[Pw_forum] problem in vc-relax convergence

l.eslami at srbiau.ac.ir l.eslami at srbiau.ac.ir
Tue Aug 29 07:07:33 CEST 2017

On 2017-08-28 11:40, Lorenzo Paulatto wrote:
> Dear Leila,
> if the total energy keeps decreasing, that it is completely normal: it
> means that you started from an energy saddle point. It may however
> indicate that your initial crystal structure was wrong.
> kind regards
> On 28/08/17 07:22, l.eslami at srbiau.ac.ir wrote:
>> Hi all,
>> We are trying to perform vc-relax calculation for a single layer ZnO
>> sheet in the presence of an oxygen vacancy in a 4*4 suppercell. We 
>> have
>> used espresso-5.4.0. with different psudopotentials such as GGA,
>> NORMCONS and non relativistic Pseudopotentials: "O.pbe-mt_fhi.upf and
>> Zn.pbe-mt_fhi.upf,  as well as GBRV pseudopotentials:
>> “o_pbe_v1.2.uspp.F.UPF” and “zn_pbe_v1.uspp.F.UPF”. The calculations 
>> do
>> not return any error. However, convergences of force and pressure do 
>> not
>> occur even after long time running of the calculations. In fact, the
>> code starts to decrease force and pressure but after a couple of
>> relaxation steps they increase. I would be grateful if anyone could
>> kindly give me any piece of advice to solve the problem.
>> many thanks,
>> Leila Eslami
>> Department of physics,
>> Islamic Azad University,
>> Tehran, Iran
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://pwscf.org/mailman/listinfo/pw_forum

Dear Lorenzo,
thank you for your kind attention. I checked the total energy of the 
system and it keeps decreasing. However, there is another ambiguous 
point in the results. In fact, as our system is a sheet and we set 15 
Angstrom distance between subsequent layers in z-direction, I expect 
that the z-components of atomic positions do not change during relation 
process. But after a couple of relation steps this happens. To me it 
doesn't seem normal!!!!

Leila Eslami
  Department of physics,
  Islamic Azad University,
  Tehran, Iran

More information about the users mailing list