Temperature increasing

Hi all.

My name is Hong Min.

I want to simulate graphene.

But its temperature increased dramatically and converged at abnormal value.

I dont know what the problem is.

I attach my input file and logfile below.

-------------input---------------------------------

dimension 3
boundary p p p
lattice sc 0.1
processors 8 1 1

units metal

atom_style charge
read_data data.graphene_rotation_run2

pair_style airebo 2.6 1 0
pair_coeff * * CH.airebo C

####################### groups / variables ###############################
group bottom id <= 1008
group upper id <> 1009 1032

group control_x1 id = 1033
group control_x2 id = 1034
group control_y1 id = 1035
group control_y2 id = 1036

group control union control_y1 control_x1 control_y2 control_x2
group U_C union upper control

neigh_modify exclude group control bottom exclude group control upper

#displace_atoms upper move 3.684 17.725 0.0 units box

variable x equal xcm(upper,x)
variable y equal xcm(upper,y)
variable z equal xcm(upper,z)
variable fx equal fcm(upper,x)
variable fy equal fcm(upper,y)
variable fz equal fcm(upper,z)
variable fsx_y1 equal f_6[1]
variable fsy_y1 equal f_6[2]
variable fsz_y1 equal f_6[3]
variable fs_y1 equal f_6[4]
variable fsx_x1 equal f_7[1]
variable fsy_x1 equal f_7[2]
variable fsz_x1 equal f_7[3]
variable fs_x1 equal f_7[4]
variable fsx_y2 equal f_8[1]
variable fsy_y2 equal f_8[2]
variable fsz_y2 equal f_8[3]
variable fs_y2 equal f_8[4]
variable fsx_x2 equal f_9[1]
variable fsy_x2 equal f_9[2]
variable fsz_x2 equal f_9[3]
variable fs_x2 equal f_9[4]

####################### velocity / fix ###############################

velocity all create 10.0 102486 mom yes rot yes dist gaussian units box

fix 1 upper npt temp 10.0 10.0 1.0 iso 1.01325 1.01325 10.0
fix 2 all langevin 10.0 10.0 1.0 699483
fix 4 control momentum 1 linear 1 1 1
fix 5 upper addforce NULL NULL -0.6491

fix 6 control_y1 spring couple upper 0.5 0 0 -1 INIT
fix 7 control_x1 spring couple upper 0.5 0 0 -1 INIT
fix 8 control_y2 spring couple upper 0.5 0 0 -1 INIT
fix 9 control_x2 spring couple upper 0.5 0 0 -1 INIT

fix 10 control momentum 1 linear 1 1 1

neigh_modify every 1 check yes

####################### output ###############################

dump 1 all atom 100 dump_rotation_run2.gra

thermo 100
thermo_style custom step ke pe etotal temp press

timestep 0.00001

run 10000
####################### run2 ###############################

fix 12 control move rotate 9.824 25.524 8.14 0 0 1 0.000001 units box
run 400000000
write_restart restart.graphene_rotation_run2

-------------------logfile-------------------------------------------------------------

KinEng PotEng TotEng Temp
1.337843 -7783.07 -7781.73 10
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13
3.44E+12 -0.00014 3.44E+12 2.57E+13

I see one problem. You probably want to put a "fix nve" command after
the "fix langevin" command.
I have been using this:

fix fxlan all langevin 300.0 300.0 1000.0 48279
fix fxnve all nve

(This is explained at
http://lammps.sandia.gov/doc/fix_langevin.html)

If that does not fix it, try reducing the complexity of your system
and your input file and see when the problem goes away. (It looks
like you already reduced the timestep.)
Good luck.

Hi all.

My name is Hong Min.

I want to simulate graphene.

But its temperature increased dramatically and converged at abnormal
value.

I dont know what the problem is.

usually bad initial coordinates or incorrect box dimensions
periodic cells. do a short run without a thermostat and
writing out positions at every step and visualize it.

that often immediately points out the problem.

axel.

axel.

I see one problem. You probably want to put a "fix nve" command after
the "fix langevin" command.
I have been using this:

fix fxlan all langevin 300.0 300.0 1000.0 48279
fix fxnve all nve

Oh dear. I just noticed that you have a fix "npt" in the line above.

Get rid of that line and try it with "fix nve" and see if the problem goes away.

I don't think you want to use "npt" here, because that fix also
effects the temperature of the system.

(The documentation says: "Likewise, this fix should not normally be
used on atoms that also have their temperature controlled by another
fix - e.g. by fix nvt or fix temp/rescale commands.")

Do you need to run your "production run" simulations at constant pressure?
If so, then perhaps "nph" would work? Try it and see.
If not, then "nve" should work fine.
Good luck.