Steve Plimpton wrote:
Both NVT and NPT require a target temperature to equilibrate to.
Tstart and Tstop simply let you change the target T over the
course of a run. If you set Tstart = Tstop, it's just like picking
a single target T.
Yes, I thought so. For example, to simulate the NVT ensemble at T=300 K, we have to set Tstart=Tstop=300. Then after enough long time step, the temperature of the system should be kept around 300 K.
However, in my solid simulation attached below, such temperature keeping at the 300 K is not observed. What do I mistake?
boundary p p p
lattice bcc 3.00
region mybox block 0 10.0 0 10.0 0 10.0
create_box 2 mybox
create_atoms 2 region mybox basis 1 1 basis 2 2
pair_coeff * * NiAl_Act.eam.alloy Al Ni
neighbor 0.3 bin
neigh_modify delay 10
fix 1 all nvt 300.0 300.0 100.0 drag 0.7
compute myTemp all temp
fix 2 all ave/time 10 10 100 compute myTemp type scalar file temp.stats ave window 20
compute myS all stress/atom
compute myE all pe/atom
thermo_style custom step temp
dump e1 all atom 10 dump.atom