Simulation of water in CNT using Moltemplate and LAMMPS softwares

LAMMPS LAMMPS Mailing List Mirror
1

Hello All,

I am working on simulation of water in CNT using LAMMPS. I have used very large number for number of run (20000000) and dt=1 femtosecond in my input file.

After visualization, I can see the water fills entire the CNT; however, the water doesn’t exit the tube even after passing long time. I have a couple of questions about my results:

1- I plotted the velocity and density/number along radial in z direction. For this purpose, the ave/spatial command was used as following:

fix 1 all ave/spatial 2 5000 10000 x center 2.0 vx vy vz density/number bound z 32.0 58.26 units box file vel1.profile

fix 11 all ave/spatial 2 5000 10000 x center 2.0 vx vy vz density/number bound z 0.0 80.0 units box file vel11.profile
fix 22 all ave/spatial 2 5000 10000 x center 2.0 vx vy vz density/number bound z 0.0 11.47 units box file vel22.profile
fix 3 all ave/spatial 2 5000 10000 x center 2.0 vx vy vz density/number bound z 11.47 32.0 units box filevel3. profile
fix 44 all ave/spatial 2 5000 10000 x center 2.0 vx vy vz density/number bound z 58.26 80.0 units box file vel4.profile

fix 2 all ave/spatial 2 5000 10000 x center 2.0 vx vy vz density/number units box file vel2.profile

I used bounds z in this command to calculate the velocity in different sections of CNT. However, the velocity profiles are the same for all of profiles and I think my obtained results are based on the whole computational domain (bounds z command doesn’t work?). How can I get accurate velocity distribution in different sections of CNT in radial direction?

2- I am thinking that velocity profile after long time should get positive steady state values in CNT as I have read in the literature. However, the velocity is always changing each Nfreq (The time step which is used to calculate velocity). How can I approach this problem?

I attached the velocity and Density/number for different number of water molecules, lengths and radii of CNT after passing long time.

I really appreciate your help.

Best regards,

Hamed.

Results.pdf (399 KB)

2

If nothing is exiting your CNT, then the velocities inside

it have to asymptote to 0.0. I’m guessing you are either

not visualizing the output correctly or are mis-interpreting

the velocity profiles. What LAMMPS does with fix ave/spatial

is pretty straightforward.

Steve

3

I sent a more detailed reply to Hamed early today. Here's the short version

I am working on simulation of water in CNT using LAMMPS. I have used very large number for number of run (20000000) and dt=1 femtosecond in my input file.

After visualization, I can see the water fills entire the CNT; however, the water doesn't exit the tube even after passing long time.

This is a consequence of the surface tension of water, and the fact
that you have a vacuum region. The system will try to minimize the
area of the air-water boundary. If water exits the other side of the
tube, this would create another air-water boundary, which would be
unfavorable.

I am assuming that you are running a simulation at constant volume
which is considerably larger than the volume of water in your
simulation, (as shown in your picture). This creates a vacuum
("bubble"). If you reduce the volume of the simulation, or run a
simulation at constant positive pressure, or fill the empty space with
more water molecules, then water should be able to freely move through
the tube.

I do not have enough experience using fix ave/chunk (or fix
ave/spatial) to comment on this. I suggest you update to the latest
version of LAMMPS, use fix ave/chunk instead of fix ave/spatial, and
also take another look at the documentation for fix ave/chunk. If it
helps, I will also send you instructions how to compile the code
yourself, and step through it with a debugger, if you think it would
help you understand what the code is doing.

Cheers
Andrew