Hi Dear all LAMMPS users
I define a binary alloy Fe-C with bcc Fe matrix and carbon in octahedral sites, but when i want to relax the system using following command i give zero for all temp, …It is a little strange that when i run the code with other computer the mentioned parameters rise very coarsely (same lammps version) and atoms lost error in the end.
pair_style eam/fs
pair_coeff * * Fe-C_Hepburn_Ackland.eam.fs Fe C
#==================COMPUTE=======================
neighbor 0.3 bin
neigh_modify delay 5
compute keng all ke/atom
#==================RELAXATION=======================
fix 1 all nvt temp 300.0 300.0 0.1
thermo 100
thermo_style custom step dt temp press pe ke etotal lx ly lz
dump mydump1 all custom 1000 dump-relax.txt id type x y z c_keng
min_style cg
minimize 1e-15 1e-15 5000 5000
timestep 0.0001
run 10000
when i use a PKA in the next step of the code, i wonder why PKA just move and other atoms are frozen.
probably i think the problem is because of the definition of the potential. Could you please help me about this problem?
Thanks
S.M.Zamzamian
Hi Dear all LAMMPS users
I define a binary alloy Fe-C with bcc Fe matrix and carbon in octahedral
sites, but when i want to relax the system using following command i give
zero for all temp, ...It is a little strange that when i run the code with
other computer the mentioned parameters rise very coarsely (same lammps
version) and atoms lost error in the end.
are you saying that the exact same input for the exact same LAMMPS version
on two different machines gives different results?
pair_style eam/fs
pair_coeff * * Fe-C_Hepburn_Ackland.eam.fs Fe C
#==================COMPUTE=======================
neighbor 0.3 bin
neigh_modify delay 5
compute keng all ke/atom
#==================RELAXATION=======================
fix 1 all nvt temp 300.0 300.0 0.1
thermo 100
thermo_style custom step dt temp press pe ke etotal lx ly lz
dump mydump1 all custom 1000 dump-relax.txt id type x y z c_keng
min_style cg
minimize 1e-15 1e-15 5000 5000
timestep 0.0001
run 10000
when i use a PKA in the next step of the code, i wonder why PKA just move
and other atoms are frozen.
probably i think the problem is because of the definition of the
potential.
it is strange that people always blame the potentials. what you describe
sounds more like a misplaced use of a thermostat. when you add energy to
your PKA, you rise the temperature of the system. the thermostat will try
to maintain the desired temperature. since you have this one extremely hot
atom, but the thermostat applies to *all* atoms, it will slow down all of
them.
Could you please help me about this problem?
the kind of simulation you describe is usually done without a thermostat.
you equilibrate the substrate to the desired temperature and then switch to
plain time integration with fix nve.
there are several other choices in your input that are dubious: your
neighborlist skin is horribly small and a neighborlist delay is not a good
idea once you create a very hot/fast atom. you may also need to check on
your time step.
axel.
With regard
Thank you Dr.Kohlmeyer for answering
I did not blame the potential, because i know that the potentials passed a massive tests and they are created by well-established researchers. I meant that “did i define the potential correctly?” or i must define it in other manner. I do switch to fix NVE after creating PKA from fix NVT, but as i mentioned the PKA will only move and other atoms are frozen. whereas, when i define a pure Fe using “Fe_2.eam.fs” potential the thermal spike correctly occurs and annealing befalls too. but when i create even just one or 2 atoms an use “Fe-C_Hepburn_Ackland.eam.fs” potential, the above problem occurs. I am very confused. I changed and replaced everything (timestep, damping, atom creating…). even i used a ZBL potential, but in that case the Fe matrix were deformed!!! it seemed that the C atom forced Fe atoms in a horrible manner. I rather exactly followed the same as mentioned in literature and papers, but i gave results for pure Fe but not for binary Fe-C. I do not what to do.here is my input:
lattice bcc 2.855 orient x 1 0 0 orient y 0 1 0 orient z 0 0 1
region myreg block 0 10 0 10 0 10 units lattice
create_box 2 myreg
create_atoms 1 box
create_atoms 2 single 1 1 0.5
mass 1 55.845
mass 2 12.011
#==================POTENTIAL=======================
pair_style eam/fs
pair_coeff * * Fe-C_Hepburn_Ackland.eam.fs Fe C
#==================COMPUTE=======================
compute keng all ke/atom
#==================RELAXATION=======================
timestep 0.0001
thermo 100
fix 1 all nvt temp 300.0 300.0 0.1
thermo 10
thermo_style custom step dt temp press pe ke etotal lx ly lz
dump mydump1 all custom 1000 dump-relax.txt id type x y z c_keng
run 20000
unfix 1
undump mydump1
#=======================PKA =================
group PKA id == 1111
velocity PKA set 99.3 297.9 496.5 sum no units box
#========================ANNEALING===================
reset_timestep 0
timestep 0.0001
fix 3 all nve
fix 4 all dt/reset 10 1.0e-5 0.001 0.1 units lattice
thermo_style custom step dt temp press pe ke etotal lx ly lz
dump mydump2 all custom 1000 dump-damage.txt id type x y z c_keng
run 800000
Best regards
S.M.Zamzamian
With regard
Thank you Dr.Kohlmeyer for answering
I did not blame the potential, because i know that the potentials passed a
massive tests and they are created by well-established researchers. I meant
that "did i define the potential correctly?" or i must define it in other
manner.
did you do it like it is described in the documentation? that is all that
matters. you can double check that yourself; your case is simple enough.
I do switch to fix NVE after creating PKA from fix NVT, but as i mentioned
the PKA will only move and other atoms are frozen. whereas, when i define a
pure Fe using "Fe_2.eam.fs" potential the thermal spike correctly occurs
and annealing befalls too. but when i create even just one or 2 atoms an
use "Fe-C_Hepburn_Ackland.eam.fs" potential, the above problem occurs. I am
very confused. I changed and replaced everything (timestep, damping, atom
creating...). even i used a ZBL potential, but in that case the Fe matrix
were deformed!!! it seemed that the C atom forced Fe atoms in a horrible
manner. I rather exactly followed the same as mentioned in literature and
papers, but i gave results for pure Fe but not for binary Fe-C. I do not
what to do.
here is my input:
this is not really asking about LAMMPS but about how to do your research.
that is something you have to discuss with your
adviser/supervisor/colleagues and also by checking the literature. for
example, have the potential parameters you are using been used for a
similar study before? LAMMPS will just do what you ask it to do in your
input script. whether that is meaningful and will result in meaningful
trajectories is a whole different matter.
axel.