Dear Lammps developers,

I’m going to do a homogenous shear test. the test specimen has a circular cross section. As the box is not triclinic, i cannot use fix deform xy. My question is that is there any way to do a homogenous shear test for a cylinder? I did such test for a block using fix deform xy and a prism region, and the results were ok.

Mnay thanks inadvance.

Best regards,

By homogeneous shear we usually mean statistically homogeneous, i.e. statistics of single point quantities don’t depend on location. This can really only happen in the special case of an infinite domain as to avoid boundary effects. We simulate this with sufficiently large periodic domains.

Ask yourself two simple questions, how would a periodic boundary with finite curvature work? And secondly, if its a homogeneous/infinite domain, why the heck would you care about what the boundary looked like?

Dear Eric,

Thanks for your reply. What I meant was that i’m going to exert quasistatic starin to the cylinder, not a differntial strain. becouse of this I used fix deform xy for a block (prism).

Regards,

Perhaps someone else gets some additional information out of this; I do not. People can mean many different things when they start using domain specific terminology. Quasistatic for me means that transient/time dependent terms are null (superfast relaxations). Which has nothing to do with your original Q.

What is your actual question? (you should take my confusion and others not answering as a clue that maybe this needs some additional thought.)