Atomic structure relaxation  [SOLVED]

Total energy, geometry optimization, DFT+U, spin....

Moderator: bguster

Locked
cgs12
Posts: 16
Joined: Thu Apr 27, 2017 6:27 pm

Atomic structure relaxation

Post by cgs12 » Fri Nov 02, 2018 8:47 pm

Hi,

I would like to ask if it is possible to perform full atomic position relaxation in Abinit. More specifically what I mean by this is that I do not have information on the space group of crystal structure in mind and wish to allow all atomic positions to relax fully so that the result converges to a particular space group corresponding to the minimum total energy at ground state. However, Abinit seems to perform relaxation based on the symmetry groups imposed in the starting structure, i.e. I am not able to start from a random space group and hope that the end result gives the correct space group.

Is it also possible to perform structure relaxation by fixing two lattice parameters and allowing the third lattice parameter to relax to minimize energy?

Thanks so much for your kind attention!

User avatar
new_986
Posts: 77
Joined: Wed Aug 04, 2010 9:27 pm

Re: Atomic structure relaxation

Post by new_986 » Mon Nov 05, 2018 11:32 am

Dear cgs12
see input variables opcell and ionmove ...
best wishes

Nawzad

cgs12
Posts: 16
Joined: Thu Apr 27, 2017 6:27 pm

Re: Atomic structure relaxation

Post by cgs12 » Mon Nov 05, 2018 4:51 pm

Dear Nawzad,

Thanks so much for your reply. What I meant was something similar to having optcell=2. However, optcell=2 allows for full relaxation subject to symmetry imposed in the initial guess structure. Here I do not know the symmetry groups of the materials of interest and wish to find those out through minimizing energy in DFT. Therefore full relaxation without symmetry imposed would be ideal.

Does it make sense for me to try perturbing all the atomic positions so that there is very minimal symmetry and nearly-full relaxation of nuclear positions can be done? Or must I resort to identifying all the potential space groups of the material of interest through experiments, other computational studies etc, and then perform DFT calculations accordingly?

Thanks!

User avatar
new_986
Posts: 77
Joined: Wed Aug 04, 2010 9:27 pm

Re: Atomic structure relaxation

Post by new_986 » Fri Nov 09, 2018 8:59 pm

cgs12 wrote:Dear Nawzad,

Thanks so much for your reply. What I meant was something similar to having optcell=2. However, optcell=2 allows for full relaxation subject to symmetry imposed in the initial guess structure. Here I do not know the symmetry groups of the materials of interest and wish to find those out through minimizing energy in DFT. Therefore full relaxation without symmetry imposed would be ideal.

Does it make sense for me to try perturbing all the atomic positions so that there is very minimal symmetry and nearly-full relaxation of nuclear positions can be done? Or must I resort to identifying all the potential space groups of the material of interest through experiments, other computational studies etc, and then perform DFT calculations accordingly?

Thanks!


Dear cgs12
give nsym input variable 0, it will give you number of symmetries and space group automatically...

beast regards
Nawzad

Boris
Posts: 128
Joined: Tue Feb 16, 2010 10:13 am
Location: France

Re: Atomic structure relaxation  [SOLVED]

Post by Boris » Mon Nov 12, 2018 1:02 pm

Hi,

If you want to remove all symmetry constraints in your calculation, just set nsym to 1.

Cheers
Boris
----------------------------------------------------------
Boris Dorado
Atomic Energy Commission
France
----------------------------------------------------------

Locked