I read some argument from HKS Abaqus document, it’s related to suitability using tetrahedral and wedge element for contact analysis. However, there’s no further test or verification cited.
Second-order tetrahedra are not suitable for the analysis of contact problems: a constant pressure on an element face produces zero equivalent loads at the corner nodes. In contact problems this makes the contact condition at the corners indeterminate, with failure of the solution likely because of excessive gap chatter. The same argument holds true for contact on triangular faces of a wedge element.
Above limitation did not notify in CalculiX documentation, so is there any comment and suggestion?
It’s more about the second order but it also depends on contact formulation. For node-to-surface contact, it’s advised to avoid second-order tetrahedrons since they converge poorly and cause contact pressure noise. Modified versions of those elements are recommended instead. Of course, first-order tetrahedrons and wedges are bad in general, they should be used only as filler elements.
Thanks for commenting, as i was thinking and guess before it’s specific for node to surface contacts. Not applicable for surface to surface contact. Abaqus documentation in element sections may lead to a user doubt since it’s not clear and notify an exception.
When subjected to uniform pressure, quadratic element have some amount of negative value (tension) at corner, but Abaqus notify as zero values. That’s lead me some doubt, also document placed at element section without except.
CalculiX seems done properly, it did not give any notify in element sections document since it’s contact type dependent. More clear at contact sections document by explanation, also improved the convergency by small amount of positive values (compression).