In the course of preparation, I would like to found out the options of beam calculations referring calculix. I would like to calculate the following. (with U1 userbeams)
A static analysis of timoshenko beam considering the sehar deformations (open section)
a linear buckling analysis of the system
The calculation of the plastic capacity load (limit load)
If I am properly informed its not possible to to a geometric nonlinear calculation with U1 elemts?
I’m doing a lot of beam calculation. but my calculations are only with solid elements (maybe also a bit with shell elements).
buckling is working with DC if you need torsional and lateral torsional buckling.
for all other calculation i use RB.
DC = Distribution Coupling (soft boundary)
RB = Rigid Body
if you like an example i can share an example beam.
I’m very interested in timo …
you guess quite right I am investigating stability problems in structural mechanic like torsional buckling.
I’m currently trying to do a comparative calculation. I have attached the task and the target results.
The auhor did a linear buckling analysis with B31OS (Abaqus, timoshenko, shear deformation) elements. To be able to compare it to Abaqus, I wanted to try it out with calculix. In code_aster there are possibilities of this simulation of this kind but I would prefer calculix
the first thing is:
in calculix it Is only possible to have a connection
with 6 dofs or only with translation (1-3) or rotation (4-6) (*COUPLING with RB)
anyway i did long time ago a similar example.
at first i would try with a very simple example (Einfeldträger)
if the buckling modes you need are available !? with beam elements
lateral torsional buckling = BDK
(drillen = torsional buckling) you don 't have these in your system.
i think lateral buckling (biegeknicken) is working !
maybe you can do the calc. for internal forces and calculix for ESV (Ersatzstabverfahren)
or the hole system with solids.
regarding capacity of global structural system (portal frames), i will consider minimum requirement for these task. most report/codes recommendations are using nonlinear material (plasticity) and geometric (large deformation) analysis including imperfection along it’s members (bowing) and trough all storey levels (eccentricity).
even these are advanced analysis, seems all results can not generalize to be fully trusted. cause of beam element used may not consider slenderness sections effect due to local buckling (web/flanges) - note: need to study further.
another uncertainties are in residual stress, local imperfections and degree of rotational restraint at beam column connections.
for a simple 2D portal frames, in my opinions: shell or solid element are best selections since more detailed analysis requirement could be performed. structured mesh also can be benefit if residual stress need to be consider.
apologize of my quick response and may too advanced. cause of thread names called capacity,
looking up this results, are Abaqus can do torsional buckling analysis of beam element? seems only lateral buckling is shown and first two modes arise for out of plane direction (minor axes).
abaqus and code aster have beam elements with 7 dofs.
one of them for warping.
i took a simple test case quickly, for beam element it can detect torsional buckling conditions also as shown in results.
compares to solid & shell element (quadratic), and refined mesh of beam element with B32 (without reduced integration). all yields nearly the same results.
many thanks for the comments and suggestions. I will continue my research and share my results and insights
In my understanding, calculix both beam and shell elements are not based on real formulations, but rather expanded as 3D elements during the solution. Please, correct me @dhondt if I’m wrong. Anyways, I would be interested to know if it is possible to account for transverse shear deformation terms and avoid shear locking.
yes, this is correct. The best results I get are with the C3D20R-elements,
which are the expansion of the S8R (or CPE8R or CPS8R or CAX8) element. By the
way, S4 elements are expanded into C3D8I elements, which are better regarding
shear locking than C3D8 elements. Just try it.
Dear @lolito ,
You can try the following test (Figure attached). This can be done either with beam or shell elements.
This cantilevered “beam” is subjected to a moment at one end while it is fixed at the other end.
This is slender, of course, say 100 elements along the longest direction while maybe one element in thickness. The curvature of the beam should follow 1/R = M/EI where R is the radius of curvature, E the young’s modulus and I the area moment of inertia. Would love to see what are the capabilities for doing such a nonlinear calculation.
Any recommendation from Guido @dhondt would be very welcome.