[CIG-SEISMO] gmsh and specfem2D

[Paul Cristini]

I had a look at your .geo and I found a small error. Your three line 
loops are not defined in the same way. The first is counterclokwise, the 
second is clockwise and the third one is counterclockwise. All should be 
defined counterclockwise.
So you need to change Line Loop(13) = {3, 5, 6, 7}; to Line Loop(13) = 
{-3, -5, -6, -7};
and it will work. I checked

Best regards

Paul Cristini

Le 06/09/2012 16:41, giuseppe.digiulio at ingv.it a écrit :
> I am a new user of specfem2D  and I am trying to generate an external 
> simple 2D mesh using gmsh; this seems  one nice external fe grid 
> generator as indicated by specfem2d user manual (version 7, page 10).
>
> My question is about Jacobian as returned by the  script 
> xcheck_quality_external_mesh.
>
> I built using gmsh a simple geometry file composed of three uniform 
> layers (see the attached  *geo file).
>
> I performed a 2D mesh using allquads following the example shown in 
> specfem2D manual  (fixing a minimum element size of 20 because 
> dimension of my model is  relatively big).
>
> gmsh 3UFscale_allquad.geo -2 -clmin 20 -order 2
>
> I set the absorbing conditions with the python script
> python ../LibGmsh2Specfem.py 3UFscale_allquad -t A -b A -r A -l A
>
> obtaining the Mesh* Material* Nodes* Surf_abs_* and Surf_free* files.
>
> Now my problem is in the quality of my mesh as returned by 
> xcheck_quality_external_mesh because it returns negative jacobian.
>
> xcheck_quality_external_mesh
> :::::::
> start reading the external node file: Nodes_3UFscale_allquad
> start reading the external mesh file: Mesh_3UFscale_allquad
> NPOIN =        12663
> NSPEC =        12462
> done reading the external files
>
> start checking if any element with a negative Jacobian is found
> element         6251  has a negative Jacobian
> STOP negative Jacobian found
> ::::::
>
> I do not know how can I improve my mesh to pass this check; I try to 
> delete the element 6251  with negative Jacobian but this is not the 
> only one. I tried different  options in the 2D meshing, but still 
> obtaining negative jacobian.
>
> And it is also strange that the  Jmin seems 0.497 when I use 
> Tool/plugins og gmsh to analyseCurveMesh (see picture in attach). If I 
> understood well from gmsh manual (pa. 66) Jmin should be the minimum 
> Jacobian and this seems not negative.
>
> I also attached the output files output_mesher.txt and 
> output_solver.txt obtained by xmeshfem2D and xspecfem2D, respectively.
> Of course xspecfem2D crashs, and all elements with negative Jacobian 
> are reported into output_solver.txt.
>
> So my question is: how can I improve my external 2D mesh built by gmsh 
> for having positive jacobian everywhere  ?
>
> If someone of developpers has any suggestions to improve my mesh 
> deleting elements with negative jacobian in my  mesh, these are 
> welcome!!!
>
> The file *tar.gz contains my trial of mesh
>
> Best regards and thank you in advance!
>
> Giuseppe Di Giulio
> ----------------------------------------------------------------
>
> Il contenuto di questa e-mail e' rivolto unicamente alle persone cui 
> e' indirizzato, e puo'
> contenere informazioni la cui riservatezza e' tutelata.
> E' proibita la copia, la divulgazione o l'uso di questo messaggio o 
> dell'informazione ivi
> contenuta da chiunque altro che non sia il destinatario indicato.
> Se avete ricevuto questa e-mail per errore, vogliate cortesemente
> comunicarlo immediatamente per telefono, fax o e-mail. Grazie.
>
> This e-mail is intended only for person or entity to which is 
> addressed and may contain
> information that is privileged, confidential or otherwise protected 
> from disclosure.
> Copying, dissemination or use of this e-mail or the information herein
> by anyone other than the intended recipient is prohibited.
> If you have received this
> e-mail by mistake, please notify us immediately by telephone, fax or 
> e-mail.
> Thank you.
>
>
>
> _______________________________________________
> CIG-SEISMO mailing list
> CIG-SEISMO at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-seismo


-- 
*--------------------------------------------------------------------------*
  Paul CRISTINI  Charge de Recherche CNRS / CNRS Research Scientist
  --------
  CNRS - Laboratoire de Mecanique et d'Acoustique (UPR 7051)

  31 chemin Joseph Aiguier, F-13402 Marseille Cedex 20 - France
  mailto: cristini at lma.cnrs-mrs.fr
  phone number: +33 (0)4 91 16 42 89
  fax number: +33 (0)4 91 16 44 81
  http://www.lma.cnrs-mrs.fr/
*--------------------------------------------------------------------------*

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geodynamics.org/pipermail/cig-seismo/attachments/20120906/4bfb516d/attachment.htm