FEM: Meshing with triangles a plate (thin shell)

[celemin]

Hello,

When the mesh of any part is done using plates (thin shell), is there any problem of meshing the plates using triangles?

I have done some calculations getting higher Von Mises stress when the mesh is done with triangles than if it is done using squares?

I would like to know if the Von Mises stress obtained when triangles have been used in the mesh are real or are values too conservatives.

Thanks in advance,

Esteban

 

[tobbe]

Hi.

Using triangles ( I assume you use linear elements) would typical give a conservative value. These elements have a constant strain element formulation and are stiffer than a four-node ectangular element, and hence they should be avoided.
I might not be able to explain this but any textbook should clearify it.

T.

 

[brad]

One can't say that triangles are "conservative"--they are overly-stiff. Overly-stiff means "less correct", but depending on the specific problem and design targets, this can translate to conservative or non-conservative.
Tobbe is correct in that first-order triangles should be avoided in favor of quadrilateral elements when possible.
Brad