[lammps-users] how to relax the pressure at x-z plane

Dear All

My question is: for an orthogonal box simulation, how to relax the pressure at x-z plane?

Some key simulation conditions during my simulation are listed below briefly:

simulation structure: fcc structure with pre-existing edge dislocation

boundary p s p

timestep 0.001

fix 1 all npt temp {Tstart} {Tend} 1.7 x 0 0 2.5 z 0 0 2.5

run 100000

minimize 1e-25 1e-25 5000 10000

As a result, pxx and pzz can be approach zero after relaxation. However, pxz cannot be eliminated. And pxz is relatively very large? How can I understand this situation? And more important, how can I eliminate the pxz to zero?
So could you kindly help understand this situation? And find a solution to eliminate the pxz for orthogonal box at boundary condition p s p?

As for reference, if I control the pxz using npt (for example: fix 1 all npt x 0 0 2.5 z 0 0 2.5 xz 0 0 2.5) to relax, it can eliminate the pxz to zero. But, correspondingly, I have to change box from orthogonal box to triclinic box. If this is the only solution to the problem above, if it will influence the pressure during subsequently shearing along xz plane?

Thanks for all of your great help in advance. Looking forward to your reply.

Best regards,

MA Shihua

City University O f HongKong