Variable moments on pipe, from 2D to 3D

[homusubi]

Hello everybody!

I’m trying to check if the walls of the composite sandwich pipe (SHELL91 elements) don’t fail by buckling. The loads input are obtained from a 2D dynamic model made in Matlab. From the 2D model I get the lateral displacement from the vertical pipe (that can be derived twice to get the bending moments) and the tension on the pipe. The problem is that I’m not sure about the best way to apply the lateral displacements or bending moments to the pipe. I was thinking in series of master nodes (/RBE3) along the length of the pipe; however, and as far as I understood, the “ring” of slave nodes will have a limited degree of freedom, meaning that the buckling will not be calculated correctly. Is that correct, or will the slave nodes be able to move freely? Do you have any suggestions on how to solve this problem? You can see the displacement and bending moment distribution in the picture, as you can see it is not constant, so applying a bending moment on the edge will not be enough.

Thanks in advance!

 

[kundalin1]

Read point 21.6.1.2 in the mechanical APDL technology demonstration guide and the help about UPGEOM command.
That example shows how to introduce initial imperfections (a combination of 10 factored mode shapes) prior to nonlinear buckling analysis.
You'll have to do the same but with just one load step and the .rst file will come from a static analysis (instead of a linear buckling analysis).
You have to create that previous static analysis with your desired displacements/moments matching the solution from matlab.

 

[homusubi]

Thank you much Kundalin1. I haven’t read the section that you recommended me… but let me see if I understood correctly (if not, correct me please!). The idea is to get the corresponding X mode shape of the pipe. Then, we will introduce the displacement of the X mode with command UPGEOM, using the proper FACTOR so that the displacement of Matlab is matched. Once the pipe is deformed, we will make a buckling analysis (Eigen value?).

Be the way, which version of the APDL technology demonstration guide are you referring to? I took a look into V13, where 21.6.1.2 is Modal Analysis with Gyroscopic Effects. I think it's not the proper one, not? I assume we are talking about the example "buckling and post-buckling of a ring=stiffened cylinder using nonlinear stabilization". I will give it a read and come back!

Thanks!