1.7677669, 0, 0.01};
> Point(15) = {1.7677669, -1.7677669, 0, 0.01};
>
> Circle(9) = {12, 2, 13};
> Circle(10) = {13, 2, 14};
> Circle(11) = {14, 2, 15};
> Circle(12) = {15, 2, 12};
>
> //blocking lines
> Line(13) = {11, 12};
> Line(14) = {4, 13};
> Line(15) = {7, 14};
> Line(16) = {9, 15};
>
> //splitting lines
> Transfinite Line {9} = 100 Using Progression 1;
> Transfinite Line {10} = 100 Using Progression 1;
> Transfinite Line {11} = 100 Using Progression 1;
> Transfinite Line {12} = 100 Using Progression 1;
> Transfinite Line {14} = 100 Using Progression 1;
> Transfinite Line {15} = 100 Using Progression 1;
> Transfinite Line {16} = 100 Using Progression 1;
> Transfinite Line {13} = 100 Using Progression 1;
> Transfinite Line {8} = 100 Using Progression 1;
> Transfinite Line {1} = 100 Using Progression 1;
> Transfinite Line {2} = 100 Using Progression 1;
> Transfinite Line {3} = 100 Using Progression 1;
> Transfinite Line {4} = 100 Using Progression 1;
> Transfinite Line {5} = 100 Using Progression 1;
> Transfinite Line {6} = 100 Using Progression 1;
> Transfinite Line {7} = 100 Using Progression 1;
> Line Loop(17) = {13, 9, -14, -1, -8};
> Plane Surface(1) = {17};
> Line Loop(19) = {15, -10, -14, 2, 3};
> Plane Surface(2) = {19};
> Line Loop(21) = {16, -11, -15, 4, 5};
> Plane Surface(3) = {21};
> Line Loop(23) = {13, -12, -16, 6, 7};
> Plane Surface(4) = {23};
> Transfinite Surface {1} = {4, 13, 12, 11};
> Transfinite Surface {2} = {7, 14, 13, 4};
> Transfinite Surface {3} = {9, 15, 14, 7};
> Transfinite Surface {4} = {11, 12, 15, 9};
>
> Recombine Surface {1};
> Recombine Surface {2};
> Recombine Surface {3};
> Recombine Surface {4};
>
> When I run this file in gmsh, I get errors like: "Error : Surface 4 cannot
> be meshed using the transfinite algo". I also get a warning stating: "Warning
> : Cannot apply Blosson: odd number of triangles (15167) in surface 4". My
> error log and .geo files are attached. Please advise on what I can do to fix
> these errors.
>
Transfinite Surface meshes require matching number of vertices on opposite
sides. Add e.g. this to your file:
N = 10;
Transfinite Line {1:8} = N;
Transfinite Line {9:12} = 2*N-1;
Transfinite Line {13:16} = 50 Using Progression 1.2;
> Thanks and Regards,
>
> Kushagra Vidyarthi
>
> (copy).geo>___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] pos high order interpolation error
> On 18 Oct 2016, at 16:21, Олег Рябков wrote: > > Dear mesh users and developers. > I got very strange error with attached pos file. It contains view with just 2 > elements and when i open it with gmsh and choose 'Adapt visualization grid' > check button gmsh crashes. The strange thing is that this 2 elements could be > visualized separatly (i,e, when i delete from view either one of them it > works). > Indeed, this is current limitation of the high-order visualization code: it cannot handle hybrid meshes (i.e. meshes with more than 1 element type). > Gmsh - 2.14.0 x64, Win 7. > > -- > С уважением, Олег > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] [Gmsh-announce] New Gmsh release: 2.14.1
> On 2 Nov 2016, at 09:24, Florian Blachère > wrote: > > On 30/10/16 09:49, Christophe Geuzaine wrote: >> New in 2.14.1: fixed regression in periodic meshes; small bug fixes and code >> cleanups. >> >> Downloads, mailing lists, etc.: http://gmsh.info >> >> >> ___ >> gmsh-announce mailing list >> gmsh-annou...@onelab.info >> http://onelab.info/mailman/listinfo/gmsh-announce > Hello, > > It seems that the 2.14.1 archive (from > http://gmsh.info/src/gmsh-2.14.1-source.tgz) is updated every day as the > nightly 'gmsh-svn-source.tgz' archive. For instance, today the folder > inside the archived is named 'gmsh-2.14.1-svn-20161102-source'. > > Is it an expected behaviour ? Hi Florian - No, it was a bug due a bad symlink. Thanks for letting us know: it should now be fixed. > > Regards, > Florian Blachère > > _______ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Question about gmsh
> On 22 Oct 2016, at 04:09, Mayank Jog wrote: > > Hello, > I have been using "simnibs" software which uses gmsh to calculate electric > fields over a 3D mesh/geometry (human head). > It outputs a final .msh file, that contains the electric field information, > and can be viewed using gmsh. > > My issue is, I want to convert this file to a list where each line = > [coord location, electric field value ] How do I go about doing it? > > The ultimate goal is to regrid this data onto a regular cartesian grid to do > some funky postprocessing (fourier transforms, followed by funky stuff). Here's an example: Merge "tutorial/view3.pos"; Plugin(CutBox).X0=0; Plugin(CutBox).Y0=0; Plugin(CutBox).Z0=0; Plugin(CutBox).X1=1; Plugin(CutBox).Y1=0; Plugin(CutBox).Z1=0; Plugin(CutBox).X2=0; Plugin(CutBox).Y2=1; Plugin(CutBox).Z2=0; Plugin(CutBox).X3=0; Plugin(CutBox).Y3=0; Plugin(CutBox).Z3=1; Plugin(CutBox).NumPointsU=20; Plugin(CutBox).NumPointsV=20; Plugin(CutBox).NumPointsW=20; Plugin(CutBox).ConnectPoints=0; Plugin(CutBox).Boundary=0; Plugin(CutBox).View=0; Plugin(CutBox).Run; Save View[1] "points.txt"; > I tried saving the .msh file as a .txt ** , and then reading it line by > line.I was thinking of replacing each triangular element with its > centroid. However, I don't know how to proceed with other elementsplease > help me out > > Thank you! > Mayank > > ** "Post processing Generic txt" option in "save as" in gmsh > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Problem about 'Plugin Crack'
> On 25 Oct 2016, at 16:12, Xiaoxin Lu wrote:
>
> Dear GMSH users and developers,
>
> I have a question about 'Plugin crack' command.
> In the manual, it is said that 'The elements touching the crack on the
> “negative” side are modified to use the newly generated vertices'.
> I was wondering how to define the 'negative' side for 2D crack. Is it defined
> by NormalX, Normal Y and Normal Z, which are the components of normal vector
> of the crack surface?
No, Normal{X,Y,Z} is only used to explicitly provide for "1D cracks" (curves)
the normal to the surface in which these cracks live. For 2D cracks the normal
is computed as usual, based on the orientation of the triangles/quadrangles.
> Thank you very much.
>
> Best ragards,
> Xiaoxin
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] 2.14.1 tarball change
> On 3 Nov 2016, at 06:09, Joseph Mingrone wrote: > > Hello, > > The FreeBSD package for gmsh is currently failing, because the tarball for > this release has changed. > > fetch: http://gmsh.info/src/gmsh-2.14.1-source.tgz: size mismatch: expected > 10862009, actual 10860112 > *** Error code 1 > > Were there changes other than the renaming of the top level directory? I > don't have an old tarball to do a diff. > The changes were minor (top level directory name change + some one-liners to fix compilation warnings). You should use the tarball currently available on the server as the reference, as the original one was lost. > Thanks, > > Joseph > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Question about gmsh
> On 3 Nov 2016, at 02:01, Mayank Jog wrote: > > Hello Jeremy and Christophe, > Thanks for the help! > > Jeremy: > The data is Current-density magnitude at nodes (attached Fig1.png). I need to > recast this onto a uniformly-spaced 3D cartesian grid (nodes are > non-uniformly spaced in 3D). > > One hack I tried (please correct me :) ), was to export the .msh file as > ascii, then interpolate the data at each element onto the regular cartesian > grid. The coordinates of each element I set as the centroid of the nodes > comprising the element. > I'm new to handling mesh datadoes that sound like a fair approximation? > > Christophe: > I tried what you suggested in the GUI. I didnt understand the first line > (Merge "tutorial/view3.pos";) , but I carried out all the rest steps. I got a > new "view", but it only had a couple of voxels. I'm enclosing Fig1.png which > shows the data, and Fig2.png which shows what I get when I carried out the > step. > You should change the points (X0, Y0, ...) to match your geometry... Interactively, use Tools->Plugins->CutBox. > , > Mayank > > > > > > > > > > > > > > > On Wed, Nov 2, 2016 at 5:24 AM, Christophe Geuzaine > wrote: > > > On 22 Oct 2016, at 04:09, Mayank Jog wrote: > > > > Hello, > > I have been using "simnibs" software which uses gmsh to calculate electric > > fields over a 3D mesh/geometry (human head). > > It outputs a final .msh file, that contains the electric field information, > > and can be viewed using gmsh. > > > > My issue is, I want to convert this file to a list where each line = > > [coord location, electric field value ] How do I go about doing it? > > > > The ultimate goal is to regrid this data onto a regular cartesian grid to > > do some funky postprocessing (fourier transforms, followed by funky stuff). > > Here's an example: > > Merge "tutorial/view3.pos"; > Plugin(CutBox).X0=0; > Plugin(CutBox).Y0=0; > Plugin(CutBox).Z0=0; > Plugin(CutBox).X1=1; > Plugin(CutBox).Y1=0; > Plugin(CutBox).Z1=0; > Plugin(CutBox).X2=0; > Plugin(CutBox).Y2=1; > Plugin(CutBox).Z2=0; > Plugin(CutBox).X3=0; > Plugin(CutBox).Y3=0; > Plugin(CutBox).Z3=1; > Plugin(CutBox).NumPointsU=20; > Plugin(CutBox).NumPointsV=20; > Plugin(CutBox).NumPointsW=20; > Plugin(CutBox).ConnectPoints=0; > Plugin(CutBox).Boundary=0; > Plugin(CutBox).View=0; > Plugin(CutBox).Run; > Save View[1] "points.txt"; > > > > > > I tried saving the .msh file as a .txt ** , and then reading it line by > > line.I was thinking of replacing each triangular element with its > > centroid. However, I don't know how to proceed with other > > elementsplease help me out > > > > Thank you! > > Mayank > > > > ** "Post processing Generic txt" option in "save as" in gmsh > > ___ > > gmsh mailing list > > gmsh@onelab.info > > http://onelab.info/mailman/listinfo/gmsh > > -- > Prof. Christophe Geuzaine > University of Liege, Electrical Engineering and Computer Science > http://www.montefiore.ulg.ac.be/~geuzaine > > Free software: http://gmsh.info | http://getdp.info | http://onelab.info > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Surface In Volume does not work GMSH for Linux
> On 3 Nov 2016, at 14:26, bill wrote: > > Hi, > But it does, strangely enough, for windows! What is wrong? Is this a > bug or do I have a setting wrong? > "Mesh.Algorithm3D = 4;" selects the 3D "Frontal" algorithm, which does not support embedded surfaces. Reset it to the default "Delaunay" (1). > Geometry file and settings dump is attached. > > Cheers, > Bill Slade > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Bug? Points in Volume only meshed for first volume
Dear Gregor, Indeed, this was only supported with our new Delaunay algorithm. I've fixed the old algorithm so that it works there, too. Thanks for the report, Christophe > On 14 Nov 2016, at 13:17, Gregor Mitscha-Baude > wrote: > > Dear all, > > I was happy to see that the new gmsh version (2.14.1) includes the > long-missing "Point In Volume" feature. > However, if multiple volumes are created in the .geo file, with points added > to each volume by "Point In Volume", then only the points lying in the > *first* volume are meshed in the output .msh file. > > I attached a simple example file input.geo with two cubes stacked on top of > each other, divided by the z=0 plane. vol1 is the lower cube (z<0) and vol2 > the upper one (z>0). A series of points of the form (0, 0, z) with z negative > and positive is added with Point in Volume. The output file was obtained by > running > > >> gmsh -3 input.geo -o out.msh > > and is also attached. As can be seen, all the embedded points with positive z > values are missing from the .msh file. Hence, only the points embedded in > vol1 were meshed. I observed this also on more complicated examples with many > different volumes: only the points embedded in volume 1 are meshed. > Is this a bug? Is there any workaround to mesh embedded points in multiple > volumes? > > Cheers, and thanks a lot, > Gregor Mitscha-Baude > _______ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Refine the mesh around the line
> On 4 Nov 2016, at 14:28, Gladiator Gladiator wrote:
>
> Hello Gmsh users,
>
> Would you please some one kindly inform me how I can refine the mesh around a
> specefic line in 2D and 3D meshes?
> For instance I want to refine a Square or Cube with unstructured quad or
> triangular mesh around its diameter (two gmsh files have been attached).
For a basic refinement, just change the mesh size associated with the points:
l=6;
h=6;
w=1;
c=0.5;
Point(1) = {0, 0, 0, c/10};
Point(2) = {l, 0, 0, c/10};
Point(3) = {0, h, 0, c};
Point(4) = {l, h, 0, c};
Line(1) = {1, 2};
Line(2) = {2, 4};
Line(3) = {4, 3};
Line(4) = {3, 1};
Line Loop(5) = {3, 4, 1, 2};
Plane Surface(6) = {5};
Physical Surface(7) = {6};
Recombine Surface {6};
For more advanced options (using "Fields"), see tutorial/t10.geo
> Thanks for your considerations,
> Look forward to your kind response,
>
>
> Best,
>
> Glad
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Big regression in MED format writing
Hello Nicolas,
Indeed! But I'm not sure this has worked in the past...
The correspondance between Gmsh/MED vertices is located in
gmsh/Geo/GModelIO_MED.cpp:
- for MED_HEXA20: map[20] = {0, 3, 2, 1, 4, 7, 6, 5, 11, 8, 16,
10, 19, 9, 18, 17, 15, 12, 14, 13};
this seems to be identical to when it was first implemented...
- for MED_HEXA27: it's actually not implemented ;-)
Could you check the ordering?
Thanks,
Christophe
> On 17 Nov 2016, at 10:06, TARDIEU Nicolas wrote:
>
> Dear Gmsh-team,
>
> I have noticed a big regression in the quadratic hexa writing in MED format.
> This can checked by re-reading with Gmsh a MED file produced by Gmsh! The
> hexa is totally twisted!
> The error appears for the complete and incomplete elements.
>
> Regards,
> Nicolas
>
>
>
> Ce message et toutes les pièces jointes (ci-après le 'Message') sont établis
> à l'intention exclusive des destinataires et les informations qui y figurent
> sont strictement confidentielles. Toute utilisation de ce Message non
> conforme à sa destination, toute diffusion ou toute publication totale ou
> partielle, est interdite sauf autorisation expresse.
>
> Si vous n'êtes pas le destinataire de ce Message, il vous est interdit de le
> copier, de le faire suivre, de le divulguer ou d'en utiliser tout ou partie.
> Si vous avez reçu ce Message par erreur, merci de le supprimer de votre
> système, ainsi que toutes ses copies, et de n'en garder aucune trace sur
> quelque support que ce soit. Nous vous remercions également d'en avertir
> immédiatement l'expéditeur par retour du message.
>
> Il est impossible de garantir que les communications par messagerie
> électronique arrivent en temps utile, sont sécurisées ou dénuées de toute
> erreur ou virus.
>
>
> This message and any attachments (the 'Message') are intended solely for the
> addressees. The information contained in this Message is confidential. Any
> use of information contained in this Message not in accord with its purpose,
> any dissemination or disclosure, either whole or partial, is prohibited
> except formal approval.
>
> If you are not the addressee, you may not copy, forward, disclose or use any
> part of it. If you have received this message in error, please delete it and
> all copies from your system and notify the sender immediately by return
> message.
>
> E-mail communication cannot be guaranteed to be timely secure, error or
> virus-free.
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Big regression in MED format writing
> On 28 Nov 2016, at 17:32, TARDIEU Nicolas wrote:
>
> Hi Christophe,
>
> It seems to me that the right order should be :
> - for MED_HEXA20: map[20] =
> {0,1,2,3,4,5,6,7,8,11,13,9,10,12,14,15,16,18,19,17 };
>
> - for MED_HEXA27: map[27] =
> {0,1,2,3,4,5,6,7,8,11,13,9,10,12,14,15,16,18,19,17,20,21,23,24,22,25,26};
>
Doesn't seem to work... I found this in the MED doc
which (if I understand correctly), would translate into
map[20] = {0, 3, 2, 1, 4, 7, 6, 5, 9, 13, 11, 8, 17, 19, 18, 16, 10, 15, 14,
12};
map[27] = {0, 3, 2, 1, 4, 7, 6, 5, 9, 13, 11, 8, 17, 19, 18, 16, 10, 15, 14,
12, 20, 22, 24, 23, 21, 25, 26};
But this also does not work (I've committed this to SVN) - not sure where I
made the mistake...
Christophe
> I can test a nightly build as soon as you integrate this patch.
>
> Regards,
> Nicolas
>
>
> Nicolas Tardieu
> Ingénieur Chercheur
> Groupe Vibrations des Structures - T63
> EDF - R&D Dpt AMA
> nicolas.tard...@edf.fr
> Tél. : 01 78 19 37 49
>
>
>
>
> -Message d'origine-
> De : cgeuza...@ulg.ac.be [mailto:cgeuza...@ulg.ac.be]
> Envoyé : samedi 26 novembre 2016 18:55
> À : TARDIEU Nicolas
> Cc : g...@geuz.org
> Objet : Re: [Gmsh] Big regression in MED format writing
>
>
> Hello Nicolas,
>
> Indeed! But I'm not sure this has worked in the past...
>
> The correspondance between Gmsh/MED vertices is located in
> gmsh/Geo/GModelIO_MED.cpp:
>
> - for MED_HEXA20: map[20] = {0, 3, 2, 1, 4, 7, 6, 5, 11, 8, 16,
> 10, 19, 9, 18, 17, 15, 12, 14, 13};
> this seems to be identical to when it was first implemented...
>
> - for MED_HEXA27: it's actually not implemented ;-)
>
> Could you check the ordering?
>
> Thanks,
>
> Christophe
>
>> On 17 Nov 2016, at 10:06, TARDIEU Nicolas wrote:
>>
>> Dear Gmsh-team,
>>
>> I have noticed a big regression in the quadratic hexa writing in MED format.
>> This can checked by re-reading with Gmsh a MED file produced by Gmsh! The
>> hexa is totally twisted!
>> The error appears for the complete and incomplete elements.
>>
>> Regards,
>> Nicolas
>>
>>
>>
>> Ce message et toutes les pièces jointes (ci-après le 'Message') sont établis
>> à l'intention exclusive des destinataires et les informations qui y figurent
>> sont strictement confidentielles. Toute utilisation de ce Message non
>> conforme à sa destination, toute diffusion ou toute publication totale ou
>> partielle, est interdite sauf autorisation expresse.
>>
>> Si vous n'êtes pas le destinataire de ce Message, il vous est interdit de le
>> copier, de le faire suivre, de le divulguer ou d'en utiliser tout ou partie.
>> Si vous avez reçu ce Message par erreur, merci de le supprimer de votre
>> système, ainsi que toutes ses copies, et de n'en garder aucune trace sur
>> quelque support que ce soit. Nous vous remercions également d'en avertir
>> immédiatement l'expéditeur par retour du message.
>>
>> Il est impossible de garantir que les communications par messagerie
>> électronique arrivent en temps utile, sont sécurisées ou dénuées de toute
>> erreur ou virus.
>>
>>
>> This message and any attachments (the 'Message') are intended solely for the
>> addressees. The information contained in this Message is confidential. Any
>> use of information contained in this Message not in accord with its purpose,
>> any dissemination or disclosure, either whole or partial, is prohibited
>> except formal approval.
>>
>> If you are not the addressee, you may not copy, forward, disclose or use any
>> part of it. If you have received this message in error, please delete it and
>> all copies from your system and notify the sender immediately by return
>> message.
>>
>> E-mail communication cannot be guaranteed to be timely secure, error or
>> virus-free.
>>
>> ___
>> gmsh mailing list
>> gmsh@onelab.info
>> http://onelab.info/mailman/listinfo/gmsh
>
> --
> Prof. Christophe Geuzaine
> University of Liege, Electrical Engineering and Computer Science
> http://www.montefiore.ulg.ac.be/~geuzaine
>
> Free software: http://gmsh.info | http://getdp.info | http://onelab.info
>
>
>
>
> Ce message et toutes les pièces jointes (ci-après le 'Message') sont établis
> à l'intention exclusive des destinataires et les informations qui y figurent
Re: [Gmsh] substring in gmsh script
> On 25 Nov 2016, at 14:43, Олег Рябков wrote:
>
>Dear mesh users and developers.
> Given some string (in gmsh script), for example, view name
> test = Str( View[0].Name );
> which can be printed using:
> Printf(test);
> I want to get substring of this string, for example first 5 characters of 5
> last ones. But i don't see any built-in functions for this purpose.
Indeed, there were none. I've added them: you can now do e.g.
test = StrSub( View[0].Name , start );
or
test = StrSub( View[0].Name , start , len);
Christophe
> None of expressions work for me as well.
> For example, there is documented syntax
> expression
> :
>
> real
> |
>
> string
> |
>
> string ~ { expression
> }
>
> string [ expression ] |
>
> But test[1] is not working (out of range index). I suppose i don't understand
> some differences between lists, strings, char-expressions and something like
> that.
>
> Is there any way to do what i want?
>
> --
> Best regards, Oleg
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] I need help with an error in gmsh using 10-node tetraeder elements
Dear Michael, Comment-out the transfinite volume command: //Transfinite Volume "*"; Without recombination (i.e. when generating tets), transfinite volume definitions need to be specified by hand (i.e. not with "*") if you want to ensure conformity between surface and volume meshes. Christophe > On 7 Dec 2016, at 12:24, Michael Groß wrote: > > Dear Christophe, > > I need your help with an error in gmsh. > > Using 10-node tetraeder elements for discretizing a cook-like cube (see > attached file), I get two nodes 36 and 37, which should not exist. These > nodes arise not in the volume elements, but in the surface elements, and > therefore in the node vector and the node sets. > > Do you have the same error? I work with the linux versions. > > With kind regards, > > -- > Prof. Michael Groß > Professur Technische Mechanik/Dynamik > Fakultät für Maschinenbau > > Technische Universität Chemnitz > Reichenhainer Straße 70 | R. W239 > 09126 Chemnitz > Germany > > www.tu-chemnitz.de > -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Installation problem in MacOS
> On 17 Dec 2016, at 20:22, 蔣承佑 wrote: > > Dear Developer, > > When I tried to install gmsh in my laptop, it doesn’t work.( Error message is > in attached files ) > > And I have no idea how solve this kind of problem. > By default on Mac we package an .app bundle. Configure with cmake -DENABLE_OS_SPECIFIC_INSTALL=0 to do a standard "unix" "make install". > Thanks! > > > > > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Is C++ API Documentation Available ???
> On 16 Dec 2016, at 15:53, jophin john wrote: > > Hi All, > > Is there any C++ API Documentation available for Gmsh? Not currently. For basic usage, cf. gmsh/utils/api_demos (e.g. mainSimple.cpp) > I wanted to generate mesh using it as a library for my project. > Thanks in advance. > > Regards, > Jophin John > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] problem in characteristic mesh value in gui
Hi Andrew - the latest version fixes this. Thanks, Christophe > On 21 Dec 2016, at 21:23, andrew wrote: > > Hi, > > When I load a file with the line 'Mesh.CharacteristicLengthFactor = 0.001;' > in the gui in Options/Mesh/Element size factor it shows as 0.00. The digits > are truncated and I have to change the value because if I forget it and I > start the meshing it goes on forever. Is there a way to have gmsh to not > truncate small values to 2 digits? > > kind regards > > andrew tsiantis > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Postprocessing high-order fields - strange colors
> On 22 Dec 2016, at 18:52, Grégory LEGRAIN > wrote: > > I have also seen this in the last version of gmsh (even for P1 fields). It > seems linked to the lighting : if you swap the viewpoint from +z to -z, > everything is OK. > Alternatively, you can also disable lighting in the options of the view. > The default lighting option changed in recent builds, from "two-side" to "single side" : beside being faster, it also allows to see when surfaces are not oriented correctly. You can restore "two-side" lighting in Options->Color->Use two-side lighting. Adaptive visualization can indeed change the orientation of the mesh: we will fix this. Christophe > Best regards, > Gregory > > Le 21/12/2016 à 19:09, Martin Vymazal a écrit : >> Hello, >> >> when I open a field stored in gmsh on high-order elements (i.e. at least P2) >> and click on Tools -> Options -> Adapt visualization grid, the colors in the >> plot suddenly become much darker (observed with recent builds of gmsh). >> These colors are then preserved when I save the plot to a file (say png). >> Could you please fix this. Thank you. >> >> Best regards, >> >> Martin >> >> ___ >> gmsh mailing list >> gmsh@onelab.info >> http://onelab.info/mailman/listinfo/gmsh > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] generating a sequence of meshes
Or you could also generate the meshes directly from the .geo file: here's how
to generate both nested or simply refined meshes (run "gmsh file.geo -"):
lc = 0.5;
Point(1) = {0, 0, 0, lc}; Point(2) = {-0.5, -0, 0, lc};
Point(3) = {0, -0.5, 0, lc}; Point(4) = {0, 0.5, 0, lc};
Point(5) = {0.5, -0, 0, lc};
Circle(1) = {2, 1, 3}; Circle(2) = {3, 1, 5};
Circle(3) = {5, 1, 4}; Circle(4) = {4, 1, 2};
Line Loop(5) = {4, 1, 2, 3};
Plane Surface(6) = {5};
Mesh.Algorithm = 6;
// uncomment this to not put points on actual boundary
// Mesh.SecondOrderLinear = 1;
Mesh 2;
Save "initial.msh";
For i In {1:5}
RefineMesh;
Save Sprintf("nested_%g.msh", i);
EndFor
For i In {1:5}
Mesh.CharacteristicLengthFactor = 1 / 2^i;
Mesh 1;
Mesh 2;
Save Sprintf("refined_%g.msh", i);
EndFor
> On 10 Jan 2017, at 19:42, Jeremy Theler wrote:
>
> What about this?
>
> gtheler@tom:~/run$ cat basic.geo.m4
> Point(1) = {0, 0, 0, lc}; Point(2) = {-0.5, -0, 0, lc};
> Point(3) = {0, -0.5, 0, lc}; Point(4) = {0, 0.5, 0, lc};
> Point(5) = {0.5, -0, 0, lc};
>
> Circle(1) = {2, 1, 3}; Circle(2) = {3, 1, 5};
> Circle(3) = {5, 1, 4}; Circle(4) = {4, 1, 2};
>
> Line Loop(5) = {4, 1, 2, 3};
> Plane Surface(6) = {5};
> gtheler@tom:~/run$ for i in `seq 1 10`; do m4 -Dlc=$i/10 basic.geo >
> basic-$i.geo; gmsh -2 basic-$i.geo; done
>
>
> --
> Jeremy Theler
> www.seamplex.com
>
>
>
> On Tue, 2017-01-10 at 15:57 +, Al wrote:
>> Hi,
>>
>>
>> I'd like to have a script that loads the file
>>
>> "mybasidgeo.geo" and generates (and stores)
>>
>> a sequence of meshes via uniform refinement.
>>
>> In Section 3.2 of the manual it is mentioned that something
>>
>> like this should be possible, but in the second line of Section 6.3
>>
>> it is written that "The actual mesh “actions ... cannot be specified
>> in the script files"
>>
>>
>> This is a bit confusing. Shall I maybe use a makefile? How?
>>
>>
>> Many thanks
>>
>> Alberto
>>
>> P.s.:
>>
>>
>>
>> The content of "mybasidgeo.geo" could be something like:
>>
>> Point(1) = {0, 0, 0, 1.0}; Point(2) = {-0.5, -0, 0, 1.0};
>> Point(3) = {0, -0.5, 0, 1.0}; Point(4) = {0, 0.5, 0, 1.0};
>> Point(5) = {0.5, -0, 0, 1.0};
>>
>> Circle(1) = {2, 1, 3}; Circle(2) = {3, 1, 5};
>> Circle(3) = {5, 1, 4}; Circle(4) = {4, 1, 2};
>>
>> Line Loop(5) = {4, 1, 2, 3};
>> Plane Surface(6) = {5};
>> ___
>> gmsh mailing list
>> gmsh@onelab.info
>> http://onelab.info/mailman/listinfo/gmsh
>
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] BRL-CAD for Geometry?
> On 3 Feb 2017, at 21:14, Todd Pierce wrote:
>
> Gorgeous Stuff Christophe,
>
> I may be a bit confused here, but it may not address the (live) command-line
> generation of geometry. I was looking into what McBain was doing with
> FreeCAD, but it looks like even what you're doing McBain, is scripting.
> Correct me if I'm wrong, but this is the same drawback I'm encountering with
> Gmsh. Gmsh won't accept piped input either.
>
> I'm looking for a package that can display graphics as the commands arrive at
> its input pipe. Yes, this is weird, but I'm trying to embed the entire
> multi-phyics suite of tools, including the generation of geometry, into a
> *live* pipeline. This is a natural language processor I am developing, so
> the user (me) will require interactivity with the whole process. Do you guys
> understand what I'm trying to do? Notice that BRL-CAD does exactly this, in
> a fashion quite similar to Gnuplot, where the graphical output is displayed
> as commands arrive at its input pipe,
As scripts can be executed in Gmsh command by command, you can definitely
create a pipe to get an interactive system.
Here's a working example where we create a separate interactive command line
program (like gnuplot) that simply feeds in scripting commands to Gmsh:
https://onelab.info/svn/gmsh/trunk/utils/solvers/c++/interactive.cpp. You can
easily create a similar driver that would use a pipe.
Christophe
> though I have no idea if BRL-CAD can really output the right file type for
> meshing, or even if you experts would advise against BRL-CAD because it's too
> primitive for other reasons.
>
> So, McBain, did I miss your point or can FreeCAD be truly interactive from
> the command line? (or in this case, its pipe)
>
> And Christophe, since you're obviously deep into these types of software, do
> you have any suggestions?
>
> -Todd .
>
> On Fri, Feb 3, 2017 at 6:49 AM, Christophe Geuzaine
> wrote:
>
> As a little teaser, just to let you know that I'm coding a direct interface
> to OpenCASCADE CAD creation right in .geo files. This will in particular
> support automatic creation of conform interfaces when you fuse solids, in
> order to handle internal boundaries.
>
> Here's what the scripts will look like (not everything is in the SVN trunk
> yet):
>
> SetFactory("OpenCASCADE");
>
> Mesh.Algorithm = 6;
> Mesh.CharacteristicLengthMin = 0.4;
> Mesh.CharacteristicLengthMax = 0.4;
>
> R = 1.4;
> s = .7;
> t = 1.25;
>
> Block(1) = {-R,-R,-R, R,R,R};
>
> Sphere(2) = {0,0,0,R*t};
>
> BooleanIntersection(3) = { Volume{1}; Delete; }{ Volume{2}; Delete; };
>
> Cylinder(4) = {-2*R,0,0, 2*R,0,0, R*s};
> Cylinder(5) = {0,-2*R,0, 0,2*R,0, R*s};
> Cylinder(6) = {0,0,-2*R, 0,0,2*R, R*s};
>
> BooleanUnion(7) = { Volume{4}; Delete; }{ Volume{5}; Delete; };
> BooleanUnion(8) = { Volume{6}; Delete; }{ Volume{7}; Delete; };
> BooleanSubtraction(9) = { Volume{3}; Delete; }{ Volume{8}; Delete; };
>
>
>
>
>
>> On 2 Feb 2017, at 23:28, G. D. McBain wrote:
>>
>>
>>
>> Sent from ProtonMail, Swiss-based encrypted email.
>>
>>
>>> Original Message
>>> Subject: [Gmsh] BRL-CAD for Geometry?
>>> Local Time: 2 février 2017 4:07 PM
>>> UTC Time: 2 février 2017 05:07
>>> From: toddcpie...@gmail.com
>>> To: gmsh@onelab.info
>>>
>>> Hi All,
>>> It's me again, wondering how I should go about integrating a command-line
>>> geometry package into OneLab (or some other chain of multi-physics tools).
>>> As I've mentioned before, although there is no shortage of tools to do
>>> meshing, finite element analysis or CFD and so forth, there is no simple
>>> way to build geometries from the command line.
>>>
>>> Has anybody ever used this BRL-CAD/Mged command line CAD package? It
>>> doesn't look bad, but I do have concerns about it exporting to formats that
>>> these meshing and physics programs use.
>>
>> In the past I have used BRL-CAD's mged command-line tool to create
>> geometries for subsequent CFD simulations. I used BRL-CAD's g-stl
>> command-line tool to export a triangulation of the bounding surface in STL
>> format. Gmsh can read this in.
>>
>> More recently though I'm using FreeCAD which can be scripted in Python and
>> so run from the command-line. Besides STL, this also outputs IGES, STEP,
>> and BRep, which can contain more information that STL. Gmsh reads all four
>> formats.
&g
Re: [Gmsh] interactive.cpp
> On 5 Feb 2017, at 00:26, Todd Pierce wrote: > > Hi All, > > I have decided to take a stab at creating an off-board command line interface > for gmsh. Twice I have been told that using the interactive.cpp program as a > template would be a good place to start. > > So, to start with, I thought I would at least compile and get the > interactive.cpp program to work. Compiling went fine but getting the program > to connect to gmsh is still a mystery. > > I start gmsh with the -listen argument, but what socket does the interactive > program take as an argument? It depends on how Gmsh is configured: look at the message console after launching Gmsh with -listen, it will tell you something like this: Info: Listening on socket '/Users/geuzaine/.gmshsock2' > Is that in gmsh's setup somewhere? > > If I get them to talk, how do I test them? > > Yes, I'm clueless, but I won't get any smarter if I don't pursue this stuff, > so help is genuinely appreciated. > > -Todd > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Geometry modification on physical entities
> On 3 Feb 2017, at 10:18, Alessandro Vicini
> wrote:
>
>
> Geometry modifications (such as a rotation) can only be applied to elementary
> entities? Is there a way to apply them on physical entities instead? Thank
> you.
No, but you can get all the elementary entities making up a physical group
using e.g. "Physical Line {tag}" (it returns a list), which you can then use in
the transformation commands.
>
> Alessandro
>
>
>
>
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] gaps
> On 26 Jan 2017, at 12:52, Kik, S.J. wrote: > > Dear, > > I have created a sphere in gmsh for a simulation in FastBEM. Fastbem however > gives some results which seem to indicate that my sphere contains a gap. > After a lot experiments I now detected that the number of nodes in the > statistics window of Gmsh differs from the number in the .msh file(attached). The number of nodes in the final .msh file depends on the Physical group definitions. (What is reported in the Statistics window is the raw number of nodes.) > Could you explain the difference and could you tell whether the sphere has > indeed a gap? > I don't see any gaps in your mesh; beware that the surfaces of the interior sphere are not oriented in a coherent way, which could lead to problems in a BEM code. (The mesh of the outer shell is also quite strange.) Christophe > Kind regards, > Sander Kik > Eindhoven University of Technology > Graduate student Building Physics and Services > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] GRegionCompound not available in python interface
> On 27 Jan 2017, at 10:36, Florian Bruckner wrote: > > the attached patch solved my problem. is it possible to add the missing > wrappers in one of the next releases? ok, patched. > greetings > Florian Bruckner > > On Thu, Jan 26, 2017 at 3:12 PM, Florian Bruckner wrote: > hello, > > i am looking for a way to create a GRegionCompound via the gmshpy interface. > It seems that #include "GRegionCompound.h" is missing in the gmshGeo.i file > for some reason. > > Is there another way to create a RegionCompound? > > greetings > Florian Bruckner > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Mesh and refine from command line
> On 25 Jan 2017, at 19:46, Jérémy Béjanin wrote: > > Hello, I would like to mesh a model, refine that mesh and save it from the > command line, in one shot, that is without having to call gmsh twice (it > seems that doing `gmsh file.geo -2 -refine -o output.msh` does not work...). > > I have found most of the options for this (eg using "Mesh 2;" followed by > "RefineMesh;" in the geo file). > > I have now two questions: > 1. How do I save the mesh using the same filename as the geo file? Is there a > way to retrieve the filename from within the geo file? I have seen the > "General.FileName" option, but typing "Save General.FileName;" at the end of > the geo file does not work. > You could do e.g. Mesh 2; Save StrCat(StrPrefix(General.FileName), "_original.msh"); RefineMesh; Save StrCat(StrPrefix(General.FileName), "_refined.msh"); > 2. I would like to run this from the command line, without the GUI appearing. > Since the mesh command is already in the geo file, I cannot just call `gmsh > -2` since that will perform the meshing twice... > Just run "gmsh file.geo -" (without "2") > Thanks, > Jeremy > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
[Gmsh] Beta-testing new Gmsh CSG modeler and boolean operations
Dear all,
An initial version of the new Gmsh geometrical modeler based on OpenCASCADE is
available on SVN and in recent nightly builds: one can now directly create
OpenCASCADE entities in the .geo files when specifying
SetFactory("OpenCASCADE");
It's still very much in beta, but now is the right time to start giving it a
try and to give me feedback.
There is no documentation yet: just explore the examples in
https://onelab.info/svn/gmsh/trunk/demos/boolean (login and passwd: gmsh)
Some caveats with the current version:
- In order to define meshing constraints or to define physical groups you need
first to "synchronise the model", using "SyncModel;"
- Not all existing primitives from the native Gmsh modeler have been ported yet
to OpenCASCADE
- The GUI cannot yet be used to create the entities interactively
Christophe
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] GMSH compilation problems with OCC7
Try "cmake -DCMAKE_PREFIX_PATH=/opt/opencascade7" > On 14 Feb 2017, at 09:39, isaac.ayala...@gmail.com wrote: > > Good evening, > > I'm experiencing difficulties compiling GMSH with OCC7. > > I have already tried the instrucions indicated in the README file about - > DCMAKE, but have had no success. > > I am currently running Arch Linux 64 bit. > > OpenCasCade is installed in /opt/opencascade/ > > I tried setting OCC_LIBS = /opt/opencascade but I did not have any luck. > > What are the exact CMAKE flags that I should use? Are they meant to link to a > specific file or just a folder? > > The opencascade folder hierarchy is this: > > /opt/opencascade7 > . > .. > bin > include > lib > share > > CMAKE outputs that it is unable to find any of the OCC libs such as TKBREP, > TKSTEP, etcetera. They are all available under /opt/opencascade7/lib > > Any input is greatly appreciated. > > -- > Isaac Ayala > ITESM Toluca > Robotics Team Coordinator (2017) > PACE RSMS Team 8 REVO - Global Student Team Leader (2015 - 2016) > > _______ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Possibility to merge matching meshes
> On 22 Feb 2017, at 18:18, vincent doyeux wrote: > > Hi, > > I was wondering if it is possible to merge together several msh files having > exactly matching surfaces. > It would be a kind of "Coherence" command but on the mesh, not the geometry. > I know it would require a complete renumbering since vertices should be > deleted. > I was wondering if such an algorithm was already implemented in gmsh. > Yes: "Coherence Mesh;" Be careful that the input numbering should not clash. > Thanks, > > Vincent > > > Post-Doctorant à l'Institut de Mécanique des Fluides de Toulouse > 2 Allée du Professeur Camille Soula 31400 TOULOUSE > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] periodic conformal surface mesh with OpenCascade factory
> On 21 Feb 2017, at 17:14, Benedikt Oswald wrote: > > Dear gmsh > > recently the OpenCascade solid modeller was made available in the > scriptin language. > We are very interested in periodic surface meshes. Is there an example > on how to > write a script, using the OCC factory for periodic surface meshes ? > with the latest nightly, cf. demos/boolean/periodic.geo > Thanks a lot & greetings, > Benedikt > > > > -- > Dr. sc. techn. Benedikt Oswald - first engineer - LSPR AG - phone +41 43 366 > 90 74 > Grubenstrasse 9, CH-8045 Zürich, benedikt.oswald@lspr.swiss - www.lspr.swiss > -- > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Unable to create 3D mesh
See attached corrected file and comments therein. Basically:
1) embedded lines/surfaces are not a substitute for correct topology! they are
only provided to embed 0D/1D/2D meshes in 1D/2D/3D ones (not e.g. 1D in 1D, or
2D in 2D, or 3D in 3D)
2) to do the intersections automatically you will want to use the new
SetFactory("OpenCASCADE") mode in recent nightly builds, and the
"BooleanFragments" operation. See gmsh/demos/boolean for examples (there is no
official documentation yet).
fixed.geo
Description: Binary data
> On 28 Feb 2017, at 18:54, adhitomo sulistyo wrote:
>
> Hi, I am trying to create a reservoir model with unstructured grids. The
> reservoir is box shaped, with two horizontal wells on one side, and one
> vertical well on the other side. The model is basically a box with three
> cylinders inside.
>
> Initially I succeeded in creating the 3D mesh. However, later I found that
> the mesh from the cylinder and the box were different and intersecting with
> each other. Based on the information from the web, I need to embed the
> surfaces inside the volume.
>
> After successfully embedding the surfaces, the 2D mesh looks good. There are
> no longer intersecting grids However, when I try to run the 3D mesh, it
> fails. I get the following error message:
>
> Info: Done meshing 2D (0.715054 s)
> Info: Meshing 3D...
> Info: Delaunay Meshing 4 volumes with 4 connected components
> Info: Meshing volume 227 (Delaunay)
> Info: 7174 points 36783 edges and 14324 faces in the initial mesh
> Info: Found void region
> Info: Found void region
> Info: Found void region
> Info: Found void region
> Error : No tetrahedra in region 227 0
> Info: 3D point insertion terminated (7204 points created):
> Info: - 0 Delaunay cavities modified for star shapeness
> Info: - 0 points could not be inserted
> Info: - 0 tetrahedra created in 0.010615 sec. (0 tets/sec.)
> Info: Meshing volume 228 (Delaunay)
> Info: 155 points 592 edges and 280 faces in the initial mesh
> Info: Found region 228
> Info: 0 points created - Worst tet radius is 1.10375 (PTS removed 0 0)
> Info: 3D point insertion terminated (7185 points created):
> Info: - 0 Delaunay cavities modified for star shapeness
> Info: - 0 points could not be inserted
> Info: - 308 tetrahedra created in 0.004804 sec. (64113 tets/sec.)
> Info: Meshing volume 229 (Delaunay)
>
> I have also attached the gmsh file.
>
> Thank you for your help.
>
> Kind regards,
> Adhitomo
>
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] bug export 2nd order mesh
Dear Pierre - Which version do you use? Nicolas Tardieu provided a patch in Gmsh 2.16 which should fix this. > On 26 Feb 2017, at 09:06, lemerle pierre wrote: > > Hello > I'd like to report a bug. > When I transform a 3D mesh into an 2nd order 3D mesh (set order 2 ; > incomplete elements options selected) the numbering of nodes is correct > saving with .unv format > but wrong saving with med format (the elements are inverted). > see http://code-aster.org/forum2/viewtopic.php?id=21490 for illustration. > Yours > Pierre Lemerle. > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Unable to create 3D mesh
> On 28 Feb 2017, at 22:33, Christophe Geuzaine wrote:
>
>
> See attached corrected file and comments therein. Basically:
>
> 1) embedded lines/surfaces are not a substitute for correct topology! they
> are only provided to embed 0D/1D/2D meshes in 1D/2D/3D ones (not e.g. 1D in
> 1D, or 2D in 2D, or 3D in 3D)
>
> 2) to do the intersections automatically you will want to use the new
> SetFactory("OpenCASCADE") mode in recent nightly builds, and the
> "BooleanFragments" operation. See gmsh/demos/boolean for examples (there is
> no official documentation yet).
PS : here's how your geometry could be defined using the new tools:
SetFactory("OpenCASCADE");
Mesh.CharacteristicLengthMin = 2;
Mesh.CharacteristicLengthMax = 2;
Block(1) = {0,0,0, 64,60,40};
Cylinder(2) = {2,0,3, 2,60,3, 1};
Cylinder(3) = {2,0,18, 2,60,18, 1};
Cylinder(4) = {62,32,40, 62,32,0, 1};
v() = BooleanFragments{ Volume{1}; Delete; }{ Volume{2,3,4}; Delete; };
>
>
>
>
>> On 28 Feb 2017, at 18:54, adhitomo sulistyo wrote:
>>
>> Hi, I am trying to create a reservoir model with unstructured grids. The
>> reservoir is box shaped, with two horizontal wells on one side, and one
>> vertical well on the other side. The model is basically a box with three
>> cylinders inside.
>>
>> Initially I succeeded in creating the 3D mesh. However, later I found that
>> the mesh from the cylinder and the box were different and intersecting with
>> each other. Based on the information from the web, I need to embed the
>> surfaces inside the volume.
>>
>> After successfully embedding the surfaces, the 2D mesh looks good. There are
>> no longer intersecting grids However, when I try to run the 3D mesh, it
>> fails. I get the following error message:
>>
>> Info: Done meshing 2D (0.715054 s)
>> Info: Meshing 3D...
>> Info: Delaunay Meshing 4 volumes with 4 connected components
>> Info: Meshing volume 227 (Delaunay)
>> Info: 7174 points 36783 edges and 14324 faces in the initial mesh
>> Info: Found void region
>> Info: Found void region
>> Info: Found void region
>> Info: Found void region
>> Error : No tetrahedra in region 227 0
>> Info: 3D point insertion terminated (7204 points created):
>> Info: - 0 Delaunay cavities modified for star shapeness
>> Info: - 0 points could not be inserted
>> Info: - 0 tetrahedra created in 0.010615 sec. (0 tets/sec.)
>> Info: Meshing volume 228 (Delaunay)
>> Info: 155 points 592 edges and 280 faces in the initial mesh
>> Info: Found region 228
>> Info: 0 points created - Worst tet radius is 1.10375 (PTS removed 0 0)
>> Info: 3D point insertion terminated (7185 points created):
>> Info: - 0 Delaunay cavities modified for star shapeness
>> Info: - 0 points could not be inserted
>> Info: - 308 tetrahedra created in 0.004804 sec. (64113 tets/sec.)
>> Info: Meshing volume 229 (Delaunay)
>>
>> I have also attached the gmsh file.
>>
>> Thank you for your help.
>>
>> Kind regards,
>> Adhitomo
>>
>>
>> ___
>> gmsh mailing list
>> gmsh@onelab.info
>> http://onelab.info/mailman/listinfo/gmsh
>
> --
> Prof. Christophe Geuzaine
> University of Liege, Electrical Engineering and Computer Science
> http://www.montefiore.ulg.ac.be/~geuzaine
>
> Free software: http://gmsh.info | http://getdp.info | http://onelab.info
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] how ensure contiguous meshing ?
> On 1 Mar 2017, at 12:54, Benedikt Oswald wrote: > > Dear gmsh > > I have the problem that I need to create a volume mesh of two blocks > whose surface touch each other > > partially, i.e. the intersection of the surfaces of these two blocks is > not empty. > > > Now, using the Opencascade factory, what is the right way to ensure that > the tetrahedral mesh > > is generated so that tetrahedra on opposite sides of the intersecting > surface match ? > BooleanFragments : cf. demos/boolean/compsolid.geo > A short hint or example is highly appreciated. > > Greetings & thanks, Benedikt > > > > -- > -- > Dr. sc. techn. Benedikt Oswald - first engineer - LSPR AG - phone +41 43 366 > 90 74 > Grubenstrasse 9, CH-8045 Zürich, benedikt.oswald@lspr.swiss > -- > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Ploting graphs in POST
> On 1 Mar 2017, at 12:43, Rami Ben-Zvi wrote: > > Dear Profs. Remacle and Geuzaine, Dear GMSH group, > > It’s been quite a long time since I asked the query below. Actually, I have > already sent a similar one to gmsh on 19/10/10, and did get > a response from Geordie McBain gdmcb...@freeshell.org, but it was unclear to > me. > > Would you please let me know if it is possible to draw graphs along a line of > nodes, and if so – how? > You could use one of the "Cut" plugins (e.g. Plugin(CutGrid)) to extract the data you want on points along a line, then plot this as a 2D graph, time-dependent or not... > Regards, > Rami > ___ > Rami Ben-Zvi > Earth and Planetary Sciences Dept. Tel. +972-8-9343397 > Weizmann Institute of ScienceFax+972-8-9344124 > Rehovot 76100, ISRAEL Mobile +972-54-6279767 > e-mail: rami.ben-...@weizmann.ac.il > ___ > > From: Rami Ben-Zvi > Sent: Tuesday, February 07, 2017 12:13 > To: 'jean-francois.rema...@uclouvain.be' > Subject: RE: Ploting graphs in POST > > Dear Prof. Remacle, > > I am writing to you because my mail to GMSH (see below) remains unanswered. > Would you please answer my query? > Could it be that I have written to a wrong address? > > Best regards, > Rami > ___ > Rami Ben-Zvi > Earth and Planetary Sciences Dept. Tel. +972-8-9343397 > Weizmann Institute of ScienceFax+972-8-9344124 > Rehovot 76100, ISRAEL Mobile +972-54-6279767 > e-mail: rami.ben-...@weizmann.ac.il > ___ > > From: Rami Ben-Zvi > Sent: Thursday, January 26, 2017 11:12 > To: 'gmsh@onelab.info' > Subject: Ploting graphs in POST > > Hello, > > I am using GMSH (v2.11.0) to plot results from my own developed 2D FEM code. > Usually I plot 2D results as contour plots. However, I wish also to show > results as graph. For example, I have a result field C(x, y, z=0, t) written > to post.msh and wish to plot C(x, y0, z0=0, t0) vs x or plot C(x0, y0, z0=0, > t) vs t. Is it possible (either with my version or the current one)? > I am using Win64 machine. > > Thank you, > Rami > ___ > Rami Ben-Zvi > Earth and Planetary Sciences Dept. Tel. +972-8-9343397 > Weizmann Institute of Science Fax+972-8-9344124 > Rehovot 76100, ISRAEL Mobile +972-54-6279767 > e-mail: rami.ben-...@weizmann.ac.il > ___ -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] How to output the master and slave node connection information when the periodic surfaces are generated by using the option of extrude + layers?
> On 24 Feb 2017, at 01:27, Junhui Liu wrote:
>
> I use Extrude command with layers to generate boundary-layer unstructured
> mesh. The recombine command is not used, so the mesh remains as unstructured.
> It appears that if the periodic surfaces are generated by using extrude
> command with layers specified, the msh (tetral mesh) data will not have the
> information of the slave and master nodes. I need to export the mesh data to
> a existing CFD software, and need to know the node connection between the
> Master and Slave periodic surfaces. The existing CFD software does not care
> if the mesh appears to be structured or unstructured. It only cares if the
> mesh is tetrahedron mesh. The following attached lines generate a box with
> structured tetrahedron mesh. The mesh data does give the Slave and Master
> surface entity numbers, but it does not give the nodes associated with them.
> Is there a way I can output the slave and master node information when
> periodic surfaces are generated by using extrude command with layers
> specified?
Hi Junhui - Not currently, but it's on the wishlist of some of our users.
>
> Thanks a lot for your help.
>
> Junhui
>
>
>
>>>>>>>>>>>>> box.geo >>>>>>>>>>>>>>>>
>
> lc = 0.1;
> Point(1) = {0.0, 0.0, 0.0, lc};
> line_extrude[] = Extrude {1.0, 0.0, 0.0} {Point{1}; Layers{10};};
>
> surface_extrude[] = Extrude {0.0, 1.0, 0.0} {Line{line_extrude[1]};
> Layers{10};};
>
> Volume_extrude[] = Extrude {0.0, 0.0, 1.0} {Surface{surface_extrude[1]};
> Layers{10};};
>
> surfMaster = surface_extrude[1];
> surfSlave = Volume_extrude[0];
>
> boundMaster[] = Boundary{Surface{surfMaster};};
> boundSlave[] = Boundary{Surface{surfSlave};};
> Periodic Surface surfSlave { boundSlave[] } = surfMaster { boundMaster[] };
>
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] surfaces from volume duplication
> On 2 Mar 2017, at 19:27, Miro Kuchta wrote:
>
> Hi,
>
> is there a way to get a list of surfaces which were created as part of
> duplicating a volume.
> In the attached code Out[0] gives me a volume (index?) of the new volume but
> I would like
> to get the surfaces which show up when the geometry is viewed in GUI (28, 33,
> 38, 43, 48, 53).
> Out[1] is not initialized.
>
Use the Boundary operator, which returns a list:
b() = Boundary{ Volume{Out[0]}; } ;
Printf("%g", Out[0]);
Printf("boundary = ", Boundary{ Volume{Out[0]};} );
> Thanks for your help, Miro
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Lloyd smoothing broken?
Hi Nico,
It's not a bug, it's a feature (for quad mesh generation). See
http://link.springer.com/chapter/10.1007%2F978-3-642-24734-7_26 for some
context.
Christophe
> On 8 Mar 2017, at 09:46, Nico Schlömer wrote:
>
> Hi everyone,
>
> When using `-optimize_lloyd`, I noticed that gmsh doesn't produce centroidal
> tesselations but for some reason favors right angles. For the simple circle
> geometry
> ```
> // Points
> p1 = newp;
> Point(p1) = {0, 0, 0, 0.1};
> p2 = newp;
> Point(p2) = {1, 0, 0, 0.1};
> p3 = newp;
> Point(p3) = {0, 1, 0, 0.1};
> p4 = newp;
> Point(p4) = {-1, 0, 0, 0.1};
> p5 = newp;
> Point(p5) = {0, -1, 0, 0.1};
> // Circle arcs
> c1 = newl;
> Circle(c1) = {p2, p1, p3};
> c2 = newl;
> Circle(c2) = {p3, p1, p4};
> c3 = newl;
> Circle(c3) = {p4, p1, p5};
> c4 = newl;
> Circle(c4) = {p5, p1, p2};
> l1 = newl;
> Compound Line(l1) = {c1,c2,c3,c4};
> ll1 = newll;
> Line Loop(ll1) = {l1};
> surf1 = news;
> Plane Surface(surf1) = {ll1};
> ```
> meshed with
> ```
> gmsh -2 circle.geo -o circle.msh -optimize_lloyd 1000
> ```
> I'm getting [1] when it should really look like [2].
>
> Any hints on what might be going wrong?
>
> Cheers,
> Nico
>
> [1] http://chunk.io/f/edd6327ec586485b9c8fe7a96ba1df9a
> [2] http://chunk.io/f/5ba441e4700a4fad8be754c4f19dd9c6
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] difference between gmsh svn nightly build
> On 13 Mar 2017, at 12:11, Benedikt Oswald wrote:
>
> Dear gmsh developers,
>
> there seems to be quite a different behavior with respect to certain
> .geo files between
>
> the svn nightly builds of 27.2.2017 and 13.3.2017. I have attached the
> .geo files.
>
> Which version is more relevant ?
The latest :-)
Boundary now returns signed entities tags (which provides orientation
information): to fix your script, just use Abs(Boundary{...})
>
> Thanks & greetings, Benedikt
>
>
>
> --
> Dr. sc. techn. Benedikt Oswald - first engineer - LSPR AG - phone +41 43 366
> 90 74
> Grubenstrasse 9, CH-8045 Zürich, benedikt.oswald@lspr.swiss
> --
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] difference between gmsh svn nightly build
> On 13 Mar 2017, at 15:35, Benedikt Oswald wrote:
>
> thanks very much!
>
> How can I specify tetrahedral meshing density
>
> using the OCC factory ?
Same as with .geo files:
- globally
- and/or at geometry vertices (with Point() or using Characteristic Length{})
- and/or using Fields
PS: mesh extrusion with OCC is now implemented
>
> Greetings Benedikt
>
>
>
> Am 13.03.17 um 14:00 schrieb Christophe Geuzaine:
>>> On 13 Mar 2017, at 12:11, Benedikt Oswald
>>> wrote:
>>>
>>> Dear gmsh developers,
>>>
>>> there seems to be quite a different behavior with respect to certain
>>> .geo files between
>>>
>>> the svn nightly builds of 27.2.2017 and 13.3.2017. I have attached the
>>> .geo files.
>>>
>>> Which version is more relevant ?
>> The latest :-)
>>
>> Boundary now returns signed entities tags (which provides orientation
>> information): to fix your script, just use Abs(Boundary{...})
>>
>>
>>> Thanks & greetings, Benedikt
>>>
>>>
>>>
>>> --
>>> Dr. sc. techn. Benedikt Oswald - first engineer - LSPR AG - phone +41 43
>>> 366 90 74
>>> Grubenstrasse 9, CH-8045 Zürich, benedikt.oswald@lspr.swiss
>>> --
>>>
>>> ___
>>> gmsh mailing list
>>> gmsh@onelab.info
>>> http://onelab.info/mailman/listinfo/gmsh
>
> --
> ------
> Dr. sc. techn. Benedikt Oswald - first engineer - LSPR AG - phone +41 43 366
> 90 74
> Grubenstrasse 9, CH-8045 Zürich, benedikt.oswald@lspr.swiss
> --
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Python wrapper in gmsh
Dear Andreas, I don't recommend using the Python wrappers for general use: they are really targeted towards Gmsh developers - so (almost) every internal Gmsh class is in the bindings, and nothing is documented/supported. In the near future we will start creating a much smaller API geared towards (advanced) users, which could be wrapped in any language. This will be documented and supported, and will mostly consist in simple functions taking integers, floating point numbers and vectors thereof. For a sneak peak, you can have a look at our new internal CAD creation routines (in Geo/GModelIO_GEO.h and Geo/GModelIO_OCC.h): we plan to wrap only some of the public functions (this is already what is now used internally in the .geo file parser). We plan to do something similar for all meshing features. Christophe > On 15 Mar 2017, at 21:21, andreas.schus...@dlr.de wrote: > > Hello, > > I’ve tried to build Gmsh from the source code with Cmake and the Microsoft > Visual C++ Build Tools 2015 to test the python wrapper functionality. > The compiling process works without problems but when generating the python > wrapper code and linking, the compiler claims: > Link error 1189: library limit of 65535 objects exceeded > To solve this general error, I could only find some advisors where separate > sub libraries should be created to separate the objects to be linked or > manually select symbols via .def files. > Both approaches seem to be very complex for non gmsh-developers. > I also tried to disable some unused external libraries from the compiling > process in Cmake to reduce the overall size of the gmsh library but this > doesn’t work either. > Are there any recommendations what to do in that special case to build gmsh > with less compiling complexity and time? > > Thanks > > —— > M.Sc. Andreas Schuster > > German Aerospace Center > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Eliminating geometric duplications in imported geometry
> On 17 Mar 2017, at 23:24, Bram Sterling
> wrote:
>
> I'm trying to mesh a 3D 2-material construct created in another program
> (FreeCAD) and brought into Gmsh by way of a BREP file. After some
> troubleshooting downstream, I determined that Gmsh is turning the two volumes
> into two disjoint meshes despite their extensive shared surface.
>
> Poking around, I suspect the reason for this is that the two volumes share no
> elementary geometry. Instead, the shared lines and surfaces are duplicated.
>
> If the original geometry had been defined in a Geo script this would seem
> straightforward to fix manually, but it seems a bit more complicated here,
> especially as this is bound to come up again with more complex geometry
> later. Is deleting the redundant geometry and rebuilding other effected
> parts manually the best option, or is there a better way?
>
With a recent nightly build, create a .geo script that does the following
(assuming the brep contains 2 volumes):
SetFactory("OpenCASCADE");
a() = ShapeFromFile("file.brep");
BooleanFragments{ Volume{a(0)}; Delete; }{ Volume{a(1)}; Delete; }
Note that you can now directly create your CAD inside Gmsh: see
demos/boolean/*.geo for examples.
Christophe
>
> Thank you,
> Bram Sterling
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Creating a spherical hole inside a volume [Files attached for testing]
Gmsh does not support intersecting STL (mesh-based) and CAD entities; at least not yet. So if you need intersecting a sphere with your STL, you'll need to use something external to intersect (a mesh of the sphere) with the STL first, before importing in Gmsh. The new OpenCASCADE stuff will not help. Christophe > On 18 Mar 2017, at 02:34, Prithivirajan V wrote: > > Thanks Guillaume ! > > I tried using the OpenCascade factory but I am getting an error, Possible due > to the way I am importing stl files. Is there a specific way of importing > them when I use > OpenCascade ? > > I am getting the following errors (GEO file is attached) : > > Info: Done reading 'Grain_Feature_1.stl' > Error : Unknown OpenCASCADE face with tag 1 > Error : Unknown OpenCASCADE surface loop with tag 1 > > Info: Done reading 'Grain_Feature_2.stl' > Error : Unknown OpenCASCADE face with tag 2 > Error : Unknown OpenCASCADE surface loop with tag 2 > Error : Unknown OpenCASCADE entity of dimension 3 with tag 1 > > > Could you or anyone help me fix this ?? > > On Fri, Mar 17, 2017 at 6:13 AM, DILASSER Guillaume > wrote: > Hello, > > > > Unfortunately, Gmsh is fairly limited when it comes to operations like > computing intersections. However, you might be able to do something using the > OpenCascade factory included in the latest nigthly build. You can download it > from gsmh.info, under the download paragraph, just under the “Development > version” mention. For example, you might want to have a look at the > Boolean.geo exemple here and there few examples on how to import geometry > from files (they will only work with the nightly, ID & password = gmsh). > Hoping that will be helpful, > > > > Faithfully Yours, > > > > Guillaume DILASSER > > Doctorant SACM / LEAS > > CEA - Centre de Saclay - Bât.123 - PC 319c > > 91191 Gif sur Yvette Cedex - France - > > > > guillaume.dilas...@cea.fr > > > > De : gmsh [mailto:gmsh-boun...@ace20.montefiore.ulg.ac.be] De la part de > Prithivirajan V > Envoyé : vendredi 17 mars 2017 03:23 > À : g...@geuz.org > Objet : [Gmsh] Creating a spherical hole inside a volume [Files attached for > testing] > > > > Hello all, > > My objective is to create the spherical hole inside a volume. I have 2 STL > files(see attached) as inputs. The complication is that the spherical hole > intersects the two other existing volumes and it is not contained entirely in > either one of the volumes. > > > > I tried two approaches both of them were a failure (geo files are attached > for both) : > > Approach 1 : > > > > I used the merge command and then I created a spherical volume and performed > meshing. The issue is the sphere is getting meshed separately without any > relation to the other 2 volumes. > > > Approach 2 : > > First , I created an STL file of the spherical void separately. Now I used > merge command on all 3 STL files. Now i am getting intersections which I am > unable to resolve. > > > I have been stuck here for a while in this, your help would be greatly > appreciated > > > > Thanks, > > Prithivi > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] getdp + gmsh High order pos files
> On 17 Mar 2017, at 14:27, Олег Рябков wrote: > >Dear gmsh/getdp users and developers. > Is there any way to output results of getdp calculaltions directly into high > order gmsh pos format. I mean we have operations like > > Print[u, OnElementsOf Omega, File "u_Dirichlet.pos", Depth 2]; > > in which we can specify elements depth subdivision, which is nice, but it > would be even nicer, if one could output high order getdp solutions directly > to high order pos format. I didn't find any clue in getdp documentation, > that's why i'm asking here. > > PS here i mean 'parsed pos', but in fact, any high order gmsh (or any other > software) post processing format is great. > It's currently not supported by GetDP - it's on our TODO list... Christophe > -- > Best regards, Oleg > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] embedded surfaces and unrolled geo files
> On 17 Mar 2017, at 10:34, Alessandro Vicini > wrote: > > > If I save the geo file as unrolled, I loose the embedded surfaces... How > could I avoid this?? Thank you. > good catch; this is now fixed in SVN r24928. > Regards, > > Alessandro > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] consistent structured mesh
> On 16 Mar 2017, at 18:59, Boehm, Christian Hermann Kurt
> wrote:
>
> Dear gmsh-community,
>
> i really need your help because otherwise i will despair. I use gmsh for my
> masterthesis in computational engineering for geometry modeling and
> finite-element meshing.
> My problem is, that i can't create a consistent structured mesh of e.g. a
> T-section. Here is a little example:
>
> Point(1)={0,0,0};
> Point(2)={10,0,0};
> Point(3)={10,10,0};
> Point(4)={0,10,0};
> Point(5)={2.5,6,0};
> Point(6)={7.5,6,0};
> Point(7)={7.5,6,10};
> Point(8)={2.5,6,10};
>
> Line(1)={1,2};
> Line(2)={2,3};
> Line(3)={3,4};
> Line(4)={4,1};
> Line(5)={5,6};
> Line(6)={6,7};
> Line(7)={7,8};
> Line(8)={8,5};
>
> Line Loop(1)={1,2,3,4};
> Line Loop(2)={5,6,7,8};
>
> Plane Surface(1)={1};
> Plane Surface(2)={2};
>
> Line{5} In Surface{1};
> Transfinite Surface{1:2};
> Recombine Surface{1:2};
>
> Is it possible to create a structured mesh, so that there will be nodes
> created on line(5) which is in surface(2). This is necessary for me, because
> the meshes of the two plates have to be consistent.
> Or is there any other way to create a consistent structured mesh for this
> example.
> I know that i can split the "ground-plate" by defining more lines and planes
> to get a structured mesh, but this is not expedient.
but it's the way to go... that's the annoying thing with structured grids ;-)
(You could actually do it without, but I don't recommend it: if you set the
number of subdivisions so that the mesh vertices match at the interface, you
can mesh the (unconnected) surfaces, save the mesh, reload it, and apply
"Coherence Mesh" on the result to fuse duplicate vertices/elements. But this is
clearly a hack.)
Christophe
> Thank you very much for your help.
>
> Kind regards,
> Christian
> Munich University of Applied Sciences
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
[Gmsh] Switch to Git
Dear all, Gmsh development has moved to Git: http://gitlab.onelab.info/gmsh/gmsh. (The old subversion repository has become read-only, and most of the material from the Trac forge has been migrated to Gitlab.) Christophe -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Segfault with Lloyd optimization
Indeed - fixed in latest git master. But don't use LLoyd: it's experimental and only for generating quad-meshes. > On 5 Apr 2017, at 14:45, Jeremy Theler wrote: > > Hi! I am obtaining a segmentation fault with latest git and the attached > geo file (it uses the OpenCASCADE factory): > > $ gdb --args gmsh -3 veeder.geo > [...] > Program received signal SIGSEGV, Segmentation fault. > lpcvt::step3 (this=this@entry=0x7fffd6b0, triangulator=..., > gf=gf@entry=0x1bb2e80) >at /home/gtheler/codigos/3ros/build/gmsh/Mesh/meshGFaceLloyd.cpp:948 > 948 s1 = borders[i].get_segment(0); > (gdb) > > > Any suggestions? > > -- > Jeremy Theler > www.seamplex.com > > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Preservation of physical surface labels after boolean operation
Hi Andrew,
It's not yet implemented, but it's clearly on our TODO list. (Currently we
already preserve the tags of the highest-dimensional entities whenever
possible: cf. demos/boolean/fragments_numbering.geo.)
Christophe
> On 15 Apr 2017, at 19:39, andrew wrote:
>
>
> Hi,
>
> The new gmsh 3 is great with the Boolean additions!.
>
> I would like to ask if there is a way to preserve the labeling of the
> physical surfaces of the objects and the tool objects before the Boolean
> operation and pass them to the new objects after the Boolean operation.
>
> For example in the following geo file there is a main object (b1[]) from
> extrusion and 3 tool objects (b2[],b3[],b4[]) created from extrusion.
>
> From these objects I define two physical surfaces (b1[2]=out - plane 5
> /b2[3]=inside - plane 11). If I perform a Boolean operation with them I lose
> the out/inside names and since the numbering of the surfaces changes from the
> creation of new objects I can't find the inside/out labels any more. Is there
> a way to preserve the names from the old objects and pass them to the new
> object? For example the newly created surfaces (plane 40 - plane 25) in the
> new object that were part of the out surface before could be preserved.
>
> Something like: Physical Surface("new_out")=FromOld("out") that could assing
> to the surfaces 40 & 25 the tag 'out' from the surface 5 since these two were
> created from the breaking of surface 5.
>
> kind regards
>
> Andrew Tsiantis
>
>
>
> _______
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] split windows camera sync
Hi Oleg, You can display different models/post-processing views in different windows or subwindows. To synchronize the viewport, use "Alt+mouse click" on the rotation and/or scale button in the status bar. Currently there's no "live update" (you need to Alt+click to sync) - I've added this to our TODO list. I've uploaded a short video to show how the current multi-window stuff works: https://youtu.be/Bxc6kaoltOA Christophe > On 22 Apr 2017, at 17:13, Олег Рябков wrote: > > Hmm.. Sorry, I forgot that it's impossible in gmsh to show different scenes > in split windows (for example, several different postprocessing fields or > geometry), so this makes my previous question pointless. In CFX it's possible > to split window and show different scenes with synced cameras (for example > temperature and velocity). I'm > afraid it's too complicated for gmsh accounting the philosophy of interface > and options. > > Anyway, gmsh is favorite tool in our lab, so we are glad that it's getting > better. > > -- > Best regards, Oleg > _______ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] boolean partitioning
Hi Andrew,
Yes, I think that a combination of BooleanFragments and embedded surfaces
should do what you want:
SetFactory("OpenCASCADE");
Block(1) = {-0, 0, 0, 1, 1, 1};
// intersecting:
Rectangle(7) = {0.4, 0.7, 0.2, 0.3, 0.3, 0};
Rectangle(8) = {-0.2, 0.3, 0.5, 0.3, 0.3, 0};
Rectangle(9) = {0.8, -0.2, 0.5, 0.3, 0.3, 0};
BooleanFragments{ Volume{1}; Delete; }{ Surface{7:9}; Delete; }
// fully inside
Rectangle(30) = {0.3, 0.3, 0.5, 0.3, 0.3, 0};
Surface {30} In Volume{1};
Christophe
> On 2 May 2017, at 19:57, andrew wrote:
>
>
> hi,
>
> In Salome which uses the opencascade model there is a Boolean operation
> called partition. With this you can create a solid from two others but with a
> shared face. A usage of that is the creation of meshes with internal walls in
> it with zero thickness aka 'baffles'. Is there a procedure in gmsh that can
> give the same results given that it has implemented the same opencascade
> model? Obviously there is a way to have such results if you give the precise
> geometry but sometimes it is very difficult to cut geometries and it is
> easier if gmsh could do that.
>
> Attached is an image of a mesh with internal baffles created with
> partitioning and the geometry that created it. I would prefer t if gmsh could
> do the meshing but I haven't found a way to load the geometry to gmsh from
> Salome.
>
> kind regards
>
> Andrew
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] boolean partitioning
Have a look at the new boolean examples, e.g. http://gitlab.onelab.info/gmsh/gmsh/blob/master/demos/boolean/compsolid.geo http://gitlab.onelab.info/gmsh/gmsh/blob/master/demos/boolean/compsolid2.geo BooleanFragments does exactly what you need. > On 3 May 2017, at 10:14, Zenker, Dr. Matthias > wrote: > > Hi, > > I would also be interested in that functionality. It would resolve the old > “duplicated surfaces” problem that forced me (and others) to leave gmsh and > do the meshing with Salomé, even though IMHO gmsh does a better job > especially in 3D meshing – in Salomé, only netgen is available as 3D > algorithm in the free version. > BTW there seems to be a macro permitting to use gmsh in Salomé, see this post > at the Salomé forum: > http://salome-platform.org/forum/forum_12/953948027#494327180 > I haven’t tested it, though. > To export a geometry from Salomé to gmsh, I have sometimes had success using > brep format. > > Matthias > > Von: andrew [mailto:armit...@gmail.com] > Gesendet: Dienstag, 2. Mai 2017 19:58 > An: gmsh@onelab.info > Betreff: [Gmsh] boolean partitioning > > > hi, > > In Salome which uses the opencascade model there is a Boolean operation > called partition. With this you can create a solid from two others but with a > shared face. A usage of that is the creation of meshes with internal walls in > it with zero thickness aka 'baffles'. Is there a procedure in gmsh that can > give the same results given that it has implemented the same opencascade > model? Obviously there is a way to have such results if you give the precise > geometry but sometimes it is very difficult to cut geometries and it is > easier if gmsh could do that. > > Attached is an image of a mesh with internal baffles created with > partitioning and the geometry that created it. I would prefer t if gmsh could > do the meshing but I haven't found a way to load the geometry to gmsh from > Salome. > > kind regards > > Andrew > > > Erbe Elektromedizin GmbH Firmensitz: 72072 Tuebingen Geschaeftsfuehrer: > Christian O. Erbe, Reiner Thede Registergericht: Stuttgart HRB 380137 > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] boolean partitioning
> On 3 May 2017, at 10:36, Zenker, Dr. Matthias > wrote: > > Thank you! > > Does it work also with STEP data imported from CAD? Yes! > > Matthias > >> -Ursprüngliche Nachricht- >> Von: Christophe Geuzaine [mailto:cgeuza...@ulg.ac.be] >> Gesendet: Mittwoch, 3. Mai 2017 10:34 >> An: Zenker, Dr. Matthias >> Cc: andrew; gmsh@onelab.info >> Betreff: Re: [Gmsh] boolean partitioning >> >> >> Have a look at the new boolean examples, e.g. >> >> https://urldefense.proofpoint.com/v2/url?u=http- >> 3A__gitlab.onelab.info_gmsh_gmsh_blob_master_demos_boolean_comps >> olid.geo&d=DwIFaQ&c=7vND7vRFB1FzfxRXc-X80O-MJSZd5Q-IVyyYW- >> Ff2gY&r=3enlI3TuQDmO9BnMCnxq2Mfm4tAO6Z3_Ae1Yc57DMr4&m=2mpe >> DRom3Noibyq90RXwKUzWF1rJRcO6ZGHMbIFv5so&s=xHnc46KE5L6ugWGTfy >> _YoQoMQy6J3MVGy7jh30SFFLo&e= >> https://urldefense.proofpoint.com/v2/url?u=http- >> 3A__gitlab.onelab.info_gmsh_gmsh_blob_master_demos_boolean_comps >> olid2.geo&d=DwIFaQ&c=7vND7vRFB1FzfxRXc-X80O-MJSZd5Q-IVyyYW- >> Ff2gY&r=3enlI3TuQDmO9BnMCnxq2Mfm4tAO6Z3_Ae1Yc57DMr4&m=2mpe >> DRom3Noibyq90RXwKUzWF1rJRcO6ZGHMbIFv5so&s=NcLs864W46WNKIoHZ >> E2UBNMZprqYawTICL8jZpa7UfM&e= >> >> BooleanFragments does exactly what you need. >> >> >>> On 3 May 2017, at 10:14, Zenker, Dr. Matthias > med.com> wrote: >>> >>> Hi, >>> >>> I would also be interested in that functionality. It would resolve the old >> “duplicated surfaces” problem that forced me (and others) to leave gmsh >> and do the meshing with Salomé, even though IMHO gmsh does a better job >> especially in 3D meshing – in Salomé, only netgen is available as 3D >> algorithm >> in the free version. >>> BTW there seems to be a macro permitting to use gmsh in Salomé, see >>> this post at the Salomé forum: >>> https://urldefense.proofpoint.com/v2/url?u=http-3A__salome- >> 2Dplatform. >>> org_forum_forum-5F12_953948027- >> 23494327180&d=DwIFaQ&c=7vND7vRFB1FzfxRX >>> c-X80O-MJSZd5Q-IVyyYW- >> Ff2gY&r=3enlI3TuQDmO9BnMCnxq2Mfm4tAO6Z3_Ae1Yc57D >>> >> Mr4&m=2mpeDRom3Noibyq90RXwKUzWF1rJRcO6ZGHMbIFv5so&s=bweZW >> T8PCF107y-xPl >>> l629T-vfFUIANxsy48ZRnkG6Q&e= >>> I haven’t tested it, though. >>> To export a geometry from Salomé to gmsh, I have sometimes had success >> using brep format. >>> >>> Matthias >>> >>> Von: andrew [mailto:armit...@gmail.com] >>> Gesendet: Dienstag, 2. Mai 2017 19:58 >>> An: gmsh@onelab.info >>> Betreff: [Gmsh] boolean partitioning >>> >>> >>> hi, >>> >>> In Salome which uses the opencascade model there is a Boolean operation >> called partition. With this you can create a solid from two others but with a >> shared face. A usage of that is the creation of meshes with internal walls >> in it >> with zero thickness aka 'baffles'. Is there a procedure in gmsh that can give >> the same results given that it has implemented the same opencascade >> model? Obviously there is a way to have such results if you give the precise >> geometry but sometimes it is very difficult to cut geometries and it is >> easier if >> gmsh could do that. >>> >>> Attached is an image of a mesh with internal baffles created with >> partitioning and the geometry that created it. I would prefer t if gmsh could >> do the meshing but I haven't found a way to load the geometry to gmsh from >> Salome. >>> >>> kind regards >>> >>> Andrew >>> >>> >>> Erbe Elektromedizin GmbH Firmensitz: 72072 Tuebingen >>> Geschaeftsfuehrer: Christian O. Erbe, Reiner Thede Registergericht: >>> Stuttgart HRB 380137 >>> >>> ___ >>> gmsh mailing list >>> gmsh@onelab.info >>> https://urldefense.proofpoint.com/v2/url?u=http- >> 3A__onelab.info_mailma >>> n_listinfo_gmsh&d=DwIFaQ&c=7vND7vRFB1FzfxRXc-X80O-MJSZd5Q- >> IVyyYW-Ff2gY >>> >> &r=3enlI3TuQDmO9BnMCnxq2Mfm4tAO6Z3_Ae1Yc57DMr4&m=2mpeDRom >> 3Noibyq90RXwK >>> UzWF1rJRcO6ZGHMbIFv5so&s=1HcJ-0ygfx48c3Da61c9yCkbIwdzENB- >> tgduaWTphlg&e >>> = >> >> -- >> Prof. Christophe Geuzaine >> University of Liege, Electrical Engineering and Computer Science >> https://urldefense.proofpoint.com/v2/url?u=http- >> 3A__www.montefiore.ulg.ac.be_- >
Re: [Gmsh] GMSH failure in meshing - Unable to recover mesh, etc.
Hi Franz, I can indeed reproduce the issue on Windows (the Mac and Linux versions work fine). We have seen problems like this (accuracy problems leading to meshing failure with large element size gradients) for a while on Windows, but have not yet found a fix. They might be related to the behavior of the robust geometrical predicates, but we haven't been able to pinpoint the issue. Christophe > On 12 May 2017, at 10:16, Bormann, F. wrote: > > Hi, > I run currently into problems when I want to mesh my model. > > I have two large equally sized connected 2D domains of total size LxH. Due to > my problem I need a high resolution at following points: > > 1. Small area around Joint line of both domains between a<=x<=L with > a>0 > 2. Small circle around point at x=L and y=H/2 > 3. Small area around line x=L and H/2-b comparison with H > > For the refinements I use the Attractor and Threshold fields with a combined > Field = Min. The minimum mesh size is 0.125 for refinement 2 and 0.25 for > refinements 1 and 3. > > I have run the meshing procedure with 5 variations. 4 of those failed. The > only parameters I varied are the: Length, height, maximum element size and > (for the successful simulation) as well the parameter a. The model size > varies between 4000x1600 and 6000x2000. And the maximum element size between > 50 and 100. > > Depending on the different variations I get different failure modes. > 1. Warning: Skipping bad triangle; Error: Unable to recover an edge. – > only one of the domains is meshed > 2. GMSH crashes > 3. For Meshing of both domains – Error: Identical points in > triangulation; Error: Unable to recover an edge. Increasing Mesh.RandomFactor > does not help > 4. Error: Unable to recover an edge. – only one of the domains is meshed > > The .geo files are attached with the suffix corresponding to the failure > mode. Suffix 5 corresponds to the working mesh generation. > > Here I used GMSH version 2.12.0 and 3.0.1. The operating system is Windows 7, > 64-bit. > > What is the reason behind the failure and is there a way to remedy the > problem? It would really help me a lot. > > Thanks already in advance. > > Cheers, > Franz > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Issues with geo file
> On 11 May 2017, at 12:05, Alessandro Vicini > wrote: > > > > Hullo everybody, > > no luck the first time, so I try again with my question... I've got two > issues with the attached geo file: > > 1) There are circle arcs defined and then rotated; this causes gmsh to give a > number of errors of the kind "control points of circle xxx are not > co-circular". Now, as the points ARE co-circular, how can I avoid these > errors? Is this a matter of some tolerance that I can set? Alessandro - You cannot rotate/translate the points after creating the circles... > > 2) The same geo file is read with no problems (apart what said at point 1) by > version 3.0 of the code, while version 2.16 crashes; why...? We're improving ;-) Christophe > > Thank you! > > Alessandro > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] this application cannot execute on your system
Does the 32bit version work? Try http://gmsh.info/bin/Windows/gmsh-git-Windows32.zip > On 9 May 2017, at 09:44, lemerle pierre wrote: > > That is the msg I get when I install gmsh on my 64b pc > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] High Order Mesh
Dear Brian, The mesh is indeed curved - but the curved triangle edges are just represented as broken by default to speed up the graphical rendering. Increase Tools->Options->Mesh->Aspect->High-order element subdivisions to improve the rendering. Christophe > On 22 May 2017, at 20:18, Brian Chubb wrote: > > To whom it may concern, > > I have a question in relation to higher order mesh generation. When I change > the mesh from order 1 to 2, it seems to refine the mesh instead of make it an > actual, second order curved mesh. Attached below are examples of what I am > having problems with. The first screenshot shows the mesh in first order, > second screenshot shows second order. Instead of curving the mesh around the > model for second order, it splits it into two linear elements. > > Can you tell me how I get curved mesh, instead of smaller linear increments? > > Thanks, > > Brian Chubb > > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Gmsh is slow after importing igs
Gmsh does not convert between CAD formats (so you cannot reliably "convert" an iges file to geo file). Also, use step files instead of iges files if you want to import data in Gmsh: step files contain the full topology, which makes producing conformal meshes much easier. Could you try importing the step version of your model in Gmsh 3.0.2 and let us know if it's still slow? If it is, could you send a representative model that exhibits this slowness? Thanks, Christophe > On 20 May 2017, at 07:14, 李宗耀 wrote: > > Dear Gmsh support team, > I have a complex geometry built in NX and I save it as .igs file. Gmsh is > slow after importing the igs file. I have tried to set > Geometry.AutoCoherence option to 0 at the end of .geo file. But it doesn't > work. Is there any tips? I wonder if I could convert .igs to .geo directly(I > don't mean using "Merge", just convert .igs file to gmsh geometry file). > Thank you very much. I am looking forward for your reply. > Good wishes! > Zongyao Li From China > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] problem 3D transfinite algorithm
Hi Danny, You specified the transfinite interpolation corners in an incompatible way between the surfaces and the volumes. For simple surfaces (with 3/4 vertices) you actually don't have to specify them by hand: here's your file, simplified: Test2.geo Description: Binary data Christophe > On 13 May 2017, at 13:39, Danny Lathouwers - TNW > wrote: > > > Hello, > > I am trying to build a geometry consisting of 3 blocks connected. > The blocks should have stretched meshes (as seen when generating 2D meshes in > the attached geo). > > I failed when attempting to do the 3D. At first I removed the interconnecting > surfaces between the blocks so that the complete volume could be meshed. > That did not work (perhaps too complex, too many surfaces involved?). > > In a second try, I added these interconnecting surfaces and generate the > volumes separately as in the Test.geo. Even the first block fails to generate > (incompatible surface message). This seems to be something I am doing wrong? > > Even if this can be made to work I will probably end up with 3 non-connected > meshes right? This woud not be what I want. Can these then be connected? > > Could you give me a hand with this geometry/mesh? > > Thanks very much ! > Danny. > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] boolean partitioning
> On 23 May 2017, at 11:22, Jeremy Theler wrote:
>
> Thanks for the example Christophe.
> How do I apply PointsOf{} to a merged BREP?
Same thing: you could do e.g.
a() = ShapeFromFile("test.brep");
p() = PointsOf{ Volume{a()}; };
>
> --
> Jeremy Theler
> www.seamplex.com
>
>
>
> On Mon, 2017-05-22 at 20:52 +0200, Christophe Geuzaine wrote:
>>
>>
>> Here's an example, for Gmsh 3.0.2:
>>
>>
>> SetFactory("OpenCASCADE");
>>
>>
>> Box(1) = {-0, 0, 0, 1, 1, 1};
>>
>> p() = PointsOf{ Volume{1}; };
>> Characteristic Length{p()} = 0.2;
>>
>> Rectangle(7) = {0.4, 0.7, 0.2, 0.3, 0.3, 0};
>> Rectangle(8) = {-0.2, 0.3, 0.5, 0.3, 0.3, 0};
>> Rectangle(9) = {0.8, -0.2, 0.5, 0.3, 0.3, 0};
>> Rectangle(10) = {0.3, 0.3, 0.5, 0.3, 0.3, 0};
>>
>> b() = BooleanFragments{ Volume{1}; Delete; }{ Surface{7:10};
>> Delete; };
>>
>> p() = PointsOf{ Surface{b({1:6})}; };
>> Characteristic Length{p()} = 0.04;
>>
>>
>>
>>
>>
>>
>>> On 4 May 2017, at 12:30, Jeremy Theler wrote:
>>>
>>>
>>>> I have found an another way to import a brep file. However I can't
>>>> understand how the characteristic sizes are applied. I want to
>>>> apply a
>>>> more dense mesh around the internal surfaces without the cost of a
>>>> field command.
>>>
>>> I don't either.
>>>
>>> But what I do is I set the characteristic length of all points to a
>>> basic lc
>>>
>>> p[] = Point "*";
>>> Characteristic Length { p[] } = lc;
>>>
>>> and then set a lower value for some of the points I want to refine
>>> around. You have to known the id of such points, though.
>>>
>>> Is there any better solution without falling back to a background
>>> field?
>>>
>>> --
>>> jeremy
>>> www.seamplex.com
>>>
>>>
>>>
>>>
>>> ___
>>> gmsh mailing list
>>> gmsh@onelab.info
>>> http://onelab.info/mailman/listinfo/gmsh
>>
>> --
>> Prof. Christophe Geuzaine
>> University of Liege, Electrical Engineering and Computer Science
>> http://www.montefiore.ulg.ac.be/~geuzaine
>>
>> Free software: http://gmsh.info | http://getdp.info |
>> http://onelab.info
>>
>>
>>
>
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] problem with surface displaying
Hi Roman, Indeed, we also noticed bugs in OpenCASCADE's surface rendering algorithm, which sometimes hangs. The only fix for now is to use the internal Gmsh surface display mode (the "cross"): just add Geometry.SurfaceType = 0; in your file to revert to this simpler display mode. Christophe > On 30 May 2017, at 15:34, Roman Przylucki wrote: > > Dear Sir. > I have a problem with a face visualization in Gmsh. > > After extruding of non plane surface (cone surface) it is imposible to show > surfaces and surface's numbers in Gmsh. > > But except that, model seems to be correct also meshing is correct. > > Do you know this problem? > Is it impossible to extrude non plane (cone) surfaces? > > Unfortunately I'm trying to prepare a complicated geometry, and I divide the > problem into several files. > > The main file to open in Gmsh is: "t1.geo" > Probem with extrusion is in the file: "t1_2_palce_wsad.geo" (sorry for such > name) in line 30. > > > Other files contain: > t1_1_dno2.geo : begining of geometry, > t1.par : geometrical parameters, > t1.m : mesh parameters, > t1.v_s : physical volumes and surfaces definition > > Gmsh version: 3.0.2, 2017 05 13 > FLTK version: 1.3.4; PETSc version 3.7.5; OCC version: 7.1.0; MED version: > 3.2.0 > Operating system: Windows 8.1 Pro 64 Polish language version > > Sincerely, > and with request for help > Roman Przylucki > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Pb with plane symmetry
Dear Marie-Claire,
Applying geometrical transformations on extruded entities with the built-in CAD
engine is not supported. Everything should work as expected however with the
new OpenCASCADE integration: try adding SetFactory("OpenCASCADE") at the
beginning of your file.
Note that extrusion meshing ("Layers{24};Recombine;") will not work after
transformations, as the structured mesh algorithm uses the original geometrical
extrusion definition.
Christophe
> On 2 Jun 2017, at 16:37, Marie-Claire Le Pape
> wrote:
>
> Hello,
>
> It seems that symmetry to a plane z=cst doesn't give what one expects ...
> I've had no problem for a plane x=cst.
> Try with the files given.
>
> Marie-Claire Le Pape
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] brep file
> On 4 Jun 2017, at 13:28, Noam ABETTAN wrote: > > Dear Mr, > > From when it will be possible to convert a brep file into .geo, please ? > Never: Gmsh's philosophy is to *not* translate between CAD formats; instead, all of Gmsh's algorithms work directly with the native CAD representations. Note that with version 3, Gmsh can import brep files and modify them (e.g. cut or transform the entities, add and join entities, etc.). You can also directly create complex geometries in the .geo files - see demos/boolean for examples. Christophe > Noam > > > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Extrude { extrude-list } Using Wire { expression-list }
> On 4 Jun 2017, at 15:45, Daniel Steinegger
> wrote:
>
> Hello I'm Daniel an i have the Problem that i wand Extrude a Surface
> along a Spline and i don't Know how to manage that. I use the command
> Extrude { extrude-list } Using Wire { expression-list } and i get
> always a Syntax error. Do you have any example for it ?
>
http://gitlab.onelab.info/gmsh/gmsh/blob/master/demos/boolean/pipe.geo
> A example code is attached
>
> best regards
> Daniel
> Steinegger___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] apply background mesh to non-planar surface
> On 4 Jun 2017, at 20:06, Jin Yao wrote: > > Hi experts, > > Is it feasible to generate background mesh to non-planar surface, such as the > surface of a sphere? > > For planar surface, it is easy to generate a background mesh, which consists > of scalar triangles. But for non-planar surface, the inner points of > triangles are not on the surface even when the vertices of scalar triangles > lie on the surface. > > Could anyone suggest a workaround if there's no direct solution? Thanks > Put your background field on a 3 grid that covers your surface: Gmsh will happily interpolate the volume data on the non-planar surface. > Yao > Sent from my phone > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Need help with boolean operations
I don't know why but the the opencascade "common" (intersection) algorithm does
not behave as expected here; however the general "fragment" operator gives the
expected result:
SetFactory("OpenCASCADE");
cl__1 = 1;
Point(1) = {0, 0, 0, 1};
Point(2) = {20, 0, 0, 1};
Point(3) = {0, 3, 0, 1};
Point(4) = {10, -4, 0, 1};
Line(1) = {3, 4};
Line(2) = {1, 2};
BooleanFragments{ Line{1}; }{ Line{2}; }
Delete{ Line{1,2}; }
Point (6) = {10, 10, 0, 1};
Line(7) = {5,6};
> On 8 Jun 2017, at 18:35, Sergio Logrosán wrote:
>
> Dear service attendant,
>
> I am using your great GMSH software for my bachelor's degree thesis. I am
> writing to you because I have a question. I need to create points of
> elementary geometry from intersections of lines or other 2D elements. I have
> tested with Boolean operations without success. Attached you will find an
> example .geo file of what I mean.
>
> My intention in the example is to create a line (3) being one of its points
> the intersection between lines 1 and 2.
>
> Thank you very much for your time.
>
> I am looking forward to your response,
>
> Kind regards,
> Sergio Logrosán
>
>
>
>
> _______
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] orient ruled surfaces
This (undocumented...) API will actually be remove quite soon, replaced by a much simpler, integer-based API that is very close to what is done in .geo files. If you feel adventurous you can give it a try: cf. Geo/GModelIO_Geo.h (for the built-in CAD kernel) and Geo/GModelIO_OCC.h (for the OpenCASCADE kernel). Christophe > On 1 Jun 2017, at 12:29, Antonio Cervone wrote: > > Hello, > when trying to use gmsh library API, I was not able to create ruled surfaces > when the edges are not properly oriented. At this link > > https://gist.github.com/capitalaslash/47c04578b8d34f42df5ff892b577b5be > > you can see an example that fails without the patch attached there. Is it the > correct way to do it? > Thanks, > > Antonio > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] gmsh issue under Ubuntu 16.04 LTS
> On 21 Jun 2017, at 18:54, paul.carr...@free.fr wrote: > > Hi (again) > > I'm trying to open a gmsh I'm working on, on my laptop under Ubuntu 16.04 > LTS, but gmsh crashes with the following message : "Instruction non permise > (core dumped)" i.e. unautorized instruction. > Is this with the binary from the gmsh web site? Could you send the output of the following command? ldd /path/to/gmsh/bin/gmsh > It is not an issue coming from gmsh a properly speaking (since it works under > Windows), but does a gmsh/ubuntu user meet such trouble? if so is there a way > to fix it. > > nb: the .gmsh* files have been removed ... > > Thanks > > Paul > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Illegal Instructions
Marc, Paul - Could you try the latest git snapshot http://gmsh.info/bin/Linux/gmsh-git-Linux64.tgz It is now linked with a less "aggressive" BLAS library, which should work on Core2 and higher. Christophe > On 22 Jun 2017, at 15:21, m.schoen...@e-und-a.ch wrote: > > Dear List, > > I get an "Illegal Instructions" error in gmsh 3.0.2 when trying to mesh 2d > models. The meshing is possible with version 2.16. > > The error only occurs on some computers on our cluster system, but on these > computers, the error is reproducible. > > Here are the processors, where the error occurs: > > AMD Phenom(tm) II X4 965 Processor > Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz > Intel(R) Xeon(R) CPU E5410 @ 2.33GHz > AMD Athlon(tm) II X2 250 Processor @ 3Ghz > > The error does not occur on the following processors: > Intel(R) Xeon(R) CPU E5-2690 v3 @ 2.60GHz > AMD Opteron(TM) Processor 6220 3.00GHz > Intel(R) Core(TM) i7-4790K CPU @ 4.00GHz > Intel Core2 Duo E6850 3.00GHz > > Best regards, > > ppa. Marc C. Schöning > Dr.-Ing. > Leiter Forschung & Entwicklung / Head of R&D > Mitglied der Geschäftsleitung / Member of the Management > > e+a > Elektromaschinen und Antriebe AG > Bachstrasse 10 > 4313 Moehlin > SWITZERLAND > > Phone +41 61 855 92 92 > Fax +41 61 855 92 99 > Mail m.schoen...@e-und-a.ch > > http://www.e-und-a.ch > > > CONFIDENTIALITY NOTICE > This communication and the information it contains is > intended for the person(s) or organisation(s) named > above and no other person(s) or organisation(s) and > may be confidential, legally privileged and protected > by law. Unauthorised use, copying or disclosure of any > of it may be unlawful. If you have received this > communication in error, please contact us immediately > by email, telephone or facsimile. We will be happy to > accept a reversed charge call (call collect). > > CONFIDENTIALITY NOTICE > This communication and the information it contains is intended for the > person(s) or organisation(s) named above and no other person(s) or > organisation(s) and may be confidential, legally privileged and protected by > law. > Unauthorised use, copying or disclosure of any of it may be unlawful. If you > have received this communication in error, please contact us immediately by > email, telephone or facsimile. We will be happy to accept a reversed charge > call (call collect).___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Antwort: Re: Illegal Instructions
Indeed - I've uploaded new official builds for both Gmsh (3.0.3) and GetDP (2.11.2) which should fix the issues. Thanks for the feedback! Christophe > On 23 Jun 2017, at 07:55, m.schoen...@e-und-a.ch wrote: > > I think getdp v2.11.1 shows same problem with Illegal Instructions, getdp v. > 2.11.0 works. > > Best regards, > > ppa. Marc C. Schöning > Dr.-Ing. > Leiter Forschung & Entwicklung / Head of R&D > Mitglied der Geschäftsleitung / Member of the Management > > e+a > Elektromaschinen und Antriebe AG > Bachstrasse 10 > 4313 Moehlin > SWITZERLAND > > Phone +41 61 855 92 92 > Fax +41 61 855 92 99 > Mail m.schoen...@e-und-a.ch > > http://www.e-und-a.ch > > > CONFIDENTIALITY NOTICE > This communication and the information it contains is > intended for the person(s) or organisation(s) named > above and no other person(s) or organisation(s) and > may be confidential, legally privileged and protected > by law. Unauthorised use, copying or disclosure of any > of it may be unlawful. If you have received this > communication in error, please contact us immediately > by email, telephone or facsimile. We will be happy to > accept a reversed charge call (call collect). > > > > Von:Christophe Geuzaine > An:m.schoen...@e-und-a.ch, paul.carr...@free.fr > Kopie:gmsh@onelab.info > Datum:22.06.2017 23:27 > Betreff:Re: [Gmsh] Illegal Instructions > > > > > Marc, Paul - Could you try the latest git snapshot > > http://gmsh.info/bin/Linux/gmsh-git-Linux64.tgz > > It is now linked with a less "aggressive" BLAS library, which should work on > Core2 and higher. > > Christophe > > > > On 22 Jun 2017, at 15:21, m.schoen...@e-und-a.ch wrote: > > > > Dear List, > > > > I get an "Illegal Instructions" error in gmsh 3.0.2 when trying to mesh 2d > > models. The meshing is possible with version 2.16. > > > > The error only occurs on some computers on our cluster system, but on these > > computers, the error is reproducible. > > > > Here are the processors, where the error occurs: > > > > AMD Phenom(tm) II X4 965 Processor > > Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz > > Intel(R) Xeon(R) CPU E5410 @ 2.33GHz > > AMD Athlon(tm) II X2 250 Processor @ 3Ghz > > > > The error does not occur on the following processors: > > Intel(R) Xeon(R) CPU E5-2690 v3 @ 2.60GHz > > AMD Opteron(TM) Processor 6220 3.00GHz > > Intel(R) Core(TM) i7-4790K CPU @ 4.00GHz > > Intel Core2 Duo E6850 3.00GHz > > > > Best regards, > > > > ppa. Marc C. Schöning > > Dr.-Ing. > > Leiter Forschung & Entwicklung / Head of R&D > > Mitglied der Geschäftsleitung / Member of the Management > > > > e+a > > Elektromaschinen und Antriebe AG > > Bachstrasse 10 > > 4313 Moehlin > > SWITZERLAND > > > > Phone +41 61 855 92 92 > > Fax +41 61 855 92 99 > > Mail m.schoen...@e-und-a.ch > > > > http://www.e-und-a.ch > > > > > > CONFIDENTIALITY NOTICE > > This communication and the information it contains is > > intended for the person(s) or organisation(s) named > > above and no other person(s) or organisation(s) and > > may be confidential, legally privileged and protected > > by law. Unauthorised use, copying or disclosure of any > > of it may be unlawful. If you have received this > > communication in error, please contact us immediately > > by email, telephone or facsimile. We will be happy to > > accept a reversed charge call (call collect). > > > > CONFIDENTIALITY NOTICE > > This communication and the information it contains is intended for the > > person(s) or organisation(s) named above and no other person(s) or > > organisation(s) and may be confidential, legally privileged and protected > > by law. > > Unauthorised use, copying or disclosure of any of it may be unlawful. If > > you have received this communication in error, please contact us > > immediately by email, telephone or facsimile. We will be happy to accept a > > reversed charge call (call > > collect).___ > >
Re: [Gmsh] Problem to create a 3d mesh for merged surfaces
Here's a .geo script that works (it removes the overlapping entities using "fragments"): airdomain_ok.geo Description: Binary data > On 7 Jul 2017, at 14:10, Jacob Abdelfatah Ndioubnane > wrote: > > Hi gmsh community; > > I´m trying to define a 3d mesh of a merged .brep surfaces for my PhD, to use > it into freefem++.When I set opencascade geometry kernel, at the first line , > surface loop creates duplicates entities, and the subsequent volume is > defined with this entities. However, the 3d mesh is seems to be correct, but > the 2d mesh has overlapping faces, and when import this into freefem I got > errors. > > http://onelab.info/pipermail/gmsh/2014/009368.html > > If I don´t use the opencascade kernel, The 2d mesh was made properly, however > when apply the 3d mesh button, gmsh gives the error ". > Info: Meshing 3D... > Info: Delaunay Meshing 1 volumes with 1 connected components > Info: Meshing volume 1 (Delaunay) > Info: 3209 points 15963 edges and 6592 faces in the initial mesh > Info: Found region 1 > Info: 0 points created - Worst tet radius is 7.55082e+16 (PTS removed 0 0) > Info: 0 points created - Worst tet radius is 8.91669 (PTS removed 0 500) > Info: 0 points created - Worst tet radius is 4.28138 (PTS removed 0 1000) > Info: 0 points created - Worst tet radius is 2.71306 (PTS removed 0 1500) > Info: 0 points created - Worst tet radius is 2.03135 (PTS removed 0 2000) > Info: 0 points created - Worst tet radius is 1.60219 (PTS removed 0 2500) > Info: 0 points created - Worst tet radius is 1.32663 (PTS removed 0 3000) > Info: 0 points created - Worst tet radius is 1.14246 (PTS removed 3 3497) > Info: 3D point insertion terminated (3209 points created): > Info: - 0 Delaunay cavities modified for star shapeness > Info: - 3973 points could not be inserted > Info: - 9413 tetrahedra created in 0.09375 sec. (100405 tets/sec.) > Info: Done meshing 3D (0.515625 s) > Error : OCC Project Point on Surface FAIL > Info: point 4.156 9.4 8.16 : Relaxation factor = 0.75 > Info: point 4.156 9.4 8.16 : Relaxation factor = 0.5625 > Info: point 4.156 9.4 8.16 : Relaxation factor = 0.421875 > Info: point 4.156 9.4 8.16 : Relaxation factor = 0.316406 > Info: point 4.156 9.4 8.16 : Relaxation factor = 0.237305 > Info: point 4.156 9.4 8.16 : Relaxation factor = 0.177979 > Info: point 4.156 9.4 8.16 : Relaxation factor = 0.133484 > Info: point 4.156 9.4 8.16 : Relaxation factor = 0.100113 > > > > When I apply the Coherence Mesh; I got the error "No tethraedra in region > 10". > > Thank you all in advance. > > Best Regards > Jacob > > > patio central.brep> exterior.brep> exteriores.brep> cristal del patio central.brep> lateral.brep> patio central.brep> interior.brep> central.brep> marcos.brep> -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Subtract elementary volume entity from physical volume using ID
> On 8 Jul 2017, at 01:39, Alex Lindsay wrote:
>
> I have a large lattice geometry generated by for loops and extrusion. I have
> a physical volume called "fuel" composed of 841 elementary volumes; addition
> to the physical group is done automatically in the for loops. The "fuel"
> geometry is attached. I want to remove one of the elementary volumes from the
> physical volume. Through the GUI, I know its ID. With the information I have,
> can I do what I want to do?
Physical Volume(1) -= {1,2,3};
> Or other suggestions?
Add the elementary ids to a list 9selectively), and only create the physical
later.
>
> Alex
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Variable Not Updating In For Loop
> On 8 Jul 2017, at 03:26, Jeremias Gonzalez wrote:
>
> Hi, below (and attached) is a simple script where, in a for loop, I make a
> mesh for a cube, output the mesh, then start over and make the mesh again
> with a changed number. For some reason, as can be seen from the printed line
> in each iteration of the loop, the variable instantiated by the for loop is
> changing as expected, but the parameter I made that sets the variable as its
> own value is stuck at the original value. Even stranger, just to be sure that
> the parameter is getting refreshed each time since it wasn't overwriting like
> I expected, at the beginning of the for loop I tried deleting the parameter,
> then allow it to be declared anew with the current for loop variable, but it
> remains stuck with the value it received from the first value from the for
> loop. Can anyone explain where things are going wrong?
cubelength = DefineNumber[ newcubelength, Name "Parameters/cubelength" ];
will assign the default value newcubelength to cubelength if the ONELAB
parameter "Parameters/cubelength" does not exist. If "Parameters/cubelength"
exists in the ONELAB database, it will use that value.
This is the mechanism that allows you to modifiy the value associated to
"Parameters/cubelength" externally (in the GUI, in other ONELAB clients), and
have Gmsh use the updated value later.
If the script must have precedence over other clients, i.e. if the script
should always assign the value of the ONELAB parameter, you can mark the
parameter as "ReadOnly" :
cubelength = DefineNumber[ newcubelength, Name "Parameters/cubelength",
ReadOnly];
> Is there a better way to do a parametric sweep and output of meshes?
>
> SetFactory("OpenCASCADE");
>
> For newcubelength In {0.1:0.25:0.05}
>
> Delete cubelength;
> cubelength = DefineNumber[ newcubelength, Name "Parameters/cubelength" ];
>
> Block(1) = {0, 0, 0, cubelength, cubelength, cubelength};
>
> Physical Volume("cubebody", 1) = {1};
>
> Mesh.SaveElementTagType=2;
> Mesh.Format=30;
> Mesh 3;
> Save Sprintf("cube%g.mesh",cubelength);
>
> Printf("%g,%g",cubelength,newcubelength);
>
> Delete Model;
> Delete Physicals;
>
> EndFor
>
> ___
> gmsh mailing list
> gmsh@onelab.info
> http://onelab.info/mailman/listinfo/gmsh
--
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine
Free software: http://gmsh.info | http://getdp.info | http://onelab.info
___
gmsh mailing list
gmsh@onelab.info
http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Ubuntu : GetDP Error : ONELAB version mismatch (server: 1.3 / client: 1.2)
Hi Hiroshi, It seems that Ubuntu is simply shipping versions of Gmsh and GetDP that contain incompatible versions of Onelab (GetDP is too old). In any case I recommend you download the Onelab bundle from onelab.info : it contains more recent (and compatible) versions of Gmsh and GetDP. Christophe > On 13 Jul 2017, at 08:27, ABE Hiroshi wrote: > > Thank you for your reply, st. > > I understand I was wrong. I thought the t1.pro file would be automatically > generated by Gmsh in some default setting. > > I followed the web page you told. > I processed onleab-MacOSX/models/waveguides/waveguide3D_rectangle.pro with > Gmsh and I found it works fine on OS X. > But Gmsh/GetDP on Ubuntu17.04 claims the same error as I mentioned. > > Should I ask this question on GetDP mailing list? > > 2017/07/13 14:31、lukshun...@gmail.com のメール： > >> On Thursday, July 13, 2017 10:11 AM, ABE Hiroshi wrote: >>> Hi, All, >>> I’m new to the list and am intresting in Gmsh/GetDP. >>> I am usually working on MacOSX and Linux boxes. For trying Gmsh/GetDP, I >>> installed Gmsh/GetDP from main repository of ubuntu17.04. >>> As a first try, open tutorial/t1.geo with Gmsh and click Solver to invoke >>> GetDP but I got error messages and fail to run the tutorial file. >> >> Hi, >> >> IIRC, there's nothing to solve in the tutorials of gmsh. They are tutorials >> for using gmsh which is only a mesh-generator and post-processor. >> >> To solve your problem, you have to create your own model (file) and provide >> it to getdp. >> >> To get feel of what will happen, try follow what this wiki page says >> >> http://onelab.info/wiki/Main_Page >> >> Regards, >> -- st > > ABE Hiroshi > from Tokorozawa, JAPAN > > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Ubuntu : GetDP Error : ONELAB version mismatch (server: 1.3 / client: 1.2)
> On 13 Jul 2017, at 11:45, ABE Hiroshi wrote: > > Hi Christophe, > > I didn’t expect Ubuntu distribution contains incompatible version of sibling > softwares. > > Okay, download from the Onelab site and build them and try. It should work. A priori you don't need to build: just download the binary. > > Thank you so much. > > Hiroshi > > 2017/07/13 16:16、Christophe Geuzaine のメール： > >> >> Hi Hiroshi, >> >> It seems that Ubuntu is simply shipping versions of Gmsh and GetDP that >> contain incompatible versions of Onelab (GetDP is too old). >> >> In any case I recommend you download the Onelab bundle from onelab.info : it >> contains more recent (and compatible) versions of Gmsh and GetDP. >> >> Christophe >> >>> On 13 Jul 2017, at 08:27, ABE Hiroshi wrote: >>> >>> Thank you for your reply, st. >>> >>> I understand I was wrong. I thought the t1.pro file would be automatically >>> generated by Gmsh in some default setting. >>> >>> I followed the web page you told. >>> I processed onleab-MacOSX/models/waveguides/waveguide3D_rectangle.pro with >>> Gmsh and I found it works fine on OS X. >>> But Gmsh/GetDP on Ubuntu17.04 claims the same error as I mentioned. >>> >>> Should I ask this question on GetDP mailing list? >>> >>> 2017/07/13 14:31、lukshun...@gmail.com のメール： >>> >>>> On Thursday, July 13, 2017 10:11 AM, ABE Hiroshi wrote: >>>>> Hi, All, >>>>> I’m new to the list and am intresting in Gmsh/GetDP. >>>>> I am usually working on MacOSX and Linux boxes. For trying Gmsh/GetDP, I >>>>> installed Gmsh/GetDP from main repository of ubuntu17.04. >>>>> As a first try, open tutorial/t1.geo with Gmsh and click Solver to invoke >>>>> GetDP but I got error messages and fail to run the tutorial file. >>>> >>>> Hi, >>>> >>>> IIRC, there's nothing to solve in the tutorials of gmsh. They are >>>> tutorials for using gmsh which is only a mesh-generator and post-processor. >>>> >>>> To solve your problem, you have to create your own model (file) and >>>> provide it to getdp. >>>> >>>> To get feel of what will happen, try follow what this wiki page says >>>> >>>> http://onelab.info/wiki/Main_Page >>>> >>>> Regards, >>>> -- st >>> >>> ABE Hiroshi >>> from Tokorozawa, JAPAN >>> >>> >>> ___ >>> gmsh mailing list >>> gmsh@onelab.info >>> http://onelab.info/mailman/listinfo/gmsh >> >> -- >> Prof. Christophe Geuzaine >> University of Liege, Electrical Engineering and Computer Science >> http://www.montefiore.ulg.ac.be/~geuzaine >> >> Free software: http://gmsh.info | http://getdp.info | http://onelab.info >> > > ABE Hiroshi > from Tokorozawa, JAPAN > -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Cannot delete entities
> On 28 Jul 2017, at 15:13, Alessandro Vicini > wrote: > > > > Hullo, can anybody tell my why Gmsh doesn't allow me to delete the points > internal to the volume in the attached geo file (for example points #32 or > #33)? I am using version 3.0.1. Good catch: "Recursive Delete" with the built-in CAD kernel did not remove the "reversed" curves (the built-in CAD kernel stores curves with both orientations), which left some points attached to invisible curves... which meant that they could not be removed. This is now fixed in the master Git branch. > > Another question: if I merge a file from the GUI, the merge command is not > appended to the script, so that if I reload the script the merge operation is > lost. Is this normal? Yes: none of the commands from the main menu are appended in the script; only the geometry construction commands from Geometry tree menu are appended to the script. > > Thank you. > > A. > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] GMSH
> On 24 Jul 2017, at 11:14, Karsten Vermeulen wrote: > > Hi, > > How are you? > > First of all - great looking rendering tool! > > I downloaded your GMSH source code just to browse through some of your code > and see how you guys did things. I have a question though - why are you guys > still using older OpenGL fixed function calls? because it was written a looong time ago :-) > Do most CAD applications still make use of the older OpenGL? > > Many thanks! > > Kind Regards, > > Karsten Vermeulen > Games Supervisor & Lecturer > SAE Institute UK > > Phone: +44 (0) 20 7923 9159 > Mobile: +44 (0) 74 3214 2602 > Web: www.sae.edu > Skype: djkarstenv > > > > > > > > > > > > > > SAE Institute is accredited by Middlesex University, London UK. Students > enrolled on a validated programme will receive a Middlesex award on > successful completion of their studies. SAE Institute is a member of Study UK > and a listed body on the register of the Department for Business, Innovation > & Skills. > > QAA checks how UK universities, colleges and other providers maintain the > standard of their higher education provision. Click here to read this > institution's latest review report. > > SAE is the trading name of SAE Education Ltd. > Registered in England No 06647488 > Registered office: SAE Oxford, Littlemore Park, Armstrong Road, Oxford, OX4 > 4FY. > > CONFIDENTIALITY CAUTION : This message is intended only for the use of the > individual or entity to whom it is addressed and contains information that is > privileged and confidential. If you, the reader of this message, are not the > intended recipient, you should not disseminate, distribute or copy this > communication. If you have received this communication in error, please > notify us immediately by return email and delete the original message. Thank > you. > > ___ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
Re: [Gmsh] Mesh ordering
> On 20 Jul 2017, at 14:07, RIVERA ROLDAN, Jorge Orlando > wrote: > > Dear members list, > > > It is possible to reorder the node numbering when meshing > No, we leave this to each solver, as each has different requirements. > Cheers > > Orlando > > > > -- > MTU Aero Engines AG > Vorstand/Board of Management: Reiner Winkler, Vorsitzender/CEO; Dr. Rainer > Martens, Michael Schreyoegg > Vorsitzender des Aufsichtsrats/Chairman of the Supervisory Board: Klaus > Eberhardt > Sitz der Gesellschaft/Registered Office: Muenchen > Handelsregister/Commercial Register: Muenchen HRB 157206 > > Diese E-Mail sowie ihre Anhaenge enthalten MTU-eigene vertrauliche oder > rechtlich geschuetzte Informationen. > Wenn Sie nicht der beabsichtigte Empfaenger sind, informieren Sie bitte den > Absender und loeschen Sie diese > E-Mail sowie die Anhaenge. Das unbefugte Speichern, Kopieren oder > Weiterleiten ist nicht gestattet. > > This e-mail and any attached documents are proprietary to MTU, confidential > or protected by law. > If you are not the intended recipient, please advise the sender and delete > this message and its attachments. > Any unauthorised storing, copying or distribution is prohibited. > _______ > gmsh mailing list > gmsh@onelab.info > http://onelab.info/mailman/listinfo/gmsh -- Prof. Christophe Geuzaine University of Liege, Electrical Engineering and Computer Science http://www.montefiore.ulg.ac.be/~geuzaine Free software: http://gmsh.info | http://getdp.info | http://onelab.info ___ gmsh mailing list gmsh@onelab.info http://onelab.info/mailman/listinfo/gmsh
