General Mechanical

General Mechanical

Convergence of nearly incompressible material – non-linear analysis

    • dzz0
      Subscriber

      Hi everyone, 


      I have a problem with the simulation of nearly incompressible material (v=.495) and i hope someone may help me with this. I have searched older topics but could not find a suitable solution.


      Specifically, this involves the simulation of a radially symmetrical silicone bond between to adherends under tension loading. For this purpose I use a USERMAT based on the Ogden model. Large deflections are activated (NLGEOM,ON).


      I want to increase the maximum displacement i can apply. I already have tested many things:


      - other elements (185,186,187,285)


      - different keyopts


      - finer mesh, which actually makes it worse somehow


      - more substeps and cumulative iterations


      - different type of symmetry and modelling adherends (flexible and rigid)


      - 2D model


       


      Does someone have another idea to increase the convergence?

      Thanks in advance!


       


      Kind regards


      Ben


       


       



       

    • peteroznewman
      Subscriber

      I recommend using/continue using:



      • Hyperelastic material model

      • 2D Axisymmetric geometry. Axis of rotation is Y axis, surface on + X side of Y axis.

      • Half symmetry at the center of the length

      • Displacement in Y on the end face, with X = 0

      • Linear elements, but not too small. PLANE182

      • Mesh with element shapes that deform to better shapes, not worse shapes. Start as wide rectangles, end up square.

      • uniform reduced integration keyopt, 1, 1

      • mixed u-p formulation keyopt, 6, 1

      • more substeps and cumulative iterations

    • dzz0
      Subscriber

      Hello Peter, 

      thank you very much for your tips. I will try this. 

      Regards
      Ben

    • peteroznewman
      Subscriber

      Ben,


      Note that with a half symmetry condition at the center, a 1 mm displacement of the end is equal to a 2 mm displacement of the full length sample.


      Regards,
      Peter

    • dzz0
      Subscriber

      Hello Peter, 

      thank you for the advice. I will consider this in the evaluation. 

      Regards,
      Ben

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