[QE-users] Molecular dynamics calculation crashing

[Elio Physics]

Dear Nicola and Pietro

Thank you for your advice and quick response. I have checked the initial supercell structure , before vc-relaxing it and after vc-relaxing it (that is before running the MD calculation). The structure looks fine. Note that the vc-relax calculation goes smoothly with no problems at all.

I will try to fix the problem, by choosing the parameters from the QE input generator that professor Nicola suggested. If the problem persists, I will post the input file on here.

Regards


________________________________
From: users <users-bounces at lists.quantum-espresso.org> on behalf of Pietro Davide Delugas <pdelugas at sissa.it>
Sent: Monday, August 8, 2022 4:55 AM
To: Quantum ESPRESSO users Forum <users at lists.quantum-espresso.org>
Subject: [QE-users] R: Molecular dynamics calculation crashing

Dear Elio

a  few  usual  things to check are:

  *   the positions and lattice description  in the supercell input. A quick glance with xcrysden is usually enough
  *    the value of the initial forces ( if the issue occurs after the first step)
  *   that you have set up correctly occupations, smearing and the density of  k-point mesh.

Apart from those it's difficult to give any other suggestion without seeing the input.
best regards
Pietro

________________________________
Da: users <users-bounces at lists.quantum-espresso.org> per conto di Elio Physics <Elio-Physics at live.com>
Inviato: lunedì 8 agosto 2022 03:02
A: Quantum Espresso users Forum <users at lists.quantum-espresso.org>
Oggetto: [QE-users] Molecular dynamics calculation crashing

Dear QE users,

I am trying to perform some molecular dynamics (MD) calculations on some 2D system I am working on. The unit cell has two atoms per unit cell. I have created a 3x3x1 supercell (18 atoms) ,  which will be the target of the MD job.
However, whatever I do, I keep on getting the famous 'too many bands not converged..' error, with the code crashing instantly afterwards. I know that this problem has been reported before. I have tried the suggested remedies:


  1.  My ecutwfc is already high (130Ry).
  2.   I have decreased the conv_thr to 10^(-10).

Neither solution seems to resolve the issue. Note that , if I use the unit cell (of two atoms) to perform the MD, the code works perfectly fine.
However, it is important to perform the MD on a supercell of the material.

Any other suggestions? my version is QE-6.7

Regards and thanks in advance

Elio Moujaess
Adjunct professor
Federal University of Rondonia
Brazil




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20220808/d9d9d0a7/attachment.html>