MD Simulation - Initial Velocity

[kmkumar]

Folks,

I wanted to perform a ab-initio MD simulation for a system with 250 atoms a NVT simulation, (at room temperature)
for 1 picosecond with the time step of 1femtosecond.

I have prepared the following INCAR file.

-------------------------------------------------
Startparameter for this Run:
LCHARG = .False
NWRITE = 2
LPETIM = F write-flag & timer
NPAR = 8

Electronic Relaxation 1
PREC = low
ENCUT = 400.00 eV
NELM = 60
ALGO = Very_Fast
EDIFF = 1E-04 stopping-criterion for ELM
LREAL = .TRUE.

Ionic Relaxation
EDIFFG = -0.01 stopping-criterion for IOM
NBLOCK = 1; KBLOCK = 100 inner block; outer block
IBRION = 0 ionic relax: 0-MD 1-quasi-New 2-CG
ISIF = 0 stress and relaxation
LCORR = T Harris-correction to forces

Molecular Dynamics
MAXMIX = 40
NSW = 1000
POTIM = 1
TEBEG = 298
SMASS = -2
NELMIN = 4
ISTART = 0
ISMEAR = 0; SIGMA=0.1
------------------------------

My first question is that if the above settings are OK
for the required simulation.

The second question is about the SMASS tag.
I set it to -2 to use the nose-hoover thermostat, but the
manual says that i need to supply the initial velocities which are
kept constant during the entire simulation.

But i performed a test simulation without any initial velocities
and the system assigned itself.
In this regard, i wanted to ask you if i can proceed without
any initial velocities.
If so, will the results obtained is reasonable.


<span class='smallblacktext'>[ Edited ]</span>

[Dr Vitaly V. G. Chaban]

In this regard, i wanted to ask you if i can proceed without
any initial velocities.
If so, will the results obtained is reasonable.

Yes, you can.

SMASS controls the behavior of the thermostat, not the initial kinetic energy.



<span class='smallblacktext'>[ Edited Wed Nov 09 2011, 06:09AM ]</span>