回复:Re: How to deal with the atom overlap in GCMC ?

Hi,

I use the nvt as the replacement of nve to equilibrated the structure and at the same time control the temperature. But from the output, I found the temperature of the system A and B is so cold and reach to zero. I cannot understand why it occured. Could you give some suggestion about temperature control ? And my system is like this: two regions A and B in the same box, temperature is the same 333K, pressure of A I set up is 50MPa, B 49MPa. I want the two region reach to the equibrium respectively and then perform the NEMD between A and B region.

Thanks! Don B.

Hi Don,
Please see the links below. You may find it helpful.

https://lammps.sandia.gov/threads/msg70362.html

https://lammps.sandia.gov/threads/msg58606.html

With the timestep decreased to 0.1 fs and gcmc 1000 seems to run smoothly (see the attachment). You could start in this way then increase the timestep.

(I have instead a doubt on the input, is this assignment of the group that you want?

variable 1 atom “type==1”
variable 2 atom “type==2”
what you group 3 dynamic 1 var 1
group 4 dynamic 2 var 2

Perhaps are you looking for a dynamical assignment or the atoms within the two region

something like “dynamic region”. It seems to me better, because you use these groups to compute the pressure…)

in.t1 (3.18 KB)

output_file (50.7 KB)

Hello everyone,

Hi Don B. Sorry to post my concern here but I too have the same problem. When I use fix gcmc in tandem with NVT, the atoms just escape simulation box though I don’t get any error message. I have tried the simulation by replacing “NVE” with “NVT+NPT”. Still the same result. Atoms escape the simulation box and I don’t have problems with attaining a particular system temperature during simulation. Btw I use lammps-12Dec18 version and run the simulations on HPC cluster. Please let me know if anyone has anything else to add to it. Thank you in advance.

variable T equal 1300.00
units metal
dimension 3
boundary p p p
atom_style atomic
read_data 0.1%P_Fe_bulk.lmp # 0.1% P in Fe bulk .
pair_style meam
pair_coeff * * library.meam Fe P FeP.meam Fe P
group Fe type 1
group P type 2
fix 1 all box/relax iso 0.0001 nreset 10
thermo 10
thermo_style custom step temp pxx pyy pzz press pe etotal lx ly lz
minimize 1.0e-12 1.0e-12 100000 100000
unfix 1

velocity all create $T 825577 mom yes rot no
reset_timestep 0

fix 6 all nve
fix 7 all langevin $T $T 0.1 343434
run 100000
unfix 6
unfix 7
dump 2 all custom 10 dump.GCMC id type xu yu zu
dump_modify 2 append yes
reset_timestep 0
timestep 0.001

fix 1 Fe gcmc 1 10 10 1 294948 $T -3.0 0.1 group Fe overlap_cutoff 1.0
fix 2 P gcmc 1 10 10 2 271828 $T -3.9 0.1 group P overlap_cutoff 1.0
compute mdtemp all temp
compute_modify mdtemp dynamic/dof yes
fix mdnvt all nvt temp $T $T 0.1
fix_modify mdnvt temp mdtemp
run 100000
print “All done”
clear

Praveen

Maybe you can use fix nve/limit ( https://lammps.sandia.gov/doc/fix_nve_limit.html ) for a few timesteps right after the GCMC to allow the “relaxation” of the inserted atoms and then switch to NVT with a normal timestep

Στις Τετ, 3 Απρ 2019 στις 11:00 π.μ., ο/η Andrea Pedrielli <[email protected]…6374…> έγραψε:

Yes I found lost atom after some time (I didn’t see it before). It seems that increasing overlap_cutoff to 1.5 is ok.
Then you could re-increase the timestep.
New version attached.

in.t1 (3.19 KB)

output_file (32.4 KB)