I am solving a tower composed of sections with different thicknesses (I am using S8R elements because the S4R didn’t work properly before).
I would like to determine the plastic collapse of this tower, assuming an ideal elasto-plastic material law (with a very small strain hardening). I cannot use displacement control since I have the actual loads and moments acting in the structure.
I have not been able to approach the plastic collapse (which is a reference value I already know) and simulations stops prematurely. I have checked the von-Mises stresses and they have not reached the yield strength
Thank you for your answer. Then, do you have any suggestion to try to improve this problem to reach the plastic collapse or at least to reach the yielding?
Yes I am. I am using an elasto-plastic material law with no strain hardening (just an slope of E/1000 for convergence)
In fact, I have already solved the problem and I have been able to reproduce the plastic collapse of the tower.
I have solved it eliminating the rigid body in the bottom edge (following the comment of xyont and other posts in the forum) and directly fixing the displacement of the nodes at the bottom edge.
i try simple case using linear solid element (incompatible) and rigid body feature, it seems to be capable in applying displacement and rotation boundary condition simultaneously.
alternatively, moment loads can be represented by eccentric position of reference point in rigid body. An offset value is ratio of moment and axial loads, collapse capacity seen at peak graph reported by section print features of surface edge at base support.
it is known that the incompatible elements with with “local” displacement or strain modes suffer from instabilities that can occur in case they are loaded in compression. The S4 in CalculiX which is derived from the C3D8I has this weakness. See, for example,
Theodore Sussman, Klaus-Jürgen Bathe, “Spurious modes in geometrically nonlinear small displacement finite elements with incompatible modes”, Computers and Structures 140 (2014) 14–22
or
Peter Wriggers, “Nichtlineare Finite-Element-Methoden”, Springer 2001, Sect. 10.4.4
I have on example with an axially compressed cylinder (elastic) that also shows some instable behavior when the S8R elements are used, obviously when the mesh is too coarse.
hi, thanks for notice and citing a paper. I forgot to mention my models using truly solid not shell element, linear hexahedral incompatible element is widely in use by much research of steel structures and still given reliable result in faster time to convergences. It’s recommended a fine mesh being used to eliminate curvature of individual element during large deformation for both surface dimension and thickness, so a volumetric change did not occur significantly.
. Nice That’s good to ear…
