[gmx-users] question about simualtion of CNT from a rookie

[Ye MEI]

Dear all,

I am trying to study an infinite length CNT in SPC water box under NPT
using gromacs 4.5.3.  The axis of CNT is along Z direction. The options
"pbc = xyz " and "periodic_molecules = yes" are included in all mdp files.
The whole system was minimized first, followed by an NVT simulation. The
structure was fine, at least when visualized with VMD.  Then I ran an NPT
simulation with weak restraint on CNT. The pressure coupling options are
set as following
; Pressure coupling is on
pcoupl          = Parrinello-Rahman     ; Pressure coupling on in NPT
pcoupltype      = semiisotropic ; uniform scaling of box vectors
tau_p           = 2.0 0.0       ; time constant, in ps
ref_p           = 1.0 0.0       ; reference pressure, in bar
compressibility = 4.5e-5 0.0    ; isothermal compressibility of water, bar^-1

And the final structure was also fine. But then I removed the constraint,
and tried to run a long time simulation under NPT, the whole system
exploded in several ps.
I have not figured out what was wrong in my calculation. Can anyone give
me a suggestion?

Ye MEI