Calculix results in paraview
- Sathish
- Topic Author
- Offline
- New Member
-
Less
More
- Posts: 4
- Thank you received: 0
11 years 3 months ago #7362
by Sathish
Calculix results in paraview was created by Sathish
Dear members
I have an interest in Finite Element Analysis. I have been using ABAQUS FE code for last 6 years. Lately, I am showing more interest in open-source software. In particular interested in CALCULIX due to its power and ITS input deck similarities with ABAQUS. However, I am a windows user all my life. Recently, I started learning Linux but I am really long way off to become a good linux programmer. I saw the power of PPA. With support a techie I managed to install CALCULIX and viewed simulation results in CGX. Before going in to further discussion I request you to view this link on how to modify CALCULIX script so as to get output in EXODUSII format
homepages.wmich.edu/~pjm8969/research/ccx_extras-dl.html
exo.fiziko.de/src/inst.html
I have lately been using PARAVIEW (for last 3 months) to view output from OPENFOAM cases. I must tell you that it is very powerful. Traditionally, people are using CGX as post-processor for CALCULIX. Although CGX is good, Paraview is much more powerful. As you are aware that paraview comes with a python console and has many advanced features. I love to use Paraview as a standard post-processor as opposed to CGX. However, my main problem is that I am not a software programmer. My understanding of Linux is limited. Therefore, I was unable to successfully carryout the installation as per the notes given in the link. In this respect, I need your support. Any help on your end is of great value to us.
A Calculix forum member (from enggprog.com) has provided a shell script that can automatically install CALCULIX on Linux. I have attached this script (document named (CXBUILD). you can down load from link
sdrv.ms/1d3NOko
However, this script is for Installation CALCULIX and its post-processor CGX. Ideally, I would like a similar script to install EXODUSII and calculix extras presented in the link, so as to view output in paraview.
I humbly request members to consider building a PPA or just a shell script. This helps many novice to use CALCULIX and paraview. Considering that you are good programmers, I would assume that it will not take much effort on your end, should you choose to build.
I look forward for your reply. If I can be of any help to you do let me know
Kind Regards
Sathish
I have an interest in Finite Element Analysis. I have been using ABAQUS FE code for last 6 years. Lately, I am showing more interest in open-source software. In particular interested in CALCULIX due to its power and ITS input deck similarities with ABAQUS. However, I am a windows user all my life. Recently, I started learning Linux but I am really long way off to become a good linux programmer. I saw the power of PPA. With support a techie I managed to install CALCULIX and viewed simulation results in CGX. Before going in to further discussion I request you to view this link on how to modify CALCULIX script so as to get output in EXODUSII format
homepages.wmich.edu/~pjm8969/research/ccx_extras-dl.html
exo.fiziko.de/src/inst.html
I have lately been using PARAVIEW (for last 3 months) to view output from OPENFOAM cases. I must tell you that it is very powerful. Traditionally, people are using CGX as post-processor for CALCULIX. Although CGX is good, Paraview is much more powerful. As you are aware that paraview comes with a python console and has many advanced features. I love to use Paraview as a standard post-processor as opposed to CGX. However, my main problem is that I am not a software programmer. My understanding of Linux is limited. Therefore, I was unable to successfully carryout the installation as per the notes given in the link. In this respect, I need your support. Any help on your end is of great value to us.
A Calculix forum member (from enggprog.com) has provided a shell script that can automatically install CALCULIX on Linux. I have attached this script (document named (CXBUILD). you can down load from link
sdrv.ms/1d3NOko
However, this script is for Installation CALCULIX and its post-processor CGX. Ideally, I would like a similar script to install EXODUSII and calculix extras presented in the link, so as to view output in paraview.
I humbly request members to consider building a PPA or just a shell script. This helps many novice to use CALCULIX and paraview. Considering that you are good programmers, I would assume that it will not take much effort on your end, should you choose to build.
I look forward for your reply. If I can be of any help to you do let me know
Kind Regards
Sathish
- Joël Cugnoni
-
- Offline
- Moderator
-
11 years 3 months ago - 11 years 3 months ago #7371
by Joël Cugnoni
Joël Cugnoni - a.k.a admin
www.caelinux.com
Replied by Joël Cugnoni on topic Re: Calculix results in paraview
Hello,
I have quickly built a package with Calculix 2.6.1 with the patch for ExodusII file output. This version is using the multithreaded SPOOLES solvers only (no patch for Cuda or Cholmod solvers) but it adds the functionnality to do direct post processing in Paraview, great!
The package is called calculix-2.6.1-mt-caelinux , file calculix-2.6.1-mt-caelinux-0.1-cae2013-amd64.deb
and it is available in CAELinux 2013 repository at sourceforge.net/projects/caelinux/files/dists/
I have also programmed a Calculix Wizard that takes a UNV mesh generated in Salome, convert it into CCX Inp format (with groups) and guides the user in the editing/pre/post processing of the analysis. This wizard is part of the caelinux-addons packages.
To use this wizard, the model has to be prepared in Salome with all necessary groups of nodes for boundary conditions / contact conditions and group of volume elements for material attribution. Also, to avoid having a lot of "shells" in Calculix, all the unnecessary face elements have to be included in one or several groups with a name starting with "X_", for example "X_faces". All these faces will be ignored in the conversion of the model, the others will be converted to shell elements.
To get on overview of what can be done, here is a snapshot of the Wizard and an example.
(seems that uploads are failing... must be a server problem...) anyway, here are the links:
www.dropbox.com/s/z48n4h5ohgv5r9s/CalculixWizard1.png
www.dropbox.com/s/vfyf2zqwnhpxqdb/CalculixWizard2.png
I have quickly built a package with Calculix 2.6.1 with the patch for ExodusII file output. This version is using the multithreaded SPOOLES solvers only (no patch for Cuda or Cholmod solvers) but it adds the functionnality to do direct post processing in Paraview, great!
The package is called calculix-2.6.1-mt-caelinux , file calculix-2.6.1-mt-caelinux-0.1-cae2013-amd64.deb
and it is available in CAELinux 2013 repository at sourceforge.net/projects/caelinux/files/dists/
I have also programmed a Calculix Wizard that takes a UNV mesh generated in Salome, convert it into CCX Inp format (with groups) and guides the user in the editing/pre/post processing of the analysis. This wizard is part of the caelinux-addons packages.
To use this wizard, the model has to be prepared in Salome with all necessary groups of nodes for boundary conditions / contact conditions and group of volume elements for material attribution. Also, to avoid having a lot of "shells" in Calculix, all the unnecessary face elements have to be included in one or several groups with a name starting with "X_", for example "X_faces". All these faces will be ignored in the conversion of the model, the others will be converted to shell elements.
To get on overview of what can be done, here is a snapshot of the Wizard and an example.
(seems that uploads are failing... must be a server problem...) anyway, here are the links:
www.dropbox.com/s/z48n4h5ohgv5r9s/CalculixWizard1.png
www.dropbox.com/s/vfyf2zqwnhpxqdb/CalculixWizard2.png
Joël Cugnoni - a.k.a admin
www.caelinux.com
Last edit: 11 years 3 months ago by Joël Cugnoni.
The following user(s) said Thank You: Sathish
- Sathish
- Topic Author
- Offline
- New Member
-
Less
More
- Posts: 4
- Thank you received: 0
11 years 3 months ago #7372
by Sathish
Replied by Sathish on topic Re: Calculix results in paraview
Dear Joel Cugnoni
It is an excellent news. You are brilliant. Well done.
It is an excellent news. You are brilliant. Well done.
- Marco Grassi
-
- Offline
- New Member
-
Less
More
- Posts: 4
- Thank you received: 0
11 years 3 months ago #7386
by Marco Grassi
Replied by Marco Grassi on topic Re: Calculix results in paraview
You are great!
As we say in Italy: "Ti devo una pizza (I owe you a pizza)!"
We also say "your appetite grows while you're eating"..so....
I don't know if I have to write this here somewhere else, but:
It would be useful to be able to easily create the *SURFACE card for calculix starting from the element groups contained in the .unv file.
In particular I'm thinking of some face of solid elements in contact (for *CONCTACT card).
If I have a .unv file containing the solid elements and the shell elements laying on the surface of the solid (already ordered in groups) it would be easy for you to create a script that read the shell elements, find the the solid which lays under the shell and evaluate the face of the solid the be put in the *SURFACE card (S1 or S2 ....S6).
For example:
In the .unv file You have 4 groups: 01_Body_A_3d_el, 02_Body_A_2d_el_for_conctact, 03_Body_B_3d_el, 04_Body_B_2d_el_for_conctact.
Read group 02, find the related nodes; from group 02 select the element related to the found nodes; find the solid faces.
Same for group 03 and 04.
Maybe it can be done only specifing the groups of shells and reading the other information on the whole model .unv.
Do you believe that it is possible?
Thanks again
Marco
As we say in Italy: "Ti devo una pizza (I owe you a pizza)!"
We also say "your appetite grows while you're eating"..so....
I don't know if I have to write this here somewhere else, but:
It would be useful to be able to easily create the *SURFACE card for calculix starting from the element groups contained in the .unv file.
In particular I'm thinking of some face of solid elements in contact (for *CONCTACT card).
If I have a .unv file containing the solid elements and the shell elements laying on the surface of the solid (already ordered in groups) it would be easy for you to create a script that read the shell elements, find the the solid which lays under the shell and evaluate the face of the solid the be put in the *SURFACE card (S1 or S2 ....S6).
For example:
In the .unv file You have 4 groups: 01_Body_A_3d_el, 02_Body_A_2d_el_for_conctact, 03_Body_B_3d_el, 04_Body_B_2d_el_for_conctact.
Read group 02, find the related nodes; from group 02 select the element related to the found nodes; find the solid faces.
Same for group 03 and 04.
Maybe it can be done only specifing the groups of shells and reading the other information on the whole model .unv.
Do you believe that it is possible?
Thanks again
Marco
- Joël Cugnoni
-
- Offline
- Moderator
-
11 years 3 months ago #7387
by Joël Cugnoni
Joël Cugnoni - a.k.a admin
www.caelinux.com
Replied by Joël Cugnoni on topic Re: Calculix results in paraview
Hi
Thanks, I really like pizza, so if I come once in Italy I may contact you
I had the idea to convert some set of surface elements into *Surface keywords, but did not had time to implement it yet. However, if you have a node set corresponding to the contact surfaces, you can import the converted INP file in CGX (run "cgx_2.6.1 -b" in a shell), and then use the following commands to generate surfaces:The surface will be written in yournodeset.sur and can be integrated in the INP manually afterwards.
Cheers
Joel
Thanks, I really like pizza, so if I come once in Italy I may contact you

I had the idea to convert some set of surface elements into *Surface keywords, but did not had time to implement it yet. However, if you have a node set corresponding to the contact surfaces, you can import the converted INP file in CGX (run "cgx_2.6.1 -b" in a shell), and then use the following commands to generate surfaces:
read yourfile.inp
comp yournodeset down
send yournodeset abq surf
Cheers
Joel
Joël Cugnoni - a.k.a admin
www.caelinux.com
- Marco Grassi
-
- Offline
- New Member
-
Less
More
- Posts: 4
- Thank you received: 0
11 years 3 months ago #7388
by Marco Grassi
Replied by Marco Grassi on topic Re: Calculix results in paraview
Dear Joel,
Thanks a lot for your sudden answer! I am sorry I bother You so much.
I'm going to follow the way that You explained to me: it is what I need.
I used your calculix wizard to convert .unv to .inp. A couple of not important notes:
- the 3 nodes shell elements have been converted to STRI35 type. Is calculix able to interpreter them?
- A *NODE FILE card is created with the request of U, S, and E but I believe the S and E are element results.
Please don't stop support CAELINUX project!!
Marco
P.S.: www.amozzarella.com/index2.html is waiting for you (and your family)....
Thanks a lot for your sudden answer! I am sorry I bother You so much.
I'm going to follow the way that You explained to me: it is what I need.
I used your calculix wizard to convert .unv to .inp. A couple of not important notes:
- the 3 nodes shell elements have been converted to STRI35 type. Is calculix able to interpreter them?
- A *NODE FILE card is created with the request of U, S, and E but I believe the S and E are element results.
Please don't stop support CAELINUX project!!
Marco
P.S.: www.amozzarella.com/index2.html is waiting for you (and your family)....
Moderators: catux
Time to create page: 0.127 seconds