help with proper barostatting water confined between two walls

Hello all,

I’ve been struggling to solve a problem that I haven’t yet been successful at it. I want to adjust the pressure of water confined between two walls at 1 bar before experiencing a Couette shear flow. Initially before applying shear flow, the pressure of water is around 150000 bar. I used npt to adjust the pressure which at the end of the MD run fluctuates around 1 bar. The visualization of the trajectory shows the formation of holes in liquid water. I have attached a snapshot of such holes. I don’t know how to set up my simulation to avoid the system from behaving like that. I wonder if anyone can help me about it.



PS. The input file is as follows:

read_restart restart.equil_nvt_298K.100000
neighbor 3.0 bin
neigh_modify delay 0 every 10 check yes

set potential function and parameters

Create Groups

group hy type 5 # H_water
group ox type 6 # O_water
set group ox charge -0.8476

set group hy charge 0.4238
group water type 5 6 # water
group slab type 1 2 3 4 # slab

pair_style hybrid/overlay rebo lj/cut 7.5 lj/cut/coul/long 8.5

pair_coeff * * rebo CH.airebo C H C H NULL NULL
pair_coeff * * lj/cut 0.0 0.0
pair_coeff * * lj/cut/coul/long 0.0 0.0
pair_coeff 1 6 lj/cut 0.013 2.71
pair_coeff 2 6 lj/cut 0.0036 1.7753
pair_coeff 3 6 lj/cut 0.013 2.71
pair_coeff 4 6 lj/cut 0.0036 1.7753
pair_coeff 6 6 lj/cut/coul/long 0.0067 3.166

bond_style harmonic
bond_coeff 1 7.669 1.0
angle_style harmonic
angle_coeff 1 1.829 109.47
kspace_style pppm 1.0e-5 #final npt relaxation

timestep 0.001

fix freeze slab setforce 0.0 0.0 0.0
compute temp_water water temp
compute press_water all pressure temp_water
thermo 500
thermo_style custom step pe etotal press temp c_press_water c_temp_water
thermo_modify norm no
fix shake_4 all shake 1e-5 20 0 b 1 a 1
fix npt_1 all npt temp 298.0 298.0 0.01 iso 1.0 1.0 1.0
fix_modify npt_1 press press_water temp temp_water
dump dcd_4 all dcd 500 equil_npt_298K.dcd
run 50000
write_restart restart.equil_npt_298K.*


Dear Farshad,

You are applying a thermostat and barostat to an
inhomogeneous/anisotropic system, and it seems that you are freezing
part of the system (the walls). While it is hard to tell exactly where
you are doing wrong from the information you provide, it is very
probable that the pressure in the water slab is not at all what your
lammps compute outputs, hence the bubble nucleation.

As a matter of fact, there is a much simpler way to impose the liquid
pressure in your geometry, i.e. in a large slab where a bulk liquid
region can be identified in the middle. In that case, you can simply
use one wall as a piston, as done in the "DPD thermostated liquid of
LJ chains sheared between (111) fcc walls" example script on lammps

Best regards,

PS: the walls on your snapshot have a strange structure, and water
molecules seem to be able to penetrate inside the lower wall. I'm not
sure this is normal behavior.

2014/1/10 Omid Omid <[email protected]>: