Force based friction test

Dear LAMMPS users,

I am doing MD simulations for frictional studies (indentation and scratch tests) between aluminum surfaces. I have used fix indent (LAMMPS indenter) command and it worked good which is displacement controlled. I am trying to simulate force based indenter to replicate the experimental setup. The indentation process looks fine and comparable to lammps indenter (displacement controlled repulsive indenter), but the scratch process is not working properly.

I have used fix aveforce command and fix lineforce command for indentation process and fix setforce (gives normal load) and fix move to scratch during scratch process.

What I am seeing is the normal force on the indenter is not stable and it is dropping down to zero. I have also not observed any wear in the indenter (group of atoms). I am not sure what mistake I am doing?. Any suggestions will be appreciated.

Here is the important portion of the code:

Indentation:

fix 4 top aveforce NULL v_force 0.0

fix 5 top lineforce 0.0 1.0 0.0

fix 6 top nve

I have used npt for substrate atoms

Scratch:

fix 4 top setforce NULL -0.08 NULL

fix 5 top move linear 0.05 0 0 units box

fix 6 top nve.

Thank you.

Regards
Raghuram

image001.jpg

Dear LAMMPS users,

I am doing MD simulations for frictional studies (indentation and scratch
tests) between aluminum surfaces. I have used fix indent (LAMMPS indenter)
command and it worked good which is displacement controlled. I am trying
to simulate force based indenter to replicate the experimental setup. The
indentation process looks fine and comparable to lammps indenter
(displacement controlled repulsive indenter), but the scratch process is
not working properly.

I have used fix aveforce command and fix lineforce command for
indentation process and fix setforce (gives normal load) and fix move to
scratch during scratch process.

What I am seeing is the normal force on the indenter is not stable and it
is dropping down to zero. I have also not observed any wear in the indenter
(group of atoms). I am not sure what mistake I am doing?. Any suggestions
will be appreciated.

​your mistake is, that ​you are time integrating the atoms in the group top
twice. check your output! LAMMPS should have warned you about it.
instead of fix move, you can make the first argument of fix setforce "0.0"
instead of "NULL" and then use "velocity top set" to give its atoms a
velocity. since the force in x is set to zero, the velocity will not change
and thus you have the same effect as using fix move, but you only use one
time integrator (fix nve).

axel.

image001.jpg

Dear Axel

Thank you very much for the response. I have tried what you said but there is other complication with it, the system explodes. I don’t know what I am doing wrong?

The scratch test works fine with the velocity set command (you have suggested) when the set force for the top group is 0.0 in all the directions. When I give constant normal force along y-direction (using set force) the system explodes because the indenter moves along y-direction continuously which is not periodic. The reason I understand is, velocity can’t be set to zero when there is force acting in that particular direction. But, I need to keep constant normal force on the indenter (in y-direction) and scratch along x-direction with constant velocity.

Fix move command works fine even tough I give constant normal force (using set force). The problem with this command is the normal force is fluctuating a lot and average is less than the given set force. For one of the cases the average is equal to the set normal force but for other cases it is not. I do not understand why this is happening?

Below is my input script and it would be great if you can help me in figuring out my mistake. Thank you.

image001.jpg

Dear Axel

Thank you very much for the response. I have tried what you said but there
is other complication with it, the system explodes. I don't know what I am
doing wrong?

The scratch test works fine with the velocity set command (you have
suggested) when the set force for the top group is 0.0 in all the
directions. When I give constant normal force along y-direction (using set
force) the system explodes because the indenter moves along y-direction
continuously which is not periodic. The reason I understand is, velocity
can't be set to zero when there is force acting in that particular
direction. But, I need to keep constant normal force on the indenter (in
y-direction) and scratch along x-direction with constant velocity.

​this is all nonsense. you have not paid proper attention ​to what i was
telling you and what you see is the result. when a system "explodes" that
is usually due to bad choice of force field, external forces or velocities.

Fix move command works fine even tough I give constant normal force
(using set force). The problem with this command is the normal force is
fluctuating a lot and average is less than the given set force. For one of
the cases the average is equal to the set normal force but for other cases
it is not. I do not understand why this is happening?

​you are *still* applying time integration to the same atoms twice. this is
wrong and thus you​ get what you deserve.
at any rate, this is now becoming an issue of understanding basic
simulation methodology and not an issue of using LAMMPS commands according
to their documentation and thus it is rather something that you need to
discuss with your adviser/supervisor. the purpose of the mailing list is
not to explain to you how to do your research. that is the concern of your
adviser/supervisor.

axel.

image001.jpg