Hot - cold sink

LAMMPS LAMMPS Mailing List Mirror
1

Dear Users,

I wrote a script to evaluate different temp. as shown below but the system was destroy, any one have idea about that

Cheers,

Alaa

set up units basic initial paramters

units real

atom_style charge

define and create simulation domain

region bbox block -100 100 -100 100 -100 100

create_box 4 bbox

place atoms in the box with(x,y,z)coordinates and type

create_atoms 1 single -1.43492000 -2.85265000 -72.84560000 units box

create_atoms 1 single 1.44508000 -2.85265000 -72.84560000 units box

create_atoms 1 single 4.32508000 -2.85265000 -72.84560000 units box

.

.

.

.

.

.

create_atoms 1 single 2.88000000 1.65500000 63.49200000 units box

#set the mass of each atom types

mass 1 196.9665

mass 2 12.00

mass 3 1.00

mass 4 32.066

#define groups of atoms

group left id <> 1 243

group right id <> 258 500

group junction id <> 244 257

define forcefield between atoms (reaxff is easy)

pair_style reax/c NULL checkqeq no

pair_coeff * * ffield_C_H_O_S_Au.reax Au C H S

setup the simulation

initiall temperature

velocity all create 600.0 4976559 rot yes dist gaussian ##4928459

fix 1 all nvt temp 300.01 300.01 100.0

velocity right create 100.0 47459 rot yes dist gaussian ##44592

fix 2 right nvt temp 100.01 100.01 100.0

velocity junction create 225.0 49753 rot yes dist gaussian ##4928459

fix 3 junction nve

relaxing (optimizing) the geometry (not mandatory) (this will run immediately)l

fix UCBOX all box/relax iso 0.0 vmax 0.001

minimize 0.0 1.0e-8 1000 10000

unfix UCBOX

there is a good default, no need it

timestep 0.5

#run 5000

#unfix 1

#fix 1 all nve

compute ke all ke/atom

variable temp atom c_ke/(1.5*0.0019872041)

fix 8 all ave/spatial 2 10000 30000 z lower 6.8 v_temp file tmp.profile units box

fix 9 all thermal/conductivity 600 z 6 swap 1

reset_timestep 100

writes out the thermodynamical variables (temperature, pressure, totlaenergy, each 100 timesteps)

thermo_style custom step time temp ke pe etotal press vol f_9

thermo 600

LAMMPS - DUMP OUT THE INITIAL GEOMETRY

dump 1 all xyz 1 optimized.xyz

run 0

undump 1

LAMMPS - DUMP OUT THE TIME EVOL GEOMETRY

dump 1 all xyz 10000 timeevol.xyz

dump d0 all cfg 100 *.cfg mass type xs ys zs vx vy vz fx fy fz

#dump i all image 200 dump.*.jpg type type

run 100000

unfix 2

fix 4 right nve

run 100000

unfix 1

#fix 5 left nve

#run 1000000

2

You have to ask more specific Qs. You can’t just

post a complicated input script and say "this doesn’t

work, help me?"

Steve

3

A quick look reveals problems with the script. For example, fix 1 and 2 are both nvt fixes setting different temperatures and seem to be acting on the group right at the same time. then there is fix 3 and 1 acting simultaneously on the group junction, etc. No surprise the system was “destroyed”…
Time to read and understand better the docs.
Carlos