Contact Analysis: Tet vs. Brick Elements 2

[score33]

Some coworkers insist that contact analyses between solids should be brick on brick...

why?

I can't find any good discussion on this and am not sure why they insist on it.

Any thoughts or references where I might learn more?

Thanks for the help.

 

[GBor]

What software and what contact algorithm? Some software uses a nodal contact algorithm that relies on the proximity of one node to a predefined contact node. Other softwares use a surface designation and check to see if a node has penetrated the surface definition. There are various forms of these and other implementations as well.

For the first of the two algorithms mentioned above, it is nice and easy to defined contact if it is clearer when the two objects will impact and, in that case, I can see brick on brick being convenient. For the second of the options, element type isn't terribly meaningful.

I'm sure other implementations rely on other definitions.

 

[score33]

For my contact analyses I typically use IDEAS, but we also have abaqus. I'll go see what I can find in the documentation for each.

thanks for the help GBor. I guess I should have started there...

 

[gwolf2]

Look at the discussions on the C3D10 and C3D10M elements in the ABAQUS manual, it spells it out. The best contact in abaqus is usually to be had with C3D8I elements.

 

[score33]

thanks gwolf, I'll go check that out.