MgO bulk structure simulation

To Whom It May Concern,

For over couple of weeks I have tried to simulate MgO bulk structure with using read_data and prepared potential.

However, I could not figure it out how to simulate that with self learning. (In my research group no one uses LAMMPS at the moment)

I know this is not appropriate to ask a broaden question but it will be very helpful for me to start LAMMPS.

Here is the input file (below) that I built and lammps_coord (https://drive.google.com/open?id=1r7K_kCNpCpJ3wuk-1TKOVQgb_Gi4ofo8), and lammps_ip (https://drive.google.com/open?id=1T0OLX1iwvLzRXGb84-Ko76w_yB8xAZOb)

------------------------ INITIALIZATION ----------------------------

units real
dimension 3
boundary p p p
atom_style charge

----------------------- ATOM DEFINITION ----------------------------

read_data lammps_coord

neighbor 0.3 bin
neigh_modify delay 10 one 10000
#neigh_modify every 1 delay 0 check yes

------------------------ FORCE FIELDS ------------------------------

pair_style hybrid/overlay coul/long 12.0 table spline 20001
kspace_style ewald 0.000001
pair_coeff * * coul/long
pair_coeff 1 2 table lammps_ip Mg_O
pair_coeff 2 2 table lammps_ip O_O

set type 1 charge 2.0 # Mg
set type 2 charge -2.0 # O

------------------------- SETTINGS ---------------------------------

variable density equal “(mass(all)*1.660538782) / (vol)”

You are trying to get the kind of help that you would get from a personal tutor. the mailing list cannot provide this and e-mail is not a good communication tool for this.
you have to realize that your problem seems to be two-fold: 1) you are not familiar with LAMMPS syntax, 2) you are not familiar with MD simulation basics. so even if there is nobody in your group that can help you with the first issue, there should be somebody that can help your with the second (and if there isn’t, then you should not be doing what you are trying to do, since nobody will be able to provide the necessary oversight and tutoring). the best way of obtaining training in issue 1) is to ask your PI to arrange for getting tutored (in person!) by a collaborator or otherwise knowledgeable person/acquaintance. otherwise you are doomed to make many mistakes and often those that are not so obvious and thus would waste a lot of (your own, but also your PI’s) time and possibly will have bogus results.

axel.

Dear Dr. Kohlmeyer

I apologise for the unsuitable questions before. Moreover, I appreciate all of your kind advice.

I figured out how to simulate nanoclusters in LAMMPS, the input as you can see in the below:

------------------------ INITIALIZATION ----------------------------

units real
dimension 3
boundary f f f
atom_style charge

----------------------- ATOM DEFINITION ----------------------------

read_data geometry

neighbor 2.0 bin
#neigh_modify delay 10 one 10000
neigh_modify every 1 delay 0 check yes

------------------------ FORCE FIELDS ------------------------------

pair_style hybrid/overlay coul/msm 12.0 buck 12.0
kspace_style msm 0.000001
pair_coeff * * coul/msm
pair_coeff 1 2 buck 21829.732 0.31813 0.0000
pair_coeff 2 2 buck 524950.1768 0.14900 642.904849

set type 1 charge 2.0 # Mg
set type 2 charge -2.0 # O

------------------------- SETTINGS ---------------------------------

variable density equal “(mass(all)*1.660538782) / (vol)”