Calculating temperature of geometric regions

LAMMPS LAMMPS Mailing List Mirror
1

I was trying to simulate a Couette flow & oscillatory Couette flow and wanted to know the thermal temperature at different regions(layers along the height of channel).
1) I was using temp/region to do it but I wasnt explicitly sure if it removed the advection velocity while providing the temperature.

2) Also I was assuming the temperature provided by default thermo command and using thermo_style temp would be the same but it comes out to be different.

Would appreciate any comments provided.

Thank you.

BR/Joseph

ps: input script for reference

.
region fluid block INF INF 2.39 52.5 INF INF units box #<<<<< CHNAGE DIMENSION CORRSPONDING TO PENETRATION DEPTH
region btm_wall block INF INF INF 0.95 INF INF units box
region top_wall block INF INF 53.86 INF INF INF units box #<<<<< CHNAGE DIMENSION CORRSPONDING TO PENETRATION DEPTH
.

# initial velocities

compute mobile flow temp
velocity flow create 1.10 482748 temp mobile units box
thermo_modify temp mobile

# thermostating

compute mobile_thermo all temp/partial 0 0 1
fix 3 all langevin 1.10 1.10 1.0 699483
fix 1 flow nve
fix 2 lower nve
fix_modify 3 temp mobile_thermo
thermo_modify temp mobile_thermo

# Couette flow

velocity lower set 0.0 0.0 0.0 units box
fix 8 upper move wiggle 10.0 0.0 0.0 20.0 units box #<<<<< CHANGE AMPLITUDE AND TIMEPERIOD
fix 4 boundary setforce 0.0 0.0 0.0

#variable pxy equal pxy

region wall_fluid_1 block INF INF 53.135 53.385 INF INF units box
#group gwall_fluid_1 region wall_fluid_1
#compute temp_tst_1 gwall_fluid_1 temp/com
compute temp_tst_1 all temp/region wall_fluid_1
.
# Run

timestep 0.002
thermo 1000
#thermo_style custom step temp evdwl ecoul epair ebond edihed eimp eangle emol etotal
#thermo_style custom step temp etotal c_temp_tst_1 c_temp_tst_2 c_temp_tst_3 c_temp_tst_4 c_temp_tst_5 c_temp_tst_6 c_temp_tst_7 c_temp_tst_8 c_temp_tst_9
.

2

no, compute temp commands like temp/region
only currently allow for one "bias" to be included
in the T calculation. In the region case, this is the
region itself (removing atoms outside the region). So
you can't combine multiple biases, like also subtracting
out a flow field.

I don't know what your question (2) means.

Steve

3

Steve,

Thank you for the clarification.

If I were to edit the source code, could you possibly suggest the best way to subtract the flow field.
Also I was thinking, for the case of oscillatory Couette flow if I were to time average the
temperature from temp/region it would probably remove the flow field because of the inherent
oscillatory nature of the problem and the fluid velocity.

In the 2nd) question that I asked I was comparing the temperature that is dumped while just using the "thermo" command along with other variables like energy etc.
and the temperature given by the argument "temp" associated with "thermo_style". I was thinking it to be the same but it was coming out to be different.

Thank you for your help.

BR/Joseph

4

You can add any code you like to compute temp/region to
subtract things. It is harder to think how to do it more
generally at a higher level.

Re: thermo temp output. You seem to be saying
you get different output if you use one of these 2 commands:

thermo_style custom temp
thermo_style custom a b c temp d e f

I don't see how that's possible. Temp is temp.

I did notice that your input script uses a thermo_modify temp ...
command somewhere. That will assign a different compute
to the temperature output - so that can change the output.

But the keyword for it in thermo_style is still temp.

Steve