[QE-users] Problems with Relax and EPC Calculations in QE

[Claudio Antonio Perottoni]

Dear Xie,

On Wed, 2025-12-17 at 10:56 +0800, 谢征奥 wrote:

Dear QE Community,

I hope you are having a productive day and that your research commitments
are not overly demanding. Allow me to introduce myself: I am Xie, a Chinese
researcher, and I am new to using QE (Quantum ESPRESSO). I have a strong
interest in computational physics, which has led me to deepen my
application of this software for my research projects. I am writing today
to ask three specific questions regarding the calculations I am conducting,
as I have encountered some difficulties and greatly value your expertise
and extraordinary experience in the field of QE, given that you are one of
the core developers of the software.


   1. When performing phonon spectrum calculations with QE, what is the
   correct sequence of steps for structural relaxation? Should I first run a
   vc-relax calculation, then increase the force and energy convergence
   thresholds before performing a subsequent relax calculation? Or is it
   sufficient to verify the convergence of the total stress from the
   vc-relax results and use them directly for subsequent calculations? In
   my case, the pressure obtained from vc-relax has met my convergence
   criteria, but when I used the optimized structure to run a relax
calculation,
   the pressure value became significantly higher: is this a reasonable
   behavior?

   A well converged vc-relax calculation should be enough. Recent QE
   versions make a final SCF calculation at the end to check for forces and
   pressures in the relaxed structure. If pressure (absolute value) is higher
   than, say, 0.5 kbar, run a second vc-relax calculation starting with the
   previously relaxed structure.

   2. When using the pw.x module, which parameters have a significant
   impact on eliminating imaginary frequencies?

   Convergence criteria for forces on atoms may be stricter than usual to
   avoid imaginary frequencies. As always, check for convergen w.r.t. ecutwfc
   and k-points grid.

   3. During the structural optimization process, is it acceptable to
   always use the parameter ibrav=0 in the relax.in input file?

   It is advisable to use ibrav according to the actual Bravais lattice of
   the compound you are calculating.

   Hope this helps!
   Claudio


Thank you in advance for the time you will dedicate to my questions and for
your valuable guidance. As a novice in QE, every suggestion will be
fundamental to advancing my learning journey and ensuring the success of my
calculations. If you require any further details about my input files or
the results obtained, I would be happy to provide all the necessary
information.

Sincerely,

Xie

Chinese Researcher


_______________________________________________________________________________
The Quantum ESPRESSO Foundation stands in solidarity with all civilians
worldwide who are victims of terrorism, military aggression, and
indiscriminate warfare.
--------------------------------------------------------------------------------
Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
users mailing list users at lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users


-- 

Claudio A. Perottoni
Universidade de Caxias do Sul
95070-560 Caxias do Sul - RS - Brazil
https://orcid.org/0000-0002-8425-845X

-- 
Enviado via UCSMail.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20251217/5367c2c4/attachment.html>