SHELL181 element size vs. thickness

[Rodrigo Benitez]

Hi guys,

I'm working with plates subjeted to concentrated pressure perpendicular to plane forces, using thick shell finite elements, in other words, mindlin-reissner SHELL181 elements. The plate is simply supported on all 4 sides, its 2.450 x 2.450 mm2 by 70 mm thickness. The concentrated pressures are applied on 150 mm x 150 mm subareas.

My question is: is it "wrong" to be using 30 mm x 30 mm element size? Is there some kind of L/t limit to this kind of finite element?

I did a mesh refinement convergence study with ESIZE = 150 mm, 75 mm, 50 mm and then 30 mm, with which I obtained the following respective max. stresses (MPa): 19,04 , 22,93, 23,77 and 24,55.

Thanks in advance.

Rodrigo

 

[STpipe]

There is nothing inherently wrong with using a finer mesh. A finer mesh typically means more accurate stress results. It's all about a balance between the level of accuracy you require in your results and the amount of run time for the solver to obtain a solution. Based on your posted results, you seem to have reached reasonable mesh convergence with the 75 mm mesh size, if stress is the governing limit state you're designing against (as opposed to shell forces and moments).

Whether the element you're using is suitable for this application would be independent of mesh size I would think and would be purely based on the element formulation itself. From a quick browsing of the SHELL181 element formulation, it seems it would be suitable given the size of the actual plate.

 

[rickfischer51]

The documentation for shell 181 says thin to moderately thick. Who knows what that means. Try modeling it with quadratic bricks and see how they compare

Rick Fischer
Principal Engineer
Argonne National Laboratory