Equilibration of two component system

Dear users,

I’m trying to model one metallic copper sample with a spherical pore inside the domain. Where the pore is contained argon gas. I’ve taken the forcefield coefficients from literature. In my model Cu-Cu have been described by EAM, where in case of Ar-Ar and Cu-Ar I use LJ potential.
During equilibration I use:

timestep 0.0001 # in metal unit
velocity all create 300 12345 mom yes rot no
fix 1 Cu nvt temp 300 300 0.01
fix 2 Ar nvt temp 300 300 0.01

for 100ps. In 2nd step to control pressure I used,

timestep 0.0001
fix 1 Cu nvt temp 300 300 0.01
fix 2 Ar nvt temp 300 300 0.01
fix 3 all press/berendsen iso 0.01 0.01 10
for another 100ps.

After this two steps temperature and pressure plot looks fine and as per the desired value assigned in the input deck. But the problem I face about the dynamics of the sample. When the thermodynamic variables hints that the sample is equilibrated, then the visualization of the steps shows that the pore which I created as spherical become distorted. I have attached the snapshot of that output.

As I want to get that pore with Ar to be spherical then what I’m doing wrong ?
Please make your important suggestion.
Thanks for your attention.

Regards,
Rajom

Cu_Ar_eql.png

Dear users,

I'm trying to model one metallic copper sample with a spherical pore inside
the domain. Where the pore is contained argon gas. I've taken the forcefield
coefficients from literature. In my model Cu-Cu have been described by EAM,
where in case of Ar-Ar and Cu-Ar I use LJ potential.
During equilibration I use:

timestep 0.0001 # in metal unit
velocity all create 300 12345 mom yes rot no
fix 1 Cu nvt temp 300 300 0.01
fix 2 Ar nvt temp 300 300 0.01

for 100ps. In 2nd step to control pressure I used,

timestep 0.0001
fix 1 Cu nvt temp 300 300 0.01
fix 2 Ar nvt temp 300 300 0.01
fix 3 all press/berendsen iso 0.01 0.01 10
for another 100ps.

After this two steps temperature and pressure plot looks fine and as per the
desired value assigned in the input deck. But the problem I face about the
dynamics of the sample. When the thermodynamic variables hints that the
sample is equilibrated, then the visualization of the steps shows that the
pore which I created as spherical become distorted. I have attached the
snapshot of that output.

As I want to get that pore with Ar to be spherical then what I'm doing wrong
?

that is really difficult to say. you are trying to do something that
can be done in several different ways and particularly, you are trying
to do multiple things at the same time. some observations from the
fragments of information you provide.

- this is mostly a problem of the science and doing MD simulations of
a complex system. don't expect to get much help from the mailing list,
since there is nothing formally wrong with what you are doing. you
probably should better discuss with your adviser and colleagues.

- you have to try and do things in smaller increments/steps so you can
more easily identify where things stop working the way you expect.

- i would first try and see, if you can build the copper system
without the argon and equilibrate it and see if there is an
instability

- you may also need to check if the lattice constant you are using for
copper is consistent with the eam potential. otherwise, you first need
to equilibrate a bulk copper system and then cut your pore from that

- your argon density looks high. more like a dense liquid. not a gas.

- i would not do a box relaxation for the entire system. probably not
even for the pore. see above about the bulk.

- isotropic box relaxation is most definitely a bad idea for a system like yours

axel.