Uniaxial/triaxial compression: cylinder rotation

[blackink]

Hi,
I've been working on this problem for quite a while now. I'm trying to simulate triaxial compression of a drained clay sample. I cannot use an axisymetric model because i have layers at odd angles that i will add different material properties to once i get the elastic, homogeneous solution correct, therefore i have to use a 3d model.

My problem is that my cylinder rotates when a confining pressure or displacement is applied. Is this a common problem? The only way I have found to keep the cylinder from rotating is by adding friction between the clay cylinder and the rigid end caps/platens. However, i need 0 friction to minimize the error for the elastic solution. I have tried many different things to fix this: surface-to-surface contact versus node-to-surface, small sliding vs finite, refined mesh, different mesh type (hybrid), nlgeom. If you look at my images closely you can see how the elements have rotated. What am i forgetting? Thanks

 

[rstupplebeen]

There must be something to resist rotation even in a symmetric model. Is there a symmetry plane in you sample? If not I expect that there is friction in the real test and I would try to model that amount or a trivial amount if it can't be calculated. Global or contact stabilization could also be used if the previous 2 suggestions don't work. I hope this helps.

Rob Stupplebeen

http://www.optimaldevice.com