Questions regarding Shake command

111291 · September 20, 2016, 2:12pm

Dear all,

I am trying to use Shake command for molecules.

First, the simulation of propane in bulk is being tried.

When I eliminated the SHAKE command, it worked and then I guess that the command which I wrote has problems.

The version of LAMMPS is Mar20 2016.

I attached my data file and molecule file and the following is a part of the log file.

Any advice would be greatly appreciated.

variable t equal 353

variable seed equal 31798.0000

units metal

atom_style full

boundary p p p

pair_style lj/cut 14.0

pair_modify mix arithmetic

pair_modify tail yes

bond_style harmonic

angle_style harmonic

read_data C3_nve.txt

orthogonal box = (0 0 0) to (50 50 50)

2 by 2 by 5 MPI processor grid

WARNING: Pair style in data file differs from currently defined pair style (…/read_data.cpp:478)

reading atoms …

12 atoms

scanning bonds …

1 = max bonds/atom

scanning angles …

1 = max angles/atom

reading bonds …

8 bonds

reading angles …

4 angles

2 = max # of 1-2 neighbors

1 = max # of 1-3 neighbors

1 = max # of 1-4 neighbors

2 = max # of special neighbors

molecule propane C3H8.txt

Read molecule propane:

3 atoms with 2 types

2 bonds with 1 types

1 angles with 1 types

0 dihedrals with 0 types

0 impropers with 0 types

group gas type 1 2

12 atoms in group gas

neighbor 1.5 bin

neigh_modify every 5 delay 0

compute mdtemp gas temp

compute_modify mdtemp dynamic yes

velocity gas create 353 31798 mom yes rot yes dist gaussian

fix const_bond gas shake 1.0e-8 200 0 b 1 a 1 t 1 2 m 15.035 14.027 mol propane

0 = # of size 2 clusters

0 = # of size 3 clusters

0 = # of size 4 clusters

4 = # of frozen angles

fix test gas nvt temp $t $t 10

fix test gas nvt temp 353 $t 10

fix test gas nvt temp 353 353 10

#fix_modify test temp mdtemp

thermo_style custom step atoms vol etotal temp press

thermo_modify format float %24.16g

thermo 1000

timestep 0.000001

run 10000

Neighbor list info …

1 neighbor list requests

update every 5 steps, delay 0 steps, check yes

max neighbors/atom: 2000, page size: 100000

master list distance cutoff = 15.5

ghost atom cutoff = 15.5

binsize = 7.75 -> bins = 7 7 7

Memory usage per processor = 7.76296 Mbytes

Step Atoms Volume TotEng Temp Press

0 12 125000 0.501845135900433 554.7142857142857 -109445658522.7831

1000 12 125000 -nan -nan nan

2000 12 125000 -nan -nan nan

3000 12 125000 -nan -nan nan

4000 12 125000 -nan -nan nan

5000 12 125000 -nan -nan nan

6000 12 125000 -nan -nan nan

7000 12 125000 -nan -nan nan

8000 12 125000 -nan -nan nan

9000 12 125000 -nan -nan nan

10000 12 125000 -nan -nan nan

Loop time of 0.593495 on 20 procs for 10000 steps with 12 atoms

Performance: 1.456 ns/day, 16.486 hours/ns, 16849.332 timesteps/s

99.6% CPU use with 20 MPI tasks x 1 OpenMP threads

MPI task timing breakdown:

C3_nve.txt (971 Bytes)

C3H8.txt (569 Bytes)

akohlmey · September 20, 2016, 3:02pm

Dear all,

I am trying to use Shake command for molecules.

First, the simulation of propane in bulk is being tried.

When I eliminated the SHAKE command, it worked and then I guess that the
command which I wrote has problems.

it works for me with LAMMPS version 15 Sep 2013

two comments:
- there isn't much of a benefit of running a 12 atoms system on 20
processors. in fact, it'll be much slower than running on 1 CPU
- your timestep is extremely small
- you may want to insert something like:

minimize 1.0e-1 1.0e-3 100 1000
reset_timestep 0

before you define fix shake, to remove some of the high potential
energy of your initial configuration.

axel.

111291 · September 20, 2016, 6:22pm

Thank you, sir.

It worked.

This time, I am trying to propane in slitpore, which consists of graphene layers.

The following is my input file.

variable t equal 353

variable seed equal 31798.0000

variable p equal 137.90

variable phi equal 0.4999

variable height1 equal 6.7001

variable height2 equal 50.550

units metal

atom_style full

boundary p p f

atom_modify map array

pair_style lj/cut 14.0

pair_modify mix arithmetic

bond_style harmonic

angle_style harmonic

read_data C3_4.4nm_2000psi.txt

molecule propane C3H8.txt

#group wall type 1

#group gas type 2 3

variable rxlo equal xlo+0.0001

variable rxhi equal xhi-0.0001

variable rylo equal ylo+0.0001

variable ryhi equal yhi-0.0001

variable rzlo equal zlo-0.0001

variable rzhi equal zhi+0.0001

region 1 block {rxlo} {rxhi} {rylo} {ryhi} {rzlo} {height1} units box side in

group wallbottom region 1

region 2 block {rxlo} {rxhi} {rylo} {ryhi} {height2} {rzhi} units box side in

group wallup region 2

region active block {rxlo} {rxhi} {rylo} {ryhi} {height1} {height2} units box side in

group wall union wallbottom wallup

set group wall type 1

group gas subtract all wall

#set group gas type 2

neighbor 1.5 bin

neigh_modify every 5 delay 0

velocity gas create t {seed} mom yes rot yes dist gaussian

compute mdtemp gas temp

compute_modify mdtemp dynamic yes

compute allperatom gas stress/atom NULL

compute allp gas reduce sum c_allperatom[1] c_allperatom[2] c_allperatom[3]

variable allpress equal -(c_allp[1]+c_allp[2]+c_allp[3])/(3*vol)

reset_timestep 0

###### EQUILIBRATE #######################################################

fix bottom wallbottom wall/lj1043 zlo ${height1} 0.00543 3.570 14.0 units box pbc no

fix_modify bottom energy yes

fix upper wallup wall/lj1043 zhi ${height2} 0.00543 3.570 14.0 units box pbc no

fix_modify upper energy yes

minimize 1.0e-1 1.0e-3 100 1000

fix const_bond gas shake 1.0e-8 200 0 b 1 a 1 t 2 3 m 15.035 14.027 mol propane

fix test gas nvt temp $t $t 100

fix_modify test temp mdtemp

thermo_style custom step atoms etotal vol c_mdtemp v_allpress press

thermo_modify format float %24.16g

for VMD

#dump vmd1 all atom 2 vmd_C3_4.4nm_2000psi.lammpstrj

#dump_modify vmd1 sort id

thermo 100

timestep 0.001

run 10000

The simulation only generated blank log files. When I removed the below ###### EQUILIBRATE ## line, it made log file.

I wonder which errors are related to blank log file (it sometimes happened) and where I should start to figure out this.

My data file and molecule file were attached.

Best regards,

Seunghwan Baek

C3_4.4nm_2000psi.txt (175 KB)

C3H8.txt (569 Bytes)

akohlmey · September 20, 2016, 10:02pm

Thank you, sir.

It worked.

This time, I am trying to propane in slitpore, which consists of graphene
layers.

no. it consists of wall fixes.

[...]

The simulation only generated blank log files. When I removed the below
###### EQUILIBRATE ## line, it made log file.

I wonder which errors are related to blank log file (it sometimes happened)
and where I should start to figure out this.

My data file and molecule file were attached.

you have to learn how to debug you inputs yourself. it is quite an
imposition to ask the mailing list to do that, whenever something
doesn't work and your email shows that you have made next to no effort
at all to understand, debug and figure out why this happens.

for a system as small and simple as this, you can easily run with a
single processor on a desktop, best with "-echo both" and careful
examination of the output.
when running in parallel, parallel libraries have a habit of
swallowing output. output to the log file is block buffered anyway and
will only be written in large chunks.

the LAMMPS documentation and the webpage give some helpful hints on
debugging as well.

most likely your configuration setup is bogus and leading to huge
forces and then the simulation will crash.

axel.