My Simulation 11.0 model of a column, comprised of BEAM188 elements with an open cross section (an I-beam shape in this case), gives the wrong linear buckling results. One end is pinned and the other is held in the two non-axial translational DOFs. At this latter end, a compressive axial force is used to impart a static preload. All of the nodes are held in the rotational DOF about the axis of the beam (ROTX). The linear buckling results do not match the prestressed modal results. Is this a known limitation?


Yes. Feature Request 63056 has been filed to address this limitation in Simulation. Essentially, Simulation 11.0 uses "ETCONTROL,SET", which activates internal nodes that are not restrained in the ROTX DOF. These unrestrained DOFs give one bad torsional buckling result per element. By de-activating Element Control with a Command Object that issues "ETCONTROL,SUGGEST", these internal nodes are not created and the results are OK. Another workaround is to turn on the warping DOF for the BEAM188 nodes. The aforementioned Feature Request 63056 should result in Simulation's setting this automatically (i.e., KEYOPT(1)=1) when an open section is detected for the BEAM188s.

A "corrected" Simulation 11.0 SP1 model is attached to demonstrate the issue.





Show Form
No comments yet. Be the first to add a comment!