Forces on the frozen atoms

LAMMPS LAMMPS Mailing List Mirror
1

Hi Steve and Axel,

In my system, i have to write real forces acting on frozen atoms. How can i do that?

Here is what i did so far: I used

fix 1 phos setforce 0.0 0.0 0.0

to fix the atom position in a group. when i dump the forces by

dump 1 phos custom force.out fx fy fz

all i have got is 0.0 . However, i need real force acting on the atom in the group . I also used

compute 1 phos property/atom fz
dump 1 phos custom 100 force.out c_1

I still got 0.0

I also tried “variable” statement by using

variable fzz equal fz[id of atom in the group]
thermo_style custom step v_fzz

still i have got 0.0.

Is there any other way to get the real force acting on frozen atom?

Best Regards
Sefa

2

Hi Steve and Axel,

In my system, i have to write real forces acting on frozen atoms. How can i
do that?

Here is what i did so far: I used

fix 1 phos setforce 0.0 0.0 0.0

to fix the atom position in a group. when i dump the forces by

dump 1 phos custom force.out fx fy fz

yup. that is what you are supposed to get since you
did set the forces to zero with fix setforce.

all i have got is 0.0 . However, i need real force acting on the atom in the
group . I also used

compute 1 phos property/atom fz
dump 1 phos custom 100 force.out c_1

same difference. again you set the forces to zero
and thus you can only output zero.

I still got 0.0

I also tried "variable" statement by using

variable fzz equal fz[id of atom in the group]
thermo_style custom step v_fzz

again. same difference. you are always accessing the same data.

still i have got 0.0.

Is there any other way to get the real force acting on frozen atom?

yes. don't include the atom in the fix for time integration.
this will have the same effect on the trajectory (if you don't
time integrate atoms, they don't move), but you don't have
to zero out the forces for it.

if you use fix setforce, all you have access to is the
total force on the group that you have zeroed out.

cheers,
    axel.

3

You can also use fix store/force and output the values that fix
stores, which are forces saved before they are zeroed (or
otherwise modified) later in the timestep.

Steve