Liquid water sheared between two walls - problems with velocity profiles

LAMMPS LAMMPS Mailing List Mirror
1

Dear Lammps User,

I have been working on liquid water sheared between 2 FCC walls in 3D dimensions. Our objective is to obtain the velocity profiles to calculate the slip. Thus, we use only NVT ensemble for our system. Wall-fluid interaction strength ratio is 2.5 . Firstly, velocity of 0.79 A/ps was set up to walls; Maxwell_Boltzmann velocity distributions at 300K, NVT and Nose-Hoover thermostat (300K) were applied for all molecules in 1ns. Secondly, velocity of 0.79 A/ps is still set up to walls to keep them moving, NVT and Nose-Hoover (300K) were applied to all molecules in next 3ns. Results were averaged in last 2ns. However, our results revealed that the velocity of the wall and liquid molecule near the walls reach ~2.5 A/ps. Anyone could support us for this problem? Thank you very much

Best Regards,
Truong Vo

2

If you are setting the wall velocity to a value, and it is not

moving at that constant velocity, then something must be
wrong with your script. You can dump the velocity of
wall atoms to a dump file to verify what their velocity is.

The initial velocity you define via the velocity command

can have box vs lattice units, which are often different.
See the velocity doc page for details. If the wall atom
velocity changes over time, then you probably need to

zero the forces on the wall, so the time integration keeps
the velocity constant.

Steve