Hello everyone,
I’m running NVE simulations of SiC with the Tersoff potential. I rescale the velocities to keep a constant temperature during a certain time, after which I remove the rescaling and let the system run under the NVE ensemble. However, I noticed that after a few steps, the temperature of the system increases. First, I tried rescaling the velocities to keep the temperature at 300 K for 70000 steps with 1 fs timestep; I then increased the time of the constant temperature run up to 700000 steps. Also, I used fix spring with the tether option to keep the system fixed in the center of mass to prevent drift. However, even with all these considerations, the temperature still increases in the constant energy run.
Any suggestions?
Thanks
Jaime Sanchez
P.S.: the input script is below
units metal
atom_style atomic
variable equitime1 index 50000
variable equitime2 index 20000
variable simtime index 6000000
variable temp index 300
variable everytime index 25
lattice diamond 4.3596
region box block -3 3 -3 3 -3 3
boundary p p p
create_box 2 box
create_atoms 2 box basis 1 2 basis 2 2 basis 3 2 basis 4 2 basis 5 1 basis 6 1 basis 7 1 basis 8 1
mass 1 28.0855
mass 2 12.011
pair_style tersoff
pair_coeff * * SiC.tersoff Si C
velocity all create ${temp} 23432534
minimize 1.0e-4 10000 100000
fix 1 all nve
fix mymass all com 100 com.dat
fix pull all spring tether 500.0 -0.326482 -0.326482 -0.326482 0.0
timestep 0.001
thermo 50
thermo_modify lost ignore
fix 2 all temp/rescale 1 {temp} {temp} 0.1 0.5
run {equitime1}
unfix 2
fix 2 all temp/rescale 1 {temp} {temp} 0.001 0.5
run {equitime2}
unfix 2
dump pos_vel all custom {everytime} pos_vel.{temp} type tag x y z vx vy vz
run ${simtime}