couette

Dear teacher, I wrote couette flow with lammps. First of all, the system is at constant temperature (temperature control of fluid and wall surface respectively). After running for a certain period of time, the wall surface is added with speed, and the constant temperature setting of fluid is removed.

The effect I want to achieve is: the wall temperature is constant. By adding the speed, the fluid in the middle region is sheared, and the fluid is heated and dissipated through the viscous dissipation. In this process, the wall temperature is constant and the cooling effect is performed, so that the fluid temperature distribution is parabolic. Form, the closer to the wall temperature, the lower the temperature, the closer to the center of the fluid.

However, I have encountered a very difficult temperature problem now. As long as the wall speed is given, the wall temperature immediately becomes the corresponding temperature at that speed. It is not clear which part of the problem. My code is as follows, beginner MD, many places have a half-solution, the problem may not be very clear, please forgive me, please help me with the help of the teacher, I am grateful!

units lj
boundary p p f
atom_style atomic
neighbor 0.3 bin
neigh_modify every 1 delay 10 check yes

############…create geometray…##################
lattice fcc 2.609804 origin 0 0 0.181991 orient x 1 -1 0 orient y 1 1 -2 orient z 1 1 1
region box block 0 17 0 11.1 -7 30 units box
create_box 3 box
mass 1 1
mass 2* 4.875

###################…define the hot wall…##################
region lower block INF INF INF INF 0 1.19

lattice fcc 2.609804 origin 0 0 0.224745 orient x 1 -1 0 orient y 1 1 -2 orient z 1 1 1
region upper block INF INF INF INF 9.55 10.65
region wall union 2 lower upper
create_atoms 2 region lower
create_atoms 3 region upper
region liquid block INF INF INF INF 2.724046 18.724046 units box
lattice fcc 0.78074 origin 0 0 0 orient x 1 0 0 orient y 0 1 0 orient z 0 0 1
create_atoms 1 region liquid
group lower region lower
group upper region upper
group wall region wall
group liquid type 1

#################…LJ potential…################

pair_style lj/cut 3.0
pair_coeff 1 1 1 1
pair_coeff 2* 2* 50 0.736872 # 12.0181 2.475
pair_coeff 1 2 0.989949 0.863436 #qin 0.14 1.0
pair_coeff 1 3 0.989949 0.863436 #qin
delete_atoms overlap 0.5 all all
min_style fire
minimize 1.0e-6 1.0e-8 10000 10000
write_data min.data
reset_timestep 0

###########################…setting…###################
timestep 0.001 #0.002329474_5fs
velocity all create 0.8267 102486 mom yes rot yes dist gaussian
compute thermal liquid temp/partial 0 1 0
compute liquid liquid temp
compute lower lower temp
compute upper upper temp
fix 1 all nve
fix 2 liquid langevin 0.8267 0.8267 0.1 498094
fix_modify 2 temp thermal
fix 0 lower langevin 0.8267 0.8267 0.1 498094
fix 00 upper langevin 0.8267 0.8267 0.1 498094
compute layers liquid chunk/atom bin/1d z center 0.2 units box
compute c1 lower chunk/atom bin/1d y center 1 units reduced
compute c2 upper chunk/atom bin/1d y center 1 units reduced
fix 4 all ave/chunk 1 200000 200000 layers vx temp file Tf.dat
fix 5 all ave/chunk 1 200000 200000 c1 vx temp file profile.lower
fix 6 all ave/chunk 1 200000 200000 c2 vx temp file profile.upper
fix 7 all ave/chunk 1 200000 200000 layers density/mass file Density.dat
thermo 2000
thermo_style custom step temp ke pe etotal c_liquid c_thermal c_lower c_upper

dump 1 all custom 2000 dump.atom id type x y z vx vy vz fx fy fz
run 500000
unfix 2
velocity upper set 1.0 0.0 0.0 sum no units box
velocity lower set -1.0 0.0 0.0 sum no units box
fix 3 wall setforce 0.0 0.0 0.0
run 4000000

Since you set the wall speeds directly and zero the force on them,
I don’t understand why you compute or care about their temperature.
They are just boundary conditions, they have no statistical temperature.
You should only monitor the temperature of the fluid that flows
between them. I suggest you look at the Couette example provided
in examples/flow as a starting point and see if you can replace its
simple pair style with the pair model you want and still get similar
behavior. Also, visualizing your flow is a good idea.

Steve