How to get rid of elements with very small, negative or zero dimension

[ladykin]

Hi There,

I'm trying to use a rigid indenter on a deformable body.
I have gotten an error with my job saying that :

2 elements have very small, negative or zero dimension/s. The elements have been identified in element set ErrElemDimSmallNegZero.

Does anyone have any idea how to get rid of elements with very small, negative or zero dimension?

 

[rstupplebeen]

Are these elements small to begin with or after compression with the indenter? Look at your ODB and there will be an element folder with the offending elements.

Basically the math breaks down with poor aspect ratios. I have inentionally skewed elements in the mesh generation to accommodate large deformation. Abaqus can also use adaptive remeshing to help with this. Some times simply increasing mesh size especially through the thickness helps.

If you post your model, pictures or provide a more detailed description we will be able to help you more.

I hope this helps.

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[mosgef]

If the indenter causes relatively small deformation, simply using *ADAPTIVE MESH, ELSET = your element set can help. I agree with Robert, that refining the mesh in the area of contact should help with initial element volumes. Further, if your simulation is dying somewhere during the run, you may be running into contact problems.
I have run into problems with convergence when indenting materials with very different moduli. In your case the rigid surface effectively has infinite stiffness, and may cause you problems.

 

[Aislingoh]

Hi all,

what, in your opinion, is a reasonable aspect ratio to apply to a highly deformable material in compression? I have seen this technique of using high aspect ratios before but haven't read any indication of how high! Some of my elements have an aspect ratio of 1000, is this completely crazy?

Aisling