General Mechanical

General Mechanical

High stress difference in between two adjacent layers in a composite tube

    • rahmanmahfuz11
      Subscriber

      Hello all, I am modeling a carbon fiber composite tube with a 4mm HDPE layer with a helical hole of 1.5mm diameter inside. The following figure shows the HDPE layer with the hole.

    • peteroznewman
      Subscriber
      I think the high stress is an artifact of the non-congruent mesh. Look at the large overlap of the one mesh with the other.
      Discard the bonded contact and use shared topology to obtain a congruent mesh so that there are no nodes that overlap into the adjacent material and I expect the stress will not be so large.
    • rahmanmahfuz11
      Subscriber
      Thank you for the reply. The bonded contact is in between the composite layers & thermoplastic solid. The composite layers are coming from ACP and the solid thermoplastic layer is coming from Mechanical Model like the picture attached below. Can I use a shared topology between two different models? If not, then how else can I overcome this issue? thank you!

    • peteroznewman
      Subscriber
      I don't know if you can use shared topology with ACP shell and normal solid model mesh. I suppose if there was a common geometry model that shared topology, and the face was meshed in ACP and the solid was meshed in normal model it might work.
      If not, you can always slice the common geometry and imprint features onto each geometry set so that mesh controls will allow a congruent mesh to be made in both models, and a node merge operation issued instead of using bonded contact.
      If you can't manage to eliminate bonded contact, does ACP permit quadratic elements? Using quadratic elements in both systems would reduce the amount of mesh penetration. If you must use linear elements, making the elements smaller will reduce the penetration.
    • rahmanmahfuz11
      Subscriber
      I have used the quadratic elements. They did reduce the penetration of mesh, thank you. But with quadratic elements, I am getting more stress than linear elements. The results in the original post are faulty for another reason. I made a mistake in the surface selection for loading. These are the original stress results with linear elements:

      And these are with quad.


Viewing 4 reply threads
  • You must be logged in to reply to this topic.