FEM of Tension Buckling

[LorenzoG]

Hi all,

I´m new here so I´d like to thank the creators of this nice forum and anyone who could help. Also sorry if i post this topic in the wrong place.

I´ll explain my issue.

I have a thin wall cylinder made of a long fibre reinforced material. The fibres are wound at a certain angle ? to the cylinder main axis, but they are NOT symmetrical (that is: they only wind at angle ?, not -? too).

Doing a mechanical tension test, I observe buckling. This is strange, but logical. Infact, the fibre reinforced material is highly anisotropic, and a tension in the main axis direction (z) causes a noticeable shear in the z? plane. In turn, constrained shearing causes buckling.


Now, I would like to simulate this in finite element code ABAQUS (6.9 version). I use an anisotropic material model, assign cylindrical material orientation, assign reasonable constraints (cylinder ends must stay plane) and apply a default axial tensional load.

After simulation, the buckling analysis just recovers the usual compression buckling modes (that is I get negative eigenvalues), but nothing else.

I thought it was an ABAQUS limitation, so to test this assumption I tried simulating the tension buckling of a plate with same material properties and constraints.

This time ABAQUS could recover positive eigenvalues (that is related to tension).


So I must imagine that I do something wrong with the modelling of the cylindrical case.... Any idea what????


Thanks in advance



Lorenzo

 

[Elabbasi]

I had a similar problem recently, also involving cylindrically wrapped fibers and used Abaqus to predict the "tensile" buckling load. You are probably running the analysis correctly since you are getting the expected buckling in the plate problem so maybe recheck the cylindrical material axes definition.

You can also try performing a postbuckling analysis (Riks method), and put some geometric or loading imperfections to trigger the buckling mode(s).

Nagi Elabbasi
Veryst Engineering