On Sun, Sep 20, 2020 at 6:39 PM Michal Krompiec <michal.kromp...@gmail.com> wrote:
> ad 1) If I understand correctly, it is because the basis set would change > in every step, so the wavefunction from the preceding step couldn’t be used > as a starting point for the scf, unless the bands were projected onto the > new PW basis. > This is just part of the story, and not the most important one. Working with a fixed basis set ensures a smoother behavior when convergence with respect to the plane-wave cutoff is not very good. In addition, it is much easier to program. Paolo Best, > Michal Krompiec > Merck KGaA > > On Sun, 20 Sep 2020 at 17:25, Robert Molt < > r.molt.chemical.phys...@gmail.com> wrote: > >> If I may ask a question to clarify a point within this thread…I have >> always had some confusion on exactly why the the # of basis vectors is not >> conserved step-to-step. This is an implicit aspect to the question being >> asked, if I understand properly. >> >> 1.) Why is the *number* of plane waves during vc-relax steps constant? >> Why is not held to a constant energy cutoff, like in an energy calculation? >> >> My understanding is that ecutrho and ecutwfc “constrain” the number of >> plane waves during a vc-relax calculation; these two variables completely >> specify the number of allowed plane waves consistent with the initial unit >> cell volume. Is this correct? >> >> My understanding is that if you wanted a constant energy cutoff, you have >> to vary the number of plane waves, since the volume is changing. Why is >> this undesirable to do? Is it because comparing different numbers of plane >> waves results in different energies, such that no fair comparison can be >> made between calculations with different numbers of basis vectors? >> >> 2.) On this specific question asked by Mohad, I am slightly confused. Is >> the essence of the problem that the pressure calculation threshold is too >> low OR is it that you should never use the “properties” other than lattice >> constants from a vc-relax for the aforementioned reason? In my head, these >> are seemingly independent issues in how the calculations are working? >> >> Dr. Robert Molt Jr. >> Indiana University Purdue University >> >> >> On Sep 20, 2020, at 6:03 AM, Lorenzo Monacelli < >> lorenzo.monace...@roma1.infn.it> wrote: >> >> >> >> >> >> >> >> >> >> If I interpret correctly the question, you are seeing a P = 6.57 >> >> kbar and not 7 as in the pw.x input press keyword because, when qe >> >> relax the unit cell, it constrains the basis defined with ecutwfc >> >> and ecutrho. This basis, as it is formed by plane waves with >> >> periodic boundaries, depends on the cell size. So the code will >> >> relax reaching 7 kbar (within the accuracy you chose in the input, >> >> usually 0.1 kbar), then it will run a new calculation in the final >> >> cell recomputing the basis with the new cell. So the value of the >> >> pressure will be slightly different. >> >> To be consistent, you should use 6.57 kbar in your plots, as it >> >> is the final pressure with the cutoffs you used for your >> >> calculation, and does not depend on the starting cell. If you want >> >> exactly 7 kbar, you can start a new vc-relax using the final >> >> structure you have at 6.57 kbar as starting point, and it will get >> >> to a much closer value. >> >> Bests, >> >> Lorenzo >> >> >> >> >> Il 20/09/20 10:54, Stefano Baroni ha >> >> scritto: >> >> >> >> >> >> >> >> >> What is difference between the frequencies computed at: i) P=0; >> >> ii) P=6.57 kbar; iii) 7 kbar? The answer to my question will >> >> imply that to yours. SB >> >> >> >> >> >> >> >> >> — >> >> >> Stefano Baroni - SISSA, Trieste - http://stefano.baroni.me >> , stefanobaroni (Skype) >> >> >> >> >> >> If the prediction that an airplane can stay up depends >> >> on the difference between Riemann and Lebesgue >> >> integration, I don’t want to fly in it [Richard W. >> >> Hammings] >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> On 20 Sep 2020, at 10:42, Mohad Abbasnejad >> >> <mohaddeseh.abbasne...@gmail.com> >> >> wrote: >> >> >> >> >> >> >> >> >> >> >> >> Hello Dear QE users >> >> >> >> >> I am studying the effect of pressure on >> >> the frequencies. >> >> >> In the experimental article, the >> >> pressure on the solid has been reported to be 7 >> >> kbar. >> >> >> >> >> >> >> >> When I put this pressure (7 kbar) on my >> >> structure and let it relax, the final pressure is >> >> calculated to be 6.57 kbar as the following. >> >> Therefore it is not exactly 7 kbar. >> >> >> >> >> >> >> >> >> >> Computing stress (Cartesian axis) and >> >> pressure >> >> >> >> >> >> total stress (Ry/bohr**3) >> >> (kbar) P= 6.57 >> >> >> 0.00004467 0.00000000 -0.00000000 >> >> 6.57 0.00 -0.00 >> >> >> 0.00000000 0.00004467 -0.00000000 >> >> 0.00 6.57 -0.00 >> >> >> -0.00000000 -0.00000000 0.00004467 >> >> -0.00 -0.00 6.57 >> >> >> >> >> >> >> >> >> >> >> >> >> MY QUESTION: Which pressure should I >> >> report on my paper; 7 kbar or 6.57 kbar? >> >> >> >> >> >> >> >> >> >> >> >> >> Any comments would be appreciated. >> >> >> >> >> >> >> >> Best regards >> >> >> >> >> Mohaddeseh >> >> >> >> >> >> >> >> >> >> >> -- >> >> >> --------------------------------------------------------- >> >> >> Mohaddeseh Abbasnejad, >> >> >> Assistant Professor of Physics, >> >> >> Faculty of Physics, >> >> >> Shahid Bahonar University of Kerman, >> >> >> Kerman, Iran >> >> >> P.O. Box 76169-133 >> >> >> Tel: +98 34 31322199 >> >> >> Fax: +98 34 33257434 >> >> >> Cellphone: +98 917 731 7514 >> >> >> E-Mail: mohaddeseh.abbasne...@gmail.com >> >> >> Website: academicstaff.uk.ac.ir/moabbasnejad >> >> >> >> >> --------------------------------------------------------- >> >> >> >> >> >> >> >> >> >> >> >> _______________________________________________ >> >> >> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) >> >> >> users mailing list users@lists.quantum-espresso.org >> >> >> https://lists.quantum-espresso.org/mailman/listinfo/users >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> _______________________________________________ >> >> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) >> >> users mailing list users@lists.quantum-espresso.org >> >> https://lists.quantum-espresso.org/mailman/listinfo/users >> >> >> >> >> >> >> >> >> >> _______________________________________________ >> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) >> users mailing list users@lists.quantum-espresso.org >> https://lists.quantum-espresso.org/mailman/listinfo/users >> >> >> _______________________________________________ >> >> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) >> >> users mailing list users@lists.quantum-espresso.org >> >> https://lists.quantum-espresso.org/mailman/listinfo/users > > _______________________________________________ > Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) > users mailing list users@lists.quantum-espresso.org > https://lists.quantum-espresso.org/mailman/listinfo/users -- Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche, Univ. Udine, via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222
_______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) users mailing list users@lists.quantum-espresso.org https://lists.quantum-espresso.org/mailman/listinfo/users