# fast cooling down

Hi,

I have used the velocity scaling temperature to control temperature of two groups of atoms(vapor and solid atoms ) and set their temperature to 160K. then switched off the the direct control and let the system run.
after that I dropped down the the solid temperature by temp/rescale and removed all previous fixes from vapor. the compute temp is used to get the vapor temperature. the problem is that the vapor temperature drops down so rapidly from 160 to 107 in 50 (ps) which isn’t correct. it should take much longer time for the vapor to exchange energy with solid surface. would you please comment on where this problem may come from?

units real
dimension 3
boundary p p p

atom_style atomic

pair_style hybrid eam/alloy lj/cut 12

pair_coeff * * eam/alloy Pt.eam.alloy NULL Pt
pair_coeff 1 1 lj/cut 0.24035 3.4
pair_coeff 1 2 lj/cut 0.475 3.085

group vapor type 1
group solid type 2

compute TV vapor temp

#--------------------Equilibrating the temperature to 160 K------------------------------

fix ts1 solid temp/rescale 1 160 160 0.05 1.0
fix tvp1 vapor temp/rescale 1 160 160 0.05 1.0

fix NVE1 all nve

fix 1 solid ave/time 10 1 10 c_ts1_temp ave running file ST1.out
fix 2 vapor ave/time 10 1 10 c_tvp1_temp ave running file tvp1.out

thermo 10
timestep 5

run 1000

# -------------Switching off the direct temperature control of vapor

unfix tvp1
unfix 2

fix 3 solid ave/time 10 1 10 c_ts1_temp ave running file ST2.out
fix 4 vapor ave/time 10 1 10 c_TV ave running file tvp2.out

run 10000

unfix ts1
unfix 1
unfix 3
unfix 4

#--------------------changing the solid temperature to a lower degree (100 k) and let the vapor cool down-----------------------

fix ts2 solid temp/rescale 1 100 100 0.05 1.0

fix 5 solid ave/time 10 1 10 c_ts2_temp ave running file ST3.out
fix 6 vapor ave/time 10 1 10 c_TV ave running file tvp3.out

run 20000

test (1.36 KB)

I don't know. Could just be basic physics.
Also, you are using an EAM potential
to model a vapor, which is unlikely
to be a good idea.

Steve