[lammps-users] density profile

Dear all,

I am doing a heat transfer problem with liquid argon inside the silver walls. In order to generate a temperature gradient in my argon region, I use langevin thermostat to control my silver walls’ temperature. One side is hot, and the other side is cold. When I plot my density profile of argon, I found that my density profile is unsymmetric if I divide the liquid argon region into many small bins. First I thought it has temperature dependence. But later I assigned the same temperature on the hot and cold walls, my density profile is still unsymmetric. I don’t know why. I think the density profile should be symmetric and if I divide my liquid argon into less bins, it is barely symmetric, but if I did it into many bins, the density profile is obviously unsymmetirc.

Can anybody give me some suggestion?



I think it should be asymmetric. More dense adjacent to the cold wall, less dense adjacent to the hot wall.

Quoting shiziyuan <[email protected]>:

But I set my both wall at the same temperature, the density profile is asymmetric too. Is there something wrong? Or it depends on the size of my bin.


Well, my second guess is that you have a sampling problem. How many particles do you have, and how long (and how many configurations) are you averaging? That you get an approximately symmetric distribution using less bins suggest your sampling is the problem. If you send the input and starting configuration I'll take a look at it.

Quoting shiziyuan <[email protected]>:

Dear Matt

I think I know where my problem is. I use ave/spatial to divide my whole domain into many bins which is not accurate for my problem. What I need to do is to divide my liquid argon region into many bins. Is there any way I can do this in lammps?


If you are only binning argon atoms, then can't you just discard the bins that don't contain any argon atoms? I don't see the problem.

Quoting shiziyuan <[email protected]>:

When I assigned hot wall and cold wall, I did get density profile as you said before. But my problem is when I assigned the walls having the same temperature, my density profile is still asymmetric. So I thought it’s because when I divide the whole domain into many bins, it’s hard to assure that my liquid region was divided into symmetric bins.

Below is my input file:

units metal
boundary f p p
atom_style full
read_restart 100eq.restart.1000000

pair_style hybrid eam lj/cut 10.0
pair_coeff 2 2 lj/cut 0.0103 3.405
pair_coeff 1 1 eam Ag_u3.eam
pair_coeff 1 2 lj/cut 0.06013 2.9895
region hol block -52 -51 INF INF INF INF units box
region col block 35 37 INF INF INF INF units box
group ho-layer region hol
group co-layer region col
group argon type 2
group hotwall id <= 648
group coldwall subtract all argon hotwall
group myall subtract all ho-layer co-layer

timestep 0.004
fix 3 hotwall langevin 100 100 0.01 8766790
fix 4 coldwall langevin 100 100 0.01 6757890
fix 5 myall nve

fix 101 argon ave/spatial 1 197000 200000 x lower 0.004 density/mass file density.profile units reduced

run 2000000

Below is my figures:

New Image.JPG

New Image 1.JPG