[lammps-users] Uniaxial Tensile test

Hi Rob,
Thanks for your reply.
I too thought in the same way, but I came across the below statement in literature.

The system should be assigned an initial temperature of 300K and is allowed to relax by holding the length of the wire constant while maintaining a constant temperature using thermostat.

Moreover, even if the length of the wire is an equilibrated value, I am getting negative stress at zeroth step and the temperature is not increasing during tension test.


Just because it’s in the literature doesn’t mean it’s correct, or what should be done with modern methods. Why wouldn’t you want to allow the length of the wire to vary, to equilibrate the stress to zero, using fix npt?


As Rob pointed out, it is not a good idea to hold wire length constant when you relax it. We’ve done a lot of mechanical elongation test of nanowires. What i did for the first equilibrium is letting the wires find themselves the equlibrium state without controlling the length. Instead, we used a nvt ensemble with bigger box size than we cut it from bulk. Then run enough long time, which depends on the wire size. It may be useful to hold one end of the wire to reduce the relaxation time. If you want, I can send you my papers regarding this issue.

Good luck,

Ajing Cao