Spreading of a LJ-type droplet onto a flat substrate

LAMMPS LAMMPS Beginners
1

Hey everyone,

I am trying to simulate a case of a droplet spreading onto a solid wall, with changing contact angles. I have initialized a wall and two LJ-type fluids. Currently, the system is running fine, but my droplet is not spreading onto the surface. I am aware that this spreading behavior comes from the value of epsilon that is chosen between the LJ fluid and the wall, but no matter the value chosen, I see no spreading happening.

Any help would be greatly appreciated!

processors 1 1 1

#System parameters
variable	rho equal 1
variable    T equal 1
variable	rc equal 2.9

# Domain size and walls
variable	x equal 20
variable	y equal 24
variable	z equal 20
variable    wallwidth equal 1.0

# Set outside of domain to be wall
variable	ylo equal -${wallwidth}

#Thermostatted region outer half of wall
variable	ythemlo equal -0.5*${wallwidth}

#Add a buffer of empty space to prevent molecules moving too far they leave the domain
variable	buf equal 2.0
variable	ylobuf equal ${ylo}-${buf}

#Basic simulation setup
units		    lj
dimension	    3
atom_style      full 
#neigh_modify	delay 0 every 1
boundary		p f p

lattice         fcc ${rho}
region          simbox block 0 $x ${ylobuf} ${y} 0 $z 
create_box      3 simbox
region		    latticeregion block INF INF ${ylo} ${y} INF INF
create_atoms    1 region latticeregion

#Set wall region
region		    lower block INF INF ${ylo} 0.0 INF INF
group		    lower region lower
set		    group lower type 2
group		    wall union lower


#Set 3rd particle region
region		    side sphere 10 5 10 5
group		    side region side
set		    group side type 3

#Create lj pair styles and set wall/site terms to zero
mass            * 1.0
pair_style      lj/cut ${rc}
pair_coeff      1 1 1.0 1.0
pair_coeff      2 2 1.0 1.0
pair_coeff      3 3 1.0 1.0
pair_coeff      2 3 4 0.93

#Set thermostatting region
region		    lowertherm block INF INF ${ylo} ${ythemlo} INF INF
group		    lowertherm region lowertherm
group           nottherm subtract all lowertherm


#Get liquid molecules to a flow group


fix             1 all nve
fix	            2 lowertherm langevin $T $T 0.1 498094 

# 4) Visualization
thermo 100
dump mydmp all atom 100 dump.lammpstrj


#Equilibration
run 100000
2

Are you sure you posted the correct input? When visualizing your system it looks like a confined system and not a droplet. It also cannot run without changing the f boundary to m.

3

Thanks for the reply.

I am not sure what you mean by confined system - I initialize the droplet with the

region side sphere 10 5 10 5

command. Do you have any other ways to make it so the droplet is not “confined”?

I also added the command
fix 5 wall setforce 0.0 0.0 0.0
so that the wall particles don’t move.
Here is the last frame of my simulation, which ran for 1e-9s

lammpsdroplet
lammpsdroplet657×846 40.3 KB

4

This is what the system created by your input looks like:

snap
snap889×848 231 KB

5

Yes, but then what I am doing wrong that is causing these particles (aside from the wall) to act in a confined way?
Is it due to the lattice command?

6

The input you provide does not model the system you describe. It is in no way how the image looks that you posted. Thus, you must be using a different input for that simulation and I cannot comment on an input I cannot see.

7

The input file I posted is just a first attempt to the case that I am trying to simulation - that of a spreading droplet.
The idea was to initialize a non-moving wall, and an LJ fluid-type droplet and let them interact - this is what the input file is meant to be. Perhaps I was not clear in my first explanation.

From the input you can see the 3 phases (wall, droplet and another lj fluid)

image
image506×602 63.5 KB

8

Up to this point your entire discussion was about a two component system. With a three component system, the situation is much more complex since it is not only the competition between the surface and the droplet, but also the surface-solvent and the droplet-solvent interactions play a significant role.

Nevertheless, the input you posted cannot be the input you are talking about and created the image from, since the posted input will crash, if not modified to avoid lost atoms and the “surface” segment in your picture looks rigid/immobile, but is not in the input you posted.

At this point, my patience is exhausted and I have no interest to continue a discussion under these circumstances. Perhaps somebody else feels differently and is willing to discuss with you.