Axisymmetric 2D analysis with elements on Y symmetry line

Greetings!

I am trying to run a simple 2D axisymmetric model. I am running into issues at location of axis-symmetric line (x=0), see picture below.

The end goal of this simulation was to simulate contact conditions between two soft hyper elastic materials (approx 1 and 10 MPa), but currently I am already running into issues with linear materials, NLGEOM=OFF.

Why is the 1st element always penetrating the harder surface, regardless of how I set the mesh? I assume it has something to do with CAX8R being expanded to a wedge element?

Thanks for helping out!

simple_2D.pmx (306.9 KB)

Quite likely because there’s no such issue when running your model in Abaqus:

You could try revolving this geometry into an equivalent narrow wedge, applying proper BCs and running with 3D elements to confirm this. Unfortunately, CalculiX’s 2D elements aren’t really robust.

Here’s a version with 3D elements:

Axisymmetric contact penetration - solid.pmx (1.8 MB)

What I’ve done is I just translated model for X=0.1 mm and I get this:

So it is indeed the wedge element that is causing issue. Physically, would it be still ok to minimally display the model for a small X and achieve OK results?

Perhaps even smaller translation would suffice. You can compare the results with the solid elements approach (e.g. from my file attached above) to make sure there’s no discrepancy.

If you plan to use contact only in the compression direction, you could try merging the nodes of the opposite parts on the axis.

Just tried that out of curiosity, but it leads to this:

Hm, so it is not only the problem of the center node, which was my impression.

Did you try both contact methods: surface-surface and node-surface?

It looks much better with:

  • first order elements (second order ones are known for having issues with contact anyway)
  • merged nodes on the axis
  • node to surface contact
  • small sliding option enabled

2 Likes