velocity command: instantaneous impulse or continuously applied "force"

LAMMPS LAMMPS Mailing List Mirror
1

Dear lammps community:

I’m using the sequence of commands shown below to bend a sample (a doubly-clamped beam - p s s boundary conditions). The first velocity command and the fix Clamp0 are meant to guarantee the atoms in group Ends remain without motion during the integration, fix NVTben. They work very well. The second velocity command is intended to initiate the movement of the atoms in group Load with a velocity (0; -10; 0) [A/ps]. My question here is: Is the velocity applied by the velocity command an instantaneous impulse applied at the first time step of the run, or an input that guaranties the velocity of the atoms in the group is the prescribed one in every time step of the run?

NVT run to simulate bending test at 300K

velocity Ends set 0.0 0.0 0.0 units box
fix Clamp0 Ends setforce 0.0 0.0 0.0
velocity Load set 0.0 -10.0 0.0 units box
fix NVTben all nvt temp 300 300 0.1

I ask because in my run, the doubly-clamped beam reaches a maximum deflection and bounces back, which seems consistent with the velocity being an instantaneous impulse. Now, if I were to required the velocity of the atoms in the group to be mantined constant at every time step, trying to force the beam to fail due to shear loading, which modification do you consider I should follow.

Additionally please notice, the system I am modeling is crystalline. Hence when the beam starts to deflect the top layers are under compression and the interatomic forces will force the beam to bounce back. I am interested in exploring options to continue slowly deflecting the beam trying to induce failure. So, I will deeply appreciate any suggestions and any experience from the point of view of implementation you would like to share.

Best regards,

2

Dear lammps community:

I'm using the sequence of commands shown below to bend a sample (a
doubly-clamped beam - p s s boundary conditions). The first velocity command
and the fix Clamp0 are meant to guarantee the atoms in group Ends remain
without motion during the integration, fix NVTben. They work very well. The
second velocity command is intended to initiate the movement of the atoms in
group Load with a velocity (0; -10; 0) [A/ps]. My question here is: Is the
velocity applied by the velocity command an instantaneous impulse applied at
the first time step of the run, or an input that guaranties the velocity of
the atoms in the group is the prescribed one in every time step of the run?

the former.

# NVT run to simulate bending test at 300K
velocity Ends set 0.0 0.0 0.0 units box
fix Clamp0 Ends setforce 0.0 0.0 0.0
velocity Load set 0.0 -10.0 0.0 units box
fix NVTben all nvt temp 300 300 0.1

I ask because in my run, the doubly-clamped beam reaches a maximum
deflection and bounces back, which seems consistent with the velocity being
an instantaneous impulse. Now, if I were to required the velocity of the
atoms in the group to be mantined constant at every time step, trying to
force the beam to fail due to shear loading, which modification do you
consider I should follow.

don't assign a velocity to those atoms and don't time integrate them,
but use fix move on them.
that will move those atoms regardless of whatever else happens.

axel.