I am trying to calculate the interfacial tension water-vacuum, using reflective walls.
I used two initial configurations, first only one water molecule and replicate (8 times in each direction), and second using a data file with 512 water molecules. I used a NVT ensamble and the Gibb’s formulation to calculate the interfacial tension.
To compare the results, I used the same input file. The results of interfacial tension obtained were very different: 80.97 mN/m ( with 512 molecules), 64.57 mN/m (using replicate). The experimental value is 72.69 mN/m
So i don’t know, What are the cause of these differences ?
I am trying to calculate the interfacial tension water-vacuum, using reflective walls.
I used two initial configurations, first only one water molecule and replicate (8 times in each direction), and second using a data file with 512 water molecules. I used a NVT ensamble and the Gibb’s formulation to calculate the interfacial tension.
To compare the results, I used the same input file. The results of interfacial tension obtained were very different: 80.97 mN/m ( with 512 molecules), 64.57 mN/m (using replicate). The experimental value is 72.69 mN/m
So i don’t know, What are the cause of these differences ?
I am trying to calculate the interfacial tension water-vacuum, using reflective walls.
I used two initial configurations, first only one water molecule and replicate (8 times in each direction), and second using a data file with 512 water molecules. I used a NVT ensamble and the Gibb’s formulation to calculate the interfacial tension.
To compare the results, I used the same input file. The results of interfacial tension obtained were very different: 80.97 mN/m ( with 512 molecules), 64.57 mN/m (using replicate). The experimental value is 72.69 mN/m
So i don’t know, What are the cause of these differences ?
I believe if it is LAMMPS question, because from two initial configurations (in theory the same), results with interfacial tension values very different. So I think that something happens in the tensor pressure calculation.
“In theory” the same does not mean they are exactly the same. Without seeing the data files, nobody can be sure.
A system with 512 atoms is very small in size and fluctuation in pressure/stress can be huge for such a small system.
Before you claim there is something wrong in the tensor pressure calculation, have you looked at other materials, preferably bulk condensed metal phase?
This is really not a LAMMPS question, but a question of how you interpret the results you obtained.
I believe if it is LAMMPS question, because from two initial configurations
(in theory the same), results with interfacial tension values very
different. So I think that something happens in the tensor pressure
calculation.
Is possible that I am wrong
yes, it is possible. this sounds more like a finite size effect, or a
difference in convergence (i.e. statistics).
not to mention that it is rarely a good idea to compare to some
experimental value, but rather you need to compare to the proper value
for the water model you chose. not all water models are equal. and
they are often differ significantly from experimental properties like
diffusivity, radial distribution functions (or structure factors),
dielectric relaxation, dielectric constant, viscosity, and so on.
so i would say, it very much is a question of the science. you have to
come up with a more convincing proof if you want to convince us that
there is a problem with LAMMPS.
In that case, This is my thought… 512 atoms is quite small. I agree with other folks.
How about you start with few other initial configurations and see what values you get, and then average over those to get better estimate.
Just to give you a perspective, I am running ~15000 atom system to calculate interfacial tension (using pressure different method) and even after 5 ns, the values are not converged (close to it though),
please in the future wait at least 24h until reposting your question.
remember that people live in different time zones and they have
different schedules. so you have to give them some time to react.