ccx + MFRONT

3rav3rav
Will the new version be able to use materials from the MFRONT library?

http://tfel.sourceforge.net/calculix.html

Regards
«1

Comments

  • VictorVictor
    My basic understanding of MFRONT is that it generates a dynamic library for a user defined material, which CCX links to. Is that right? I imagine you could already use it with Mecway by entering the material card in CCX -> Custom model definition and installing the dll (?) by hand.
  • 3rav3rav
    That's right.

    Defining material:
    *MATERIAL, NAME=@CALCULIXBEHAVIOUR_ELASTICITY@2
    *USER MATERIAL, CONSTANTS=2
    200000,0.3

    It would be nice if Mecway had such a library at the start with a few materials (Drucker-Prager, Johnson-Cook, Chaboche).

    Regards
  • FERNANDOFERNANDO
    I'm trying to apply the MFRONT library and I can't. I have copied the dll into the directory and enter the material type in CCX>, but it doesn't work. Has anyone managed to work with it?
  • disladisla
    How did you install the Library.?
    What's the error message?
  • FERNANDOFERNANDO
    The message es this:
    *ERROR: invalid material name '@CALCULIXBEHAVIOUR_RAMBERGOSGOODNONLINEARELASTICITY' (unable to load library)
  • FERNANDOFERNANDO
    It would be very interesting for geotechnics to have models such as Hardening soil model or Soft soil model, included in programs such as PLAXIS, and specific models for concrete such as those found in ABAQUS or ANSYS.
  • FERNANDOFERNANDO
    Other useful models would be cam clay or hoek braun for rocks.
  • disladisla
    Hi Fernando,

    Rafal.brzegowy will try to compile a new portable version by the end of this week.
    Let's see if that helps.

    Mean time you could try the old portable one. It works for me, at least for the most basic material models. Just put all those files inside your ccx 2.21 folder.

    https://calculix.discourse.group/t/calculix-mfront/205/6

    This liml file works for me.



  • FERNANDOFERNANDO
    I have tried, and It doesn't work. Can't load de library
  • disladisla
    edited January 11
    Have you set up ccx_dynamic.exe ver 2.21 as your ccx solver.?
    It should look like this.


  • FERNANDOFERNANDO
    This is the answer

  • FERNANDOFERNANDO
    It seems that ccx_dynamic.exe is not working
  • disladisla
    edited January 12
    Well, That's some progress. :)
    Could you push on the Details button and post to see what’s the problem.?

  • FERNANDOFERNANDO
    No output.
  • SergioSergio
    Just a newbie question, why the MFRONT library of materials (and others) is not directly integrated in the standard CCX distribution? There are a lot of post here and in the CCX forum on how to install/compile this feature
  • 3rav3rav
    @FERNANDO Since this is a dynamic version, do you have the mkl/PARDISO libraries connected?
  • FERNANDOFERNANDO
    I have a version compiled with mkl pardiso in 2.20. I have tried and it calculates but it stops on the first step.
  • FERNANDOFERNANDO

  • 3rav3rav
    Does your analysis work with regular material (not MFornt, but e.g. *ELASTIC)?
  • disladisla
    edited January 12
    Oficial ccx_dynamic.exe ver 2.21 as your ccx solver.
    http://www.dhondt.de/calculix_2.21_4win.zip


    @Sergio
    Most of the available material models I have tested doesn't or provide nonsense results. Each material behavior .mfront need the user revision and a deep study of how it works and how to define the constants. I think there are many people reluctant to include Mfront material libraries into the solver as it opens the door to many questions and bug reports that would not be directly related to the solver.

    MFRONT is sold as El DORADO and it could, but I think it's better to check and compile the behavior one by one and only if you are sure what it does.

    Some of them have different unit system or constants already predefined inside for test purposes. Some call constants with different names . You may think you are giving a constant parameter, and it is completely override. Others deliver displacements but not stresses. ….Would be Pandora’s box. Calculix allow you to call MFRONT but another thing is to include their material behaviors.

  • 3rav3rav
    It is possible that due to the lack of these: librarieslibgcc_s_seh-1.dll, libwinpthread-1.dll, libstdc++-6.dll, MFront with the Mecway or PrePoMax could not work (bconverged has these libraries in the bin folder).

    https://calculix.discourse.group/t/calculix-mfront/205/43?u=rafal.brzegowy
  • FERNANDOFERNANDO
    @3rav, the program works with internal materials.
  • FERNANDOFERNANDO
    with the compiled version ccx_MKL.exe works but does not converge
  • FERNANDOFERNANDO
    I have finally managed to get the Drucker Prager model to work
  • FERNANDOFERNANDO
    Although the program appears to calculate with the MFRONT models, the results do not seem consistent. It would be necessary to analyze them well with simple cases.
  • disladisla
    Great,
    Are you using the old portable version of MFRONT or the new published?
  • FERNANDOFERNANDO
    The old one. Where can I download the new one?
  • FERNANDOFERNANDO
    I already have it from the calculix forum. Thank you
  • FERNANDOFERNANDO
    edited January 15
    Has anyone done checks with the results of any material from the MFRONT library? Testing with the Druker Prager model, the results do not seem reliable.
  • 3rav3rav
    @FERNANDO

    If you can, compare ccx_2.20+MFront vs ccx_2.21+MFront with Drucker Prager
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