[QE-users] vc-relax for orthorhombic systems
Aleksandra Oranskaia
aleksandra.oranskaia at kaust.edu.sa
Mon Apr 8 12:41:16 CEST 2019
Dear Giuseppe and Lorenzo,
Thank you both for sharing your knowledge, I truly appreciate all your comments and will think on them.
Let me reply:
to Giuseppe:
> Why do you need to be so tightly close to experimental lattice parameters?
Yes, I do try to validate setups for using them in evolutionary crystallography methods (A.Oganov methods) that are based on large scale optimizations (for them one needs settings as “cheap” as possible, but still reasonably correct).
In general, I would always trust functional/pseudopotentials that give close to experimental structural parameters (I try to use as a reference low temperature XRD for lattice constants, like obtained at 5K) as well as band gaps (as reasonable as DFT can do it, I also check them using higher levels of theory that are too expensive for optimizations).
Fixing experimental lattice constants introduce artifice forces/stresses, so I avoid this approach when possible.
to Lorenzo:
> Also, keep in mind that vc-relax is done at fixed number of plane-waves, not constant cutoff. You should always repeat it starting from the final configuration and
Could you please rewrite the last sentence?
> Correcting for the lack of VdW interaction with a simple Grimme-D2 method could be enough to get a better agreement.
Thanks for the suggestion, I did tested all vdw corrections (for PBE and PBEsol) and majority of vdw functionals (btw, for my systems D2 gave crazy numbers).
to everybody:
3 days in a row I was testing everything that my imagination allowed to and came up to a conclusion that with soft materials, with orthorhombic cells of > 20 atoms there are TOO many degrees of freedom to expect vc-relax to give good numbers for cell parameters.
Within the same third decimal digit in total energy (-xxxx.xxX Ry) one might came to slightly(?!) different cells depending on starting points and general logic of relaxation (say, you start from orthorhombic space group doing relax and for a corresponding ibrav continue with vc-relax Or you start from the closest tetragonal space group and continue with xyz vc-relax).
I will also do a few variable-cell NEB (coupled with QE for ab initio part) between the orthorhombic and tetragonal systems, maybe it will help me more to better understand the problem.
PS I have a full right to be wrong I am learning and will be thankful for any feedback ;-)
Best,
Alex.
___
Aleksandra Oranskaia (M.Sc.)
ChemS PhD student, KAUST
Phone: +966 50 1335254
> On Apr 8, 2019, at 1:00 PM, users-request at lists.quantum-espresso.org wrote:
>
> Send users mailing list submissions to
> users at lists.quantum-espresso.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
> https://lists.quantum-espresso.org/mailman/listinfo/users
> or, via email, send a message with subject or body 'help' to
> users-request at lists.quantum-espresso.org
>
> You can reach the person managing the list at
> users-owner at lists.quantum-espresso.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> Today's Topics:
>
> 1. Re: Compilation on a server ends with error 1 (Fabio Costa)
> 2. Pseudopotential type is different in the website description
> and in the file (Rajesh Raju)
> 3. Re: Compilation on a server ends with error 1
> (Pietro Davide Delugas)
> 4. Ab-Initio Thermal Transport (Arihant Bhandari)
> 5. (no subject) (as gj)
> 6. Re-use of charge density don't give exactely the same result
> in scf (JAY Antoine)
> 7. Re: Re-use of charge density don't give exactely the same
> result in scf (Pietro Delugas)
> 8. Re: Re-use of charge density don't give exactely the same
> result in scf (Pietro Delugas)
> 9. Re: Re-use of charge density don't give exactely the same
> result in scf (Paolo Giannozzi)
> 10. Re: vc-relax for orthorhombic systems (Aleksandra Oranskaia)
> 11. Re: vc-relax for orthorhombic systems (Lorenzo Paulatto)
> 12. Re: vc-relax for orthorhombic systems (Giuseppe Mattioli)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 7 Apr 2019 23:53:39 +0000
> From: Fabio Costa <fabiocosta4 at hotmail.com>
> To: Quantum Espresso users Forum <users at lists.quantum-espresso.org>
> Subject: Re: [QE-users] Compilation on a server ends with error 1
> Message-ID:
> <CP2PR80MB192453E4DE82D4BE5A5A060CF2530 at CP2PR80MB1924.lamprd80.prod.outlook.com>
>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Dear Paolo and Pietro
>
> Thanks for the assistance, and sorry for the late reply on your advice.
>
> I contacted the system administrator, and he explained to me that as there are several other applications running under this old gfortran version, and because of that he could not just issue "sudo apt update gfortran", on the risk of messing up with all the other stuff currently running on the system. His advice to me was to manually install a newer version inside my workspace.
>
> I did that, and now I have both the native 4.4.6 and the 5.4.0 gfortran compilers. I also added to my bashrc the line "export F90=/path/to/gfortran5.4.0/". When i run configure, it detects the gfortran v5.4.0, so I think that I did this right. At first, make all still crashes at phcg.x. I proceeded changing the lines in Makefile as Pietro sugested, and when compiling again it goes a bit further, but still crashes. To avoid these crashes, I tryed to compile just the packages I'm intending to use (pw, ph, w90 and epw). By issuing make pw, followed by make ph, make w90, the task is succesful, but It craches again when I issue make epw. The print screen of the error message is in the link: https://imgur.com/CerJStE .
>
> Also, I'm still getting the same errors with the compiled executables. About Paolo's advice, in my current situation, where I cant just update the whole thing, and kinda have two version of gfortran. I'm not sure if the crashes are because when the executables run, they are still linking to the gfortran 4.4.6. What course of action would you sugest?
>
> Thank you for the assistance
>
> F?bio Costa
> MSc student in physics
> Federal University of Bahia, Brazil
>
> ________________________________
> De: users <users-bounces at lists.quantum-espresso.org> em nome de Paolo Giannozzi <p.giannozzi at gmail.com>
> Enviado: quinta-feira, 28 de mar?o de 2019 11:40
> Para: Quantum Espresso users Forum
> Assunto: Re: [QE-users] Compilation on a server ends with error 1
>
> There are two unrelated problems:
> - the linker is unable to find the main program in PHonon/Gamma. I have never seen this happen before, but the link is done in a slightly different way from all other cases and this will be fixed
> - the executable doesn't work likely due to a known problem of many old compilers with unallocated optional variables
>
> Paolo
>
>
> On Thu, Mar 28, 2019 at 3:47 AM Fabio Costa <fabiocosta4 at hotmail.com<mailto:fabiocosta4 at hotmail.com>> wrote:
> Dear all
>
> I'm trying to install QE-6.4 on a server, by doing the usual procedure, ./configure and make all commands.
>
> The configure step seems to be successful, but the compilation fails after a couple minutes, with a message "make: *** [ph] Error 1". Even with this interruption, a few .x files are compiled, such as pw.x, ph.x, etc., but when I try to run a calculation, It also results in an error message.
>
> Appologies for my not-so-much enlightening description, as I myself could not extract many info from the error messages, so I hope the print screens will be more helpfull than I could.
>
> https://imgur.com/HeT8JMM
> https://imgur.com/ZyyVGt3
> https://imgur.com/DyYlO4R
>
> Thanks in advance for any assistance
>
> F?bio Costa
> MSc student in physics
> Federal University of Bahia, Brazil
> [https://i.imgur.com/DyYlO4R.png?fb]<https://imgur.com/DyYlO4R>
>
> Imgur<https://imgur.com/DyYlO4R>
> Compilation error
> imgur.com<http://imgur.com>
>
>
> [https://i.imgur.com/ZyyVGt3.png?fb]<https://imgur.com/ZyyVGt3>
>
> Imgur<https://imgur.com/ZyyVGt3>
> configure output
> imgur.com<http://imgur.com>
>
>
> [https://i.imgur.com/HeT8JMM.png?fb]<https://imgur.com/HeT8JMM>
>
> Imgur<https://imgur.com/HeT8JMM>
> pw.x error
> imgur.com<http://imgur.com>
>
>
>
>
> _______________________________________________
> users mailing list
> users at lists.quantum-espresso.org<mailto:users at 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
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190407/af265843/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Mon, 8 Apr 2019 07:12:45 +0000
> From: Rajesh Raju <R.Raju at bham.ac.uk>
> To: "users at lists.quantum-espresso.org"
> <users at lists.quantum-espresso.org>
> Subject: [QE-users] Pseudopotential type is different in the website
> description and in the file
> Message-ID:
> <05F2283D2ED72C4DA1426BD0EE0E6E6833F29512 at EX14.adf.bham.ac.uk>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Hi users,
>
> I am new to QE and searching pseudopotential for my system which contains Cu and C. I learned the naming scheme and found some mismatch
>
> For example: For Cu, the first entry in the PS library list is in the website shows the following information
> Cu.pbe-dn-kjpaw_psl.1.0.0.UPF<https://www.quantum-espresso.org/upf_files/Cu.pbe-dn-kjpaw_psl.1.0.0.UPF>
>
> Origin: PS Library
> Author: ADC
> Generated using "atomic" code by A. Dal Corso v.6.3
>
> Pseudopotential type: USPP
> Functional type: PBE
> Non Linear Core Correction
> Scalar relativistic
>
>
> When I opened the file, it shows Pseudopotential type as PAW.
>
> <UPF version="2.0.1">
> <PP_INFO>
> Generated using "atomic" code by A. Dal Corso v.6.3
> Author: ADC
> Generation date: 6Sep2018
> Pseudopotential type: PAW
> Element: Cu
>
> Another entry in the PSLibrary Cu.pbe-dn-kjpaw_psl.0.2.UPF<https://www.quantum-espresso.org/upf_files/Cu.pbe-dn-kjpaw_psl.0.2.UPF> mention Pseudopotentail type PAW in the website and in the file.
>
> Cu.pbe-dn-kjpaw_psl.0.2.UPF<https://www.quantum-espresso.org/upf_files/Cu.pbe-dn-kjpaw_psl.0.2.UPF>
>
> Origin: PS Library
> Author: ADC
> Generated using "atomic" code by A. Dal Corso v.5.0.2 svn rev. 9415
>
> Pseudopotential type: PAW
> Functional type: PBE
> Non Linear Core Correction
> Scalar relativistic
>
> Similarly for other elements such as C and O.
>
> Is it just an error in the website ?? Or is there any physical meaning in these changes??
>
> Thanks
>
> Rajesh
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/8d9b6d59/attachment-0001.html>
>
> ------------------------------
>
> Message: 3
> Date: Mon, 8 Apr 2019 09:17:00 +0200
> From: Pietro Davide Delugas <pdelugas at sissa.it>
> To: users at lists.quantum-espresso.org
> Subject: Re: [QE-users] Compilation on a server ends with error 1
> Message-ID: <3f1fd403-cee6-2b21-4e8a-c8fbabaa9401 at sissa.it>
> Content-Type: text/plain; charset="windows-1252"; Format="flowed"
>
> Dear Paolo
>
> :) many compliments to your system administrator ...
>
> this is yet an other issue with a too compiler version. It is very
> likely the case that your mpif90 is still using fortran 4.4.6 just check
> with mpif90 --version? and see what it prints out.
> If mpif90 is using 5.4 the simpler thing to do is to oblige the
> compiler, and change line 58 of PP/src/plotband.f90 from
>
> CHARACTER(len=:), ALLOCATABLE :: line
>
> to
>
> CHARACTER(len=2048), ALLOCATABLE :: line
>
> and anything should be working. You can meet other errors of this kind,
> read carefully the error messages and try to fix them.
>
> In the case ( I am afraid it is actually the case) that mpif90 is using
> the old fortran version, you need to install an? mpi library that uses
> your desired gfortran version.
>
> It is a little bit complicated but not very difficult, just takes a bit
> of time and patience,? you have just to download the source, better a
> not too old or not too new one , for example this one should work:
>
> https://download.open-mpi.org/release/open-mpi/v2.0/openmpi-2.0.4.tar.gz
>
> follow configure instructions taking care to enable f90 API and to??
> point? to the right fortran compiler.
> It's up to you to decide whether it takes less time doing this? by
> yourself,? or convince the system administrator to help you ...
> best wishes
> Pietro
>
>
>
>
>
> On 04/08/2019 01:53 AM, Fabio Costa wrote:
>> Dear Paolo and Pietro
>>
>> Thanks for the assistance, and sorry for the late reply on your advice.
>>
>> I contacted the system administrator, and he explained to me that as
>> there are several other applications running under this old gfortran
>> version, and because of that he could not just issue "sudo apt update
>> gfortran", on the risk of messing up with all the other stuff
>> currently running on the system. His advice to me was to manually
>> install a newer version inside my workspace.
>>
>> I did that, and now I have both the native 4.4.6 and the 5.4.0
>> gfortran compilers. I also added to my bashrc the line "export
>> F90=/path/to/gfortran5.4.0/". When i run configure, it detects the
>> gfortran v5.4.0, so I think that I did this right. At first, make all
>> still crashes at phcg.x. I proceeded changing the lines in Makefile as
>> Pietro sugested, and when compiling again it goes a bit further, but
>> still crashes. To avoid these crashes, I tryed to compile just the
>> packages I'm intending to use (pw, ph, w90 and epw). By issuing make
>> pw, followed by make ph, make w90, the task is succesful, but It
>> craches again when I issue make epw. The print screen of the error
>> message is in the link: https://imgur.com/CerJStE?.
>>
>> Also, I'm still getting the same errors with the compiled executables.
>> About Paolo's advice, in my current situation, where I cant just
>> update the whole thing, and kinda have two version of gfortran. I'm
>> not sure if the crashes are because when? the executables run, they
>> are still linking to the gfortran 4.4.6. What course of action would
>> you sugest?
>>
>> Thank you for the assistance
>>
>> F?bio Costa
>> MSc student in physics
>> Federal University of Bahia, Brazil
>>
>> ------------------------------------------------------------------------
>> *De:* users <users-bounces at lists.quantum-espresso.org> em nome de
>> Paolo Giannozzi <p.giannozzi at gmail.com>
>> *Enviado:* quinta-feira, 28 de mar?o de 2019 11:40
>> *Para:* Quantum Espresso users Forum
>> *Assunto:* Re: [QE-users] Compilation on a server ends with error 1
>> There are two unrelated problems:
>> - the linker is unable to find the main program in PHonon/Gamma. I
>> have never seen this happen before, but the link is done in a slightly
>> different way from all other cases and this will be fixed
>> - the executable doesn't work likely due to a known problem of many
>> old compilers with unallocated optional variables
>>
>> Paolo
>>
>>
>> On Thu, Mar 28, 2019 at 3:47 AM Fabio Costa <fabiocosta4 at hotmail.com
>> <mailto:fabiocosta4 at hotmail.com>> wrote:
>>
>> Dear all
>>
>> I'm trying to install QE-6.4 on a server, by doing the usual
>> procedure, ./configure and make all commands.
>>
>> The configure step seems to be successful, but the compilation
>> fails after a couple minutes, with a message "make: *** [ph] Error
>> 1". Even with this interruption, a few .x files are compiled, such
>> as pw.x, ph.x, etc., but when I try to run a calculation, It also
>> results in an error message.
>>
>> Appologies for my not-so-much enlightening description, as I
>> myself could not extract many info from the error messages, so I
>> hope the print screens will be more helpfull than I could.
>>
>> https://imgur.com/HeT8JMM
>> https://imgur.com/ZyyVGt3
>> https://imgur.com/DyYlO4R
>>
>> Thanks in advance for any assistance
>>
>> F?bio Costa
>> MSc student in physics
>> Federal University of Bahia, Brazil
>> <https://imgur.com/DyYlO4R>
>>
>> Imgur <https://imgur.com/DyYlO4R>
>> Compilation error
>> imgur.com <http://imgur.com>
>>
>>
>> <https://imgur.com/ZyyVGt3>
>>
>> Imgur <https://imgur.com/ZyyVGt3>
>> configure output
>> imgur.com <http://imgur.com>
>>
>>
>> <https://imgur.com/HeT8JMM>
>>
>> Imgur <https://imgur.com/HeT8JMM>
>> pw.x error
>> imgur.com <http://imgur.com>
>>
>>
>>
>>
>> _______________________________________________
>> users mailing list
>> users at lists.quantum-espresso.org
>> <mailto:users at 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
>>
>>
>>
>> _______________________________________________
>> users mailing list
>> users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/aeb373b9/attachment-0001.html>
>
> ------------------------------
>
> Message: 4
> Date: Mon, 8 Apr 2019 12:36:40 +0530
> From: Arihant Bhandari <arihantb at iitk.ac.in>
> To: users at lists.quantum-espresso.org
> Subject: [QE-users] Ab-Initio Thermal Transport
> Message-ID:
> <CAO8a-Kn9GSvN7VNbj3a6zHR_j_AUtr02KOZQ6CJVWd53+NhxnA at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Respected QE users and developers,
>
> I read the paper on calculating the thermal conductivity from the Density
> Functional Theory (https://www.nature.com/articles/nphys3509). I am trying
> to calculate the thermal conductivity for different systems using this
> method. I do not have knowledge of how to do it with QE codes. I am unable
> to find a detailed description of the thermal conductivity calculation in
> the QE documentation.
>
> The paper states that
> "The methodology presented above has been implemented in the Quantum
> ESPRESSO suite of computer codes: a Car?Parrinello (CP) AIMD trajectory is
> first generated using the cp.x code; the energy flux is then evaluated
> along this trajectory by an add-on to the pw.x code implemented using
> several density-functional perturbation theory routines borrowed from the
> ph.x code; the thermal conductivity is finally computed from the GK
> relation, or the equivalent Einstein relation."
>
> I am unable to follow this and do the computation. I will be very
> grateful if someone can guide (for an amateur) to how to go about doing the
> calculation or can provide a detailed documentation on how to do
> the calculation (hands-on).
>
> Thanks and regards
> Arihant Bhandari
> PhD student
> Indian Institute of Technology Kanpur
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/429d7235/attachment-0001.html>
>
> ------------------------------
>
> Message: 5
> Date: Mon, 8 Apr 2019 09:37:23 +0200
> From: as gj <quant.es.as01 at gmail.com>
> To: users at lists.quantum-espresso.org
> Subject: [QE-users] (no subject)
> Message-ID:
> <CALuz=ExjQzDXtp0q2s29URm9QXBPgvvk4GWpkNH4Np+9Ozy0fQ at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Dear all QE users,
> i am new user;
> i would like to calculate InSb band structure by GW PBE. i tried it with
> GWL and i have a result on gamma point only and i would all the band
> structure.
> So i have a question : GWL treats only gamma point or it could done
> calculation for various kpoints ? If yes how it does ?
>
> in another part,i am trying SternheimerGW application linked with QE.
> currently, i work with QE6.2.0. And when i launch gw.x after scf
> calculation (which runs and converges well) from tutorial for Si, i have
> the following error (in attached fil i put the gw.out):
>
> evaluate self energy for k = ( 0.0000 0.0000 0.0000 )
>
> Error in routine sigma.f90 (2):
> error opening ./tmp/_gw0/si.coul1
> stopping ...
> i obtained the same for the second Li example and the another found in
> examples folder.
>
> I tried also on QE6.3 with the same Si tutorial abd i have the following
> error which is different and i don't know why ?
>
> ######## T E R N H E I M E R
> http://www.sternheimergw.org/
> ### ###
> ### ## #######
> # #
> ### ## ### #
> ## #
> ### ## # #
> ## #
> ##### # # # #
> #
> /// ##### \\\ # # # #
> #
> /// #### \\\ # # # # #
> /// ### \\\ # ######### # # # #
> ((( ### ))) # # # # # #
> \\\ ### ### /// # # # # # #
> \\\ ### ### /// # # # # # #
> \\\ ### ### /// # ## ### ###
> ########## ########### # #
>
>
> Program SternheimerGW v.0.15 starts on 5Apr2019 at 10:52:17
>
> This program is part of the open-source Quantum ESPRESSO suite
> for quantum simulation of materials; please cite
> "P. Giannozzi et al., J. Phys.:Condens. Matter 21 395502 (2009);
> "P. Giannozzi et al., J. Phys.:Condens. Matter 29 465901 (2017);
> URL http://www.quantum-espresso.org",
> in publications or presentations arising from this work. More details
> at
> http://www.quantum-espresso.org/quote
>
> Serial version
>
> Please cite SternheimerGW as:
> M. Schlipf, H. Lambert, N. Zibouche, and F. Giustino,
> SternheimerGW,
> paper in preparation, URL http://www.sternheimergw.org
>
> To increase the reproducibility of your results you can mention the
> git description of this version: unknown
>
> other relevant papers:
> H. Lambert and F. Giustino, Phys. Rev. B 88, 075117 (2013)
> F. Giustino, M. L. Cohen, and S. G. Louie, Phys. Rev. B 81, 115105
> (201
> 0)
>
> Reading data from directory:
> ./tmp/si.save/
>
> IMPORTANT: XC functional enforced from input :
> Exchange-correlation = PZ ( 1 1 0 0 0 0)
> Any further DFT definition will be discarded
> Please, verify this is what you really want
>
>
> G-vector sticks info
> --------------------
> sticks: dense smooth PW G-vecs: dense smooth PW
> Sum 235 235 85 2277 2277 531
>
>
>
> Coulomb Perturbations for ( 4, 4, 4,) uniform grid of q-points
> ( 8q-points):
> N xq(1) xq(2) xq(3)
> 1 0.000000000 0.000000000 0.000000000
> 2 -0.250000000 0.250000000 -0.250000000
> 3 0.500000000 -0.500000000 0.500000000
> 4 0.000000000 0.500000000 0.000000000
> 5 0.750000000 -0.250000000 0.750000000
> 6 0.500000000 0.000000000 0.500000000
> 7 0.000000000 -1.000000000 0.000000000
> 8 -0.500000000 -1.000000000 0.000000000
>
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> Error in routine ggens (2):
> mismatch in number of G-vectors
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>
>
> someone can explain at what these errors are due to ?
>
> thanks in advance for your answer,
> Anne-Sophie
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/f1ef479c/attachment-0001.html>
>
> ------------------------------
>
> Message: 6
> Date: Mon, 08 Apr 2019 09:41:17 +0200
> From: "JAY Antoine" <Antoine.JAY at isae-supaero.fr>
> To: "Quantum Espresso users Forum" <users at lists.quantum-espresso.org>
> Subject: [QE-users] Re-use of charge density don't give exactely the
> same result in scf
> Message-ID: <ecf-5caafb00-3f-244aec4 at 99267285>
> Content-Type: text/plain; charset="utf-8"
>
>
> Dear all,
> After a first scf calculation and saving charge density and wave functions,
> I launch the same scf with these wave functions and charge density as an initialization (using startingpot and startingwfc).
> I go from 27 scf cycles in the first run to only 4 at in the second run.
> However, I was expecting that after the 2 first scf cycles, the energy difference is smaller than the conv_thr because it has already been converged during the first run.
> I suppose that this phenomenon is due to the accuracy (number of digits and number of points) used to saved the electronic density and the wfcs.
> Is it possible to confirm my assumption and to change this accuracy with a magic parameter?
>
> Best regards,
>
> Antoine Jay
> ISAE-SUPAERO
> Toulouse, France
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/62e654e6/attachment-0001.html>
>
> ------------------------------
>
> Message: 7
> Date: Mon, 8 Apr 2019 09:48:54 +0200
> From: Pietro Delugas <pdelugas at sissa.it>
> To: users at lists.quantum-espresso.org
> Subject: Re: [QE-users] Re-use of charge density don't give exactely
> the same result in scf
> Message-ID: <43e12ad7-1b9f-5511-415a-b332e03585d3 at sissa.it>
> Content-Type: text/plain; charset="utf-8"; Format="flowed"
>
> Hello
>
> just a plain?? parameter from the boring manual, but it should work
>
> ? https://www.quantum-espresso.org/Doc/INPUT_PW.html#idm735
>
> Pietro
>
> On 08/04/19 09:41, JAY Antoine wrote:
>> Dear all,
>> After a first scf calculation and saving charge density and wave
>> functions,
>> I launch the same scf with these wave functions and charge density as
>> an initialization (using startingpot and startingwfc).
>> I go from 27 scf cycles in the first run to only 4 at in the second run.
>> However, I was expecting that after the 2 first scf cycles, the energy
>> difference is smaller than the conv_thr because it has already been
>> converged during the first run.
>> I suppose that this phenomenon is due to the accuracy (number of
>> digits and number of points) used to saved the electronic density and
>> the wfcs.
>> Is it possible to confirm my assumption and to change this accuracy
>> with a magic parameter?
>>
>> Best regards,
>>
>> Antoine Jay
>> ISAE-SUPAERO
>> Toulouse, France
>>
>> _______________________________________________
>> users mailing list
>> users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/1c89cf76/attachment-0001.html>
>
> ------------------------------
>
> Message: 8
> Date: Mon, 8 Apr 2019 09:53:57 +0200
> From: Pietro Delugas <pdelugas at sissa.it>
> To: users at lists.quantum-espresso.org
> Subject: Re: [QE-users] Re-use of charge density don't give exactely
> the same result in scf
> Message-ID: <c95b22d6-64b0-47e8-07cd-011f5990f430 at sissa.it>
> Content-Type: text/plain; charset="utf-8"; Format="flowed"
>
> ops sorry, I sent the message without checking the link ...
>
> the parameter to? set the threshod of the diagonilization is
>
> diago_thr_init
>
> and the right link to the manual for it is:
>
> https://www.quantum-espresso.org/Doc/INPUT_PW.html#idm783
>
> Pietro
>
> On 08/04/19 09:48, Pietro Delugas wrote:
>>
>> Hello
>>
>> just a plain?? parameter from the boring manual, but it should work
>>
>> https://www.quantum-espresso.org/Doc/INPUT_PW.html#idm735
>>
>> Pietro
>>
>> On 08/04/19 09:41, JAY Antoine wrote:
>>> Dear all,
>>> After a first scf calculation and saving charge density and wave
>>> functions,
>>> I launch the same scf with these wave functions and charge density as
>>> an initialization (using startingpot and startingwfc).
>>> I go from 27 scf cycles in the first run to only 4 at in the second run.
>>> However, I was expecting that after the 2 first scf cycles, the
>>> energy difference is smaller than the conv_thr because it has already
>>> been converged during the first run.
>>> I suppose that this phenomenon is due to the accuracy (number of
>>> digits and number of points) used to saved the electronic density and
>>> the wfcs.
>>> Is it possible to confirm my assumption and to change this accuracy
>>> with a magic parameter?
>>>
>>> Best regards,
>>>
>>> Antoine Jay
>>> ISAE-SUPAERO
>>> Toulouse, France
>>>
>>> _______________________________________________
>>> users mailing list
>>> users at lists.quantum-espresso.org
>>> https://lists.quantum-espresso.org/mailman/listinfo/users
>>
>> _______________________________________________
>> users mailing list
>> users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/e062000b/attachment-0001.html>
>
> ------------------------------
>
> Message: 9
> Date: Mon, 8 Apr 2019 10:05:15 +0200
> From: Paolo Giannozzi <p.giannozzi at gmail.com>
> To: Quantum Espresso users Forum <users at lists.quantum-espresso.org>
> Subject: Re: [QE-users] Re-use of charge density don't give exactely
> the same result in scf
> Message-ID:
> <CAPMgbCuSg5sthftdFc=EORThYuVLKA-U1K1nryKyPs9h4SW+Wg at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> On Mon, Apr 8, 2019 at 9:41 AM JAY Antoine <Antoine.JAY at isae-supaero.fr>
> wrote:
>
> I suppose that this phenomenon is due to the accuracy (number of digits and
>> number of points) used to saved the electronic density and the wfcs.
>>
>
> no, it's not: the charge density and Kohn-Sham orbitals are saved to file
> with full precision
>
> Paolo
> -
> 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
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/f0fab0b9/attachment-0001.html>
>
> ------------------------------
>
> Message: 10
> Date: Mon, 8 Apr 2019 11:52:21 +0300
> From: Aleksandra Oranskaia <aleksandra.oranskaia at kaust.edu.sa>
> To: Quantum Espresso users Forum <users at lists.quantum-espresso.org>
> Subject: Re: [QE-users] vc-relax for orthorhombic systems
> Message-ID: <8574DD36-39D5-4F5F-A96A-4A3D9D6A4D47 at kaust.edu.sa>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Right now I am testing some desperate approaches like increasing k-points, decreasing mixing_beta together with trust_radii to force systems to change sightly fro step to step in vc-relax, etc.
>
> I also thought that more precise calculation of stresses (in vc-relx routine) might help, does anyone have experience with parameters involved in stress calculations like ecfixed, qcutz, q2sigma?
>
>
>
> Best,
> Alex.
> ___
> Aleksandra Oranskaia (M.Sc.)
> ChemS PhD student, KAUST
> Phone: +966 50 1335254
>
>
>
>
>
>
>> On Apr 6, 2019, at 1:00 PM, users-request at lists.quantum-espresso.org wrote:
>>
>> Send users mailing list submissions to
>> users at lists.quantum-espresso.org
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>> https://lists.quantum-espresso.org/mailman/listinfo/users
>> or, via email, send a message with subject or body 'help' to
>> users-request at lists.quantum-espresso.org
>>
>> You can reach the person managing the list at
>> users-owner at lists.quantum-espresso.org
>>
>> When replying, please edit your Subject line so it is more specific
>> than "Re: Contents of users digest..."
>>
>>
>> Today's Topics:
>>
>> 1. vc-relax for orthorhombic systems (Aleksandra Oranskaia)
>>
>>
>> ----------------------------------------------------------------------
>>
>> Message: 1
>> Date: Fri, 5 Apr 2019 19:39:39 +0300
>> From: Aleksandra Oranskaia <aleksandra.oranskaia at kaust.edu.sa>
>> To: Quantum Espresso users Forum <users at lists.quantum-espresso.org>
>> Subject: [QE-users] vc-relax for orthorhombic systems
>> Message-ID: <8FCA3967-9817-4748-B8D3-34EC6F0295A2 at kaust.edu.sa>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> Dear QE users and developers,
>>
>> 1 When dealing with orthorhombic systems, I start with space_group super-tight relaxation of the atomic positions (keeping right balance in force and scf accuracy, increasing ecutrho plus 100 Ry than my system actually requires, etc), continue with vc-relax using dofree = ibrav (and press_thr = 0.0), and Always obtain at least one lattice constant screwed up (with an error like +/- 0.1 A as compared to the experimental constant).
>> I tried 1000 setups for a few systems (varying functionals, vdw corrections, different pseudo libraries, etc) and have a feeling that sth is not right with vc-relax: for other than orthorhombic systems the results match the experimental lattice constants with an acceptable error.
>>
>> 2 I was wondering how vc-relax work for big orthorhombic unit cells. I took alpha-S8 (space group 70, so 128 atoms turn into 32 within ibrav = 10 with 8 symmetry elements) and am not satisfied with any lattice constants: on PBEsol level, error in volume is <1% as compared to the experimental unit cell, but lattice constants are like 10.602 (exp: 10.465), 12.937 (exp: 12.866), 24.286 (exp: 24.486) A that do not look acceptable at all :-(
>>
>> I would hugely appreciate any comments,
>>
>>
>> Thanks,
>> Alex.
>> ___
>> Aleksandra Oranskaia (M.Sc.)
>> ChemS PhD student, KAUST
>> Phone: +966 50 1335254
>>
>>
>>
>>
>>
>>
>>
>> --
>>
>> This message and its contents, including attachments are intended solely
>> for the original recipient. If you are not the intended recipient or have
>> received this message in error, please notify me immediately and delete
>> this message from your computer system. Any unauthorized use or
>> distribution is prohibited. Please consider the environment before printing
>> this email.
>> -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190405/7ab54d9a/attachment-0001.html>
>>
>> ------------------------------
>>
>> Subject: Digest Footer
>>
>> _______________________________________________
>> users mailing list
>> users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
>>
>> ------------------------------
>>
>> End of users Digest, Vol 141, Issue 6
>> *************************************
>
>
> --
>
> This message and its contents, including attachments are intended solely
> for the original recipient. If you are not the intended recipient or have
> received this message in error, please notify me immediately and delete
> this message from your computer system. Any unauthorized use or
> distribution is prohibited. Please consider the environment before printing
> this email.
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/5f8c8063/attachment-0001.html>
>
> ------------------------------
>
> Message: 11
> Date: Mon, 8 Apr 2019 11:16:31 +0200
> From: Lorenzo Paulatto <paulatz at gmail.com>
> To: users at lists.quantum-espresso.org
> Subject: Re: [QE-users] vc-relax for orthorhombic systems
> Message-ID: <35e816d3-a48f-82b3-8439-e345a4898414 at gmail.com>
> Content-Type: text/plain; charset=windows-1252; format=flowed
>
>> actually requires, etc), continue with vc-relax using dofree = ibrav
>
> Hello Aleksandra, I've implemented that constraint in a very simple way,
> as long as the final pressure/strain is zero, there should be no
> problem, but better use it with a grain of salt.
>
> Also, keep in mind that vc-relax is done at fixed number of plane-waves,
> not constant cutoff. You should always repeat it starting from the final
> configuration and
>
>
>> 2 I was wondering how vc-relax work for big orthorhombic unit cells. I
>> took alpha-S8 (space group 70, so 128 atoms turn into 32 within ibrav =
>> 10 with 8 symmetry elements) and am not satisfied with any lattice
>> constants: on PBEsol level, error in volume is <1% as compared to the
>> experimental unit cell, but lattice constants are like 10.602 (exp:
>> 10.465), 12.937 (exp: 12.866), 24.286 (exp: 24.486) A that do not look
>> acceptable at all :-(
>
> Do you have any reason to expect that DFT should perform better than
> this? It does not look like a very big discrepancy. 0.1 A is the order
> of the difference between PBE GGA and PZ LDA for very simple materials.
> Correcting for the lack of VdW interaction with a simple Grimme-D2
> method could be enough to get a better agreement.
>
> I did some calculations in orthorhombic cells myself, and found it very
> difficult to be completely sure about the final relaxation. Because the
> material can be sort of soft, it is possible that a deformation that
> changes the angles and the cell sides while keeping the volume about
> constants has a very small energy gradient.
>
> I would also guess that thermal effects could be important, hence
> considering the Free Energy (Enthalpy - TS) could be necessary.
> Unfortunately, this is quite expensive and complicated, even at the
> Quasi-Harmonic level.
>
>
>
>> I would hugely appreciate any comments,
>>
>>
>> Thanks,
>> Alex.
>> ___
>> Aleksandra Oranskaia (M.Sc.)
>> ChemS PhD student, KAUST
>> Phone: +966 50 1335254
>>
>>
>>
>>
>>
>>
>>
>> ------------------------------------------------------------------------
>> This message and its contents, including attachments are intended solely
>> for the original recipient. If you are not the intended recipient or
>> have received this message in error, please notify me immediately and
>> delete this message from your computer system. Any unauthorized use or
>> distribution is prohibited. Please consider the environment before
>> printing this email.
>>
>> _______________________________________________
>> users mailing list
>> users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
>>
>
> --
> Lorenzo Paulatto - Paris
>
>
> ------------------------------
>
> Message: 12
> Date: Mon, 08 Apr 2019 11:41:15 +0200
> From: Giuseppe Mattioli <giuseppe.mattioli at ism.cnr.it>
> To: Quantum Espresso users Forum <users at lists.quantum-espresso.org>
> Subject: Re: [QE-users] vc-relax for orthorhombic systems
> Message-ID:
> <20190408114115.Horde.84yvxdqZF_L753Et6ysyiuE at webmail.sic.rm.cnr.it>
> Content-Type: text/plain; charset=utf-8; format=flowed; DelSp=Yes
>
>
> Dear Alex
> Only a comment: Why do you need to be so tightly close to experimental
> lattice parameters? They are generally not the final output of a bunch
> of simulations. As you are not trying to validate a new functional or
> some new correction, your results seem to be close enough to
> experimental data to ensure the calculation of further parameters
> (DOS, band structure, absorption spectroscopy, conductivity, ...). If
> you need *exactly* to work with experimental lattice parameters for
> some reason instead, then use experimental values and do not perform
> vc-relax but relax calculations. Otherwise, I would start asking which
> experimental parameters are you trying to reproduce, e.g., at which
> temperature, how many defects were present in the experimental sample,
> ... And I would answer that a "true experimental value" probably does
> not exist...
>
> As you need practical answers and not philosophical ones, I would be
> satisfied of a result within 1% from "some measurements"
>
> HTH
> Giuseppe
>
>
> Aleksandra Oranskaia <aleksandra.oranskaia at kaust.edu.sa> ha scritto:
>
>> Right now I am testing some desperate approaches like increasing
>> k-points, decreasing mixing_beta together with trust_radii to force
>> systems to change sightly fro step to step in vc-relax, etc.
>>
>> I also thought that more precise calculation of stresses (in vc-relx
>> routine) might help, does anyone have experience with parameters
>> involved in stress calculations like ecfixed, qcutz, q2sigma?
>>
>>
>>
>> Best,
>> Alex.
>> ___
>> Aleksandra Oranskaia (M.Sc.)
>> ChemS PhD student, KAUST
>> Phone: +966 50 1335254
>>
>>
>>
>>
>>
>>
>>> On Apr 6, 2019, at 1:00 PM, users-request at lists.quantum-espresso.org wrote:
>>>
>>> Send users mailing list submissions to
>>> users at lists.quantum-espresso.org
>>>
>>> To subscribe or unsubscribe via the World Wide Web, visit
>>> https://lists.quantum-espresso.org/mailman/listinfo/users
>>> or, via email, send a message with subject or body 'help' to
>>> users-request at lists.quantum-espresso.org
>>>
>>> You can reach the person managing the list at
>>> users-owner at lists.quantum-espresso.org
>>>
>>> When replying, please edit your Subject line so it is more specific
>>> than "Re: Contents of users digest..."
>>>
>>>
>>> Today's Topics:
>>>
>>> 1. vc-relax for orthorhombic systems (Aleksandra Oranskaia)
>>>
>>>
>>> ----------------------------------------------------------------------
>>>
>>> Message: 1
>>> Date: Fri, 5 Apr 2019 19:39:39 +0300
>>> From: Aleksandra Oranskaia <aleksandra.oranskaia at kaust.edu.sa>
>>> To: Quantum Espresso users Forum <users at lists.quantum-espresso.org>
>>> Subject: [QE-users] vc-relax for orthorhombic systems
>>> Message-ID: <8FCA3967-9817-4748-B8D3-34EC6F0295A2 at kaust.edu.sa>
>>> Content-Type: text/plain; charset="iso-8859-1"
>>>
>>> Dear QE users and developers,
>>>
>>> 1 When dealing with orthorhombic systems, I start with space_group
>>> super-tight relaxation of the atomic positions (keeping right
>>> balance in force and scf accuracy, increasing ecutrho plus 100 Ry
>>> than my system actually requires, etc), continue with vc-relax
>>> using dofree = ibrav (and press_thr = 0.0), and Always obtain at
>>> least one lattice constant screwed up (with an error like +/- 0.1 A
>>> as compared to the experimental constant).
>>> I tried 1000 setups for a few systems (varying functionals, vdw
>>> corrections, different pseudo libraries, etc) and have a feeling
>>> that sth is not right with vc-relax: for other than orthorhombic
>>> systems the results match the experimental lattice constants with
>>> an acceptable error.
>>>
>>> 2 I was wondering how vc-relax work for big orthorhombic unit
>>> cells. I took alpha-S8 (space group 70, so 128 atoms turn into 32
>>> within ibrav = 10 with 8 symmetry elements) and am not satisfied
>>> with any lattice constants: on PBEsol level, error in volume is <1%
>>> as compared to the experimental unit cell, but lattice constants
>>> are like 10.602 (exp: 10.465), 12.937 (exp: 12.866), 24.286 (exp:
>>> 24.486) A that do not look acceptable at all :-(
>>>
>>> I would hugely appreciate any comments,
>>>
>>>
>>> Thanks,
>>> Alex.
>>> ___
>>> Aleksandra Oranskaia (M.Sc.)
>>> ChemS PhD student, KAUST
>>> Phone: +966 50 1335254
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> --
>>>
>>> This message and its contents, including attachments are intended solely
>>> for the original recipient. If you are not the intended recipient or have
>>> received this message in error, please notify me immediately and delete
>>> this message from your computer system. Any unauthorized use or
>>> distribution is prohibited. Please consider the environment before printing
>>> this email.
>>> -------------- next part --------------
>>> An HTML attachment was scrubbed...
>>> URL:
>>> <http://lists.quantum-espresso.org/pipermail/users/attachments/20190405/7ab54d9a/attachment-0001.html>
>>>
>>> ------------------------------
>>>
>>> Subject: Digest Footer
>>>
>>> _______________________________________________
>>> users mailing list
>>> users at lists.quantum-espresso.org
>>> https://lists.quantum-espresso.org/mailman/listinfo/users
>>>
>>> ------------------------------
>>>
>>> End of users Digest, Vol 141, Issue 6
>>> *************************************
>>
>>
>> --
>>
>> This message and its contents, including attachments are intended solely
>> for the original recipient. If you are not the intended recipient or have
>> received this message in error, please notify me immediately and delete
>> this message from your computer system. Any unauthorized use or
>> distribution is prohibited. Please consider the environment before printing
>> this email.
>
>
>
> GIUSEPPE MATTIOLI
> CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
> Via Salaria Km 29,300 - C.P. 10
> I-00015 - Monterotondo Scalo (RM)
> Mob (*preferred*) +39 373 7305625
> Tel + 39 06 90672342 - Fax +39 06 90672316
> E-mail: <giuseppe.mattioli at ism.cnr.it>
>
>
>
> ------------------------------
>
> Subject: Digest Footer
>
> _______________________________________________
> users mailing list
> users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
> ------------------------------
>
> End of users Digest, Vol 141, Issue 7
> *************************************
>
--
This message and its contents, including attachments are intended solely
for the original recipient. If you are not the intended recipient or have
received this message in error, please notify me immediately and delete
this message from your computer system. Any unauthorized use or
distribution is prohibited. Please consider the environment before printing
this email.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190408/3c944f53/attachment.html>
More information about the users
mailing list