I am trying to do uniaxial deformation of Gold with non-periodic boundary conditions. The documentation says that for deformation of solids, it is generally better to remap the atomic positions while the fix nvt/sllod only works with remapping of velocities. Ideally I would want to use fix npt for the deformation, but since I am using non-periodic conditions, I have to use nvt. Should I go for remapping velocities or choose some other time integration scheme? Is the nvt/sllod fix only meant for fluids? What is the best time integration scheme for such a simulation? I am attaching the source code for reference
yes, fix nvt/sllod is only meant for fluids. Fix deform also
makes little sense with a non-periodic system as it is the periodic
images of atoms that enforce the deformation. What deformation
are you doing that fix deform allows with a non-periodic system?
I wanted to do a deformation study of gold nanowires. I used periodic conditions in the z direction whereas the x and y directions are non-periodic. I am pulling along the z axis which is periodic. Is this approach correct? Also, what kind of ensemble do you recommend with these settings? I was thinking of nvt as plastic deformation is a volume conserving process (neglecting the initial elastic part). Is there any way to use npt in this case?
You can pull in z with fix deform since its periodic and leave x,y non-periodic.
You can thermostat with fix langevin (my choice when perturbing a system),
or fix nvt if you're careful about pull rates and damping constants.