NON-LINEAR BEHAVIOUR OF SHELLS

JJJJJJ
Hello collages. I have done a simple test, where I have modeled a square hollow section 100x3, submited to a uniformally load on the top. I have checked results of this 2m simple beam with classical formulas, and the results of the internal solver of Mecway are very close. Calculix, I think because the methodology to expand from shells to solid, has more than 60% diference.
Is there intention to develop nonlinear capabilities for the internal solver? I want to evaluate stainless steel material analysis, for example, or non-linear buckling of shells.

Thank you for your time.

JJ

Comments

  • VictorVictor
    Hello JJJ

    Nonlinear shells for the internal solver are not a priority, sorry.

    I'm surprised at that 60% error. You might need more refinement with the CCX shells. Also make sure the mesh is quad-dominant, not linear triangles.

    You can also use the OpenRadioss solver which has more traditional nonlinear shells. Use damping to obtain a static solution.
  • JJJJJJ
    Hi Victor,

    Thank you very much for your reply, and of course sincerity. Bad news from you about that... I need to work with stainless in some projects, and as you know the Ramberg-Osgood model is the material law to be considered.
    Aditional options could be:
    1- Posibility about Opensees as solver https://opensees.berkeley.edu/
    2- Posibility about OOFEM as solver https://www.oofem.org/doku.php?id=en:oofem
    What do you think about these options? Would be very nice.

    I have used the same mesh for both softwares, and it is quad-dominant. I am not confident with this procedure
  • JJJJJJ
    Hi Victor, could you see this question I did? On the other hand, CODE-ASTER is a recognised software in industry. I don´t know how it works, but are you thinking that Mecway could be a pre and post processor for that?
  • VictorVictor
    Yes, if I add another solver, I think it would probably be Code Aster. Not sure about Opensees or OOFEM because I haven't heard about them very much.

    Using the same mesh could be the problem and you might need to refine it for CCX.
  • MishalMishal
    @JJJ I have used Opensees, it is pretty easy to use and the manual is very detailed. I came up with python scripts which automate model development. The only issue is that the python packages gives some issues so I have opted to use the exe where data is directly fed into the exe file
  • JJJJJJ
    @Victor
    Very good news, to know that you are thinking about to include Code Aster as an option. An industry recognised software, be possible one of the most (if not the most) on the Open Source world.
    OOFEM is a software implemented in NextFEM.
    OpenSees I think would be a very interesting option: it is very recognised for seismic analysis. Today is implemented by different pre and post processors (NextFEM and STKO), but also in commercial software (Cype).
    And what about Nastran: this is also a very recognised software in industry. After your implementation of Mystran, I think won´t be very difficult to do that
    @Mishal yes you are right, documentation is very good. But graphic interface is always wellc@Victoroi
  • FERNANDOFERNANDO
    edited July 15
    Para un ingeniero civil una opción magnífica sería OPENSEES.
  • JJJJJJ
    No sólo para un ingeniero civil. Fíjate Cypeconnect... se amplían muchísimo las opciones disponibles.
  • polymathicpolymathic
    Yes, if I add another solver, I think it would probably be Code Aster


    After several attempts over the years to get a handle on Code Aster, with great success lately (thanks to the help of LLMs to get through the language barrier and near-vertical learning curve), I can say the juice is well and truly worth the squeeze. Brilliant solver.
  • sofien_73sofien_73
    Code Aster should be a great solver, Victor. Is it planed for next update of MECWAY???
  • KundertdeWitKundertdeWit
    edited October 10
    Hi all,

    I have been long time looking for a nice GUI for Code-Aster. I have been using Salome-Meca, but always find it too tedious to use. In the first half of this year I established the working principle of a bridge between Mecway as pre-/post processor and Code-Aster as a solver. That structure proved to work. After that I didn’t have time to spent on it until last week. My son is programming a lot with Claude Code, and this week he didn’t have time to use it because of his homework 😊. So I was allowed to spent his weekly limit. Unfortunately, it worked out quite well and I decided to unlock more of Code-Aster than I normally use. As a result, the MecwayAsterBridge grew quiet large in capabilities. You can find it here: https://github.com/trednuK/MecwayAsterBridge

    I have added a lot of validation checks and documentation. However, since it is build by AI, it still needs human verification to find edge cases which only appear in daily use.

    Enjoy using it and please let me know if you have any suggestions or find some bugs.

    Kind regards,
    Kundert
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In Register