Materials

Materials

CONTACT_TIEBREAK_NODES_ONLY

    • Agathe Leclerc
      Subscriber

      Hello!

      I am studying the influence of the meshing on a model using CONTACT_TIEBREAK_NODES_ONLY, starting from a model I found on LS-Dyna examples. When refining the mesh, it appears that I don't manage to still having the same surface of contact. The set of elements on one side of the contact represent the same area but have more elements than with the initial mesh. And the set of nodes considers all the nodes above that mush be in contact at the beginning before failure. When I launch the simulation, only one random node is connected to the other surface. I also tried to create 4 different areas of contact, but still one node gets connected.

      It is possible to have several nodes connected to several elements with this fonction? If yes, how can I do it?  Thank you!!

      (The second picture is the initial area of contact with the first meshing and the first picture is the same area with more elements with a refined mesh, where I want the nodes above to be connectef to have the same surface contact.)

    • Reno Genest
      Ansys Employee

      Hello Agathe,

      Have you tried adjusting SAST and SBST to negative values to be larger than the gap between the elements in contact? Also, have you tried increasing MAXPAR?

       

      Reno.

      • Agathe Leclerc
        Subscriber

        Dear Reno, 

        Thank you for your answer, however I tried but it doesn't seem to work better. I connected nodes to segments. In your case you created segment in both sides. Do you think it could have an influence? I might try to do this.

    • Reno Genest
      Ansys Employee

      Hello Agathe,

      I tried with a node set for SURFA and it works the same:

       

      In the image above, I have the 4 green nodes in the node set for SURFA. How many nodes do you have in the node set for SURFA? How many nodes are tied and how many are not tied? Could you show the results of the Model Checking:

      Could you show the elements in your model and post a picture (similar to the pictures I post)? I am not sure I fully see the problems.

       

      Thanks.

       

      Reno.

      • Agathe Leclerc
        Subscriber

        Thank you very much for your answer. However I found my mistake. The normals of my segments were not in the good direction so the contact wasn't made properly.

        Thank you for your time!

    • Reno Genest
      Ansys Employee

      Hello Agathe,

      Good catch! You may want to consider the newer AUTOMATIC contacts which do not depend on the normal direction (among other things):

      Here is more information on contacts:

      https://ftp.lstc.com/anonymous/outgoing/support/FAQ_docs/contact.pdf

      https://ftp.lstc.com/anonymous/outgoing/support/FAQ/contact.overview

       

      Reno.

       

       

Viewing 3 reply threads
  • You must be logged in to reply to this topic.