LS Dyna

LS Dyna

Simulating carboard with *MAT_PAPER. Guidance needed!

    • Sami Farag
      Subscriber

      Hello,

      I am new to Ls-dyna, and am working on my graduation project for making a model to simulate the creasing process of the folding line in corrugated cardboard. I am trying to work with the *MAT_PAPER for the layers of the cardboard, and I have some questions. I would really appreciate if I can get some guidance.

      I am not concerned with delamination between layers.

      Q1. What would be a suitable and neat connection method between the three layers? I have merged the duplicate nodes between each layer, but as the mesh is not fine enough, merging the duplicate nodes results in a bad connection geometry.

      For the contacts, I have defined the following contacts

      ·        *CONTACT_AUTOMATIC_SURFACE_TO_SURFACE between (Kraft, Lower), (Flute, Test), (Flute, Kraft), (Test, Upper)

      ·        *CONTACT_AUTOMATIC_SINGLE_SURFACE for the Flute

      Q2. Do I also need to define contact for (Kraft, Upper) , (Test, Lower), (Flute, Upper), (Flute, Lower)?

      Q3. Which one should be master and which one should be slave? Does it matter?

      Q4. Are these contact types good enough for my application or is there better contact types?

       

      Q5. If I to make the model of shell elements with thickness instead of solid elements. How should I account for the thicknesses? Should there be a spacing between each shell surface to account for the thickness? Or should they be modeled without spacing and touching each other?

      Q6. The material model *MAT_PAPER works only with explicit method, how can I do this process a quasistatically in explicit? 

      I have added a picture of the whole setup, a picture showing the merged nodes.

      I hope I formulated my questions clearly. I would really appreciate any help. :)

    • Andreas Koutras
      Ansys Employee

      Hello,

      Q1: You can try with *CONTACT_TIED_SURF_TO_SURF to tie the two parts, instead of merging nodes.

      Q2: Contact will be required to transfer contact force between any parts that are not connected through the mesh.

      Q3: In AUTOMATIC_SURFACE_TO_SURFACE contact the order of master and slave surfaces does not matter. In the SINGLE_SURFACE contact, all the parts coming into contact or self-contact are included in the slave side.

      Q4: Yes, those are the most commonly used sliding contacts. You can start with setting SOFT=1, which is a node-to-segment contact. SOFT=2 will activate the (more detailed but also more costly) segment-to-segment contact algorithm.

      Q5: AUTOMATIC contacts account for the actual shell thickness offset by default, therefore, the nodes of the shells can be placed along the mid-thickness of the shells.

      Q6: To achieve quasi-static loading in explicit, you will need to apply the loading with a slow enough rate. You can judge about the contribution of the dynamic effects by looking at the kinetic energy in the GLSTAT output file. I common rule of thumb to characterize the behavior as quasi-static is that the kinetic energy is not more than 1% of the internal energy for most of the analysis time.

Viewing 1 reply thread
  • You must be logged in to reply to this topic.