Dear Steve,
We define part of system as a rigid body and give a constant velocity to it. Then we dump the sum of the forces on the rigid body during the simulation. Instead of giving an average steady state 0 value, it gives a constant value. That doesn’t make sense, since the whole system is moving together with the velocity given to the rigid body
Please see the below script we used:
units metal
time:ps; distance: Angstrom; energy: eV; force:eV/A=1.6 nN;
atom_style hybrid atomic
boundary p p p
read_data lammps.dat
timestep 0.001
group bot type = 1
group therm1 type = 2
group film type = 3
group tipdown type = 4
group therm2 type = 5
group tiptop type = 6
group tip union tipdown therm2 tiptop
group sub union bot therm1 film
group mobile subtract all bot
group subnve subtract all tiptop
group tem union therm2 therm1
group dump union tip film
pair_style hybrid/overlay tersoff table linear 1000
pair_coeff * * tersoff SiC.tersoff C C C C C C
pair_coeff 3 4*5 table lj.table CC 10.0
fix 1 bot setforce 0 0 0
velocity mobile create 300.0 23456789 units box
fix 2 mobile langevin 300.0 300.0 0.01 699483
compute 1 tiptop reduce sum fx
compute 2 tiptop reduce sum fz
compute new3d mobile temp
fix 13 tiptop ave/time 1 7350 7350 c_1 file tiptop-fx1
fix 14 tiptop ave/time 1 7350 7350 c_2 file tiptop-fz1
dump 1 dump cfg 7350 dump.*.cfg id type xs ys zs x y z id type
thermo_style custom step temp etotal press vol lx ly lz
thermo 7350
thermo_modify temp new3d
Cutting speed: 1.0 A/ps = 100 m/s
#fix 3 tiptop nve/noforce
#fix 4 subnve nve
#unfix 3
#unfix 4
fix 21 tiptop rigid/nve single force * on on on torque * off off off
fix 22 subnve nve
velocity tiptop set 0.2 0 0 units box
run 1000000
Thanks
Zhendong