From jcwu.suda at hotmail.com Wed Sep 1 05:24:44 2010 From: jcwu.suda at hotmail.com (wujianchun) Date: Wed, 1 Sep 2010 11:24:44 +0800 Subject: [Pw_forum] Which smearing scheme is best for the semiconductor DOS and PDOS calculations Message-ID: Dear All, I want to plot the DOS and PDOS for the semiconductors ( the CdS with point defect ). In the pw.x calculation, I set the occupation = 'fixed', For the DOS and PDOS, there are several smearing schemes: 0 Simple Gaussian 1 Methfessel-Paxton of order 1 -1 Marzari-Vanderbilt "cold smearing" -99 Fermi-Dirac function Which one is the best for the semiconductor DOS and PDOS? Actually I intended to adopt the occupation= "tetrahedra" for the DOS calculation, but " The tetrahedron method is presently not implemented" in PDOS calculation, so it's hard to match the DOS and PDOS exactly if I choose the different smearing scheme for the PDOS. Any advice will be appreciated Thank you. Rgds Jianchun Wu Department of physics Soochow university,China 215006 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100901/9bcbaa93/attachment.htm From robguerra at unimore.it Wed Sep 1 11:24:32 2010 From: robguerra at unimore.it (Roberto Guerra) Date: Wed, 1 Sep 2010 11:24:32 +0200 Subject: [Pw_forum] inverse participation ratio Message-ID: <201009011124.33000.robguerra@unimore.it> Hello, I wish calculate the inverse participation ratio (IPR) as post processing of pwscf. this is the formula: http://ulisse.fisica.unimo.it/ipr.png where "i" is the index of the volume element. The problem (beside my unaptness about fortran) is that the wave functions are in the real space, and therefore I guess I have to call an inverse fft for each IPR calculation. What do you think is the best way to code such a thing? As a separate program or together with pp or some other tool? Is there any source where to take routines suitable for this task? Thank you, Roberto. From giannozz at democritos.it Wed Sep 1 12:19:55 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 01 Sep 2010 12:19:55 +0200 Subject: [Pw_forum] inverse participation ratio In-Reply-To: <201009011124.33000.robguerra@unimore.it> References: <201009011124.33000.robguerra@unimore.it> Message-ID: <4C7E28CB.1090101@democritos.it> Roberto Guerra wrote: > this is the formula: > http://ulisse.fisica.unimo.it/ipr.png > where "i" is the index of the volume element. I guess a multiplication by the volume element (Omega/N if Omega is the volume of the unit cell) is missing. For norm-conserving pseudopotentials, the denominator is always equal to 1. For ultrasoft pseudopotentials youy should verify whether such a formula is still valid. > What do you think is the best way to code such a thing? the "monkey way" is the best: modify one of the postprocessing codes. For instance, case 7 in code PP/pp.x. Copy what you need for instance from PP/localdos.f90 (ignore what you do not need) P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Wed Sep 1 12:21:40 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 01 Sep 2010 12:21:40 +0200 Subject: [Pw_forum] Which smearing scheme is best for the semiconductor DOS and PDOS calculations In-Reply-To: References: Message-ID: <4C7E2934.9080903@democritos.it> wujianchun wrote: > For the DOS and PDOS, there are several smearing schemes: if all you need is a plot, use simple gaussian P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From baroni at sissa.it Wed Sep 1 21:46:38 2010 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 1 Sep 2010 21:46:38 +0200 Subject: [Pw_forum] Example07 In-Reply-To: <323824.87985.qm@web65707.mail.ac4.yahoo.com> References: <19DFC083856541A1ABEC9F67A44F15CA@w2agzhp9280> <884199.15225.qm@web65715.mail.ac4.yahoo.com> <8EFA38A34840482D906C3A2548B64F42@w2agzhp9280> <323824.87985.qm@web65707.mail.ac4.yahoo.com> Message-ID: <170EDFB5-3917-4CD0-9019-C53AA4BD2815@sissa.it> Hi Eyvaz: On Aug 30, 2010, at 7:47 PM, Eyvaz Isaev wrote: > Hi Paul, > > Enclosed file contains lambda.f90, as well as matdyn.f90 etc. files. In fact, matdyn now contains > all ingredients to calculate lambda and T_c. In this sense lambda.f90 is now obsolete, nevertheless, you can use it. I think it would be best to leave advanced features out of "basic" routines/codes. Personally, I would prefer to have many small apllications each doing clearly a well defined task, rather than a general-purpose code doing everything according to the input keyword ... ... just a matter of personal preference (but also of programming hygiene fro those who will have to soild their hands in the codes written by others ...) This being said, I am most happy that you (and a few others) take the time to contribute with their codes to effort of this small but lively community. Best wishes, Stefano > > This files were used within QE 4.2. Any problem, please contact me. > > Bests, > Eyvaz. > > From: w2agz > To: PWSCF Forum > Sent: Mon, August 30, 2010 3:30:16 AM > Subject: Re: [Pw_forum] Example07 > > Eyvaz, thanks for getting back. Good to hear from you again. -Paul > > From: Eyvaz Isaev > Sent: Sunday, August 29, 2010 3:36 PM > To: PWSCF Forum > Subject: Re: [Pw_forum] Example07 > > Hi Paul, > > No, this is a (g)fortran issue. I will fix it asap. > > Eyvaz. > > ------------------------------------------------------------------- > Prof. Eyvaz Isaev, > Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden > Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, > isaev at ifm.liu.se, eyvaz_isaev at yahoo.com > > > From: w2agz > To: pw_forum at pwscf.org > Sent: Sun, August 29, 2010 11:48:45 PM > Subject: [Pw_forum] Example07 > > To All QE Gurus: > > A few days ago, I downloaded de-tared, configured and made espresso-4.2.1, both on my trusty old linux box, and new Win 7 one running VirtualBox. On the former I use an old Kubuntu 7-something and the latter the most recent release, 10-something. Both use bash shell scripting. I started running the examples, and everything was going swimmingly until I got to my favorite, Example07, and on both boxes got the same error on the "lambda.x" step (I inserted the relevant .f90 code) below: > **************************************************************** > running matdyn for a2F(omega) calculation... done > running lambda.x for lambda calculation...At line 98 of file lambda.f90 > Fortran runtime error: Constant string in input format > (5x,'Gaussian Broadening: ',f7.3,' Ry, ngauss=',i4) > ^ > Error condition encountered during test: exit status = 2 > Aborting > > (Fortran code) > read (iuelph,9000) degauss1, ngauss1 > 9000 format(5x,'Gaussian Broadening: ',f7.3,' Ry, ngauss=',i4) > > $ECHO " running lambda.x for lambda calculation...\c" > $LAMBDA_COMMAND < lambda.in > lambda.out > check_failure $? > ****************************************************************** > I thought it might be a "check_failure" thing (I've had trouble with that before), so I commented out all references to the exec...but it still occurred. Maybe a bash issue? > > -Paul > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100901/99e7122f/attachment-0001.htm From jcwu.suda at hotmail.com Thu Sep 2 03:13:54 2010 From: jcwu.suda at hotmail.com (wujianchun) Date: Thu, 2 Sep 2010 09:13:54 +0800 Subject: [Pw_forum] Which smearing scheme is best for the semiconductor DOS and PDOS calculations In-Reply-To: <4C7E2934.9080903@democritos.it> References: , <4C7E2934.9080903@democritos.it> Message-ID: Dear Dr. Paolo Giannozzi, Thank you for your kind help. Rgds Jianchun > Date: Wed, 1 Sep 2010 12:21:40 +0200 > From: giannozz at democritos.it > To: pw_forum at pwscf.org > Subject: Re: [Pw_forum] Which smearing scheme is best for the semiconductor DOS and PDOS calculations > > wujianchun wrote: > > > For the DOS and PDOS, there are several smearing schemes: > > if all you need is a plot, use simple gaussian > > P. > -- > Paolo Giannozzi, Democritos and University of Udine, Italy > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100902/d6d70e5e/attachment.htm From wongkc81 at gmail.com Thu Sep 2 05:50:53 2010 From: wongkc81 at gmail.com (Wong Kelvin) Date: Thu, 2 Sep 2010 11:50:53 +0800 Subject: [Pw_forum] HCP and FCC site energy difference on Cu 111 Message-ID: Dear all, I calcuated the total energy of Cu (111) surface {4 atom per layer, six layer with bottom 3 layer fixed, 11 A vacuum layer) with one additional Cu atom on FCC and HCP sites. (using Cu-PBE-d-rrkjus.UPF, 30 and 300 Ry for energy and rho cutoff. 7x7x1 0 0 0 auto kmesh.) The difference in total energy of whole cells is only about 6 meV which seems to be too small. Pease comment, on the validity of this result. Wong KC NTU singapore -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100902/5e5f9de4/attachment.htm From degironc at sissa.it Thu Sep 2 09:27:48 2010 From: degironc at sissa.it (Stefano de Gironcoli) Date: Thu, 02 Sep 2010 09:27:48 +0200 Subject: [Pw_forum] HCP and FCC site energy difference on Cu 111 In-Reply-To: References: Message-ID: <4C7F51F4.1040209@sissa.it> Dear Wong Kelvin, In general I would expect the FCC vs HCP site energy difference to be small... how small is hard to say.. what would be the energy difference per atom if you do bulk FCC vs bulk HCP ? Is there some experimental number you can compare with? being the difference so small every convergence parameter (cutoff, k-point, smearing width) may affect the result... and the overall accuracy of the XC functional may be questionable... probably the final conclusion will be that the two sites are almost degenerate stefano Wong Kelvin wrote: > Dear all, > > I calcuated the total energy of Cu (111) surface {4 atom per layer, six > layer with bottom 3 layer fixed, 11 A vacuum layer) with one additional Cu > atom on FCC and HCP sites. > (using Cu-PBE-d-rrkjus.UPF, 30 and 300 Ry for energy and rho cutoff. 7x7x1 0 > 0 0 auto kmesh.) > The difference in total energy of whole cells is only about 6 meV which > seems to be too small. Pease comment, on the validity of this result. > > Wong KC > NTU singapore > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From cyrille.barreteau at cea.fr Thu Sep 2 09:07:56 2010 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Thu, 02 Sep 2010 09:07:56 +0200 Subject: [Pw_forum] HCP and FCC site energy difference on Cu 111 In-Reply-To: <4C7F51F4.1040209@sissa.it> References: <4C7F51F4.1040209@sissa.it> Message-ID: <4C7F4D4C.2020705@cea.fr> Dear Wong Kelvin In copper the difference of energy for an adatom in fcc or hcp is very small, see for example: http://prb.aps.org/abstract/PRB/v70/i7/e075415 6 mev seems a very reasonable number. cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 IRAMIS, SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://iramis.cea.fr/Pisp/cyrille.barreteau/ ================================================================== Stefano de Gironcoli wrote: >Dear Wong Kelvin, > > In general I would expect the FCC vs HCP site energy difference to be >small... how small is hard to say.. > what would be the energy difference per atom if you do bulk FCC vs >bulk HCP ? Is there some experimental number you can compare with? > being the difference so small every convergence parameter (cutoff, >k-point, smearing width) may affect the result... and the overall >accuracy of the XC functional may be questionable... probably the final >conclusion will be that the two sites are almost degenerate >stefano > >Wong Kelvin wrote: > > >>Dear all, >> >>I calcuated the total energy of Cu (111) surface {4 atom per layer, six >>layer with bottom 3 layer fixed, 11 A vacuum layer) with one additional Cu >>atom on FCC and HCP sites. >>(using Cu-PBE-d-rrkjus.UPF, 30 and 300 Ry for energy and rho cutoff. 7x7x1 0 >>0 0 auto kmesh.) >>The difference in total energy of whole cells is only about 6 meV which >>seems to be too small. Pease comment, on the validity of this result. >> >>Wong KC >>NTU singapore >> >> >>------------------------------------------------------------------------ >> >>_______________________________________________ >>Pw_forum mailing list >>Pw_forum at pwscf.org >>http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From jperaltac at gmail.com Thu Sep 2 15:38:55 2010 From: jperaltac at gmail.com (joaquin peralta) Date: Thu, 2 Sep 2010 08:38:55 -0500 Subject: [Pw_forum] Improve convergence, slab and electric field. In-Reply-To: <1283237963.27836.13.camel@walk.ijs.si> References: <1283164236.2883.42.camel@walk.ijs.si> <1283237963.27836.13.camel@walk.ijs.si> Message-ID: Thanks, i obtain more fast convergence for 1V/A, but ... If I use the previous results from 1V/A in order to calculate an increase in the Field i obtain fluctuation in the energy. total energy = -261.46819272 Ry total energy = -261.46815798 Ry total energy = -261.46816720 Ry total energy = -261.46260676 Ry total energy = -261.37365516 Ry total energy = -261.46820680 Ry total energy = -261.46821935 Ry total energy = -261.46807352 Ry total energy = -261.46825528 Ry total energy = -261.46820900 Ry total energy = -261.46803348 Ry for my case, is recommendable reduce the precision by default of the energy convergence from 1D-4 to 1D-3? i mean with 3 decimal is good for me, but i don't know if a similar scheme happen at different electric field. Best Regards Joaqu?n Peralta On Tue, Aug 31, 2010 at 1:59 AM, Tone Kokalj wrote: > On Mon, 2010-08-30 at 14:43 -0500, joaquin peralta wrote: > > > > Sorry, I have now a slab put in the middle of the cell in the Z axis, > > from z=32 to z=50 (maximum). > > I see. If the slab is located in the middle of the cell (i.e. at 0.5 in > crystal units) then the dipole layer at 0.0 is OK! > > In this case here is the suggestion: > Your 1 V/Angs is rather strong field. First perform SCF with lower field > or no field (what Paolo suggested) and then use that as an input guess. > > Regards, Tone > > -- > Anton Kokalj > J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia > (tel: +386-1-477-3523 // fax:+386-1-477-3822) > > Please, if possible, avoid sending me Word or PowerPoint attachments. > See: http://www.gnu.org/philosophy/no-word-attachments.html > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- ---------------------------------------------------- Group of NanoMaterials ---------------------------------------------------- http://www.gnm.cl ---------------------------------------------------- Joaqu?n Andr?s Peralta Camposano ---------------------------------------------------- http://www.lpmd.cl/jperalta In a world without frontiers, who needs Gates and Win. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100902/29ac0808/attachment.htm From jperaltac at gmail.com Thu Sep 2 16:59:09 2010 From: jperaltac at gmail.com (joaquin peralta) Date: Thu, 2 Sep 2010 09:59:09 -0500 Subject: [Pw_forum] Improve convergence, slab and electric field. In-Reply-To: <20100902154430.2gn7q63b1coko8oo@webmail.sissa.it> References: <1283164236.2883.42.camel@walk.ijs.si> <1283237963.27836.13.camel@walk.ijs.si> <20100902154430.2gn7q63b1coko8oo@webmail.sissa.it> Message-ID: Sorry , my mistake 'again' I wrote bad units I'm using : 1V/nm = 0.0019447 [au] (Input for espresso) with this the convergence is ok. When i increase the electric field to (using previous charge density) 3V/nm = 0.00583431 [au] The convergence energy present the jumps. total energy = -261.46819272 Ry total energy = -261.46815798 Ry total energy = -261.46816720 Ry total energy = -261.46260676 Ry total energy = -261.37365516 Ry total energy = -261.46820680 Ry total energy = -261.46821935 Ry total energy = -261.46807352 Ry total energy = -261.46825528 Ry total energy = -261.46820900 Ry total energy = -261.46803348 Ry I'm using as reference the paper of Sanchez (Molecular Physics, 102: 9, 1045-1055). In this study they use from 0 to 2 ev/A (this units are right now) Any hint I really appreciate it, Thanks in advance Joaquin Peralta On Thu, Sep 2, 2010 at 8:44 AM, Stefano de Gironcoli wrote: > 1 eV/A is a already a Huge electric field... > how thick is your slab ? how many eV is the potential drop from one side to > the other ? is that a reasonable situation ? > stefano > > > Quoting joaquin peralta : > > Thanks, i obtain more fast convergence for 1V/A, but ... >> >> If I use the previous results from 1V/A in order to calculate an increase >> in >> the Field i obtain fluctuation in the energy. >> >> total energy = -261.46819272 Ry >> total energy = -261.46815798 Ry >> total energy = -261.46816720 Ry >> total energy = -261.46260676 Ry >> total energy = -261.37365516 Ry >> total energy = -261.46820680 Ry >> total energy = -261.46821935 Ry >> total energy = -261.46807352 Ry >> total energy = -261.46825528 Ry >> total energy = -261.46820900 Ry >> total energy = -261.46803348 Ry >> >> for my case, is recommendable reduce the precision by default of the >> energy >> convergence from 1D-4 to 1D-3? i mean with 3 decimal is good for me, but i >> don't know if a similar scheme happen at different electric field. >> >> Best Regards >> Joaqu?n Peralta >> >> On Tue, Aug 31, 2010 at 1:59 AM, Tone Kokalj wrote: >> >> On Mon, 2010-08-30 at 14:43 -0500, joaquin peralta wrote: >>> > >>> > Sorry, I have now a slab put in the middle of the cell in the Z axis, >>> > from z=32 to z=50 (maximum). >>> >>> I see. If the slab is located in the middle of the cell (i.e. at 0.5 in >>> crystal units) then the dipole layer at 0.0 is OK! >>> >>> In this case here is the suggestion: >>> Your 1 V/Angs is rather strong field. First perform SCF with lower field >>> or no field (what Paolo suggested) and then use that as an input guess. >>> >>> Regards, Tone >>> >>> -- >>> Anton Kokalj >>> J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia >>> (tel: +386-1-477-3523 // fax:+386-1-477-3822) >>> >>> Please, if possible, avoid sending me Word or PowerPoint attachments. >>> See: http://www.gnu.org/philosophy/no-word-attachments.html >>> >>> >>> _______________________________________________ >>> Pw_forum mailing list >>> Pw_forum at pwscf.org >>> http://www.democritos.it/mailman/listinfo/pw_forum >>> >>> >> >> >> -- >> ---------------------------------------------------- >> Group of NanoMaterials >> ---------------------------------------------------- >> http://www.gnm.cl >> ---------------------------------------------------- >> Joaqu?n Andr?s Peralta Camposano >> ---------------------------------------------------- >> http://www.lpmd.cl/jperalta >> >> In a world without frontiers, >> who needs Gates and Win. >> >> > > > ---------------------------------------------------------------- > SISSA Webmail https://webmail.sissa.it/ > Powered by Horde http://www.horde.org/ > > > -- ---------------------------------------------------- Group of NanoMaterials ---------------------------------------------------- http://www.gnm.cl ---------------------------------------------------- Joaqu?n Andr?s Peralta Camposano ---------------------------------------------------- http://www.lpmd.cl/jperalta In a world without frontiers, who needs Gates and Win. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100902/c3880068/attachment.htm From Trinh.Vo at jpl.nasa.gov Thu Sep 2 20:17:17 2010 From: Trinh.Vo at jpl.nasa.gov (Vo, Trinh (388C)) Date: Thu, 2 Sep 2010 11:17:17 -0700 Subject: [Pw_forum] Nscf and scf Message-ID: Dear PWSCF Users, I run the calculation for a system of 216 atoms (SiP). The CPU time for the calculation of 4x4x4 kpt mesh, which actually did calculation for 24 kpts due to symmetry, is 1d4h18m (~28h). When I performed a nscf calculation for 41kpts along 001 direction, the CPU time is 3d5h (~77h). Thus, the time for one kpt in scf calculation is ~1.17h/kpt, and 1.87h in nscf calculation. I am wondering how can it take more time for nscf calculation than for scf calculation. I don?t understand why. Could you explain to me what can cause this to happen. I use QE-4.1.2. Thank you very much, Trinh Vo CalTech/JPL -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100902/a9f674a9/attachment-0001.htm From tone.kokalj at ijs.si Fri Sep 3 11:05:18 2010 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Fri, 03 Sep 2010 11:05:18 +0200 Subject: [Pw_forum] Improve convergence, slab and electric field. In-Reply-To: References: <1283164236.2883.42.camel@walk.ijs.si> <1283237963.27836.13.camel@walk.ijs.si> Message-ID: <1283504718.4010.63.camel@walk.ijs.si> On Thu, 2010-09-02 at 08:38 -0500, joaquin peralta wrote: > Thanks, i obtain more fast convergence for 1V/A, but ... > > If I use the previous results from 1V/A in order to calculate an > increase in the Field i obtain fluctuation in the energy. > > total energy = -261.46819272 Ry > total energy = -261.46815798 Ry > total energy = -261.46816720 Ry > total energy = -261.46260676 Ry > total energy = -261.37365516 Ry > total energy = -261.46820680 Ry > total energy = -261.46821935 Ry > total energy = -261.46807352 Ry > total energy = -261.46825528 Ry > total energy = -261.46820900 Ry > total energy = -261.46803348 Ry > > for my case, is recommendable reduce the precision by default of the > energy convergence from 1D-4 to 1D-3? To what energy threshold are you referring: conv_thr (in &electrons) or etot_conv_thr (in &control). According to the value of 1d-3/1d-4 I would assume the latter; note however that the latter has nothing to do with the SCF convergence, as it is an optimization threshold criteria for geometry optimizer. If you use 1d-4 for the conv_thr, then the value is much too high, it should be 1d-6 or, even better, lower than that. After how many SCF cycles these "fluctuations" appear? You may try to reduce mixing_beta to 5%, ... The following considerations will speed-up your calculations, hence you will be able to do more SCF convergence tests: from the previous emails of yours, I notice that you use over 50 Angs of vacuum, which is HUGE (I would assume that about 25 Angs should be sufficient; in this case there will be more than 10 Angs spacing from the slab to the dipole-layer). I also noticed that ecutwfc/ecutrho of 30/500 ry is a very large number for Aluminum. Regards, Tone -- Anton Kokalj J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia (tel: +386-1-477-3523 // fax:+386-1-477-3822) Please, if possible, avoid sending me Word or PowerPoint attachments. See: http://www.gnu.org/philosophy/no-word-attachments.html From giannozz at democritos.it Fri Sep 3 15:42:20 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 03 Sep 2010 15:42:20 +0200 Subject: [Pw_forum] Nscf and scf In-Reply-To: References: Message-ID: <4C80FB3C.2060803@democritos.it> Vo, Trinh (388C) wrote: > I am wondering how can it take more time for nscf calculation > than for scf calculation. in the scf calculation, at each scf step, you perform iterative diagonalizations starting from the wavefunctions of the previous scf step, using a convergence criterion for diagonalization that is very loose at the beginning, gets tighter as you approach self-consistency. In this way, you perform several scf steps, each one requiring diagonalizations that converge quickly. In the non-scf calculation, you make a single step but you have to start from superposition of atomic orbitals and use a tight convergence criterion for diagonalization. Another factor is the number of bands used: in scf calculations one typically uses only occupied bands; in non-scf calculation, one typically is interested in empty bands as well. This makes each diagonalization much costlier in the non-scf case than in the scf case. So typically a non-scf calculation may be faster (per k-point) than a scf one, but not by much; sometimes not at all, or even slower. Have a look at the number of H*psi made in the two calculations and at the time spent there. H+psi is a basic ingredient of iterative diagonalization, and typically takes the bigger slice of computer time. The implementation of non-scf calculations is currently non-optimal, because we throw away sll the information from scf except the potential, but since non-scf is seldom the dominant part of a realistic calculation, little effort has been devoted to its further optimization. P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From asafis at yahoo.com.br Fri Sep 3 17:14:42 2010 From: asafis at yahoo.com.br (=?iso-8859-1?Q?=C1lvaro_Alves?=) Date: Fri, 3 Sep 2010 08:14:42 -0700 (PDT) Subject: [Pw_forum] Precision Message-ID: <127503.43420.qm@web52302.mail.re2.yahoo.com> Hello PWscf users. I'm investigating a magnetic system where magnetic coupling (J = E1-E2) is about 0.00001 Ry. E1 and E2 are the total energies of ferromagnetic and antiferromagnetic configurations, respectively. I always used to four decimal places the total energy. Is it correct to use more than four decimal places? What is the maximum precision that allows this kind of calculus? Sincerely, A. S. Santos Univesidade Federal Fluminense Niter?i-Rio de Janeiro-Brasil -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100903/7927580b/attachment.htm From Trinh.Vo at jpl.nasa.gov Fri Sep 3 21:20:46 2010 From: Trinh.Vo at jpl.nasa.gov (Vo, Trinh (388C)) Date: Fri, 3 Sep 2010 12:20:46 -0700 Subject: [Pw_forum] Nscf and scf In-Reply-To: <4C80FB3C.2060803@democritos.it> References: <4C80FB3C.2060803@democritos.it> Message-ID: <63E24D720DCE7246ABAA722B52D23CDF015492F4F661@ALTPHYEMBEVSP10.RES.AD.JPL> Dear Paolo, Thank you very much for detailed explanation. Trinh -----Original Message----- From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Paolo Giannozzi Sent: Friday, September 03, 2010 6:42 AM To: PWSCF Forum Subject: Re: [Pw_forum] Nscf and scf Vo, Trinh (388C) wrote: > I am wondering how can it take more time for nscf calculation > than for scf calculation. in the scf calculation, at each scf step, you perform iterative diagonalizations starting from the wavefunctions of the previous scf step, using a convergence criterion for diagonalization that is very loose at the beginning, gets tighter as you approach self-consistency. In this way, you perform several scf steps, each one requiring diagonalizations that converge quickly. In the non-scf calculation, you make a single step but you have to start from superposition of atomic orbitals and use a tight convergence criterion for diagonalization. Another factor is the number of bands used: in scf calculations one typically uses only occupied bands; in non-scf calculation, one typically is interested in empty bands as well. This makes each diagonalization much costlier in the non-scf case than in the scf case. So typically a non-scf calculation may be faster (per k-point) than a scf one, but not by much; sometimes not at all, or even slower. Have a look at the number of H*psi made in the two calculations and at the time spent there. H+psi is a basic ingredient of iterative diagonalization, and typically takes the bigger slice of computer time. The implementation of non-scf calculations is currently non-optimal, because we throw away sll the information from scf except the potential, but since non-scf is seldom the dominant part of a realistic calculation, little effort has been devoted to its further optimization. P. -- Paolo Giannozzi, Democritos and University of Udine, Italy _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From modaresi.mohsen at gmail.com Sat Sep 4 09:29:44 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Sat, 4 Sep 2010 00:29:44 -0700 Subject: [Pw_forum] A problem with pwcond.x Message-ID: Dear Users, I tried to calculate conduction of CNT22 with "pwcond.x". scf calculation had been done successfully. But when i run pwcond.x calculation stoped and i get this out put. Could you help me, please? Sincerely yours, Mohsen Modarresi Ferdowsi University of Mashhad -------------- next part -------------- A non-text attachment was scrubbed... Name: cond.out Type: application/octet-stream Size: 48975 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100904/ce60dbf9/attachment-0001.obj From barfi.koa at gmail.com Sat Sep 4 12:04:43 2010 From: barfi.koa at gmail.com (barfi koa) Date: Sat, 4 Sep 2010 14:34:43 +0430 Subject: [Pw_forum] examples 01 & 05 and finding orbitals Message-ID: Dear PWScf Users In example 01 and 05, there are some input files for calculating the band structure of silicon that I ran them and drew the band structure successfully. But I can not find any information about orbitals (1s, 2s, 2p, ...) in the output files to ascribe them to the bands. How can I find the atomic orbital or molecular orbital origin of each bands by Quantum Espresso. Many Thanks Truly Yours Koa -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100904/832f5b33/attachment.htm From quantumdft at gmail.com Sat Sep 4 17:32:10 2010 From: quantumdft at gmail.com (vega lew) Date: Sat, 04 Sep 2010 23:32:10 +0800 Subject: [Pw_forum] examples 01 & 05 and finding orbitals In-Reply-To: References: Message-ID: <4C82667A.8090709@gmail.com> I think projwfc.x should be used to calculated the pdos. Then you could get the individual contribution of certain orbital. hope helps. vega On 09/04/10 18:04, barfi koa wrote: > > Dear PWScf Users > > In example 01 and 05, there are some input files for calculating the > band structure of silicon that I ran them and drew the band structure > successfully. But I can not find any information about orbitals (1s, > 2s, 2p, ...) in the output files to ascribe them to the bands. How can > I find the atomic orbital or molecular orbital origin of each bands by > Quantum Espresso. Many Thanks > > Truly Yours > Koa > > > > > > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- ================================================ Vega Lew ( weijia liu) Graduate student State Key Laboratory of Materials-oriented Chemical Engineering Nanjing University of Technology, 210009, Nanjing, China *************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, China *************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100904/857561a0/attachment.htm From mohnish.iitk at gmail.com Sun Sep 5 07:42:47 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Sun, 5 Sep 2010 11:12:47 +0530 Subject: [Pw_forum] PBE functional for Mg Message-ID: Dear QE users, Does anybody has PBE exchange functional for Mg? Please give me if you have it. Thanks in advance. -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100905/6c2765d1/attachment.htm From sclauzer at sissa.it Mon Sep 6 10:23:53 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Mon, 06 Sep 2010 10:23:53 +0200 Subject: [Pw_forum] A problem with pwcond.x In-Reply-To: References: Message-ID: <4C84A519.8030300@sissa.it> Dear Mohsen, hard to say from your output without any further information. You should submit the input for scf and cond steps as well. First you could try to recompile pwcond.x using some debug flags (like "-g -check all -traceback" for the ifort compiler). By the way, are you sure that you need to use a PP for C with beta projectors for the 3d and 4f states? GS On 09/04/2010 09:29 AM, mohsen modaresi wrote: > Dear Users, > I tried to calculate conduction of CNT22 with "pwcond.x". > scf calculation had been done successfully. But when i run pwcond.x > calculation stoped and i get this out put. > Could you help me, please? > > Sincerely yours, > > Mohsen Modarresi > Ferdowsi University of Mashhad > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100906/5b008c2e/attachment.htm From nazari at iasbs.ac.ir Mon Sep 6 11:02:04 2010 From: nazari at iasbs.ac.ir (nazari at iasbs.ac.ir) Date: Mon, 6 Sep 2010 13:32:04 +0430 (IRDT) Subject: [Pw_forum] Spin polarized calculation In-Reply-To: References: Message-ID: <49779.161.116.73.99.1283763724.squirrel@mail.iasbs.ac.ir>
Dear All, In Vasp there is two options for doing spin polarized computations : ISPIN=2 which is done without any restriction. And ISPIN=2 and NUPDOWAN=2 or 4. In the later casse? the spin polarized caculation is done with imposing restriction on the number of alpha and beta electrons. I wish two konw what are the equivalent cases?for spin polarized calculation in ?PWSCF ? Regards Fariba Nazari IASBS? -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100906/9ae487c9/attachment.htm From barfi.koa at gmail.com Mon Sep 6 14:26:10 2010 From: barfi.koa at gmail.com (barfi koa) Date: Mon, 6 Sep 2010 16:56:10 +0430 Subject: [Pw_forum] ldos and pdos Message-ID: Dear PWScf Users I have read the forum (about example 08) but I have confused the definition and meaning of dosup(E), dosdw(E) , Int dos(E), ldosup(E), ldosdw(E) , pdosup(E), pdosdw(E). I would appreciate if one explains them more clearly. Is up/dw related to spin up and down electrons ? If one adds dosup(E) to dosdw(E) what if the result means ? in the latter case what does Int dos(E) mean ? what is the difference between ldos and pdos in below PDOS section and in which way I can ascribe them to d(xy) , d(xz) and ... orbitals ? The results in the below are some parts of example 08 of Quantum Espresso: DOS: # E (eV) dosup(E) dosdw(E) Int dos(E) 5.749 0.0000E+00 0.0000E+00 0.0000E+00 5.849 0.9326E-03 0.4523E-03 0.1385E-03 5.949 0.3730E-02 0.2683E-02 0.7799E-03 6.049 0.1327E-01 0.6778E-02 0.2785E-02 6.149 0.1422E-01 0.1860E-01 0.6067E-02 PDOS: # E (eV) ldosup(E) ldosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) 5.000 -0.280E-10 -0.109E-10 -0.933E-12 -0.348E-12 -0.872E-11 -0.338E-11 -0.872E-11 -0.338E-11 -0.933E-12 -0.348E-12 -0.872E-11 -0.338E-11 5.100 -0.328E-09 -0.139E-09 -0.142E-10 -0.579E-11 -0.999E-10 -0.424E-10 -0.999E-10 -0.424E-10 -0.142E-10 -0.579E-11 -0.999E-10 -0.424E-10 5.200 -0.285E-08 -0.132E-08 -0.160E-09 -0.716E-10 -0.844E-09 -0.394E-09 -0.844E-09 -0.394E-09 -0.160E-09 -0.716E-10 -0.844E-09 -0.394E-09 5.300 -0.183E-07 -0.933E-08 -0.134E-08 -0.655E-09 -0.519E-08 -0.267E-08 -0.519E-08 -0.267E-08 -0.134E-08 -0.655E-09 -0.519E-08 -0.267E-08 5.400 -0.844E-07 -0.478E-07 -0.811E-08 -0.439E-08 -0.227E-07 -0.130E-07 -0.227E-07 -0.130E-07 -0.811E-08 -0.439E-08 -0.227E-07 -0.130E-07 5.500 -0.271E-06 -0.173E-06 -0.349E-07 -0.210E-07 -0.670E-07 -0.436E-07 -0.670E-07 -0.436E-07 -0.349E-07 -0.210E-07 -0.670E-07 -0.436E-07 5.600 -0.546E-06 -0.411E-06 -0.101E-06 -0.693E-07 -0.114E-06 -0.907E-07 -0.114E-06 -0.907E-07 -0.101E-06 -0.693E-07 -0.114E-06 -0.907E-07 Sincerely Yours B. Koa -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100906/29629830/attachment.htm From ttduyle at gmail.com Mon Sep 6 15:46:16 2010 From: ttduyle at gmail.com (Duy Le) Date: Mon, 6 Sep 2010 09:46:16 -0400 Subject: [Pw_forum] Spin polarized calculation In-Reply-To: <49779.161.116.73.99.1283763724.squirrel@mail.iasbs.ac.ir> References: <49779.161.116.73.99.1283763724.squirrel@mail.iasbs.ac.ir> Message-ID: There should be something called constrained_magnetization and fixed_magnetization in the manual. I haven't used them so I have no further comment. -------------------------------------------------- Duy Le PhD Student Department of Physics University of Central Florida. "Men don't need hand to do things" On Mon, Sep 6, 2010 at 5:02 AM, wrote: > Dear All, > > In Vasp there is two options for doing spin polarized computations : ISPIN=2 > which is done without any restriction. And ISPIN=2 and NUPDOWAN=2 or 4. In > the later casse? the spin polarized caculation is done with imposing > restriction on the number of alpha and beta electrons. I wish two konw what > are the equivalent cases?for spin polarized calculation in ?PWSCF ? > > Regards > > Fariba Nazari > > IASBS > > -- > This message has been scanned for viruses and > dangerous content by MailScanner, and is > believed to be clean. > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > From giannozz at democritos.it Mon Sep 6 15:56:00 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 6 Sep 2010 15:56:00 +0200 Subject: [Pw_forum] ldos and pdos In-Reply-To: References: Message-ID: <80CE54C5-43AB-41C2-8417-5098A648B641@democritos.it> On Sep 6, 2010, at 14:26 , barfi koa wrote: > I have confused the definition and meaning of dosup(E), dosdw(E) , > Int dos(E), ldosup(E), ldosdw(E) , pdosup(E), pdosdw(E). > I would appreciate if one explains them more clearly. see files Doc/INPUT_PROJWFC.* and the last few slides here: http://www.fisica.uniud.it/~giannozz/QE-Tutorial/tutorial_magn.pdf P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From matteo at umn.edu Mon Sep 6 16:17:30 2010 From: matteo at umn.edu (Matteo Cococcioni) Date: Mon, 6 Sep 2010 09:17:30 -0500 Subject: [Pw_forum] Spin polarized calculation In-Reply-To: <49779.161.116.73.99.1283763724.squirrel@mail.iasbs.ac.ir> References: <49779.161.116.73.99.1283763724.squirrel@mail.iasbs.ac.ir> Message-ID: Dear Fariba, there is a number of options to constrain the magnetization (e.g., imposing a specific multiplicity). Please take a look to the Doc/INPUT_PW file of the distribution. regards, Matteo On Mon, Sep 6, 2010 at 4:02 AM, wrote: > Dear All, > > In Vasp there is two options for doing spin polarized computations : > ISPIN=2 which is done without any restriction. And ISPIN=2 and NUPDOWAN=2 or > 4. In the later casse the spin polarized caculation is done with imposing > restriction on the number of alpha and beta electrons. I wish two konw what > are the equivalent cases for spin polarized calculation in PWSCF ? > > Regards > > Fariba Nazari > > IASBS > > -- > This message has been scanned for viruses and > dangerous content by *MailScanner* , and is > believed to be clean. > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Matteo Cococcioni Department of Chemical Engineering and Materials Science, University of Minnesota 421 Washington Av. SE Minneapolis, MN 55455 Tel. +1 612 624 9056 Fax +1 612 626 7246 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100906/f83a1b70/attachment.htm From degironc at sissa.it Mon Sep 6 17:05:38 2010 From: degironc at sissa.it (Stefano de Gironcoli) Date: Mon, 06 Sep 2010 17:05:38 +0200 Subject: [Pw_forum] Nscf and scf In-Reply-To: References: Message-ID: <4C850342.5020901@sissa.it> how many iterations does the iterative diagonalization takes in the nscf calculation and how many in the scf one? stefano Vo, Trinh (388C) wrote: > Dear PWSCF Users, > > I run the calculation for a system of 216 atoms (SiP). The CPU time for the calculation of 4x4x4 kpt mesh, which actually did calculation for 24 kpts due to symmetry, is 1d4h18m (~28h). When I performed a nscf calculation for 41kpts along 001 direction, the CPU time is 3d5h (~77h). Thus, the time for one kpt in scf calculation is ~1.17h/kpt, and 1.87h in nscf calculation. I am wondering how can it take more time for nscf calculation than for scf calculation. I don?t understand why. Could you explain to me what can cause this to happen. I use QE-4.1.2. > > Thank you very much, > > Trinh Vo > CalTech/JPL > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From mohnish.iitk at gmail.com Mon Sep 6 18:57:00 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Mon, 6 Sep 2010 22:27:00 +0530 Subject: [Pw_forum] Mg pseudopotential without semi-core state Message-ID: Dear QE users, Does anybody have Mg Pseudopotential without semi-core state? If somebody has then please give me. Thanks a lot in advance -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100906/ce288755/attachment.htm From modaresi.mohsen at gmail.com Tue Sep 7 11:17:45 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Tue, 7 Sep 2010 13:47:45 +0430 Subject: [Pw_forum] DOS problem Message-ID: Dear all, "dos.x" calculate "density of state per unit cell" or "density of state per site (atom)"? Thanks for your reply Mohsen Modarresi Ferdowsi University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100907/a606185b/attachment.htm From degironc at sissa.it Tue Sep 7 11:40:41 2010 From: degironc at sissa.it (Stefano de Gironcoli) Date: Tue, 07 Sep 2010 11:40:41 +0200 Subject: [Pw_forum] DOS problem In-Reply-To: References: Message-ID: <4C860899.2000507@sissa.it> density of states per unit cell. Stefano de Gironcoli - SISSA and DEMOCRITOS mohsen modaresi wrote: > Dear all, > "dos.x" calculate "density of state per unit cell" or "density of state per > site (atom)"? > Thanks for your reply > Mohsen Modarresi > Ferdowsi University > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From asmogunov at gmail.com Tue Sep 7 11:46:19 2010 From: asmogunov at gmail.com (Alex Smogunov) Date: Tue, 7 Sep 2010 11:46:19 +0200 Subject: [Pw_forum] A problem with pwcond.x In-Reply-To: <4C84A519.8030300@sissa.it> References: <4C84A519.8030300@sissa.it> Message-ID: Moreover, pwcond is not implemented for f orbitals. But for CNTs, I agree with Gabriel, I am not sure you really need f projectors ... Alexander. 2010/9/6 Gabriele Sclauzero > Dear Mohsen, > > hard to say from your output without any further information. You should > submit the input for scf and cond steps as well. First you could try to > recompile pwcond.x using some debug flags (like "-g -check all -traceback" > for the ifort compiler). > By the way, are you sure that you need to use a PP for C with beta > projectors for the 3d and 4f states? > > GS > > On 09/04/2010 09:29 AM, mohsen modaresi wrote: > > Dear Users, > I tried to calculate conduction of CNT22 with "pwcond.x". > scf calculation had been done successfully. But when i run pwcond.x > calculation stoped and i get this out put. > Could you help me, please? > > Sincerely yours, > > Mohsen Modarresi > Ferdowsi University of Mashhad > > > > _______________________________________________ > Pw_forum mailing listPw_forum at pwscf.orghttp://www.democritos.it/mailman/listinfo/pw_forum > > > > -- > > Gabriele Sclauzero, EPFL SB ITP CSEA > PH H2 462, Station 3, CH-1015 Lausanne > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100907/2515c3e3/attachment.htm From mighfar at jncasr.ac.in Tue Sep 7 12:31:48 2010 From: mighfar at jncasr.ac.in (Mighfar Imam) Date: Tue, 7 Sep 2010 16:01:48 +0530 (IST) Subject: [Pw_forum] DOS problem In-Reply-To: References: Message-ID: dos.x does it for the whole system (i.e. unit cell). projwfc.x can do it for the individual atoms or their electronic states. -Mighfar Imam JNCASR, Bangalore. > Dear all, > "dos.x" calculate "density of state per unit > cell" or "density of state per > site (atom)"? > Thanks for your reply > Mohsen Modarresi > Ferdowsi University > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From modaresi.mohsen at gmail.com Tue Sep 7 14:09:45 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Tue, 7 Sep 2010 16:39:45 +0430 Subject: [Pw_forum] A problem with pwcond.x In-Reply-To: References: <4C84A519.8030300@sissa.it> Message-ID: Dear Gabriele, I dont use ifort i use Gfortran. I change PP but the problem does not solve. Best Regards, Mohsen On Tue, Sep 7, 2010 at 2:16 PM, Alex Smogunov wrote: > Moreover, pwcond is not implemented for f orbitals. > But for CNTs, I agree with Gabriel, I am not sure you > really need f projectors ... > Alexander. > > > 2010/9/6 Gabriele Sclauzero > > Dear Mohsen, >> >> hard to say from your output without any further information. You >> should submit the input for scf and cond steps as well. First you could try >> to recompile pwcond.x using some debug flags (like "-g -check all >> -traceback" for the ifort compiler). >> By the way, are you sure that you need to use a PP for C with beta >> projectors for the 3d and 4f states? >> >> GS >> >> On 09/04/2010 09:29 AM, mohsen modaresi wrote: >> >> Dear Users, >> I tried to calculate conduction of CNT22 with "pwcond.x". >> scf calculation had been done successfully. But when i run pwcond.x >> calculation stoped and i get this out put. >> Could you help me, please? >> >> Sincerely yours, >> >> Mohsen Modarresi >> Ferdowsi University of Mashhad >> >> >> >> _______________________________________________ >> Pw_forum mailing listPw_forum at pwscf.orghttp://www.democritos.it/mailman/listinfo/pw_forum >> >> >> >> -- >> >> Gabriele Sclauzero, EPFL SB ITP CSEA >> PH H2 462, Station 3, CH-1015 Lausanne >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100907/7894d52d/attachment.htm From sclauzer at sissa.it Tue Sep 7 16:13:53 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Tue, 7 Sep 2010 16:13:53 +0200 Subject: [Pw_forum] A problem with pwcond.x In-Reply-To: References: <4C84A519.8030300@sissa.it> Message-ID: <19743958-3B60-41F1-AB38-6129592CD5BC@sissa.it> Il giorno 07/set/2010, alle ore 14.09, mohsen modaresi ha scritto: > Dear Gabriele, > I dont use ifort i use Gfortran. Then you may want to find out the equivalent debugging options for that compiler. > I change PP but the problem does not solve. You haven't submitted any input file or other information, so I suppose that you don't need further help. Regards, Gabriele > > Best Regards, > > Mohsen > > > On Tue, Sep 7, 2010 at 2:16 PM, Alex Smogunov wrote: > Moreover, pwcond is not implemented for f orbitals. > But for CNTs, I agree with Gabriel, I am not sure you > really need f projectors ... > Alexander. > > > 2010/9/6 Gabriele Sclauzero > > Dear Mohsen, > > hard to say from your output without any further information. You should submit the input for scf and cond steps as well. First you could try to recompile pwcond.x using some debug flags (like "-g -check all -traceback" for the ifort compiler). > By the way, are you sure that you need to use a PP for C with beta projectors for the 3d and 4f states? > > GS > > On 09/04/2010 09:29 AM, mohsen modaresi wrote: >> Dear Users, >> I tried to calculate conduction of CNT22 with "pwcond.x". >> scf calculation had been done successfully. But when i run pwcond.x >> calculation stoped and i get this out put. >> Could you help me, please? >> >> Sincerely yours, >> >> Mohsen Modarresi >> Ferdowsi University of Mashhad >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> > > > -- > > Gabriele Sclauzero, EPFL SB ITP CSEA > PH H2 462, Station 3, CH-1015 Lausanne ? Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100907/33d6b80b/attachment.htm From dimpy.sharma at tyndall.ie Tue Sep 7 20:20:31 2010 From: dimpy.sharma at tyndall.ie (Dimpy Sharma) Date: Tue, 7 Sep 2010 19:20:31 +0100 Subject: [Pw_forum] How to choose the energy0 value while doing a Ballistic conductance using the code pwcond.x Message-ID: Hi there, My warm greetings to all of you. I would like to know how do we choose the energy0 (inital energy ) while doing a ballistic conductance using pwcond.x.As from the example and manual it mentions as the inital energy,which energy does it refer to?is this the total energy of the system ?I could not follow the term nenergy as well.Any advice would be appreciated. Thanks and regards Dimpy Dimpy Sharma ETG-Group Tyndall National Institute Cork -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100907/53c073ac/attachment-0001.htm From modaresi.mohsen at gmail.com Tue Sep 7 20:55:34 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Tue, 7 Sep 2010 23:25:34 +0430 Subject: [Pw_forum] How to choose the energy0 value while doing a Ballistic conductance using the code pwcond.x In-Reply-To: References: Message-ID: Dear Dimpy Sharma, pwcond.x give the Transmission probablity as a function of energy. energy0 determine the minimum of energy in the energy axis. Mohsen Modarresi Ferdowsi University On Tue, Sep 7, 2010 at 10:50 PM, Dimpy Sharma wrote: > > Hi there, > > My warm greetings to all of you. I would like to know how do we choose the > energy0 (inital energy ) while doing a ballistic conductance using > pwcond.x.As from the example and manual it mentions > as the inital energy,which energy does it refer to?is this the total energy > of the system ?I could not follow the term nenergy as well.Any advice would > be appreciated. > > Thanks and regards > > Dimpy > Dimpy Sharma > ETG-Group > Tyndall National Institute > Cork > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100907/8e89d3cf/attachment.htm From modaresi.mohsen at gmail.com Wed Sep 8 00:19:12 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Wed, 8 Sep 2010 02:49:12 +0430 Subject: [Pw_forum] DOS problem In-Reply-To: References: Message-ID: Thanks for your answers. I have a question that it might not crelated to Q. E Is it true to fit DOS which had been calculated with a tight binding model with DFT one to get the parameters in the tight binding model? Thanks for your help Mohen Modarresi On Tue, Sep 7, 2010 at 3:01 PM, Mighfar Imam wrote: > dos.x does it for the whole system (i.e. unit > cell). projwfc.x can do it for the individual > atoms or their electronic states. > > -Mighfar Imam > JNCASR, Bangalore. > > > Dear all, > > "dos.x" calculate "density of state per unit > > cell" or "density of state per > > site (atom)"? > > Thanks for your reply > > Mohsen Modarresi > > Ferdowsi University > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100908/8a378e19/attachment.htm From manoj at phys.ufl.edu Tue Sep 7 23:52:44 2010 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Tue, 7 Sep 2010 17:52:44 -0400 (EDT) Subject: [Pw_forum] How to choose the energy0 value while doing a Ballistic conductance using the code pwcond.x In-Reply-To: Message-ID: Hello, I just wanted to stress the fact that when you use energy0=0, your calculation is being done at Fermi energy. You can look up the code in do_cond.f90 for more details. nenergy is number of energy steps, denergy is energy of each step. Again from do_cond.f90 ! the array of energies is automatically formed DO ien = 1, nenergy earr(ien) = energy0 + (ien-1)*denergy tran_tot(ien) = 0.d0 ENDDO HTH Manoj Srivastava Department of Physics, University of Florida On Tue, 7 Sep 2010, mohsen modaresi wrote: > Dear Dimpy Sharma, > pwcond.x give the Transmission probablity as a function of energy. energy0 > determine the minimum of energy in the energy axis. > > Mohsen Modarresi > Ferdowsi University > > > > On Tue, Sep 7, 2010 at 10:50 PM, Dimpy Sharma wrote: > > > > > Hi there, > > > > My warm greetings to all of you. I would like to know how do we choose the > > energy0 (inital energy ) while doing a ballistic conductance using > > pwcond.x.As from the example and manual it mentions > > as the inital energy,which energy does it refer to?is this the total energy > > of the system ?I could not follow the term nenergy as well.Any advice would > > be appreciated. > > > > Thanks and regards > > > > Dimpy > > Dimpy Sharma > > ETG-Group > > Tyndall National Institute > > Cork > > > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > From modaresi.mohsen at gmail.com Wed Sep 8 08:45:18 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Wed, 8 Sep 2010 11:15:18 +0430 Subject: [Pw_forum] A problem with pwcond.x In-Reply-To: <19743958-3B60-41F1-AB38-6129592CD5BC@sissa.it> References: <4C84A519.8030300@sissa.it> <19743958-3B60-41F1-AB38-6129592CD5BC@sissa.it> Message-ID: Dear Gabriele, Thanks for your answer. We neglect d and f orbitals. But we faced with this error: "%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from check_stop_now : error # 1 check_stop not initialized %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Thanks again for your help. Mohsen Modarresi -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100908/8bdb3f93/attachment.htm From cyrille.barreteau at cea.fr Wed Sep 8 07:51:08 2010 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Wed, 08 Sep 2010 07:51:08 +0200 Subject: [Pw_forum] DOS problem In-Reply-To: References: Message-ID: <4C87244C.7080406@cea.fr> Dear Mohen Modarresi, If you want to determine the parameters (hopping integrals and onsite elements) of a tight-binding model you should rather try to fit the ab-initio band structure than the total dos since there is much more information in the E(k) relation. cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 IRAMIS, SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://iramis.cea.fr/Pisp/cyrille.barreteau/ ================================================================== mohsen modaresi wrote: > Thanks for your answers. > I have a question that it might not crelated to Q. E > Is it true to fit DOS which had been calculated with a tight > binding model with DFT one to get the parameters in the tight binding > model? > Thanks for your help > > Mohen Modarresi > On Tue, Sep 7, 2010 at 3:01 PM, Mighfar Imam > wrote: > > dos.x does it for the whole system (i.e. unit > cell). projwfc.x can do it for the individual > atoms or their electronic states. > > -Mighfar Imam > JNCASR, Bangalore. > > > Dear all, > > "dos.x" calculate "density of state per unit > > cell" or "density of state per > > site (atom)"? > > Thanks for your reply > > Mohsen Modarresi > > Ferdowsi University > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > >------------------------------------------------------------------------ > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From sclauzer at sissa.it Wed Sep 8 09:24:46 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Wed, 08 Sep 2010 09:24:46 +0200 Subject: [Pw_forum] A problem with pwcond.x In-Reply-To: References: <4C84A519.8030300@sissa.it> <19743958-3B60-41F1-AB38-6129592CD5BC@sissa.it> Message-ID: <4C873A3E.3090305@sissa.it> On 09/08/2010 08:45 AM, mohsen modaresi wrote: > Dear Gabriele, > Thanks for your answer. We neglect d and f orbitals. But we faced with > this error: > > "%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from check_stop_now : error # 1 > check_stop not initialized > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% This is strange, it should not happen. Which version of QE are you using? If it is lower than 4.2.2 please upgrade your QE distribution and retry. If the problem persists, please send me the input files (scf+pwcond). If you don't want to upgrade (which, however, I strongly suggest to do in any case), check if tran_save is initialized to FALSE in the ELSE statement here below: IF (ikind.NE.0 .AND. tran_prefix.NE.' ') THEN ! tran_save = .TRUE. ... ELSE ! tran_save = .FALSE. IF (recover) call errore('do_cond',& 'you must specify tran_prefix in order to restart',1) ENDIF If not, add that line to the code and recompile. HTH GS > Thanks again for your help. > > Mohsen Modarresi > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100908/70e4c2ef/attachment.htm From sclauzer at sissa.it Wed Sep 8 09:34:21 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Wed, 08 Sep 2010 09:34:21 +0200 Subject: [Pw_forum] How to choose the energy0 value while doing a Ballistic conductance using the code pwcond.x In-Reply-To: References: Message-ID: <4C873C7D.5090808@sissa.it> On 09/07/2010 11:52 PM, Manoj Srivastava wrote: > Hello, > I just wanted to stress the fact that when you use energy0=0, your > calculation is being done at Fermi energy. You can look up the > code in do_cond.f90 for more details. nenergy is number of energy steps, > denergy is energy of each step. Again from do_cond.f90 > > ! the array of energies is automatically formed > DO ien = 1, nenergy > earr(ien) = energy0 + (ien-1)*denergy > tran_tot(ien) = 0.d0 > ENDDO > > HTH > Manoj Srivastava > Department of Physics, > University of Florida > > On Tue, 7 Sep 2010, mohsen > modaresi wrote: > > >> Dear Dimpy Sharma, >> pwcond.x give the Transmission probablity as a function of energy. energy0 >> determine the minimum of energy in the energy axis. >> Actually, it is the maximum energy, since denergy (in the piece of code reported here above by Manoj) should be negative. This is important, since energy0 is also needed to perform the basis set reduction (through ewind and epsproj parameters, see Doc/INPUT_PWCOND). This basis set reduction will result in an accuracy equal to or higher than that specified by these parameters only for energies <= energy0 (when ewind and epsproj are in a reasonable range). HTH GS >> Mohsen Modarresi >> Ferdowsi University >> >> >> >> On Tue, Sep 7, 2010 at 10:50 PM, Dimpy Sharmawrote: >> >> >>> Hi there, >>> >>> My warm greetings to all of you. I would like to know how do we choose the >>> energy0 (inital energy ) while doing a ballistic conductance using >>> pwcond.x.As from the example and manual it mentions >>> as the inital energy,which energy does it refer to?is this the total energy >>> of the system ?I could not follow the term nenergy as well.Any advice would >>> be appreciated. >>> >>> Thanks and regards >>> >>> Dimpy >>> Dimpy Sharma >>> ETG-Group >>> Tyndall National Institute >>> Cork >>> >>> >>> _______________________________________________ >>> Pw_forum mailing list >>> Pw_forum at pwscf.org >>> http://www.democritos.it/mailman/listinfo/pw_forum >>> >>> >>> >> > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne From bisu123 at platinum.materials.iisc.ernet.in Wed Sep 8 09:54:05 2010 From: bisu123 at platinum.materials.iisc.ernet.in (bisu123 at platinum.materials.iisc.ernet.in) Date: Wed, 8 Sep 2010 13:24:05 +0530 (IST) Subject: [Pw_forum] calculation of lattice parameters[ i.e. 'a' and 'c/a' ratio] for HCP metals say Mg. Message-ID: <3917.10.194.30.14.1283932445.squirrel@platinum.materials.iisc.ernet.in> Dear all, Can someone tell me the correct methodology to start with lattice parameters estimation for HCP metals. I shall tell what I did for my earlier simulations involving Magnesium. First, I fixed the ecut value and then used the same ecut value to fix the ecutrho value. Then using the fixed ecut and ecutrho, I varied the k-points to obtain the K-points. in my case I obtained [12 12 9]. Then for lattice parameter calculation, [ having fixed ecut, ecutrho, k-points of course] I fixed 'c/a' ratio initially[ c/a= sqrt(8/3)] and varied the 'a' across the ideal value of 'a'[ a=3.21 Angstroms =6.06602 at. units] and obtained the value for 'a' corresponding for the minimum energy configuration. Then using the obtained 'a' in the previous step, I varied the 'c/a' ratio across the ideal value to obtain c/a=1.58, corresponding to the minimum energy configuration. However this estimate involves significant error since the ideal value for c/a( for Mg=sqrt(8/3)=1.624). Then I used the value for 'c/a' in the previous step[ i.e 1.58] and varied 'a' to find the value of 'a'. Finally, I obtained a=6.1 a.u=3.23 angstroms, c/a=1.58. The psuedopotential I used for Mg is Mg.pw91-np-van.UPF . Thanks all, regards, Biswaranjan Dash Research Scholar, Deformation Mechanisms Modeling Group, C/0- Asst Prof S.Karthikeyan Indian Institute of Science,Bangalore India -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From modaresi.mohsen at gmail.com Wed Sep 8 12:35:28 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Wed, 8 Sep 2010 15:05:28 +0430 Subject: [Pw_forum] A problem with pwcond.x In-Reply-To: <4C873A3E.3090305@sissa.it> References: <4C84A519.8030300@sissa.it> <19743958-3B60-41F1-AB38-6129592CD5BC@sissa.it> <4C873A3E.3090305@sissa.it> Message-ID: Dear Gabriele, We used Q. E 4.2. We upgrade it to 4.2.1 and the problem solved. Thanks for your guide. Mohsen On Wed, Sep 8, 2010 at 11:54 AM, Gabriele Sclauzero wrote: > On 09/08/2010 08:45 AM, mohsen modaresi wrote: > > Dear Gabriele, > Thanks for your answer. We neglect d and f orbitals. But we faced with this > error: > > > "%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from check_stop_now : error # 1 > check_stop not initialized > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > > This is strange, it should not happen. Which version of QE are you using? > If it is lower than 4.2.2 please upgrade your QE distribution and retry. If > the problem persists, please send me the input files (scf+pwcond). > > If you don't want to upgrade (which, however, I strongly suggest to do in > any case), check if tran_save is initialized to FALSE in the ELSE statement > here below: > > IF (ikind.NE.0 .AND. tran_prefix.NE.' ') THEN > ! > tran_save = .TRUE. > > ... > > ELSE > ! > tran_save = .FALSE. > IF (recover) call errore('do_cond',& > 'you must specify tran_prefix in order to restart',1) > ENDIF > > > If not, add that line to the code and recompile. > > > HTH > > GS > > > Thanks again for your help. > > Mohsen Modarresi > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.orghttp://www.democritos.it/mailman/listinfo/pw_forum > > > > -- > > Gabriele Sclauzero, EPFL SB ITP CSEA > PH H2 462, Station 3, CH-1015 Lausanne > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100908/4c2503fe/attachment.htm From marsamos at democritos.it Wed Sep 8 16:34:09 2010 From: marsamos at democritos.it (marsamos at democritos.it) Date: Wed, 08 Sep 2010 16:34:09 +0200 Subject: [Pw_forum] deadline extension Message-ID: <20100908163409.gg9ur9ju4oscs4sc@mail.democritos.it> Dear QE users, the dealine for subscribing to the Hands on Workshop NANOEXCITE2010 has been extended to Septembre the 13th. Please visit the workshop webpage at http://www.democritos.it/nanoexcite2010/ . bests Layla -- Dr. L. Martin-Samos tel. +39 040 3787 429 CNR-DEMOCRITOS and International School for Advanced Studies (ISAS-SISSA) via Beirut 2-4 34151 Trieste Italy ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From mohnish.iitk at gmail.com Wed Sep 8 17:01:50 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Wed, 8 Sep 2010 20:31:50 +0530 Subject: [Pw_forum] calculation of lattice parameters[ i.e. 'a' and 'c/a' ratio] for HCP metals say Mg. In-Reply-To: <3917.10.194.30.14.1283932445.squirrel@platinum.materials.iisc.ernet.in> References: <3917.10.194.30.14.1283932445.squirrel@platinum.materials.iisc.ernet.in> Message-ID: Dear Biswaranjan! First try to get the experimental (or theoretical value done by somebody else) value of the lattice parameter of your system. If you are using ultrasoft pseudopotential then ecutrho between 8-10 times of ecut will suffice. For fixing 'ecutwfc' you have to do "vc-relax" calculation (see doc-input-pw for more detail) at different 'ecutwfc' then see the convergence with respect to ecut. When the difference between energies for two consecutive ecutwfc is below the minimum limit of accuracy you want then choose lower value(difference of 5 Ry. between two consecutive ecut will suffice). The same procedure you have to repeat for convergence with respect to k-points. You can write a script for doing all the calculations in one shot. On Wed, Sep 8, 2010 at 1:24 PM, wrote: > Dear all, > Can someone tell me the correct methodology to start with > lattice parameters estimation for HCP metals. I shall tell > what I did for my earlier simulations involving Magnesium. > > First, I fixed the ecut value and then used the same ecut value to fix the > ecutrho value. Then using the fixed ecut and ecutrho, I varied the > k-points to obtain the K-points. in my case I obtained [12 12 9]. Then > for lattice parameter calculation, [ having fixed ecut, ecutrho, k-points > of course] I fixed 'c/a' ratio initially[ c/a= sqrt(8/3)] and varied the > 'a' across the ideal value of 'a'[ a=3.21 Angstroms =6.06602 at. units] > and obtained the value for 'a' corresponding for the minimum energy > configuration. Then using the obtained 'a' in the previous step, I varied > the 'c/a' ratio across the ideal value to obtain c/a=1.58, corresponding > to the minimum energy configuration. However this estimate involves > significant error since the ideal value for c/a( for Mg=sqrt(8/3)=1.624). > Then I used the value for 'c/a' in the previous step[ i.e 1.58] and varied > 'a' to find the value of 'a'. > > Finally, I obtained a=6.1 a.u=3.23 angstroms, c/a=1.58. The > psuedopotential I used for Mg is Mg.pw91-np-van.UPF . > > Thanks all, > > regards, > Biswaranjan Dash > Research Scholar, > Deformation Mechanisms Modeling Group, > C/0- Asst Prof S.Karthikeyan > Indian Institute of Science,Bangalore > India > > > > > > -- > This message has been scanned for viruses and > dangerous content by MailScanner, and is > believed to be clean. > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100908/f0d8475b/attachment.htm From Trinh.Vo at jpl.nasa.gov Thu Sep 9 03:24:42 2010 From: Trinh.Vo at jpl.nasa.gov (Vo, Trinh (388C)) Date: Wed, 8 Sep 2010 18:24:42 -0700 Subject: [Pw_forum] Nscf and scf In-Reply-To: <4C850342.5020901@sissa.it> Message-ID: Dear Stefano, For scf calculation I see the avg number of iterations varies from 3 to 4 per scf step, and there are total 22 scf steps. For nscf one, it is about 59.6. Thanks, Trinh On 9/6/10 8:05 AM, "Stefano de Gironcoli" wrote: > how many iterations does the iterative diagonalization takes in the nscf > calculation and how many in the scf one? > stefano > > Vo, Trinh (388C) wrote: >> Dear PWSCF Users, >> >> I run the calculation for a system of 216 atoms (SiP). The CPU time for the >> calculation of 4x4x4 kpt mesh, which actually did calculation for 24 kpts due >> to symmetry, is 1d4h18m (~28h). When I performed a nscf calculation for >> 41kpts along 001 direction, the CPU time is 3d5h (~77h). Thus, the time for >> one kpt in scf calculation is ~1.17h/kpt, and 1.87h in nscf calculation. I >> am wondering how can it take more time for nscf calculation than for scf >> calculation. I don?t understand why. Could you explain to me what can cause >> this to happen. I use QE-4.1.2. >> >> Thank you very much, >> >> Trinh Vo >> CalTech/JPL >> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From rezashidpoor at gmail.com Thu Sep 9 13:14:29 2010 From: rezashidpoor at gmail.com (reza shidpoor) Date: Thu, 9 Sep 2010 15:44:29 +0430 Subject: [Pw_forum] Re [5] : PP-US of Au Message-ID: Dear Gabriele and Dear Paolo Thank you for previous comments. I used your hints and I run small cluster with only 3 atoms but the previous problem has remained. Input file &control calculation='relax' title='Au3Cluster' restart_mode='from_scratch' outdir='./tmp' prefix='Au3' etot_conv_thr=1.0D-4 pseudo_dir = './' wf_collect=.true. / &system ibrav= 2, celldm(1) = 26,nat= 3, ntyp= 1, ecutwfc =72.0, ecutrho=560.0, occupations='smearing', smearing='fd', degauss=0.01 nspin=1, / &electrons electron_maxstep=100, diagonalization='cg', conv_thr=1.0D-6, mixing_beta=0.1, / &ions ion_dynamics='bfgs' / &cell / ATOMIC_SPECIES Au 196.97 Au.pbe-nd-rrkjus.UPF ATOMIC_POSITIONS {angstrom} Au 0.000 0.000 0.000 Au 4.110 0.000 0.000 Au 0.000 4.110 0.000 K_POINTS {automatic} 2 2 2 1 1 1 .... and The Error : forrtl: severe (174): SIGSEGV, segmentation fault occurred Image PC Routine Line Source pw.x 000000000052089C addusdens_ 27 addusdens.f90 pw.x 00000000004DCB76 sum_band_ 153 sum_band.f90 pw.x 000000000044B7D7 electrons_ 287 electrons.f90 pw.x 000000000040BA6E MAIN__ 92 pwscf.f90 pw.x 000000000040B8DC Unknown Unknown Unknown libc.so.6 00002AE29DCF2ABD Unknown Unknown Unknown pw.x 000000000040B7D9 Unknown Unknown Unknown Question : Is it possible that the source of error is related to be semiconducting while I identified metal in input file with smearing keyword ? Best Regards, Reza.Shidpour SUT -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100909/666e63a9/attachment.htm From dimpy.sharma at tyndall.ie Thu Sep 9 13:34:10 2010 From: dimpy.sharma at tyndall.ie (Dimpy Sharma) Date: Thu, 9 Sep 2010 12:34:10 +0100 Subject: [Pw_forum] segmentation fault in PWcond calcuation Message-ID: Hi there, I was trying to calculate transmisson using pwcond,The system contains 282 atoms,my scf calculation for both the lead and scattering run successfully but while running the Pwcond,the calculation stops without giving CRASH file.The output file crashes with the following 6_656: (392.304688) net_send: could not write to fd=5, errno = 32 Segmentation fault segmentation fault. Any advice welcome. Thanks and regards Dimpy Dimpy Sharma ETG-Group Tyndall National Institute Cork -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100909/6cb041e8/attachment.htm From sclauzer at sissa.it Thu Sep 9 14:03:24 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Thu, 9 Sep 2010 14:03:24 +0200 Subject: [Pw_forum] Re [5] : PP-US of Au In-Reply-To: References: Message-ID: Il giorno 09/set/2010, alle ore 13.14, reza shidpoor ha scritto: > > Dear Gabriele and Dear Paolo > > Thank you for previous comments. > > I used your hints and I run small cluster with only 3 atoms But your cell is still quite big, and the PW cutoffs are large for a US PP. Unless you have some special need, I would suggest to use something like 32 and 320 Ry for the wavefunctions and the charge density, resp. To localize the problem you could try to reduce ecutrho first, since from your "debugging" output it seems that the problems comes from the charge density augmentation. > but the previous problem has remained. > > Input file > > &control > calculation='relax' > title='Au3Cluster' > restart_mode='from_scratch' > outdir='./tmp' > prefix='Au3' > etot_conv_thr=1.0D-4 > pseudo_dir = './' > wf_collect=.true. > / > &system > ibrav= 2, celldm(1) = 26,nat= 3, ntyp= 1, > ecutwfc =72.0, ecutrho=560.0, > occupations='smearing', smearing='fd', degauss=0.01 > nspin=1, > / > &electrons > electron_maxstep=100, > diagonalization='cg', > conv_thr=1.0D-6, > mixing_beta=0.1, > / > &ions > ion_dynamics='bfgs' > / > &cell > / > > ATOMIC_SPECIES > Au 196.97 Au.pbe-nd-rrkjus.UPF > ATOMIC_POSITIONS {angstrom} > > Au 0.000 0.000 0.000 > Au 4.110 0.000 0.000 > Au 0.000 4.110 0.000 > > K_POINTS {automatic} > 2 2 2 1 1 1 > > .... > > and > > The Error : > > forrtl: severe (174): SIGSEGV, segmentation fault occurred > Image PC Routine Line Source > pw.x 000000000052089C addusdens_ 27 addusdens.f90 > pw.x 00000000004DCB76 sum_band_ 153 sum_band.f90 > pw.x 000000000044B7D7 electrons_ 287 electrons.f90 > pw.x 000000000040BA6E MAIN__ 92 pwscf.f90 > pw.x 000000000040B8DC Unknown Unknown Unknown > libc.so.6 00002AE29DCF2ABD Unknown Unknown Unknown > pw.x 000000000040B7D9 Unknown Unknown Unknown > > > Question : Is it possible that the source of error is related to be semiconducting while I identified metal in input file with smearing keyword ? Impossible, I would say... moreover, a 3 Au atoms cluster has an odd number of valence electrons and then needs to be treated as a metallic system (unless you want to fix the occupations). HTH GS > > > > Best Regards, > Reza.Shidpour > SUT > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum ? Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100909/a8d5675b/attachment.htm From sclauzer at sissa.it Thu Sep 9 14:07:33 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Thu, 9 Sep 2010 14:07:33 +0200 Subject: [Pw_forum] segmentation fault in PWcond calcuation In-Reply-To: References: Message-ID: Il giorno 09/set/2010, alle ore 13.34, Dimpy Sharma ha scritto: > > Hi there, > > I was trying to calculate transmisson using pwcond,The system contains 282 atoms, > It's a lot of atoms for the memory requirements of PWcond. Have you tried with a smaller system? > my scf calculation for both the lead and scattering run successfully but while running the Pwcond,the calculation stops without giving CRASH file.The output file crashes with the following > > 6_656: (392.304688) net_send: could not write to fd=5, errno = 32 > Segmentation fault > > segmentation fault. > Most probably you have run out of memory. Try with a smaller system or reduce the PW cutoffs or provide more memory per process (e.g. by running with less processes than the total number of CPUs available per node). HTH GS > > Any advice welcome. > Thanks and regards > Dimpy > Dimpy Sharma > ETG-Group > Tyndall National Institute > Cork > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum ? Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100909/eea32f57/attachment.htm From mjfmartins at ice.ufjf.br Thu Sep 9 19:28:25 2010 From: mjfmartins at ice.ufjf.br (=?ISO-8859-1?Q?MATEUS_JOS=C9_FERNANDES_MARTINS?=) Date: Thu, 9 Sep 2010 14:28:25 -0300 Subject: [Pw_forum] wfc files. Message-ID: Hello all, That question is more interesting for developers of the code or people who are interested in software development. I'm trying to develop a software post-processing that needs to read the wave function (wfc files). Already looked into the source code of the programs and can not find the piece of code that formats the wavefunction in the current format. What part of the source code I can read the file wfc before the formatting? Thank you. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100909/ec545725/attachment-0001.htm From mjfmartins at ice.ufjf.br Thu Sep 9 19:29:05 2010 From: mjfmartins at ice.ufjf.br (=?ISO-8859-1?Q?MATEUS_JOS=C9_FERNANDES_MARTINS?=) Date: Thu, 9 Sep 2010 14:29:05 -0300 Subject: [Pw_forum] wfc files. Message-ID: Hello everyone, That question is more interesting for developers of the code or people who are interested in software development. I'm trying to develop a software post-processing that needs to read the wave function (wfc files). Already looked into the source code of the programs and can not find the piece of code that formats the wavefunction in the current format. What part of the source code I can read the file wfc before the formatting? Thank you. Mateus -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100909/106be568/attachment.htm From lcqsigi at gmail.com Fri Sep 10 04:55:37 2010 From: lcqsigi at gmail.com (S. Sanchez) Date: Thu, 9 Sep 2010 22:55:37 -0400 Subject: [Pw_forum] Are the units of the "efield" tag in the PW.x and CP.x codes different? Message-ID: Dear all, I have a question regarding the units of "finite electric fields" reported in the input webpage of the PW.x code: http://www.quantum-espresso.org/input-syntax/INPUT_PW.html For example, in that webpage, the "eamp" tag reports the units of the electric field in "Hartree" (51.4220632*10^10 V/m). This value for the atomic units of the electric field makes complete sense; however, I do not quite understand the units reported for "efield" tag. For example, I understand the fact that the value of the "efield" tag should be smaller than that of "eamp" tag; indeed, the value of the "finite electric field" appears to be that of the "eamp" tag divided by the square root of 2. That is: (51.4221*10^10 V/m)/SQRT(2)=36.3609*10^10 V/m. What I do not understand is the fact that the label "efiled" in the INPUT_PW.html webpage reports the following information: Amplitude of the finite electric field (in Ry a.u.; 1 a.u. = 36.3609*10^10 V/m) If I follow the above expression, i.e., (51.4221*10^10 V/m)/SQRT(2)=36.3609*10^10 V/m, the units of the "finite electric field" should be Hartrees and not in Rydbergs, right? Also, in the INPUT_CP.html webpage, the same label "efield" reports the following: Amplitude of the finite electric field (in a.u.; 1 a.u. = 36.3609*10^10 V/m); where, one assumes a.u. to be equal to Hartree. So, I would like to know if the units of the "finite electric field" in the PW.x code are given in Hartrees or in Rydbergs? Thanks a lot for your help, S. Sanchez -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100909/9ec3f018/attachment.htm From giannozz at democritos.it Fri Sep 10 11:20:38 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 10 Sep 2010 11:20:38 +0200 Subject: [Pw_forum] wfc files. In-Reply-To: References: Message-ID: <4C89F866.9020006@democritos.it> MATEUS JOS? FERNANDES MARTINS wrote: > I'm trying to develop a software post-processing that needs > to read the wave function (wfc files) take any existing post-processing code that reads the wavefunctions and modify it. It will be much simpler than trying to figure out where and how wavefunctions are read in the code. P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Fri Sep 10 12:35:58 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 10 Sep 2010 12:35:58 +0200 Subject: [Pw_forum] Re [5] : PP-US of Au In-Reply-To: References: Message-ID: <4C8A0A0E.90505@democritos.it> It works for me - P. -- Paolo Giannozzi, Democritos and University of Udine, Italy -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: 421.out Url: http://www.democritos.it/pipermail/pw_forum/attachments/20100910/e9b9ed47/attachment.txt From giannozz at democritos.it Fri Sep 10 12:50:44 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 10 Sep 2010 12:50:44 +0200 Subject: [Pw_forum] Are the units of the "efield" tag in the PW.x and CP.x codes different? In-Reply-To: References: Message-ID: <4C8A0D84.6070204@democritos.it> S. Sanchez wrote: > What I do not understand is the fact that the label "efiled" in the > INPUT_PW.html webpage reports the following information: > Amplitude of the finite electric field (in Ry a.u.; 1 a.u. = > 36.3609*10^10 V/m) > > If I follow the above expression, i.e., (51.4221*10^10 > V/m)/SQRT(2)=36.3609*10^10 V/m, the units of the "finite electric field" > should be Hartrees and not in Rydbergs, right? "eamp" is in "Hartree" atomic units, i.e. 51.4221*10^10 V/m "efield" is in "Rydberg" atomic units, i.e. 36.3609*10^10 V/m, i.e. 51.4221*10^10 V/m/sqrt(2) (where sqrt(2)=e, in Ry a.u.) > Also, in the INPUT_CP.html webpage, the same label "efield" reports the > following: > Amplitude of the finite electric field (in a.u.; 1 a.u. = 36.3609*10^10 V/m) this is actually incorrect (I think): in CP the electric field is in "Hartree" atomic units = 51.4221*10^10 V/m P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From orondo at MIT.EDU Fri Sep 10 14:41:14 2010 From: orondo at MIT.EDU (Peter O Orondo) Date: Fri, 10 Sep 2010 08:41:14 -0400 Subject: [Pw_forum] Single Cell vs Super-Cell Message-ID: <00f701cb50e5$76b35c30$641a1490$@edu> Hi- I am trying to calculate the chemical potential of H in a bunch of metals. When I use a single cell, I get results that compare to experiment. However, when I use a 2x2x2 super-cell, I expect to get a more accurate result since the super-cell should model the physical properties more accurately. However, the super-cell results don't even have the right shape. The results look rather random. Any insights into what could be wrong? Here are more details of what I am doing: To get the chemical potential, I put the a number of H atoms at random into the available O-sites. Because there can be a large number H-atoms configurations, I take an average of sample SCF energies. I build the super-cell using the same techniques as in this forum - and I check the structure and configuration with Xcrysden before calculating its energy. Because the lattice constant changes depending on the number of absorbed H, I relax it at each concentration. My reference configuration is a H2 molecule in a large box (same size as the supercell) Thanks in advance Peter Student, EECS, MIT -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100910/f0053bc5/attachment.htm From modaresi.mohsen at gmail.com Fri Sep 10 15:25:03 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Fri, 10 Sep 2010 17:55:03 +0430 Subject: [Pw_forum] Single Cell vs Super-Cell In-Reply-To: <00f701cb50e5$76b35c30$641a1490$@edu> References: <00f701cb50e5$76b35c30$641a1490$@edu> Message-ID: I have the same question about bandstructure in big unit cells: I want to calculate band structructure of a big unit cell but i get some parallel line . My computing resources are limited and i can increse one of the "ecut","ether" , "number of K-point in the scf and band structure calculation" "self consistent accuracy". What is your suggestion? thanks Mohsen Modarresi Ferdowsi University of Mashhad, Iran. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100910/eeaed1b9/attachment-0001.htm From tkaloni at gmail.com Fri Sep 10 18:28:39 2010 From: tkaloni at gmail.com (Thaneshwor Kaloni) Date: Fri, 10 Sep 2010 22:13:39 +0545 Subject: [Pw_forum] Hi Message-ID: Dear all user, I have a question , how one can plot bonding charge density, how one can plot the wave function using data created by scf or nscf files?-- I will be appreciate any type of support. Best Regards! Thaneshwor P Kaloni -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100910/12c3c0c7/attachment.htm From elie.moujaes at hotmail.co.uk Fri Sep 10 22:37:27 2010 From: elie.moujaes at hotmail.co.uk (Elie Moujaes) Date: Fri, 10 Sep 2010 21:37:27 +0100 Subject: [Pw_forum] Bilayer graphene with electric field applied (revisited) Message-ID: Dear all, I am still having problems with the convergence concerning a system of bilayer graphene under the effect of an electric field. The band structure graph is very jiggy. Below you will find the calculation= `bands` input and part of this output will also be shown. Please can anyone advise me on that. Thanks very much. I used all types of pseudopotentials with the correct ecutrho and ecutwfc. Elie Moujaes University of Nottingham NG7 2RD UK INPUT: &control prefix='bi elgraphene', calculation='bands', restart_mode='from_scratch', lelfield = .true., pseudo_dir = '/espresso-4.2/pseudo/', outdir='/tmp/results_MOUJAES/' / &system ibrav= 4, celldm(1) =4.608737, celldm(3)=5.330410, nat=4, ntyp= 1, ecutwfc = 37.D0,ecutrho = 147.D0, occupations='smearing',london =.true., smearing='methfessel-paxton', degauss=0.01 / &electrons conv_thr=1.D-5, mixing_beta=0.3D0, mixing_mode='local-TF' efield_cart(1) = 0.0, efield_cart(2) = 0.0, efield_cart(3) = 0.001, startingwfc = 'random' / ATOMIC_SPECIES C 12.0107 C.lda-paw_kj.UPF ATOMIC_POSITIONS crystal C 0.000000 0.000000 0.000000 C 0.000000 0.000000 0.257692 C 0.333333 -0.33333 0.000000 C -0.333333 0.33333 0.257692 K_POINTS automatic 130 130 1 0 0 0 PART OF THE OUTPUT: The potential is recalculated from file : /tmp/results_MOUJAES/bi elgraphene.save/charge-density.dat negative rho (up, down): 0.637E-05 0.000E+00 Starting wfc are random total cpu time spent up to now is 1233.06 secs per-process dynamical memory: 17.2 Mb Band Structure Calculation Davidson diagonalization with overlap c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged ...... -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100910/8a8df98b/attachment.htm From bipulrr at gmail.com Sat Sep 11 05:55:20 2010 From: bipulrr at gmail.com (Bipul Rakshit) Date: Sat, 11 Sep 2010 09:25:20 +0530 Subject: [Pw_forum] Low q phonon mode. Message-ID: I am doing phonon calculation of graphene. I am using 16x16x1 q-point grid. Which results in 30 high symm points. The kinetic energy and charge density are 40 and 480 Ry res. The k-point grid is also 16x16x1. so i just want to ask, whether is it meaningful to do calculation with higher precision to obtain the dynamical matrices *.dyn1 and *.dyn2 using 36x36x1 and 100, 2800 Ry. i want higher precision for just low q modes. But calculating the whole phonon dispersion with the later parameters is computationaly too much expensive. Any suggesion will be helpful. Thanks -- Dr. Bipul Rakshit Research Associate, S N Bose Centre for Basic Sciences, Salt Lake, Kolkata 700 098 India From alexeftimiades at gmail.com Sat Sep 11 20:15:31 2010 From: alexeftimiades at gmail.com (Alexander Eftimiades) Date: Sat, 11 Sep 2010 14:15:31 -0400 Subject: [Pw_forum] Quantum Espresso installed, but examples not running Message-ID: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> I hope I came to the right place to ask this question. I have compiled quantum espresso using this set of instructions: http://how-to.wikia.com/wiki/How_to_install_PWscf_on_a_Debian_or_Debian_based_operating_system I made two modifications to these instructions: I installed version 4.2.1 and put the folder together using these instructions:http://qe-forge.org/frs/shownotes.php?release_id=64. These instructions did not say anything about "PLUMED" plugin so I did not download it. I have gfortran installed so I did not do the first part of the instructions (although I later tried to install g95 using the link provided and it kept saying "authentication failed") I have openmpi installed but I only have one cpu at the moment (I intend to set up a network later.) Everything seemed to work ok so I tried to run one of the examples. I got this: /opt/espresso-4.2.1/examples/example08 : starting This example shows how to use pw.x to calculate the DOS of Ni. executables directory: /opt/espresso-4.2.1/bin pseudo directory: /opt/espresso-4.2.1/pseudo temporary directory: /tmp checking that needed directories and files exist... done running pw.x as: /opt/espresso-4.2.1/bin/pw.x running dos.x as: /opt/espresso-4.2.1/bin/dos.x running projwfc.x as: /opt/espresso-4.2.1/bin/projwfc.x running kvecs_FS.x as: /opt/espresso-4.2.1/bin/kvecs_FS.x running bands_FS.x as: /opt/espresso-4.2.1/bin/bands_FS.x cleaning /tmp... done running the scf calculation for Ni...[debian:13270] *** Process received signal *** [debian:13270] Signal: Segmentation fault (11) [debian:13270] Signal code: (128) [debian:13270] Failing at address: (nil) [debian:13270] [ 0] [0xb779040c] [debian:13270] [ 1] /usr/lib/libfftw3.so.3 [0xb6f30d5c] [debian:13270] [ 2] /usr/lib/libfftw3.so.3 [0xb6f30eca] [debian:13270] [ 3] /usr/lib/libfftw3.so.3 [0xb6f2e684] [debian:13270] [ 4] /usr/lib/libfftw3.so.3 [0xb6f33618] [debian:13270] [ 5] /usr/lib/libfftw3.so.3(dfftw_execute_dft_+0x5c) [0xb6fa6b7c] [debian:13270] [ 6] /opt/espresso-4.2.1/bin/ pw.x(__fft_scalar_MOD_cft_1z+0x912) [0x81d0b62] [debian:13270] [ 7] /opt/espresso-4.2.1/bin/ pw.x(__fft_parallel_MOD_tg_cft3s+0x7d3) [0x81ce9d3] [debian:13270] [ 8] /opt/espresso-4.2.1/bin/pw.x(cft3_+0x52) [0x807ccd2] [debian:13270] [ 9] /opt/espresso-4.2.1/bin/pw.x(interpolate_+0x408) [0x8176848] [debian:13270] [10] /opt/espresso-4.2.1/bin/pw.x(set_vrs_+0x145) [0x80f3c25] [debian:13270] [11] /opt/espresso-4.2.1/bin/pw.x(potinit_+0xc1d) [0x80cf10d] [debian:13270] [12] /opt/espresso-4.2.1/bin/pw.x(init_run_+0x44d) [0x809ceed] [debian:13270] [13] /opt/espresso-4.2.1/bin/pw.x(MAIN__+0xa2) [0x8074b82] [debian:13270] [14] /opt/espresso-4.2.1/bin/pw.x(main+0x39) [0x834acc9] [debian:13270] [15] /lib/i686/cmov/libc.so.6(__libc_start_main+0xe5) [0xb67d9455] [debian:13270] [16] /opt/espresso-4.2.1/bin/pw.x [0x8074a41] [debian:13270] *** End of error message *** ./run_example: line 174: 13270 Segmentation fault $PW_COMMAND < ni.scf.in > ni.scf.out Error condition encountered during test: exit status = 139 Aborting ---------rest of email continues here---------- Similar messages come up for other examples. Could someone please tell me what is going on? And--yes I read the user manual instructions (and yes chapter 2). If there is a pointer in there I missed please tell me. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100911/d4977c8c/attachment.htm From eyvaz_isaev at yahoo.com Sat Sep 11 21:36:48 2010 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sat, 11 Sep 2010 12:36:48 -0700 (PDT) Subject: [Pw_forum] Low q phonon mode. In-Reply-To: References: Message-ID: <316922.22938.qm@web65704.mail.ac4.yahoo.com> Hi, From: Bipul Rakshit > The kinetic energy and charge density are 40 and 480 Ry > using 36x36x1 and 100, 2800 Ry. Please keep in mind the cutoff energy is a characteristic feature of a pseudopotential, especially for US and PAW. "ecutrho" can be obtained doing some test calculations. Anyway, you should do some tests, if you are not sure which ecutwfc and ecutrho are to be applied. > so i just want to ask, whether is it meaningful to do calculation with higher >precision to obtain the dynamical matrices *.dyn1 and *.dyn2 using 36x36x1 and >100, 2800 Ry. I think you will kill your supercomputer with these parameters. May be it is worth doing several phonon tests for a particular small q-point varying ecutwfc, ecutrho and k-mesh. Then decide, which parameters are better for you. Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com From eyvaz_isaev at yahoo.com Sat Sep 11 22:07:18 2010 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sat, 11 Sep 2010 13:07:18 -0700 (PDT) Subject: [Pw_forum] Quantum Espresso installed, but examples not running In-Reply-To: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> References: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> Message-ID: <599726.53728.qm@web65709.mail.ac4.yahoo.com> Hi, Usually "Segmentation fault" tells about insufficient memory problem, but it might be due to a compiler/lib error. I can start pw.x on my Ubuntu 10.04 (in fact, Debian-based, and quite buggy) with gfortran. Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com ________________________________ From: Alexander Eftimiades To: pw_forum at pwscf.org Sent: Sat, September 11, 2010 8:15:31 PM Subject: [Pw_forum] Quantum Espresso installed, but examples not running I hope I came to the right place to ask this question. I have compiled quantum espresso using this set of instructions: http://how-to.wikia.com/wiki/How_to_install_PWscf_on_a_Debian_or_Debian_based_operating_system I made two modifications to these instructions: 1. I installed version 4.2.1 and put the folder together using these instructions:http://qe-forge.org/frs/shownotes.php?release_id=64. These instructions did not say anything about "PLUMED" plugin so I did not download it. 2. I have gfortran installed so I did not do the first part of the instructions (although I later tried to install g95 using the link provided and it kept saying "authentication failed") I have openmpi installed but I only have one cpu at the moment (I intend to set up a network later.) Everything seemed to work ok so I tried to run one of the examples. I got this: /opt/espresso-4.2.1/examples/example08 : starting This example shows how to use pw.x to calculate the DOS of Ni. executables directory: /opt/espresso-4.2.1/bin pseudo directory: /opt/espresso-4.2.1/pseudo temporary directory: /tmp checking that needed directories and files exist... done running pw.x as: /opt/espresso-4.2.1/bin/pw.x running dos.x as: /opt/espresso-4.2.1/bin/dos.x running projwfc.x as: /opt/espresso-4.2.1/bin/projwfc.x running kvecs_FS.x as: /opt/espresso-4.2.1/bin/kvecs_FS.x running bands_FS.x as: /opt/espresso-4.2.1/bin/bands_FS.x cleaning /tmp... done running the scf calculation for Ni...[debian:13270] *** Process received signal *** [debian:13270] Signal: Segmentation fault (11) [debian:13270] Signal code: (128) [debian:13270] Failing at address: (nil) [debian:13270] [ 0] [0xb779040c] [debian:13270] [ 1] /usr/lib/libfftw3.so.3 [0xb6f30d5c] [debian:13270] [ 2] /usr/lib/libfftw3.so.3 [0xb6f30eca] [debian:13270] [ 3] /usr/lib/libfftw3.so.3 [0xb6f2e684] [debian:13270] [ 4] /usr/lib/libfftw3.so.3 [0xb6f33618] [debian:13270] [ 5] /usr/lib/libfftw3.so.3(dfftw_execute_dft_+0x5c) [0xb6fa6b7c] [debian:13270] [ 6] /opt/espresso-4.2.1/bin/pw.x(__fft_scalar_MOD_cft_1z+0x912) [0x81d0b62] [debian:13270] [ 7] /opt/espresso-4.2.1/bin/pw.x(__fft_parallel_MOD_tg_cft3s+0x7d3) [0x81ce9d3] [debian:13270] [ 8] /opt/espresso-4.2.1/bin/pw.x(cft3_+0x52) [0x807ccd2] [debian:13270] [ 9] /opt/espresso-4.2.1/bin/pw.x(interpolate_+0x408) [0x8176848] [debian:13270] [10] /opt/espresso-4.2.1/bin/pw.x(set_vrs_+0x145) [0x80f3c25] [debian:13270] [11] /opt/espresso-4.2.1/bin/pw.x(potinit_+0xc1d) [0x80cf10d] [debian:13270] [12] /opt/espresso-4.2.1/bin/pw.x(init_run_+0x44d) [0x809ceed] [debian:13270] [13] /opt/espresso-4.2.1/bin/pw.x(MAIN__+0xa2) [0x8074b82] [debian:13270] [14] /opt/espresso-4.2.1/bin/pw.x(main+0x39) [0x834acc9] [debian:13270] [15] /lib/i686/cmov/libc.so.6(__libc_start_main+0xe5) [0xb67d9455] [debian:13270] [16] /opt/espresso-4.2.1/bin/pw.x [0x8074a41] [debian:13270] *** End of error message *** ./run_example: line 174: 13270 Segmentation fault $PW_COMMAND < ni.scf.in > ni.scf.out Error condition encountered during test: exit status = 139 Aborting ---------rest of email continues here---------- Similar messages come up for other examples. Could someone please tell me what is going on? And--yes I read the user manual instructions (and yes chapter 2). If there is a pointer in there I missed please tell me. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100911/7acb5b40/attachment.htm From giannozz at democritos.it Sat Sep 11 22:08:47 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sat, 11 Sep 2010 22:08:47 +0200 Subject: [Pw_forum] Quantum Espresso installed, but examples not running In-Reply-To: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> References: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> Message-ID: <97CB9382-7F6E-4059-971B-55EB2EB073E1@democritos.it> On Sep 11, 2010, at 20:15 , Alexander Eftimiades wrote: > I hope I came to the right place to ask this question. you did > If there is a pointer in there I missed please tell me. 2.7.1 [...] If you get mysterious errors in the provided tests and examples: your compiler, or your mathematical libraries, or MPI libraries, or a combination thereof, is very likely buggy. Although the presence of subtle bugs in QUANTUM ESPRESSO that are not revealed during the testing phase can never be ruled out, it is very unlikely that this happens on the provided tests and examples. 10.1.0.9 pw.x crashes with segmentation fault or similarly obscure messages Possible reasons: * too much RAM memory or stack requested (see next item). * if you are using highly optimized mathematical libraries, verify that they are designed for your hardware. * If you are using aggressive optimization in compilation, verify that you are using the appropriate options for your machine * The executable was not properly compiled, or was compiled on a different and incompatible environment. * buggy compiler or libraries: this is the default explanation if you have problems with the provided tests and examples. P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at democritos.it Sat Sep 11 22:17:11 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sat, 11 Sep 2010 22:17:11 +0200 Subject: [Pw_forum] Quantum Espresso installed, but examples not running In-Reply-To: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> References: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> Message-ID: On Sep 11, 2010, at 20:15 , Alexander Eftimiades wrote: > [debian:13270] [ 1] /usr/lib/libfftw3.so.3 [0xb6f30d5c] if this message appears for all errors, compile internal FFTW instead of using external FFTW v.3: in make.sys replace -D__FFTW3 with -D__FFTW and remove -lfftw3 from libraries, delete all objects that depend upon such options (rm Modules/fft_scalar.o clib/fft_stick.o should be enough), recompile P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From alexeftimiades at gmail.com Sat Sep 11 22:19:43 2010 From: alexeftimiades at gmail.com (Alexander Eftimiades) Date: Sat, 11 Sep 2010 16:19:43 -0400 Subject: [Pw_forum] Quantum Espresso installed, but examples not running In-Reply-To: <97CB9382-7F6E-4059-971B-55EB2EB073E1@democritos.it> References: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> <97CB9382-7F6E-4059-971B-55EB2EB073E1@democritos.it> Message-ID: I should then mention that everything used to compile and execute Quantum Espresso was built for Debian because it was installed using synaptic package manager. The large library of packages Synaptic provides is exactly the reason I use Debian. I just updated all my software recently so I am quite sure I am using the latest version of MPI, gfortran, etc Synaptic has to offer. If you still think my compilers are the problem--any ideas on how I could go about fixing this. I scoured the web for free ab initio quantum chemistry software that can calculate phonon dispersion relations and this definitely seemed to be one of the best. I would really like to get this running--especially since my projects are not very different than some of the examples provided. On Sep 11, 2010, at 4:08 PM, Paolo Giannozzi wrote: > On Sep 11, 2010, at 20:15 , Alexander Eftimiades wrote: > >> I hope I came to the right place to ask this question. > > you did > >> If there is a pointer in there I missed please tell me. > > 2.7.1 > [...] If you get mysterious errors in the provided tests and > examples: your compiler, or your mathematical > libraries, or MPI libraries, or a combination thereof, is very likely > buggy. Although the presence of subtle > bugs in QUANTUM ESPRESSO that are not revealed during the testing > phase can never be ruled out, > it is very unlikely that this happens on the provided tests and > examples. > > 10.1.0.9 pw.x crashes with segmentation fault or similarly obscure > messages > Possible reasons: > > * too much RAM memory or stack requested (see next item). > * if you are using highly optimized mathematical libraries, > verify that they are designed for your hardware. > * If you are using aggressive optimization in compilation, > verify that you are using the appropriate options for your machine > * The executable was not properly compiled, or was compiled on a > different and incompatible environment. > * buggy compiler or libraries: this is the default explanation > if you have problems with the provided tests and examples. > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From alexeftimiades at gmail.com Sat Sep 11 22:45:55 2010 From: alexeftimiades at gmail.com (Alexander Eftimiades) Date: Sat, 11 Sep 2010 16:45:55 -0400 Subject: [Pw_forum] Quantum Espresso installed, but examples not running In-Reply-To: References: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> Message-ID: I am not sure what you mean by "internal." I installed libfftw3 and dev-libfftw3 from Synaptic Package manager. I am sorry, but I am still relatively new to linux, what should I do to compile "internal" FFTW? On Sep 11, 2010, at 4:17 PM, Paolo Giannozzi wrote: > > On Sep 11, 2010, at 20:15 , Alexander Eftimiades wrote: > >> [debian:13270] [ 1] /usr/lib/libfftw3.so.3 [0xb6f30d5c] > > if this message appears for all errors, compile internal > FFTW instead of using external FFTW v.3: in make.sys > replace -D__FFTW3 with -D__FFTW and remove -lfftw3 > from libraries, delete all objects that depend upon such > options (rm Modules/fft_scalar.o clib/fft_stick.o should > be enough), recompile > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From modaresi.mohsen at gmail.com Sat Sep 11 22:49:28 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Sun, 12 Sep 2010 01:19:28 +0430 Subject: [Pw_forum] Quantum Espresso installed, but examples not running In-Reply-To: References: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> Message-ID: Perhaps it is not the best way, but i think if you uninstall FFTW from Synaptic esperesso use internal FFTW. On Sun, Sep 12, 2010 at 1:15 AM, Alexander Eftimiades < alexeftimiades at gmail.com> wrote: > I am not sure what you mean by "internal." I installed libfftw3 and > dev-libfftw3 from Synaptic Package manager. > > I am sorry, but I am still relatively new to linux, what should I do > to compile "internal" FFTW? > > On Sep 11, 2010, at 4:17 PM, Paolo Giannozzi wrote: > > > > > On Sep 11, 2010, at 20:15 , Alexander Eftimiades wrote: > > > >> [debian:13270] [ 1] /usr/lib/libfftw3.so.3 [0xb6f30d5c] > > > > if this message appears for all errors, compile internal > > FFTW instead of using external FFTW v.3: in make.sys > > replace -D__FFTW3 with -D__FFTW and remove -lfftw3 > > from libraries, delete all objects that depend upon such > > options (rm Modules/fft_scalar.o clib/fft_stick.o should > > be enough), recompile > > > > P. > > --- > > Paolo Giannozzi, Dept of Physics, University of Udine > > via delle Scienze 208, 33100 Udine, Italy > > Phone +39-0432-558216, fax +39-0432-558222 > > > > > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100912/48eb7f40/attachment.htm From alexeftimiades at gmail.com Sat Sep 11 22:56:35 2010 From: alexeftimiades at gmail.com (Alexander Eftimiades) Date: Sat, 11 Sep 2010 16:56:35 -0400 Subject: [Pw_forum] Quantum Espresso installed, but examples not running In-Reply-To: References: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> Message-ID: If I uninstall FFTW from synaptic then synaptic will uninstall quite a few other packages. What about: Paolo Giannozzi's suggestion-- If this message appears for all errors, compile internal FFTW instead of using external FFTW v.3: in make.sys replace -D__FFTW3 with -D__FFTW and remove -lfftw3 from libraries, delete all objects that depend upon such options (rm Modules/fft_scalar.o clib/fft_stick.o should be enough), recompile I still do not know what you mean by "internal"? Do you mean Quantum Espresso comes with FFTW? Do I have to download/install anything or do I just set these variables in the make.sys file and compile as usual? On Sep 11, 2010, at 4:49 PM, mohsen modaresi wrote: > > Perhaps it is not the best way, but i think if you uninstall FFTW > from Synaptic esperesso use internal FFTW. > > > On Sun, Sep 12, 2010 at 1:15 AM, Alexander Eftimiades > wrote: > I am not sure what you mean by "internal." I installed libfftw3 and > dev-libfftw3 from Synaptic Package manager. > > I am sorry, but I am still relatively new to linux, what should I do > to compile "internal" FFTW? > > On Sep 11, 2010, at 4:17 PM, Paolo Giannozzi wrote: > > > > > On Sep 11, 2010, at 20:15 , Alexander Eftimiades wrote: > > > >> [debian:13270] [ 1] /usr/lib/libfftw3.so.3 [0xb6f30d5c] > > > > if this message appears for all errors, compile internal > > FFTW instead of using external FFTW v.3: in make.sys > > replace -D__FFTW3 with -D__FFTW and remove -lfftw3 > > from libraries, delete all objects that depend upon such > > options (rm Modules/fft_scalar.o clib/fft_stick.o should > > be enough), recompile > > > > P. > > --- > > Paolo Giannozzi, Dept of Physics, University of Udine > > via delle Scienze 208, 33100 Udine, Italy > > Phone +39-0432-558216, fax +39-0432-558222 > > > > > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100911/515c250e/attachment.htm From giannozz at democritos.it Sat Sep 11 23:02:38 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sat, 11 Sep 2010 23:02:38 +0200 Subject: [Pw_forum] Quantum Espresso installed, but examples not running In-Reply-To: References: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> Message-ID: <67941736-2191-4C96-9537-770DA3FE1CA0@democritos.it> On Sep 11, 2010, at 22:45 , Alexander Eftimiades wrote: > I am sorry, but I am still relatively new to linux, what should I do > to compile "internal" FFTW? you should do what was written in my message: >> in make.sys replace -D__FFTW3 with -D__FFTW and >> remove -lfftw3 from libraries, delete all objects that >> depend upon such options ("rm Modules/fft_scalar.o clib/fft_stick.o" maybe you also need a "rm clib/clib.a", just in case >> should be enough), recompile "make pwall" or "make all" depending on what you need. P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From alexeftimiades at gmail.com Sat Sep 11 23:26:43 2010 From: alexeftimiades at gmail.com (Alexander Eftimiades) Date: Sat, 11 Sep 2010 17:26:43 -0400 Subject: [Pw_forum] Quantum Espresso installed, but examples not running In-Reply-To: <67941736-2191-4C96-9537-770DA3FE1CA0@democritos.it> References: <88490EEB-8E4F-4837-BD06-E4D6A779D55A@gmail.com> <67941736-2191-4C96-9537-770DA3FE1CA0@democritos.it> Message-ID: <00CD185B-C0D2-41EA-8ADE-E6BB45B92735@gmail.com> Thank you, that seemed to work well. On Sep 11, 2010, at 5:02 PM, Paolo Giannozzi wrote: > > On Sep 11, 2010, at 22:45 , Alexander Eftimiades wrote: > >> I am sorry, but I am still relatively new to linux, what should I do >> to compile "internal" FFTW? > > you should do what was written in my message: > >>> in make.sys replace -D__FFTW3 with -D__FFTW and >>> remove -lfftw3 from libraries, delete all objects that >>> depend upon such options ("rm Modules/fft_scalar.o clib/fft_stick.o" > > maybe you also need a "rm clib/clib.a", just in case > >>> should be enough), recompile > > > "make pwall" or "make all" depending on what you need. > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From yanjiaxu at gmail.com Sun Sep 12 08:45:44 2010 From: yanjiaxu at gmail.com (jiaxu yan) Date: Sun, 12 Sep 2010 14:45:44 +0800 Subject: [Pw_forum] q2r.x error Message-ID: Dear PWscf users, I performed an electron-phonon coupling calculation and I encountered the error when I executed q2r.x. The error message appears:'At line 356 of file q2r.f90 Fortran runtime error: Bad real number in item 1 of list input' Could you tell me why? Thank you very much Yan Jiaxu School of Physical and Mathematical Sciences Nanyang Technological University, Singapore -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100912/b46324f5/attachment.htm From giannozz at democritos.it Sun Sep 12 09:18:24 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sun, 12 Sep 2010 09:18:24 +0200 Subject: [Pw_forum] q2r.x error In-Reply-To: References: Message-ID: On Sep 12, 2010, at 8:45 , jiaxu yan wrote: > The error message appears:'At line 356 of file q2r.f90 > Fortran runtime error: Bad real number in item 1 of list input' > Could you tell me why? the code already told you why (expecting a real number as first item in a line, found something else) and where (line 356 fo q2r.f90). You should look atline 356 of q2r.f90, understand which file is read there and which line does not look good P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From yanjiaxu at gmail.com Sun Sep 12 10:12:39 2010 From: yanjiaxu at gmail.com (jiaxu yan) Date: Sun, 12 Sep 2010 16:12:39 +0800 Subject: [Pw_forum] q2r.x error Message-ID: Hi, Paolo Thank you for such fast reply. The line 356 fo q2r.f90 is " READ(filea2F,*) deg, ef, dosscf" I have checked the file 'a2Fq2r.51' and it seems to be all right. The first lines of the file is: 5.000000000000000E-003 -0.107770458638552 8.08004901171817 1 Dynamical Matrix in cartesian axes q = ( 0.000000000 0.000000000 0.000000000 ) 1 1 1.62795683 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 1.62795683 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 1.05310588 0.00000000 1 2 -3.01515344 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 -3.01515344 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 1.27442418 0.00000000 1 3 1.38764451 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 1.38764451 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 -2.33154662 0.00000000 .......... .......... .......... .......... ........... .......... The odd thing is that all goes well when I prefromed the same calculation with more coarse parameters(less K points in scf calculation and small Q in el-phon calculation). I am very bemused.... Yan Jiaxu School of Physical and Mathematical Sciences Nanyang Technological University, Singapore On Sep 12, 2010, at 8:45 , jiaxu yan wrote: > The error message appears:'At line 356 of file q2r.f90 > Fortran runtime error: Bad real number in item 1 of list input' > Could you tell me why? the code already told you why (expecting a real number as first item in a line, found something else) and where (line 356 fo q2r.f90). You should look atline 356 of q2r.f90, understand which file is read there and which line does not look good P. --- Paolo Giannozzi, Dept of Physics, University of 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://www.democritos.it/pipermail/pw_forum/attachments/20100912/eab72d48/attachment.htm From zdw2000 at gmail.com Sun Sep 12 13:22:38 2010 From: zdw2000 at gmail.com (Wei Zhou) Date: Sun, 12 Sep 2010 19:22:38 +0800 Subject: [Pw_forum] dos anyone has a potassium pseudopotential with PBE and 9 valences Message-ID: the pseduo potential cannot find in the pwscf psuedo libary. I need it work correctly at a pressure of 300GPa and the PBE exchange-correlation function is needed. any help will be appreciated -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100912/0f4b6a25/attachment.htm From quantumdft at gmail.com Sun Sep 12 15:11:23 2010 From: quantumdft at gmail.com (vega lew) Date: Sun, 12 Sep 2010 21:11:23 +0800 Subject: [Pw_forum] How to fix certain bond length in NEB calculation Message-ID: <4C8CD17B.70402@gmail.com> Dear all, I am calculating the bond activation process on oxide surface. There are only one bond broken in this process. When I try to calculate the activation barrier of this process using normal NEB calculation, I failed in the convergence of the calculation. The task is not converged even after hundreds of NEB steps. So I want to fix the bond length of certain bond in different images to see how the energy changes as a function of bond length. This might be very convenient to use internal coordinate to fix bond length of certain bond of adsorbate on the surface. I just wondering whether I could fix certain bond length in NEB calculation by a simple command in PWSCF. thank you for reading. best wishes, vega -- ================================================ Vega Lew ( weijia liu) Graduate student State Key Laboratory of Materials-oriented Chemical Engineering Nanjing University of Technology, 210009, Nanjing, China *************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, China *************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100912/32df8091/attachment.htm From barfi.koa at gmail.com Sun Sep 12 20:11:16 2010 From: barfi.koa at gmail.com (B. Koa) Date: Sun, 12 Sep 2010 22:41:16 +0430 Subject: [Pw_forum] dosup and pdosup in pdos_tot file Message-ID: Dear PWScf Users It seems that "dosup" and "pdosup" which are in the pdos_tot file must be equal and really they are equal in many cases. But in some cases they differ a very bit with each other. I think that the origin of this very small difference is only the numerical procedure and not any physical one. Am I correct ? Many Thanks B. Koa -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100912/5e90580c/attachment.htm From rezashidpoor at gmail.com Mon Sep 13 11:14:03 2010 From: rezashidpoor at gmail.com (reza shidpoor) Date: Mon, 13 Sep 2010 13:44:03 +0430 Subject: [Pw_forum] Re [6] : US-PP of Au Message-ID: Dear Paolo Thank you for reply. With ecut = 72 Ry , I repeated the simulation five times. The simulation proceeded several scf iterations and several bfgs iterations but the jobs were never finished. ( I attached one of the out puts. ) Is it possible the ecut is so great that RAM is not sufficient for 3-atoms cluster ? Best regards. Reza.Shidpour SUT -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100913/09d10f15/attachment-0001.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: Au3_relax_6.out Type: application/octet-stream Size: 53837 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100913/09d10f15/attachment-0002.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: Au3_relax.in Type: application/octet-stream Size: 772 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100913/09d10f15/attachment-0003.obj From Lorenzo.Paulatto at impmc.upmc.fr Mon Sep 13 11:29:15 2010 From: Lorenzo.Paulatto at impmc.upmc.fr (Lorenzo Paulatto) Date: Mon, 13 Sep 2010 11:29:15 +0200 Subject: [Pw_forum] Re [6] : US-PP of Au In-Reply-To: References: Message-ID: On Mon, 13 Sep 2010 11:14:03 +0200, reza shidpoor wrote: > Dear Paolo > > Thank you for reply. > > With ecut = 72 Ry , I repeated the simulation five times. The simulation > proceeded > several scf iterations and several bfgs iterations but the jobs were > never > finished. ( I attached one of the out puts. ) > > Is it possible the ecut is so great that RAM is not sufficient for > 3-atoms > cluster ? Not really, because after the first BFGS iteration the amount of RAM memory required does not increase any more. From your output it is not clear what is that causes the stop, is it always a segmentation fault? To me it looks like everything is going well as the forces are decreasing; however are you sure of your structure? I've had a look at it with xcrysden and it looks like there is a whole lot of vacuum: the gold atoms almost form a chain along one of the cell axis... regards > > Best regards. > > Reza.Shidpour > SUT -- Lorenzo Paulatto post-doc @ IMPMC/UPMC - Universit? Paris 6 phone: +33 (0)1 44 27 74 89 www: http://www-int.impmc.upmc.fr/~paulatto/ previously (take note of the change!): phd student @ SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 www: http://people.sissa.it/~paulatto/ From elie.moujaes at hotmail.co.uk Mon Sep 13 14:38:30 2010 From: elie.moujaes at hotmail.co.uk (Elie Moujaes) Date: Mon, 13 Sep 2010 13:38:30 +0100 Subject: [Pw_forum] non existent or non writable directory Message-ID: Dear all, I have a small problem. I was trying to run a scf calculation for my system and I used the same outdir I usually use in my calculations but then after a few seconds the program stopped with an error: from outdir: error # 1 /exp/home/caiapo/........./Moujaes-Results/ non existent or non writable I made another directory and when I re-ran the calculation i got the same problem (although that new directory is empty).. How can I resolve this? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100913/69bafdbd/attachment.htm From siyouber at yahoo.fr Mon Sep 13 15:33:13 2010 From: siyouber at yahoo.fr (Bertrand SITAMTZE) Date: Mon, 13 Sep 2010 13:33:13 +0000 (GMT) Subject: [Pw_forum] Re : non existent or non writable directory In-Reply-To: Message-ID: <81122.95975.qm@web26501.mail.ukl.yahoo.com> Dear Elie, I guess your problem is because you do not act as root on your machine.? Make sure the directory exist. Then you should type the appropriate root command. I mean for example "sudo" for Ubuntu system; in front of the quantum espresso command. This should probably solve the problem. ********************************** Bertrand SITAMTZE YOUMBI Department of Physics University of Yaounde I-Cameroon ************************* --- En date de?: Lun 13.9.10, Elie Moujaes a ?crit?: De: Elie Moujaes Objet: [Pw_forum] non existent or non writable directory ?: pw_forum at pwscf.org Date: Lundi 13 septembre 2010, 14h38 Dear all, ? I have a small problem. I was trying to run a scf calculation for my system and I used the same outdir I usually use in my calculations but then after a few seconds the program stopped with an error: ? from outdir: error #???? 1 /exp/home/caiapo/........./Moujaes-Results/ non existent or non writable ? I made another directory and when I re-ran the calculation i got the same problem (although that new directory is empty).. How can I resolve this? ? ? -----La pi?ce jointe associ?e suit----- _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100913/25510ba8/attachment.htm From giannozz at democritos.it Mon Sep 13 15:46:41 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 13 Sep 2010 15:46:41 +0200 Subject: [Pw_forum] Re : non existent or non writable directory In-Reply-To: <81122.95975.qm@web26501.mail.ukl.yahoo.com> References: <81122.95975.qm@web26501.mail.ukl.yahoo.com> Message-ID: <4C8E2B41.4010109@democritos.it> Bertrand SITAMTZE wrote: > I mean for example "sudo" for Ubuntu system; in front of the > quantum espresso command. This should probably solve the problem. running a user code like QE with root privileges is not a great idea. Nothing nasty should happen, but if it does, your machine is at risk. > from outdir: error # 1 > /exp/home/caiapo/........./Moujaes-Results/ non existent or non writable if the code says that "/some/nonexistent/or/nonwritable/directory/" is nonexistent or not writable, it is nonexistent or non writable. The code reprints the full directory name so that one can easily check what is wrong, for instance issuing commands "ls /some/nonexistent/or/nonwritable/directory/" or "touch /some/nonexistent/or/nonwritable/directory/any-file-name". In most cases, it is just a misspell. P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From elie.moujaes at hotmail.co.uk Mon Sep 13 18:41:36 2010 From: elie.moujaes at hotmail.co.uk (Elie Moujaes) Date: Mon, 13 Sep 2010 17:41:36 +0100 Subject: [Pw_forum] non existent or non writable directory In-Reply-To: References: Message-ID: Dear Profs Gianozzi and Sitamtze, Thanks for the help. It was easy. I just missed a subdirectory but strangely enough I have not realised that before..Sorry. Thanks for the help Elie Moujaes University of Nottingham NG7 2RD uk From: elie.moujaes at hotmail.co.uk To: pw_forum at pwscf.org Date: Mon, 13 Sep 2010 13:38:30 +0100 Subject: [Pw_forum] non existent or non writable directory Dear all, I have a small problem. I was trying to run a scf calculation for my system and I used the same outdir I usually use in my calculations but then after a few seconds the program stopped with an error: from outdir: error # 1 /exp/home/caiapo/........./Moujaes-Results/ non existent or non writable I made another directory and when I re-ran the calculation i got the same problem (although that new directory is empty).. How can I resolve this? _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100913/d0c6b59b/attachment.htm From rezashidpoor at gmail.com Mon Sep 13 20:49:55 2010 From: rezashidpoor at gmail.com (reza shidpoor) Date: Mon, 13 Sep 2010 23:19:55 +0430 Subject: [Pw_forum] Re [7] : US-PP of Au Message-ID: Dear Gabriele Thank you for your comments in pwscf forum. I performed your suggestion that ecut is set to 32 Ry. The structure was relaxed and I repeated with ecut=20 , 22 Ry. It seems that the structure did not relaxed with this point " * ............. ........... the Fermi energy is -5.0636 ev ! total energy = -262.94392529 Ry Harris-Foulkes estimate = -262.94392533 Ry estimated scf accuracy < 0.00000006 Ry The total energy is the sum of the following terms: one-electron contribution = -323.90644330 Ry hartree contribution = 178.85727642 Ry xc contribution = -104.02662243 Ry ewald contribution = -13.85257067 Ry smearing contrib. (-TS) = -0.01556530 Ry convergence has been achieved in 2 iterations Forces acting on atoms (Ry/au): negative rho (up, down): 0.998E-04 0.000E+00 atom 1 type 1 force = 0.00000000 0.00223711 0.00000000 atom 2 type 1 force = -0.00165978 -0.00111856 0.00000000 atom 3 type 1 force = 0.00165978 -0.00111856 0.00000000 Total force = 0.003608 Total SCF correction = 0.000988 SCF correction compared to forces is too large, reduce conv_thr The maximum number of steps has been reached. End of BFGS Geometry Optimization Begin final coordinates ATOMIC_POSITIONS (angstrom) Au 0.000000000 0.248435531 0.000000000 Au -2.339313337 1.325352234 0.000000000 Au 2.339313337 1.325352234 0.000000000 End final coordinates* ........................ ........................ Questions : 1) The electron_maxstep is equaled 100. What is purpose of "number of states" ? 2) Should I consider system to be semiconducting If I want to simulate 4-atoms cluster? Best Regards, Reza.Shidpour SUT -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100913/57452288/attachment.htm From parwana.habibi at cea.fr Tue Sep 14 10:29:40 2010 From: parwana.habibi at cea.fr (HABIBI Parwana) Date: Tue, 14 Sep 2010 10:29:40 +0200 Subject: [Pw_forum] problem using ld1.x Message-ID: Dear all, I am trying to build a relativistic USPP in order to check the spin-orbit effect on chromium. Here is attached my input to create the USPP. However, when I run ld1.x, I get an error: Program LD1 v.4.1 starts ... Today is 10Sep2010 at 15:50:24 !!!!!!!!!!!!!!!!!!! WARNING !!!!!!!!!!!!!!!!!!!!!!! Message from routine ld1_readin: US requires at least two energies per channel 4S !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from ld1_setup : error # 1 all electron wfc corresponding to pseudo-state 4P not found %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... Could you enlight me on that error? Thank you, Parwana HABIBI Ph.D student, CEA-Saclay, DSM/IRAMIS/SPCSI -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100914/89b89b12/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: Cr_us_pbe_6rel.in.parwana Type: application/octet-stream Size: 896 bytes Desc: Cr_us_pbe_6rel.in.parwana Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100914/89b89b12/attachment.obj From Lorenzo.Paulatto at impmc.upmc.fr Tue Sep 14 11:24:21 2010 From: Lorenzo.Paulatto at impmc.upmc.fr (Lorenzo Paulatto) Date: Tue, 14 Sep 2010 11:24:21 +0200 Subject: [Pw_forum] problem using ld1.x In-Reply-To: References: Message-ID: Dear Parwana, if you set rel_dist to 'average', for a full-relativistic configuration, then the states with j=l-1/2 and those with j=l+1/2 will be filled with the same number of electrons **per orbital**. More precisely (I'm reading from the code) if you 0<=N<=6 electrons in n orbital with l=L they will be split betwee the -1/2 and the +1/2 in the following way: occ(L-1/2) = N * (2*L) / (2*(2*L+1)) occ(L+1/2) = N * (2*(2*L+1)) / (2*L) which in your specific case gives 1.6 and 2.4 respectively. The meaning is not really clear to me, but I also don't see the point of using rel_dist='average', so maybe it is clear to you. regards -- Lorenzo Paulatto post-doc @ IMPMC/UPMC - Universit? Paris 6 phone: +33 (0)1 44 27 74 89 www: http://www-int.impmc.upmc.fr/~paulatto/ previously (take note of the change!): phd student @ SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 www: http://people.sissa.it/~paulatto/ From dimpy.sharma at tyndall.ie Tue Sep 14 13:13:07 2010 From: dimpy.sharma at tyndall.ie (Dimpy Sharma) Date: Tue, 14 Sep 2010 12:13:07 +0100 Subject: [Pw_forum] pwcond.x;nz1,ewind and epsproj Message-ID: Hi, In the Input File Description for pwcond.x,what does the term nz1,ewind and epsproj means and how do we choose their values in the example,I have seen that for a 7 unit cells, the nz1 value taken is 11 and for aluminum chain nz=1.My system contains 282 atoms with 7 unit cell,how shall I choose nz,ewind and epsproj values? Thanks and regards DImpy Dimpy Sharma ETG-Group Tyndall National Institute Cork -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100914/02809650/attachment.htm From neto.baldini at gmail.com Tue Sep 14 13:22:19 2010 From: neto.baldini at gmail.com (ettore) Date: Tue, 14 Sep 2010 08:22:19 -0300 Subject: [Pw_forum] Problems with pw Message-ID: <4C8F5AEB.6060006@gmail.com> Hi there, I'm new with Quantum Espresso and I'm trying to run the examples in the examples_dir. But I always got an error, namely: Error condition encountered during test: exit status = 2 Aborting I taught it was about the pseudopotentials, but everything seems to be all right. Could someone help, me? Please best regards Ettore From somesh.kb at gmail.com Tue Sep 14 15:06:47 2010 From: somesh.kb at gmail.com (Somesh Kumar Bhattacharya) Date: Tue, 14 Sep 2010 15:06:47 +0200 Subject: [Pw_forum] prob with cppp.x in QE-4.2.1 Message-ID: Hi All, Am facing a problem with the cppp.x. in version QE-4.2.1 installed in cineca sp6. I have a negatively charged system n(H2O) + excess electron. After electronic minimization, when I tried to calculate the spin charge density with cppp.x, am getting the following error : task # 13 from cppp : error # 1 IONS not found in data-file.xml My input file for cppp.x is : &INPUTPP prefix = 'h2o-EE', fileout = 'chg-spin-elec', output = 'xsf', outdir = './', lcharge = .TRUE. ndr = 52, charge_density = 'spin', atomic_number(1) = 8, atomic_number(2) = 1, / Can anybody tell me what is the source of the problem? Regards Somesh ----------------------------------------------------------- Somesh Kr. Bhattacharya Post Doctoral Fellow Room No. 263, Leonardo Building, The Abdus Salam International Centre for Theoretical Physics Strada Costiera, 11 I-34014 Trieste Italy Phone: +39-040-2240399 http://portal.ictp.it/cmsp/members/postdoctoral-fellows/somesh-kumar-bhattacharya/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100914/a4a5ea83/attachment.htm From sclauzer at sissa.it Tue Sep 14 17:48:36 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Tue, 14 Sep 2010 17:48:36 +0200 Subject: [Pw_forum] Re [7] : US-PP of Au In-Reply-To: References: Message-ID: Dear Reza, Le 13-set-2010 ? 20:49, reza shidpoor a ?crit : > > Dear Gabriele > > Thank you for your comments in pwscf forum. > > I performed your suggestion that ecut is set to 32 Ry. The structure was relaxed So you can confirm that with this cutoff it works, do you? > and I repeated with ecut=20 , 22 Ry. Why do you want to do so? If the cutoff id too low the forces are not described correctly and the ralaxation algorithm might fail to converge. > It seems that the structure did not relaxed with this point " > > ............. > ........... > the Fermi energy is -5.0636 ev > > ! total energy = -262.94392529 Ry > Harris-Foulkes estimate = -262.94392533 Ry > estimated scf accuracy < 0.00000006 Ry > > The total energy is the sum of the following terms: > > one-electron contribution = -323.90644330 Ry > hartree contribution = 178.85727642 Ry > xc contribution = -104.02662243 Ry > ewald contribution = -13.85257067 Ry > smearing contrib. (-TS) = -0.01556530 Ry > > convergence has been achieved in 2 iterations > > Forces acting on atoms (Ry/au): > > > negative rho (up, down): 0.998E-04 0.000E+00 > atom 1 type 1 force = 0.00000000 0.00223711 0.00000000 > atom 2 type 1 force = -0.00165978 -0.00111856 0.00000000 > atom 3 type 1 force = 0.00165978 -0.00111856 0.00000000 > > Total force = 0.003608 Total SCF correction = 0.000988 > SCF correction compared to forces is too large, reduce conv_thr Besides increasing the cutoff, you might need to decrease this threshold on order for the forces to be perfectly consistent with the total energy. > > The maximum number of steps has been reached. > > End of BFGS Geometry Optimization > Begin final coordinates > > ATOMIC_POSITIONS (angstrom) > Au 0.000000000 0.248435531 0.000000000 > Au -2.339313337 1.325352234 0.000000000 > Au 2.339313337 1.325352234 0.000000000 > End final coordinates > > ........................ > ........................ > > Questions : > > 1) The electron_maxstep is equaled 100. What is purpose of "number of states" ? ??? You are confusing number of states with number of electrpnic steps and perhaps also with ionic steps. Please spend some time reading the manual and some examples or tutorials. > > 2) Should I consider system to be semiconducting If I want to simulate 4-atoms cluster? If you are not sure, you can treat your system as metallic with a small smearing and then, if it turns out to have a gap, reduce/remove the smearing. GS > > > Best Regards, > Reza.Shidpour > SUT > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100914/016cd9ca/attachment-0001.htm From mohnish.iitk at gmail.com Tue Sep 14 18:01:48 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Tue, 14 Sep 2010 21:31:48 +0530 Subject: [Pw_forum] "segmentation fault" in new version of ubuntu Message-ID: Dear QE users! Previously I was using QE on UBUNTU-9.04, it was working fine with that. But now I am using UBUNTU -10.04 . I started doing vc-relax calculation but it is 'segmentation fault'. Can anybody please help me. Thanks alot in advance -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100914/3cf79e7c/attachment.htm From sclauzer at sissa.it Tue Sep 14 18:12:37 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Tue, 14 Sep 2010 18:12:37 +0200 Subject: [Pw_forum] pwcond.x;nz1,ewind and epsproj In-Reply-To: References: Message-ID: Please have a look at this post in the forum archive: http://www.democritos.it/pipermail/pw_forum/2006-August/004826.html Since you are an experienced user of this forum you should know that you may find answer to your questions by browsing the forum archives. Indeed you can find the link above in a few seconds by typing ewind epsproj site:democritos.it on google. HTH GS Envoy? de mon iPhone Le 14-set-2010 ? 13:13, "Dimpy Sharma" a ?crit : > Hi, > In the Input File Description for pwcond.x,what does the term nz1,ewind and epsproj means and how do we choose their values in the example,I have seen that for a 7 unit cells, the nz1 value taken is 11 and for aluminum chain nz=1.My system contains 282 atoms with 7 unit cell,how shall I choose nz,ewind and epsproj values? > > Thanks and regards > > DImpy > > Dimpy Sharma > ETG-Group > Tyndall National Institute > Cork > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100914/1dca17e3/attachment.htm From giannozz at democritos.it Tue Sep 14 20:36:58 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 14 Sep 2010 20:36:58 +0200 Subject: [Pw_forum] "segmentation fault" in new version of ubuntu In-Reply-To: References: Message-ID: <4E7A157E-D836-4C24-8B61-D6B4DD4D52D4@democritos.it> On Sep 14, 2010, at 18:01 , mohnish pandey wrote: > 'segmentation fault'. Can anybody please help me. no later than a few days ago: http://www.democritos.it/pipermail/pw_forum/2010-September/018118.html P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at democritos.it Tue Sep 14 20:42:42 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 14 Sep 2010 20:42:42 +0200 Subject: [Pw_forum] Problems with pw In-Reply-To: <4C8F5AEB.6060006@gmail.com> References: <4C8F5AEB.6060006@gmail.com> Message-ID: On Sep 14, 2010, at 13:22 , ettore wrote: > I'm new with Quantum Espresso you are new with computers, judging from your question > Error condition encountered during test: exit status = 2 > Aborting you need to look into output file and see what the error message is. P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From asafis at yahoo.com.br Wed Sep 15 04:28:23 2010 From: asafis at yahoo.com.br (=?iso-8859-1?Q?=C1lvaro_Alves?=) Date: Tue, 14 Sep 2010 19:28:23 -0700 (PDT) Subject: [Pw_forum] input dist.x Message-ID: <313109.3042.qm@web52303.mail.re2.yahoo.com> Dear QE users!Can someone give me an input of the tool dist.x ? Sincerely, A. S. Santos Univesidade Federal Fluminense Niter?i-Rio de Janeiro-Brasil A. S. Santos -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100914/f6bd05c4/attachment.htm From didi5158 at gmail.com Wed Sep 15 05:05:07 2010 From: didi5158 at gmail.com (Wang Di) Date: Wed, 15 Sep 2010 11:05:07 +0800 Subject: [Pw_forum] vc-relax problem when calculate lattice parameters Message-ID: Hello all: i wish to calculate the Raman spectra of the crystal with espresso 4.1.1 package. However, the calculated frequencies included negative value. According to the pwscf manual, the unstable structure or the bad scf calculation cause the negative value, I know the structure of the material is stable below the melt point, so I perform the relax of the atom position, and change difference PP (GGA, LDA) to test. Unlucky, the changes still exist the negative frequencies and the calculated frequencies different from the experimental value totally. recently, according to some references about the Raman spectra calculation with espresso, the optimization of lattice parameters is needed, I performing vc-relax calculation, but the job stop with the error that %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from scale_h : error # 1 Not enough space allocated for radial FFT: try restarting with a larger cell_factor. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% according previous questions in mail list, the error means the G value change too much. From the out file, the first step scf calculation, the total force was so small ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ Total force = 0.000027 Total SCF correction = 0.000004 SCF correction compared to forces is too large, reduce conv_thr entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P=-3539.67 -0.02405899 0.00000000 0.00000000 -3539.20 0.00 0.00 0.00000000 -0.02405899 0.00000000 0.00 -3539.20 0.00 0.00000000 0.00000000 -0.02406857 0.00 0.00 -3540.61 BFGS Geometry Optimization number of scf cycles = 1 number of bfgs steps = 0 enthalpy new = -573.6572611922 Ry new trust radius = 0.2000000000 bohr new conv_thr = 1.0E-10 Ry new unit-cell volume = 1258.11889 a.u.^3 ( 186.43398 Ang^3 ) CELL_PARAMETERS (alat) 0.963490012 0.000000000 0.000000000 -0.481745006 0.834406827 0.000000000 0.000000000 0.000000000 1.854979314 ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- the bfge calculation go on in 3 step and stop with the above error. the input file: ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- &CONTROL calculation = 'vc-relax' , restart_mode = 'from_scratch' , wf_collect = .true. , outdir = '/lustre/AIOFM/gxgu/workLCAF/temp/' , wfcdir = '/lustre/AIOFM/gxgu/workLCAF/wftemp/' , pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , prefix = 'LCAF' , etot_conv_thr = 1.0D-7 , forc_conv_thr = 1.0D-6 , nstep = 100 , tstress = .true. , tprnfor = .true. , dipfield = .true. , / &SYSTEM ibrav = 4, celldm(1) = 9.4490, celldm(3) = 1.9253, nat = 18, ntyp = 4, ecutwfc = 30 , ecutrho = 120 , nosym = .true. , / &ELECTRONS electron_maxstep = 100, conv_thr = 1.0D-10 , / &IONs ion_dynamics = 'bfgs' , / &CELL cell_dynamics = 'bfgs' , / ATOMIC_SPECIES Li 6.94100 03-Li.LDA.fhi.UPF Al 26.98154 13-Al.LDA.fhi.UPF Ca 40.07800 20-Ca.LDA.fhi.UPF F 18.99840 09-F.LDA.fhi.UPF ATOMIC_POSITIONS crystal Li 0.333333333 0.666666667 0.250000000 Li 0.666666667 0.333333333 0.750000000 F 0.375595671 0.008455746 0.142528376 F 0.991544254 0.367139925 0.142528376 F 0.632860075 0.624404329 0.142528376 F 0.991544254 0.624404329 0.357471624 F 0.632860075 0.008455746 0.357471624 F 0.375595671 0.367139925 0.357471624 F 0.624404329 0.991544254 0.857471624 F 0.008455746 0.632860075 0.857471624 F 0.367139925 0.375595671 0.857471624 F 0.008455746 0.375595671 0.642528376 F 0.367139925 0.991544254 0.642528376 F 0.624404329 0.632860075 0.642528376 Al 0.666666667 0.333333333 0.250000000 Al 0.333333333 0.666666667 0.750000000 Ca 0.000000000 0.000000000 0.000000000 Ca 0.000000000 0.000000000 0.500000000 K_POINTS automatic 16 16 16 0 0 0 ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Do you have some advises for the solution the case or elimination the Raman negative frequencies, any help is greatly appreciated. Thanks in advance Wang Di -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100915/6e1d0aa5/attachment-0001.htm From mohnish.iitk at gmail.com Wed Sep 15 05:37:22 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Wed, 15 Sep 2010 09:07:22 +0530 Subject: [Pw_forum] "segmentation fault" in new version of ubuntu In-Reply-To: <4E7A157E-D836-4C24-8B61-D6B4DD4D52D4@democritos.it> References: <4E7A157E-D836-4C24-8B61-D6B4DD4D52D4@democritos.it> Message-ID: Thanks a lot Dr. Paolo it is working now. On Wed, Sep 15, 2010 at 12:06 AM, Paolo Giannozzi wrote: > > On Sep 14, 2010, at 18:01 , mohnish pandey wrote: > > > 'segmentation fault'. Can anybody please help me. > > no later than a few days ago: > http://www.democritos.it/pipermail/pw_forum/2010-September/018118.html > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100915/1910c971/attachment.htm From mohnish.iitk at gmail.com Wed Sep 15 05:44:31 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Wed, 15 Sep 2010 09:14:31 +0530 Subject: [Pw_forum] vc-relax problem when calculate lattice parameters In-Reply-To: References: Message-ID: Dear Wang, This is a common error when you are very far from equilibrium. Start with the experimental value of the lattice parameter. Or you can take the lattice parameter from the point where the calculation crashes, this may also work(but might fail when you are very far from equilibrium). And next time please mention your affiliation. On Wed, Sep 15, 2010 at 8:35 AM, Wang Di wrote: > Hello all: > i wish to calculate the Raman spectra of the crystal with espresso 4.1.1 > package. However, the calculated frequencies included negative value. > According to the pwscf manual, the unstable structure or the bad scf > calculation cause the negative value, I know the structure of the material > is stable below the melt point, so I perform the relax of the atom position, > and change difference PP (GGA, LDA) to test. Unlucky, the changes still > exist the negative frequencies and the calculated frequencies different from > the experimental value totally. recently, according to some references > about the Raman spectra calculation with espresso, the optimization of > lattice parameters is needed, I performing vc-relax calculation, but the job > stop with the error that > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from scale_h : error # 1 > Not enough space allocated for radial FFT: try restarting with a > larger cell_factor. > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > according previous questions in mail list, the error means the G value > change too much. From the out file, the first step scf calculation, the > total force was so small > > > ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ > Total force = 0.000027 Total SCF correction = 0.000004 > SCF correction compared to forces is too large, reduce conv_thr > > > entering subroutine stress ... > > total stress (Ry/bohr**3) (kbar) > P=-3539.67 > -0.02405899 0.00000000 0.00000000 -3539.20 0.00 0.00 > 0.00000000 -0.02405899 0.00000000 0.00 -3539.20 0.00 > 0.00000000 0.00000000 -0.02406857 0.00 0.00 -3540.61 > > > BFGS Geometry Optimization > > number of scf cycles = 1 > number of bfgs steps = 0 > > enthalpy new = -573.6572611922 Ry > > new trust radius = 0.2000000000 bohr > new conv_thr = 1.0E-10 Ry > > new unit-cell volume = 1258.11889 a.u.^3 ( 186.43398 Ang^3 ) > > CELL_PARAMETERS (alat) > 0.963490012 0.000000000 0.000000000 > -0.481745006 0.834406827 0.000000000 > 0.000000000 0.000000000 1.854979314 > > ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- > > the bfge calculation go on in 3 step and stop with the above error. the > input file: > > > ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- > &CONTROL > calculation = 'vc-relax' , > restart_mode = 'from_scratch' , > wf_collect = .true. , > outdir = '/lustre/AIOFM/gxgu/workLCAF/temp/' , > wfcdir = '/lustre/AIOFM/gxgu/workLCAF/wftemp/' , > pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , > prefix = 'LCAF' , > etot_conv_thr = 1.0D-7 , > forc_conv_thr = 1.0D-6 , > nstep = 100 , > tstress = .true. , > tprnfor = .true. , > dipfield = .true. , > / > &SYSTEM > ibrav = 4, > celldm(1) = 9.4490, > celldm(3) = 1.9253, > nat = 18, > ntyp = 4, > ecutwfc = 30 , > ecutrho = 120 , > nosym = .true. , > / > &ELECTRONS > electron_maxstep = 100, > conv_thr = 1.0D-10 , > / > &IONs > ion_dynamics = 'bfgs' , > / > &CELL > cell_dynamics = 'bfgs' , > / > ATOMIC_SPECIES > Li 6.94100 03-Li.LDA.fhi.UPF > Al 26.98154 13-Al.LDA.fhi.UPF > Ca 40.07800 20-Ca.LDA.fhi.UPF > F 18.99840 09-F.LDA.fhi.UPF > ATOMIC_POSITIONS crystal > Li 0.333333333 0.666666667 0.250000000 > Li 0.666666667 0.333333333 0.750000000 > F 0.375595671 0.008455746 0.142528376 > F 0.991544254 0.367139925 0.142528376 > F 0.632860075 0.624404329 0.142528376 > F 0.991544254 0.624404329 0.357471624 > F 0.632860075 0.008455746 0.357471624 > F 0.375595671 0.367139925 0.357471624 > F 0.624404329 0.991544254 0.857471624 > F 0.008455746 0.632860075 0.857471624 > F 0.367139925 0.375595671 0.857471624 > F 0.008455746 0.375595671 0.642528376 > F 0.367139925 0.991544254 0.642528376 > F 0.624404329 0.632860075 0.642528376 > Al 0.666666667 0.333333333 0.250000000 > Al 0.333333333 0.666666667 0.750000000 > Ca 0.000000000 0.000000000 0.000000000 > Ca 0.000000000 0.000000000 0.500000000 > K_POINTS automatic > 16 16 16 0 0 0 > > ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- > > Do you have some advises for the solution the case or elimination the Raman > negative frequencies, any help is greatly appreciated. > > Thanks in advance > Wang Di > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100915/02958e6c/attachment.htm From giannozz at democritos.it Wed Sep 15 07:59:24 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 15 Sep 2010 07:59:24 +0200 Subject: [Pw_forum] input dist.x In-Reply-To: <313109.3042.qm@web52303.mail.re2.yahoo.com> References: <313109.3042.qm@web52303.mail.re2.yahoo.com> Message-ID: <17593293-E204-4019-AFC1-3606B32EECE9@democritos.it> On Sep 15, 2010, at 4:28 , ?lvaro Alves wrote: > Dear QE users! Can someone give me an input of the tool dist.x ? it is described in the header of pwtools/dist.f P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From asafis at yahoo.com.br Wed Sep 15 08:19:56 2010 From: asafis at yahoo.com.br (=?iso-8859-1?Q?=C1lvaro_Alves?=) Date: Tue, 14 Sep 2010 23:19:56 -0700 (PDT) Subject: [Pw_forum] Res: input dist.x In-Reply-To: <17593293-E204-4019-AFC1-3606B32EECE9@democritos.it> References: <313109.3042.qm@web52303.mail.re2.yahoo.com> <17593293-E204-4019-AFC1-3606B32EECE9@democritos.it> Message-ID: <262060.22151.qm@web52301.mail.re2.yahoo.com> Thanks, Giannozzi. A. S. Santos Univesidade Federal Fluminense Niter?i-Rio de Janeiro-Brasil ________________________________ De: Paolo Giannozzi Para: PWSCF Forum Enviadas: Quarta-feira, 15 de Setembro de 2010 2:59:24 Assunto: Re: [Pw_forum] input dist.x On Sep 15, 2010, at 4:28 , ?lvaro Alves wrote: > Dear QE users! Can someone give me an input of the tool dist.x ? it is described in the header of pwtools/dist.f P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100914/46b2f29c/attachment.htm From rezashidpoor at gmail.com Wed Sep 15 11:43:09 2010 From: rezashidpoor at gmail.com (reza shidpoor) Date: Wed, 15 Sep 2010 14:13:09 +0430 Subject: [Pw_forum] Re [8] : US-PP of Au Message-ID: Dear Lorenzo Thank you to reply. It is always segmentation fault. My system is 3-atoms Au cluster. this cluster forms a chain of atoms but the calculations never finished because of segmentation fault. Dear Gabriele the work was completely don with kinetic-energy cutoff=32.0000 Ry and charge density cutoff = 320.0000 Ry . *Forces acting on atoms (Ry/au): atom 1 type 1 force = -0.00009579 -0.00009579 0.00000000 atom 2 type 1 force = 0.00013265 -0.00003686 0.00000000 atom 3 type 1 force = -0.00003686 0.00013265 0.00000000 Total force = 0.000237 Total SCF correction = 0.000290 SCF correction compared to forces is too large, reduce conv_thr bfgs converged in 24 scf cycles and 19 bfgs steps (criteria: energy < 0.10E-03, force < 0.10E-02) End of BFGS Geometry Optimization Final energy = -262.9046471560 Ry Begin final coordinates ATOMIC_POSITIONS (angstrom) Au 1.381049485 1.381049485 0.000000000 Au 3.868137287 -1.139186772 0.000000000 Au -1.139186772 3.868137287 0.000000000 End final coordinates* Best wishes. Reza.Shidpour Institute for Nanoscienec and Nanotechnology, SUT -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100915/97843a52/attachment.htm From giannozz at democritos.it Wed Sep 15 11:44:30 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 15 Sep 2010 11:44:30 +0200 Subject: [Pw_forum] Re [8] : US-PP of Au In-Reply-To: References: Message-ID: <4C90957E.7060808@democritos.it> reza shidpoor wrote: > the calculations never finished because of segmentation fault. upgrade your compiler to something that works P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From rixs at uni-mainz.de Wed Sep 15 13:21:31 2010 From: rixs at uni-mainz.de (Stephan Rix) Date: Wed, 15 Sep 2010 13:21:31 +0200 Subject: [Pw_forum] problems with converge of geometric optimization Message-ID: I am trying to calculate an H-center in CaF2, and I am having trouble with the geometry optimization. I have tried adjusting various parameters such as ion dynamics, trust radius, convergence tresholds, but it seems that convergence cannot be improved. Using bfgs I always get the following error: %%%%%%%%%%%%%%%%%%%%%%%%%% from bfgs : error # 1 bfgs history already reset at previous step %%%%%%%%%%%%%%%%%%%%%%%%%% Alternatively the electonic convergence gets extremely slow. In addition, no matter what parameters I adjust the total forces in the system never converge, e.g. Total force = 0.026394 Total SCF correction = 0.000314 Total force = 0.018379 Total SCF correction = 0.000282 Total force = 0.018824 Total SCF correction = 0.000120 Total force = 0.018857 Total SCF correction = 0.000155 Total force = 0.020215 Total SCF correction = 0.000193 Total force = 0.058765 Total SCF correction = 0.000241 Total force = 0.016998 Total SCF correction = 0.000205 Total force = 0.042967 Total SCF correction = 0.000238 Total force = 0.106226 Total SCF correction = 0.000073 Total force = 0.137142 Total SCF correction = 0.000071 Total force = 0.109549 Total SCF correction = 0.000108 Total force = 0.068723 Total SCF correction = 0.000103 Total force = 0.076599 Total SCF correction = 0.000168 Total force = 0.033233 Total SCF correction = 0.000342 Total force = 0.026883 Total SCF correction = 0.000248 Total force = 0.025846 Total SCF correction = 0.000378 Total force = 0.025596 Total SCF correction = 0.000225 Total force = 0.025482 Total SCF correction = 0.000275 Total force = 0.025508 Total SCF correction = 0.000128 I am thankful for any suggestions. Stephan Rix Institute for Inorganic und Analytical Chemistry Johannes Gutenberg-University Mainz, Germany -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100915/b78f2dd0/attachment.htm From glapenna at iccom.cnr.it Wed Sep 15 15:19:53 2010 From: glapenna at iccom.cnr.it (Giovanni La Penna) Date: Wed, 15 Sep 2010 15:19:53 +0200 (CEST) Subject: [Pw_forum] Occupations in CP Message-ID: Dear all, I tried to perform CP-MD simulations moving electrons beyond the highest KS state. In theory, this can be done using occupations='from_input' and playing with the OCCUPATIONS card, using a number of bands larger than half of the number of electrons. Here is an example for isolated H2 molecule: # electron steepest-descent &control calculation='cp', restart_mode='from_scratch', prefix='elsd2', pseudo_dir='/home/lapenna/espresso_pseudo/', outdir='/home/lapenna/deloc/H2/./tmp/', ekin_conv_thr =1.e-06, nstep=50, iprint=10, isave=100, dt=2.0, ndr=50, ndw=60, / &system ibrav= 1, celldm(1)= 11, nat=2, ntyp=1, tot_charge=0., ! nelec=2, nspin=1, nbnd=2, ecutwfc = 25.0, ecutrho = 250.0, nr1b = 20, nr2b = 20, nr3b = 20, ! occupations='fixed', occupations='from_input', / &electrons emass = 400., electron_dynamics = 'sd', ! electron_dynamics = 'damp', orthogonalization = 'ortho', ortho_max = 1000, / &ions ion_dynamics = 'none', ion_positions = 'from_input', ! ion_dynamics = 'damp', / ATOMIC_SPECIES H 1.008 H.pbe-rrkjus.UPF ATOMIC_POSITIONS {bohr} H 0.472420E+01 0.472438E+01 0.476847E+01 H 0.472443E+01 0.472425E+01 0.619194E+01 OCCUPATIONS 0.0 2.0 In the output (QE-4.2) I read: ... Wave Initialization: random initial wave-functions Occupation number from init nbnd = 2 0.00 0.00 ... Same occupation numbers every time they are printed, inconsistently with the charge of the system that is zero. Continuing with the dynamics, the H2 dissociation occurs as expected, while with occupations 2. 0. the correct minimum is satisifed. Maybe the comparison between different occupations in CP-MD is nonsense, but this is another story. Why the occupations are all zero beyond the last non-zero from input? Is it a safety measure preventing such hazardous experiments? Also in this latter case, the measure does not work, except in the case of a careful output reading... PW prints occupations as in input. Thank you in advance, Giovanni ============================================================ Giovanni La Penna - National research council (Cnr) Institute for chemistry of organo-metallic compounds (Iccom) via Madonna del Piano 10, I-50019 Sesto Fiorentino, Firenze, Italy tel.: +39 055 522-5264, fax: +39 055 522-5203 e-mail: glapenna at iccom.cnr.it - http://www.iccom.cnr.it/lapenna skype: giovannilapenna ============================================================ From jiachen at princeton.edu Wed Sep 15 16:33:09 2010 From: jiachen at princeton.edu (jia chen) Date: Wed, 15 Sep 2010 10:33:09 -0400 Subject: [Pw_forum] problems with converge of geometric optimization Message-ID: Hi Stephan, If the only problem is bfgs, you can change ion_dynamics to damp. Regards -- Jia Chen Dept of Chemistry Princeton University Princeton, NJ 08544 From degironc at sissa.it Wed Sep 15 17:10:07 2010 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 15 Sep 2010 17:10:07 +0200 Subject: [Pw_forum] problems with converge of geometric optimization In-Reply-To: References: Message-ID: <4C90E1CF.6040400@sissa.it> This happens when the energy landscape become noisy compared with the expected energy decrease in the relaxation steps. This can happen when you are very close to the minimum. How large are your forces ? How tight is your force/energy threshold ? stefano Stephan Rix wrote: > I am trying to calculate an H-center in CaF2, and I am having trouble with > the geometry optimization. > I have tried adjusting various parameters such as ion dynamics, trust > radius, convergence tresholds, but it seems that convergence cannot be > improved. > > Using bfgs I always get the following error: > > %%%%%%%%%%%%%%%%%%%%%%%%%% > from bfgs : error # 1 > bfgs history already reset at previous step > %%%%%%%%%%%%%%%%%%%%%%%%%% > > Alternatively the electonic convergence gets extremely slow. > > In addition, no matter what parameters I adjust the total forces in the > system never converge, e.g. > Total force = 0.026394 Total SCF correction = 0.000314 > Total force = 0.018379 Total SCF correction = 0.000282 > Total force = 0.018824 Total SCF correction = 0.000120 > Total force = 0.018857 Total SCF correction = 0.000155 > Total force = 0.020215 Total SCF correction = 0.000193 > Total force = 0.058765 Total SCF correction = 0.000241 > Total force = 0.016998 Total SCF correction = 0.000205 > Total force = 0.042967 Total SCF correction = 0.000238 > Total force = 0.106226 Total SCF correction = 0.000073 > Total force = 0.137142 Total SCF correction = 0.000071 > Total force = 0.109549 Total SCF correction = 0.000108 > Total force = 0.068723 Total SCF correction = 0.000103 > Total force = 0.076599 Total SCF correction = 0.000168 > Total force = 0.033233 Total SCF correction = 0.000342 > Total force = 0.026883 Total SCF correction = 0.000248 > Total force = 0.025846 Total SCF correction = 0.000378 > Total force = 0.025596 Total SCF correction = 0.000225 > Total force = 0.025482 Total SCF correction = 0.000275 > Total force = 0.025508 Total SCF correction = 0.000128 > > I am thankful for any suggestions. > > Stephan Rix > Institute for Inorganic und Analytical Chemistry > Johannes Gutenberg-University Mainz, Germany > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From cvasse at hotmail.com Wed Sep 15 17:31:56 2010 From: cvasse at hotmail.com (Vasse chis) Date: Wed, 15 Sep 2010 17:31:56 +0200 Subject: [Pw_forum] Bi - pseudopotential (fully relativistic) Message-ID: Dear users, Searching the forum, a time ago I saw a discussion on how to generate a Bi PP (fully relativistic) (http://www.democritos.it/pipermail/pw_forum/2006-May/004181.html). I was wondering how that worked out, the PP is not listed on QE pseudopotentials database. I'm looking for that kind of a PP. Please publish it in the database or share it in some way. Thank you in advance. Vasse Chis Department of Physics University of Central Florida -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100915/9ce9ba58/attachment.htm From giannozz at democritos.it Wed Sep 15 17:52:42 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 15 Sep 2010 17:52:42 +0200 Subject: [Pw_forum] Single Cell vs Super-Cell In-Reply-To: <00f701cb50e5$76b35c30$641a1490$@edu> References: <00f701cb50e5$76b35c30$641a1490$@edu> Message-ID: <4C90EBCA.1010405@democritos.it> Peter O Orondo wrote: > I am trying to calculate the chemical potential of H in a bunch of > metals. When I use a single cell, I get results that compare to > experiment. However, when I use a 2x2x2 super-cell, I expect to get a > more accurate result since the super-cell should model the physical > properties more accurately. However, the super-cell results don?t even > have the right shape. The results look rather random. > > Any insights into what could be wrong? hard to say. The fact that you get results that compare well to experiments doesn't imply that those results are correct. In simple metals, if you put one H atom per unit cell you simulate a hydride rather than H dissolved into a metal. If you increase the size of the supercell, you should notice that your results (e.g. the formation energy) converges to some value, which should be good for a isolated impurity. Convergence may be quite slow, though, depending upon what you calculate and in which system Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From didi5158 at gmail.com Thu Sep 16 04:42:19 2010 From: didi5158 at gmail.com (Wang Di) Date: Thu, 16 Sep 2010 10:42:19 +0800 Subject: [Pw_forum] vc-relax problem when calculate lattice parameters In-Reply-To: References: Message-ID: Dear MOHNISH First, thank you very much for your advices. The initial crystal structure were referenced to the experimental value. Using the vc-relax calculation is to find the structure with minimum energy for Raman frequencies calculation. Beside the change of the initial lattice parameter and atom position, is there another means to solve the case such as changing some threshold value? Best regards Wang ----------------------------------------------------------------- *Wang Di** The Crystal Lab,* *AnHui Institute of Optics and Fine Mechanics, *** *Chinese Academy of Sciences,** No350. , Shushanhu Road, Hefei, 230031, China**, E-mail: didi5158 @gmail.ocm** ,* ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100916/b0b45204/attachment-0001.htm From mohnish.iitk at gmail.com Thu Sep 16 04:47:39 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Thu, 16 Sep 2010 08:17:39 +0530 Subject: [Pw_forum] vc-relax problem when calculate lattice parameters In-Reply-To: References: Message-ID: Dear Wang, To the best of my knowledge there is not any other way..But you can try increasing the smearing parameter.. This will help in convergence but I dont know whether it will solve your problem or not... On Thu, Sep 16, 2010 at 8:12 AM, Wang Di wrote: > Dear MOHNISH > > First, thank you very much for your advices. The initial crystal structure > were referenced to the experimental value. Using the vc-relax calculation is > to find the structure with minimum energy for Raman frequencies calculation. > Beside the change of the initial lattice parameter and atom position, is > there another means to solve the case such as changing some threshold value? > > Best regards > Wang > > > ----------------------------------------------------------------- > *Wang Di** > The Crystal Lab,* > *AnHui Institute of Optics and Fine Mechanics, *** > *Chinese Academy of Sciences,** > No350. , Shushanhu Road, Hefei, 230031, China**, > E-mail: didi5158 @gmail.ocm** ,* > ----------------------------------------------------------------- > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100916/56937243/attachment.htm From hqzhou at nju.edu.cn Thu Sep 16 10:09:47 2010 From: hqzhou at nju.edu.cn (Huiqun Zhou) Date: Thu, 16 Sep 2010 16:09:47 +0800 Subject: [Pw_forum] Re [8] : US-PP of Au References: Message-ID: <02445BF0AECD41BBAE3B0DA86E1E6E91@solarflare> reza, Another possibility is that you have flawed memory chip(s). I trid your original (actually, changed to scf calculation) 13 Au cluster on my machine without anyproblem. huiqun zhou @earth sciences, nanjing university, china ----- Original Message ----- From: reza shidpoor To: pw_forum at pwscf.org Sent: Wednesday, September 15, 2010 5:43 PM Subject: [Pw_forum] Re [8] : US-PP of Au Dear Lorenzo Thank you to reply. It is always segmentation fault. My system is 3-atoms Au cluster. this cluster forms a chain of atoms but the calculations never finished because of segmentation fault. Dear Gabriele the work was completely don with kinetic-energy cutoff=32.0000 Ry and charge density cutoff = 320.0000 Ry . Forces acting on atoms (Ry/au): atom 1 type 1 force = -0.00009579 -0.00009579 0.00000000 atom 2 type 1 force = 0.00013265 -0.00003686 0.00000000 atom 3 type 1 force = -0.00003686 0.00013265 0.00000000 Total force = 0.000237 Total SCF correction = 0.000290 SCF correction compared to forces is too large, reduce conv_thr bfgs converged in 24 scf cycles and 19 bfgs steps (criteria: energy < 0.10E-03, force < 0.10E-02) End of BFGS Geometry Optimization Final energy = -262.9046471560 Ry Begin final coordinates ATOMIC_POSITIONS (angstrom) Au 1.381049485 1.381049485 0.000000000 Au 3.868137287 -1.139186772 0.000000000 Au -1.139186772 3.868137287 0.000000000 End final coordinates Best wishes. Reza.Shidpour Institute for Nanoscienec and Nanotechnology, SUT ------------------------------------------------------------------------------ _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100916/48bc9b07/attachment-0001.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: au_13.scf-ec72.out Type: application/octet-stream Size: 26781 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100916/48bc9b07/attachment-0001.obj From rixs at uni-mainz.de Thu Sep 16 10:31:20 2010 From: rixs at uni-mainz.de (Stephan Rix) Date: Thu, 16 Sep 2010 10:31:20 +0200 Subject: [Pw_forum] problems with converge of geometric optimization Message-ID: @Stefano I have listed the forces from the last bfgs iteration. The forces are in the same range for the other iterations. The etot_conv_thr is set to 1.0D-4, the forc_conv_thr to 1.0D-3. @Jia I've tried setting ion_dynamics to damp. However, then I run into difficulty with the electronic convergence. Thanks again. Forces from last bfgs iteration: atom 1 type 2 force = 0.00674604 0.00401572 -0.00315882 atom 2 type 2 force = -0.00707328 -0.00401102 -0.00417376 atom 3 type 1 force = 0.00015167 -0.00003914 0.00011509 atom 4 type 2 force = -0.00016229 -0.00052862 0.00042288 atom 5 type 2 force = 0.00015237 -0.00005063 -0.00033087 atom 6 type 2 force = 0.00097198 0.00025690 0.00085239 atom 7 type 2 force = -0.00020876 0.00063517 0.00036422 atom 8 type 2 force = 0.00057032 0.00054804 0.00024175 atom 9 type 2 force = 0.00096382 -0.00038407 0.00067724 atom 10 type 2 force = -0.00027981 -0.00030741 -0.00017944 atom 11 type 2 force = 0.00014642 0.00002012 -0.00002955 atom 12 type 1 force = -0.00022861 0.00033720 0.00044263 atom 13 type 1 force = 0.00034998 -0.00016970 0.00062187 atom 14 type 1 force = -0.00008588 -0.00012786 0.00006090 atom 15 type 1 force = 0.00189702 0.00094614 -0.00025442 atom 16 type 2 force = 0.00050876 0.00005659 0.00049815 atom 17 type 2 force = 0.00088281 -0.00040997 0.00090335 atom 18 type 2 force = 0.00134491 0.00032421 -0.00078394 atom 19 type 2 force = 0.00066250 0.00025287 -0.00049312 atom 20 type 2 force = -0.00011040 -0.00050559 -0.00006915 atom 21 type 2 force = 0.00095934 -0.00035701 -0.00081817 atom 22 type 2 force = -0.00038611 -0.00007281 -0.00001118 atom 23 type 2 force = 0.00115667 0.00046339 -0.00028953 atom 24 type 1 force = 0.00114906 -0.00047371 0.00049679 atom 25 type 1 force = -0.00090091 0.00041464 0.00047287 atom 26 type 1 force = -0.00127753 -0.00043198 -0.00016529 atom 27 type 1 force = 0.00092469 -0.00029180 0.00027208 atom 28 type 2 force = -0.00004433 -0.00023892 0.00041675 atom 29 type 2 force = 0.00100550 -0.00031420 0.00113785 atom 30 type 2 force = 0.00000884 -0.00111712 0.00068738 atom 31 type 2 force = 0.00003832 -0.00009958 0.00076318 atom 32 type 2 force = 0.00056036 0.00058104 0.00185944 atom 33 type 2 force = 0.00069824 0.00004986 0.00197662 atom 34 type 2 force = 0.00007865 -0.00031354 -0.00024944 atom 35 type 2 force = 0.00040935 -0.00065039 0.00079960 atom 36 type 1 force = 0.00054453 -0.00050520 -0.00081090 atom 37 type 1 force = -0.00014881 0.00019052 -0.00056561 atom 38 type 1 force = -0.00069173 0.00086720 0.00049374 atom 39 type 1 force = -0.00032728 -0.00048677 0.00049346 atom 40 type 2 force = 0.00005298 -0.00003038 0.00057008 atom 41 type 2 force = 0.00052392 0.00031805 0.00092472 atom 42 type 2 force = 0.00053091 0.00148619 0.00034716 atom 43 type 2 force = -0.00139237 -0.00074654 -0.00130242 atom 44 type 2 force = -0.00015835 0.00027715 -0.00041422 atom 45 type 2 force = 0.00098086 -0.00018579 -0.00095187 atom 46 type 2 force = -0.00064724 -0.00000916 0.00016756 atom 47 type 2 force = 0.00368611 -0.00002795 0.00045003 atom 48 type 1 force = 0.00031498 -0.00033410 -0.00002164 atom 49 type 1 force = -0.00072047 0.00063453 0.00013858 atom 50 type 1 force = 0.00054850 0.00029128 0.00078326 atom 51 type 1 force = -0.00018902 0.00061944 0.00105489 atom 52 type 2 force = -0.00043475 0.00002882 0.00006407 atom 53 type 2 force = -0.00056709 0.00012075 0.00041822 atom 54 type 2 force = -0.00056741 0.00024715 0.00023894 atom 55 type 2 force = -0.00041280 0.00059675 0.00160623 atom 56 type 2 force = 0.00014651 0.00028960 0.00159954 atom 57 type 2 force = -0.00171450 0.00002096 0.00247953 atom 58 type 2 force = -0.00017183 0.00085781 0.00007596 atom 59 type 2 force = -0.00102680 0.00054910 -0.00040500 atom 60 type 1 force = -0.00009379 0.00014378 -0.00127470 atom 61 type 1 force = 0.00012531 -0.00025665 -0.00112033 atom 62 type 1 force = 0.00080427 -0.00056558 0.00101772 atom 63 type 1 force = -0.00498284 0.00033298 -0.00350540 atom 64 type 2 force = -0.00086830 0.00066374 0.00055719 atom 65 type 2 force = -0.00073070 -0.00057656 0.00125530 atom 66 type 2 force = -0.00083096 -0.00034313 -0.00035115 atom 67 type 2 force = 0.00106363 -0.00040306 -0.00077587 atom 68 type 2 force = 0.00071954 0.00035647 -0.00132880 atom 69 type 2 force = 0.00017170 0.00042807 -0.00121455 atom 70 type 2 force = -0.00025769 -0.00395602 0.00174604 atom 71 type 2 force = -0.00129773 -0.00071931 0.00039821 atom 72 type 1 force = -0.00301732 -0.00065251 0.00435206 atom 73 type 1 force = 0.00329464 0.00030046 0.00470849 atom 74 type 1 force = 0.00484552 -0.00038429 -0.00313326 atom 75 type 1 force = 0.00022645 -0.00012545 0.00117296 atom 76 type 2 force = -0.00037719 0.00045763 -0.00319435 atom 77 type 2 force = -0.00040814 0.00008839 0.00005440 atom 78 type 2 force = 0.00033602 -0.00001990 0.00046392 atom 79 type 2 force = 0.00028389 -0.00067385 0.00205674 atom 80 type 2 force = -0.00004079 0.00009821 -0.00339820 atom 81 type 2 force = -0.00030699 -0.00038710 0.00080237 atom 82 type 2 force = 0.00008637 -0.00026351 -0.00008527 atom 83 type 2 force = -0.00156333 0.00023433 0.00055603 atom 84 type 1 force = -0.00025828 -0.00004507 0.00074608 atom 85 type 1 force = 0.00057079 -0.00060726 -0.00014710 atom 86 type 1 force = -0.00011808 -0.00005027 0.00081874 atom 87 type 1 force = -0.00282270 0.00131933 -0.00472331 atom 88 type 2 force = -0.00158720 0.00067537 0.00032135 atom 89 type 2 force = -0.00039389 -0.00111775 0.00005539 atom 90 type 2 force = -0.00056433 -0.00005048 -0.00167901 atom 91 type 2 force = -0.00014824 0.00003539 -0.00627238 atom 92 type 2 force = -0.00205334 0.00037334 -0.00044205 atom 93 type 2 force = -0.00025693 0.00472185 0.00131091 atom 94 type 2 force = -0.00366386 0.00012575 0.00058219 atom 95 type 1 force = -0.00339605 -0.00077984 0.00213525 atom 96 type 1 force = 0.00430043 0.00036977 0.00195456 atom 97 type 1 force = 0.00347151 -0.00182346 -0.00493084 Total force = 0.025508 Total SCF correction = 0.000128 Stephan Rix Institute for Inorganic und Analytical Chemistry Johannes Gutenberg-University Mainz, Germany Date: Wed, 15 Sep 2010 17:10:07 +0200 From: Stefano de Gironcoli Subject: Re: [Pw_forum] problems with converge of geometric optimization To: PWSCF Forum Message-ID: <4C90E1CF.6040400 at sissa.it> Content-Type: text/plain; charset=ISO-8859-1; format=flowed This happens when the energy landscape become noisy compared with the expected energy decrease in the relaxation steps. This can happen when you are very close to the minimum. How large are your forces ? How tight is your force/energy threshold ? stefano Stephan Rix wrote: > I am trying to calculate an H-center in CaF2, and I am having trouble with > the geometry optimization. > I have tried adjusting various parameters such as ion dynamics, trust > radius, convergence tresholds, but it seems that convergence cannot be > improved. > > Using bfgs I always get the following error: > > %%%%%%%%%%%%%%%%%%%%%%%%%% > from bfgs : error # 1 > bfgs history already reset at previous step > %%%%%%%%%%%%%%%%%%%%%%%%%% > > Alternatively the electonic convergence gets extremely slow. > > In addition, no matter what parameters I adjust the total forces in the > system never converge, e.g. > Total force = 0.026394 Total SCF correction = 0.000314 > Total force = 0.018379 Total SCF correction = 0.000282 > Total force = 0.018824 Total SCF correction = 0.000120 > Total force = 0.018857 Total SCF correction = 0.000155 > Total force = 0.020215 Total SCF correction = 0.000193 > Total force = 0.058765 Total SCF correction = 0.000241 > Total force = 0.016998 Total SCF correction = 0.000205 > Total force = 0.042967 Total SCF correction = 0.000238 > Total force = 0.106226 Total SCF correction = 0.000073 > Total force = 0.137142 Total SCF correction = 0.000071 > Total force = 0.109549 Total SCF correction = 0.000108 > Total force = 0.068723 Total SCF correction = 0.000103 > Total force = 0.076599 Total SCF correction = 0.000168 > Total force = 0.033233 Total SCF correction = 0.000342 > Total force = 0.026883 Total SCF correction = 0.000248 > Total force = 0.025846 Total SCF correction = 0.000378 > Total force = 0.025596 Total SCF correction = 0.000225 > Total force = 0.025482 Total SCF correction = 0.000275 > Total force = 0.025508 Total SCF correction = 0.000128 > > I am thankful for any suggestions. > > Stephan Rix > Institute for Inorganic und Analytical Chemistry > Johannes Gutenberg-University Mainz, Germany > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100916/b4051765/attachment.htm From degironc at sissa.it Thu Sep 16 10:50:43 2010 From: degironc at sissa.it (Stefano de Gironcoli) Date: Thu, 16 Sep 2010 10:50:43 +0200 Subject: [Pw_forum] problems with converge of geometric optimization In-Reply-To: References: Message-ID: <4C91DA63.5030805@sissa.it> except for the first 2 atoms and a few more in the cell the forces are below threshold... The energy criterion is probably already satisfied (it usually converge faster than the forces). Are the atoms with comparatively large forces close to each others and alway the same or are they a random set that change from one iteration to the other ? Do they form a cluster where something special is expected to happen ? You could try to take the last configuration and restart the relaxation from scratch freezing all the atoms whose forces are below threshold so as to reduce the dimension of the subspace that bfgs is dealing with... However the fact that damped dynamics finds problems to converge may mean that there is something unstable in your calculation.. are you sure that the cutoffs (wfc and rho) are sufficient, the BZ sampling is appropriate for the smearing width you are using... stefano Stephan Rix wrote: > @Stefano > I have listed the forces from the last bfgs iteration. The forces are in the > same range for the other iterations. > The etot_conv_thr is set to 1.0D-4, the forc_conv_thr to 1.0D-3. > > @Jia > I've tried setting ion_dynamics to damp. However, then I run into difficulty > with the electronic convergence. > > Thanks again. > > Forces from last bfgs iteration: > atom 1 type 2 force = 0.00674604 0.00401572 -0.00315882 > atom 2 type 2 force = -0.00707328 -0.00401102 -0.00417376 > atom 3 type 1 force = 0.00015167 -0.00003914 0.00011509 > atom 4 type 2 force = -0.00016229 -0.00052862 0.00042288 > atom 5 type 2 force = 0.00015237 -0.00005063 -0.00033087 > atom 6 type 2 force = 0.00097198 0.00025690 0.00085239 > atom 7 type 2 force = -0.00020876 0.00063517 0.00036422 > atom 8 type 2 force = 0.00057032 0.00054804 0.00024175 > atom 9 type 2 force = 0.00096382 -0.00038407 0.00067724 > atom 10 type 2 force = -0.00027981 -0.00030741 -0.00017944 > atom 11 type 2 force = 0.00014642 0.00002012 -0.00002955 > atom 12 type 1 force = -0.00022861 0.00033720 0.00044263 > atom 13 type 1 force = 0.00034998 -0.00016970 0.00062187 > atom 14 type 1 force = -0.00008588 -0.00012786 0.00006090 > atom 15 type 1 force = 0.00189702 0.00094614 -0.00025442 > atom 16 type 2 force = 0.00050876 0.00005659 0.00049815 > atom 17 type 2 force = 0.00088281 -0.00040997 0.00090335 > atom 18 type 2 force = 0.00134491 0.00032421 -0.00078394 > atom 19 type 2 force = 0.00066250 0.00025287 -0.00049312 > atom 20 type 2 force = -0.00011040 -0.00050559 -0.00006915 > atom 21 type 2 force = 0.00095934 -0.00035701 -0.00081817 > atom 22 type 2 force = -0.00038611 -0.00007281 -0.00001118 > atom 23 type 2 force = 0.00115667 0.00046339 -0.00028953 > atom 24 type 1 force = 0.00114906 -0.00047371 0.00049679 > atom 25 type 1 force = -0.00090091 0.00041464 0.00047287 > atom 26 type 1 force = -0.00127753 -0.00043198 -0.00016529 > atom 27 type 1 force = 0.00092469 -0.00029180 0.00027208 > atom 28 type 2 force = -0.00004433 -0.00023892 0.00041675 > atom 29 type 2 force = 0.00100550 -0.00031420 0.00113785 > atom 30 type 2 force = 0.00000884 -0.00111712 0.00068738 > atom 31 type 2 force = 0.00003832 -0.00009958 0.00076318 > atom 32 type 2 force = 0.00056036 0.00058104 0.00185944 > atom 33 type 2 force = 0.00069824 0.00004986 0.00197662 > atom 34 type 2 force = 0.00007865 -0.00031354 -0.00024944 > atom 35 type 2 force = 0.00040935 -0.00065039 0.00079960 > atom 36 type 1 force = 0.00054453 -0.00050520 -0.00081090 > atom 37 type 1 force = -0.00014881 0.00019052 -0.00056561 > atom 38 type 1 force = -0.00069173 0.00086720 0.00049374 > atom 39 type 1 force = -0.00032728 -0.00048677 0.00049346 > atom 40 type 2 force = 0.00005298 -0.00003038 0.00057008 > atom 41 type 2 force = 0.00052392 0.00031805 0.00092472 > atom 42 type 2 force = 0.00053091 0.00148619 0.00034716 > atom 43 type 2 force = -0.00139237 -0.00074654 -0.00130242 > atom 44 type 2 force = -0.00015835 0.00027715 -0.00041422 > atom 45 type 2 force = 0.00098086 -0.00018579 -0.00095187 > atom 46 type 2 force = -0.00064724 -0.00000916 0.00016756 > atom 47 type 2 force = 0.00368611 -0.00002795 0.00045003 > atom 48 type 1 force = 0.00031498 -0.00033410 -0.00002164 > atom 49 type 1 force = -0.00072047 0.00063453 0.00013858 > atom 50 type 1 force = 0.00054850 0.00029128 0.00078326 > atom 51 type 1 force = -0.00018902 0.00061944 0.00105489 > atom 52 type 2 force = -0.00043475 0.00002882 0.00006407 > atom 53 type 2 force = -0.00056709 0.00012075 0.00041822 > atom 54 type 2 force = -0.00056741 0.00024715 0.00023894 > atom 55 type 2 force = -0.00041280 0.00059675 0.00160623 > atom 56 type 2 force = 0.00014651 0.00028960 0.00159954 > atom 57 type 2 force = -0.00171450 0.00002096 0.00247953 > atom 58 type 2 force = -0.00017183 0.00085781 0.00007596 > atom 59 type 2 force = -0.00102680 0.00054910 -0.00040500 > atom 60 type 1 force = -0.00009379 0.00014378 -0.00127470 > atom 61 type 1 force = 0.00012531 -0.00025665 -0.00112033 > atom 62 type 1 force = 0.00080427 -0.00056558 0.00101772 > atom 63 type 1 force = -0.00498284 0.00033298 -0.00350540 > atom 64 type 2 force = -0.00086830 0.00066374 0.00055719 > atom 65 type 2 force = -0.00073070 -0.00057656 0.00125530 > atom 66 type 2 force = -0.00083096 -0.00034313 -0.00035115 > atom 67 type 2 force = 0.00106363 -0.00040306 -0.00077587 > atom 68 type 2 force = 0.00071954 0.00035647 -0.00132880 > atom 69 type 2 force = 0.00017170 0.00042807 -0.00121455 > atom 70 type 2 force = -0.00025769 -0.00395602 0.00174604 > atom 71 type 2 force = -0.00129773 -0.00071931 0.00039821 > atom 72 type 1 force = -0.00301732 -0.00065251 0.00435206 > atom 73 type 1 force = 0.00329464 0.00030046 0.00470849 > atom 74 type 1 force = 0.00484552 -0.00038429 -0.00313326 > atom 75 type 1 force = 0.00022645 -0.00012545 0.00117296 > atom 76 type 2 force = -0.00037719 0.00045763 -0.00319435 > atom 77 type 2 force = -0.00040814 0.00008839 0.00005440 > atom 78 type 2 force = 0.00033602 -0.00001990 0.00046392 > atom 79 type 2 force = 0.00028389 -0.00067385 0.00205674 > atom 80 type 2 force = -0.00004079 0.00009821 -0.00339820 > atom 81 type 2 force = -0.00030699 -0.00038710 0.00080237 > atom 82 type 2 force = 0.00008637 -0.00026351 -0.00008527 > atom 83 type 2 force = -0.00156333 0.00023433 0.00055603 > atom 84 type 1 force = -0.00025828 -0.00004507 0.00074608 > atom 85 type 1 force = 0.00057079 -0.00060726 -0.00014710 > atom 86 type 1 force = -0.00011808 -0.00005027 0.00081874 > atom 87 type 1 force = -0.00282270 0.00131933 -0.00472331 > atom 88 type 2 force = -0.00158720 0.00067537 0.00032135 > atom 89 type 2 force = -0.00039389 -0.00111775 0.00005539 > atom 90 type 2 force = -0.00056433 -0.00005048 -0.00167901 > atom 91 type 2 force = -0.00014824 0.00003539 -0.00627238 > atom 92 type 2 force = -0.00205334 0.00037334 -0.00044205 > atom 93 type 2 force = -0.00025693 0.00472185 0.00131091 > atom 94 type 2 force = -0.00366386 0.00012575 0.00058219 > atom 95 type 1 force = -0.00339605 -0.00077984 0.00213525 > atom 96 type 1 force = 0.00430043 0.00036977 0.00195456 > atom 97 type 1 force = 0.00347151 -0.00182346 -0.00493084 > Total force = 0.025508 Total SCF correction = 0.000128 > > Stephan Rix > Institute for Inorganic und Analytical Chemistry > Johannes Gutenberg-University Mainz, Germany > > > Date: Wed, 15 Sep 2010 17:10:07 +0200 > From: Stefano de Gironcoli > Subject: Re: [Pw_forum] problems with converge of geometric > optimization > To: PWSCF Forum > Message-ID: <4C90E1CF.6040400 at sissa.it> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > This happens when the energy landscape become noisy compared with the > expected energy decrease in the relaxation steps. > This can happen when you are very close to the minimum. > How large are your forces ? How tight is your force/energy threshold ? > stefano > > Stephan Rix wrote: > >> I am trying to calculate an H-center in CaF2, and I am having trouble with >> the geometry optimization. >> I have tried adjusting various parameters such as ion dynamics, trust >> radius, convergence tresholds, but it seems that convergence cannot be >> improved. >> >> Using bfgs I always get the following error: >> >> %%%%%%%%%%%%%%%%%%%%%%%%%% >> from bfgs : error # 1 >> bfgs history already reset at previous step >> %%%%%%%%%%%%%%%%%%%%%%%%%% >> >> Alternatively the electonic convergence gets extremely slow. >> >> In addition, no matter what parameters I adjust the total forces in the >> system never converge, e.g. >> Total force = 0.026394 Total SCF correction = 0.000314 >> Total force = 0.018379 Total SCF correction = 0.000282 >> Total force = 0.018824 Total SCF correction = 0.000120 >> Total force = 0.018857 Total SCF correction = 0.000155 >> Total force = 0.020215 Total SCF correction = 0.000193 >> Total force = 0.058765 Total SCF correction = 0.000241 >> Total force = 0.016998 Total SCF correction = 0.000205 >> Total force = 0.042967 Total SCF correction = 0.000238 >> Total force = 0.106226 Total SCF correction = 0.000073 >> Total force = 0.137142 Total SCF correction = 0.000071 >> Total force = 0.109549 Total SCF correction = 0.000108 >> Total force = 0.068723 Total SCF correction = 0.000103 >> Total force = 0.076599 Total SCF correction = 0.000168 >> Total force = 0.033233 Total SCF correction = 0.000342 >> Total force = 0.026883 Total SCF correction = 0.000248 >> Total force = 0.025846 Total SCF correction = 0.000378 >> Total force = 0.025596 Total SCF correction = 0.000225 >> Total force = 0.025482 Total SCF correction = 0.000275 >> Total force = 0.025508 Total SCF correction = 0.000128 >> >> I am thankful for any suggestions. >> >> Stephan Rix >> Institute for Inorganic und Analytical Chemistry >> Johannes Gutenberg-University Mainz, Germany >> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From bipulrr at gmail.com Thu Sep 16 14:35:09 2010 From: bipulrr at gmail.com (Bipul Rakshit) Date: Thu, 16 Sep 2010 18:05:09 +0530 Subject: [Pw_forum] Error Message-ID: Dear PWSCF users, I am running some scf/relax calculation in AIX machine, and it shows the following error from read_pseudo_mesh : error # 1 Reading pseudo file (R) for K the same calculation I am running in LINUX machines, and it is running without any error Both the calculations I am running in parallel processors I am sending you the input file and the error file .... please have a look and suggest me -- Bipul Rakshit Research Fellow, S N Bose Centre for Basic Sciences, Salt Lake, Kolkata 700 098 India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100916/b69bac0c/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: KO2.scf.in Type: application/octet-stream Size: 2544 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100916/b69bac0c/attachment.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: err.file Type: application/octet-stream Size: 3080 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100916/b69bac0c/attachment-0001.obj From giannozz at democritos.it Thu Sep 16 17:12:32 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 16 Sep 2010 17:12:32 +0200 Subject: [Pw_forum] Error In-Reply-To: References: Message-ID: <4C9233E0.4030901@democritos.it> Bipul Rakshit wrote: > I am running some scf/relax calculation in AIX machine, and it shows the > following error > > from read_pseudo_mesh : error # 1 > Reading pseudo file (R) for K remove DOS control-M characters from K pseudopotential file: cat "file-with-ctrl-M" | tr -d "^V^M" > "file-without-ctrl-M" where ^V^M means "press control+V followed by control+M" P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From orondo at MIT.EDU Thu Sep 16 21:55:17 2010 From: orondo at MIT.EDU (Peter O Orondo) Date: Thu, 16 Sep 2010 15:55:17 -0400 Subject: [Pw_forum] Single Cell vs Super-Cell Message-ID: <02a601cb55d9$17514f60$45f3ee20$@edu> Thanks Prof. for that insight. Peter O Orondo wrote: > I am trying to calculate the chemical potential of H in a bunch of > metals. When I use a single cell, I get results that compare to > experiment. However, when I use a 2x2x2 super-cell, I expect to get a > more accurate result since the super-cell should model the physical > properties more accurately. However, the super-cell results don't even > have the right shape. The results look rather random. > > Any insights into what could be wrong? hard to say. The fact that you get results that compare well to experiments doesn't imply that those results are correct. In simple metals, if you put one H atom per unit cell you simulate a hydride rather than H dissolved into a metal. If you increase the size of the supercell, you should notice that your results (e.g. the formation energy) converges to some value, which should be good for a isolated impurity. Convergence may be quite slow, though, depending upon what you calculate and in which system Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100916/54b6dc4c/attachment.htm From jdburton1 at gmail.com Thu Sep 16 22:22:40 2010 From: jdburton1 at gmail.com (J. D. Burton) Date: Thu, 16 Sep 2010 15:22:40 -0500 Subject: [Pw_forum] where does pp.x get degauss from for plot_num = 3? Message-ID: <004601cb55dc$ea9e2570$bfda7050$@gmail.com> Greetings, I'm trying to use plot_num = 3 to image the metal induced gap states in a metal/insulator heterostructure and what I'm getting doesn't make much sense: there should be a roughly exponential decay into the insulator from both interfaces, but I find the Fermi-level ldos in the insulator is roughly the same as in the metal, and there are some regions in the metal where there is essentially zero Fermi-level ldos. Anyway, my question is: Where does the pp.x program get the value of degauss that it uses to smear out the eigenvalues for the ldos calculation? I want to try changing this to see what happens to the result. Cheers, J. D. Burton ************************************ J. D. Burton, Ph.D. jdburton1 at gmail.com University of Nebraska Lincoln Physics and Astronomy Office Ph. (402) 472 2499 Mobile Ph. (402) 419 9918 310A Jorgensen Hall ************************************ "The job of a scientist is to generate wrong ideas as fast as possible." -- Murray Gell-Mann -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100916/dae507ec/attachment-0001.htm From elie.moujaes at hotmail.co.uk Fri Sep 17 17:25:50 2010 From: elie.moujaes at hotmail.co.uk (Elie Moujaes) Date: Fri, 17 Sep 2010 16:25:50 +0100 Subject: [Pw_forum] quantum capacitance Message-ID: Dear all, I am interested in extracting the quantum capacitance of gated bilayer graphene system (system under the effect of an external electric field inclusing the resulting screening effects)that I am working on. Is there a possibility to do so using quantum espresso (maybe from the calculation of the bandstructure). Thanks Elie Moujaes University of Nottingham NG7 2RD UK -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100917/02b635dc/attachment.htm From kirtinandan07 at gmail.com Sat Sep 18 18:33:54 2010 From: kirtinandan07 at gmail.com (vicky singh) Date: Sat, 18 Sep 2010 22:03:54 +0530 Subject: [Pw_forum] no. of nbnd Message-ID: Hi I am studying a supercell for Ni with 64 atoms. while specifying nbnd how much i select i am not sure. if i go by document PW.html it says that for metal it should be nelec/2 + 20 %. then should i take nbnd = 400. also it says minimum 4 more i.e. should i take nbnd = 330. Thanks in advance. with regards vicky singh research student Bangalore, India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100918/8c0708d6/attachment.htm From ttduyle at gmail.com Sat Sep 18 18:48:31 2010 From: ttduyle at gmail.com (Duy Le) Date: Sat, 18 Sep 2010 12:48:31 -0400 Subject: [Pw_forum] no. of nbnd In-Reply-To: References: Message-ID: As the PW.html said, just count number of electrons then divide it by 2, then add 20% of that number. -------------------------------------------------- Duy Le PhD Student Department of Physics University of Central Florida. "Men don't need hand to do things" On Sat, Sep 18, 2010 at 12:33 PM, vicky singh wrote: > Hi > I am studying a supercell for Ni with 64 atoms. while specifying nbnd how > much i select i am not sure. if i go by document PW.html it says that for > metal it should be nelec/2 + 20 %. then should i take nbnd = 400. also it > says minimum 4 more i.e. should i take nbnd = 330. Thanks in advance. > > with regards > > vicky singh > research student > Bangalore, India > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > From giannozz at democritos.it Sat Sep 18 18:53:40 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sat, 18 Sep 2010 18:53:40 +0200 Subject: [Pw_forum] no. of nbnd In-Reply-To: References: Message-ID: <0AC966F9-992E-4F8F-A2B0-AD36D9BE981C@democritos.it> On Sep 18, 2010, at 18:33 , vicky singh wrote: > while specifying nbnd how much i select i am not sure the criterion is as follows: for each k-points, all occupied and partially occupied bands must be included among the nbnd lowest bands P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From marzari at MIT.EDU Sat Sep 18 19:01:24 2010 From: marzari at MIT.EDU (Nicola Marzari) Date: Sat, 18 Sep 2010 18:01:24 +0100 Subject: [Pw_forum] no. of nbnd In-Reply-To: References: Message-ID: <4C94F064.5010908@mit.edu> vicky singh wrote: > Hi > I am studying a supercell for Ni with 64 atoms. while specifying nbnd > how much i select i am not sure. if i go by document PW.html it says > that for metal it should be nelec/2 + 20 %. then should i take nbnd = > 400. also it says minimum 4 more i.e. should i take nbnd = 330. Thanks > in advance. Dear Vicky, short answer: it depends on "degauss". You are studying a metallic system, and introducing some fictitious temperature for the electronic degrees of freedom (this is the "broadening" or "smearing") in order to improve the quality of your k-point sampling (otherwise at zero temperature there are sharp discontinuities in the integrals over the brillouin zone). you need to have enough bands so that the highest bands are close to empty - having more would be a waste of cpu, having less is somewhat incorrect because you force the higher bands to be more occupied than they would like to be according to the broadening you have chosen. In a gaussian/m-p/m-v scheme, it is safe to say that the occupations becomes negligible around 3 times degauss above the fermi energy; check for yourself (by asking the code to print occupations) that the highest bands have occupation less than 0.01 . If you were to use Fermi-Dirac, that has longer tails, you would need more bands (hence better not to use it, unless you are keen on real fermi-dirac statistics). nicola -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From giannozz at democritos.it Sat Sep 18 19:20:23 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sat, 18 Sep 2010 19:20:23 +0200 Subject: [Pw_forum] where does pp.x get degauss from for plot_num = 3? In-Reply-To: <004601cb55dc$ea9e2570$bfda7050$@gmail.com> References: <004601cb55dc$ea9e2570$bfda7050$@gmail.com> Message-ID: <09CD4729-BEE8-4B4E-A4CB-CE95525A9AE5@democritos.it> On Sep 16, 2010, at 22:22 , J. D. Burton wrote: > Where does the pp.x program get the value of degauss that it uses > to smear out the eigenvalues for the ldos calculation? I want to > try changing this to see what happens to the result. > the value of degauss (and of "ngauss", used to select the kind of smearing) > is read from the xml data file, if present. Rather than trying to locate where it is read, take inspiration from PP/dos.f90 (search for "gauss") P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From quantumdft at gmail.com Sat Sep 18 19:30:29 2010 From: quantumdft at gmail.com (vega lew) Date: Sun, 19 Sep 2010 01:30:29 +0800 Subject: [Pw_forum] Do you think we could fix certain bond length during the relax calculation Message-ID: <4C94F735.9070902@gmail.com> Dear all, I want to calculate the energy barrier for breaking certain bond on solid surface. The NEB is most straightforward method to do this. But my system is very large, the NEB is very time consuming method. So I want to fix the bond to a series of length, such as 1.0A, 1.1A, 1.2A, 1.3A ......, and plot the energy profile during the bond stretch process. But I found the Q-E have no direct command to control the bond length of the system. How could I cope with this problem of Q-E? thank you in advance. Any suggestion is greatly appreciated. best wishes, vega -- ================================================ Vega Lew ( weijia liu) Graduate student State Key Laboratory of Materials-oriented Chemical Engineering Nanjing University of Technology, 210009, Nanjing, China *************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, China *************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100919/f690790a/attachment.htm From mcmahon.physics at gmail.com Sun Sep 19 00:26:03 2010 From: mcmahon.physics at gmail.com (Jeffrey McMahon) Date: Sat, 18 Sep 2010 17:26:03 -0500 Subject: [Pw_forum] question about press_conv_thr Message-ID: Dear all, Which is more important for a vc-relax calculation with external pressure, the absolute or relative magnitude of press_conv_thr? I am thinking the latter, because if I were to perform a vc-relax at a pressure of 2000 kbar it would seem as if the default value of press_conv_thr=0.5 is fine. However, if the pressure were 20000 kbar then press_conv_thr=0.5 seems rather small. So at very high pressures is a relative convergence value, such as 1e-4*P (e.g., 2 at 20000 kbar), a more appropriate quantity? Thanks in advance. Jeffrey M. McMahon The Institute for Condensed Matter Theory Department of Physics University of Illinois at Urbana-Champaign Urbana, IL 61801-3080 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100918/f58d4056/attachment.htm From kirtinandan07 at gmail.com Sun Sep 19 03:10:28 2010 From: kirtinandan07 at gmail.com (vicky singh) Date: Sun, 19 Sep 2010 06:40:28 +0530 Subject: [Pw_forum] no. of nbnd In-Reply-To: <4C94F064.5010908@mit.edu> References: <4C94F064.5010908@mit.edu> Message-ID: thank you very much for the reply. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100919/7c9aeddb/attachment.htm From degironc at sissa.it Sun Sep 19 10:52:39 2010 From: degironc at sissa.it (Stefano de Gironcoli) Date: Sun, 19 Sep 2010 10:52:39 +0200 Subject: [Pw_forum] question about press_conv_thr In-Reply-To: References: Message-ID: <4C95CF57.8010709@sissa.it> dear Jeffrey McMahon, the bulk modulus is defined as K = V dP/dV, compressibility is its inverse. hence a given absolute tolerance on the pressure gives a relative tolerance on the volume determination proportional to the compressibility of the material. I think that for "normal pressures" where bulk moduli (or compressibilities) do not change by order of magnitude as you increase the pressure then the absolute threshold is a better criterion. If you deal with ultra-high pressure and the bulk modulus becomes proportional to the external pressure you may be right and a relative criterion may be what you actually want. Anyway it is something that you can tune on input on the basis of an estimated value of compressibility and the relative accuracy on volumes that you desire. best, stefano Jeffrey McMahon wrote: > Dear all, > > Which is more important for a vc-relax calculation with external pressure, > the absolute or relative magnitude of press_conv_thr? I am thinking the > latter, because if I were to perform a vc-relax at a pressure of 2000 kbar > it would seem as if the default value of press_conv_thr=0.5 is fine. > However, if the pressure were 20000 kbar then press_conv_thr=0.5 seems > rather small. So at very high pressures is a relative convergence value, > such as 1e-4*P (e.g., 2 at 20000 kbar), a more appropriate quantity? Thanks > in advance. > > Jeffrey M. McMahon > > The Institute for Condensed Matter Theory > Department of Physics > University of Illinois at Urbana-Champaign > Urbana, IL 61801-3080 > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From giannozz at democritos.it Sun Sep 19 11:03:03 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sun, 19 Sep 2010 11:03:03 +0200 Subject: [Pw_forum] Do you think we could fix certain bond length during the relax calculation In-Reply-To: <4C94F735.9070902@gmail.com> References: <4C94F735.9070902@gmail.com> Message-ID: <4A636B18-C407-4387-A0DA-A6D5AC70EC1D@democritos.it> On Sep 18, 2010, at 19:30 , vega lew wrote: > But I found the Q-E have no direct command to control the > bond length of the system. there is an implementation of contrained optimization, using damped MD (not bfgs). See card CONTRAINTS and file Doc/constraints_HOWTO.tex P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From m_pazoki at physics.sharif.edu Sun Sep 19 16:21:32 2010 From: m_pazoki at physics.sharif.edu (meysam pazoki) Date: Sun, 19 Sep 2010 18:51:32 +0430 Subject: [Pw_forum] Problem in plotting the Charge Density Message-ID: Dear PWSCF users I have a problem in plotting the 3D charge density in espresso.I ran my scf job.and now i want to use pp.x to plot charge density in 3 dimensions.I use *plot_num=0*,and use various parameters of *output_format(=4,5,6)* for 3d plotting of charge density but xcrysden cant show any meaningful data about charge density.Am i wrong in my pp.in file? can Anyone help me in this problem? Thanks in advance meysam pazoki SUT -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100919/49a89b9b/attachment.htm From juanlopez415 at gmail.com Sun Sep 19 16:42:26 2010 From: juanlopez415 at gmail.com (Juan Manuel Lopez) Date: Sun, 19 Sep 2010 10:42:26 -0400 Subject: [Pw_forum] Problem in plotting the Charge Density In-Reply-To: References: Message-ID: Hi Meysam, After you open the XCrysDen file (show the atomic structure), you need to go to the menu Tools-Data Grid (in order to update the charge density values), after that in the dialog box you can manipulate and plot your charge density. Note: For XCrysDen format you need output_format(3 or 5). On Sun, Sep 19, 2010 at 10:21 AM, meysam pazoki wrote: > Dear PWSCF users > > I have a problem in plotting the 3D charge density in espresso.I ran my scf > job.and now i want to > use pp.x to plot charge density in 3 dimensions.I use *plot_num=0*,and use > various parameters of *output_format(=4,5,6)* for 3d plotting of charge > density but xcrysden cant show any meaningful data about charge density.Am i > wrong in my pp.in file? > can Anyone help me in this problem? > Thanks in advance > meysam pazoki > SUT > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Best regards, -------------------------------------- Juan M. Lopez Encarnacion Research associate, Institute of Physics, Universidad Aut?noma de Santo Domingo, Dominican Republic Graduate Student Doctoral Program in Chemical Physics, University of Puerto Rico, Rio Piedras Campus P.O.Box 23343 San Juan, PR 00931-3343 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100919/7d432655/attachment.htm From didi5158 at gmail.com Mon Sep 20 08:07:30 2010 From: didi5158 at gmail.com (Wang Di) Date: Mon, 20 Sep 2010 14:07:30 +0800 Subject: [Pw_forum] question about the stress in relax process Message-ID: Dear all I perform the relax calculate to the crystal structure. After 11 step bfgs, the total force and total scf correction had reach to the criterion, however, the total stress was so big P=-3443.87. I have no idea about deal with the case, could anyone give some advices? Thanks in advance. the input file : &CONTROL calculation = 'scf' , restart_mode = 'from_scratch' , wf_collect = .true. , outdir = '/lustre/AIOFM/gxgu/workLCAF/5/temp/' , wfcdir = '/lustre/AIOFM/gxgu/workLCAF/5/wftemp/' , pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , prefix = 'LCAF' , etot_conv_thr = 1.0D-7 , forc_conv_thr = 1.0D-6 , nstep = 100 , tstress = .true. , tprnfor = .true. , dipfield = .true. , / &SYSTEM ibrav = 4, celldm(1) = 9.42432151, celldm(3) = 1.94804220, nat = 18, ntyp = 4, ecutwfc = 30 , ecutrho = 120 , nosym = .false. , / &ELECTRONS electron_maxstep = 100, conv_thr = 1.0D-10 , / ATOMIC_SPECIES Li 6.94100 03-Li.LDA.fhi.UPF Al 26.98154 13-Al.LDA.fhi.UPF Ca 40.07800 20-Ca.LDA.fhi.UPF F 18.99840 09-F.LDA.fhi.UPF ATOMIC_POSITIONS crystal Li 0.333333333 0.666666667 0.250000000 Li -0.333333333 -0.666666667 -0.250000000 F 0.389997 0.045970 0.148326 F -0.045970 0.344027 0.148326 F -0.344027 -0.389997 0.148326 F -0.045970 -0.389997 0.351674 F -0.344027 0.045970 0.351674 F 0.389997 0.344027 0.351674 F -0.389997 -0.045970 -0.148326 F 0.045970 -0.344027 -0.148326 F 0.344027 0.389997 -0.148326 F 0.045970 0.389997 0.648326 F 0.344027 -0.045970 0.648326 F -0.389997 -0.344027 0.648326 Al 0.666666667 0.333333333 0.250000000 Al -0.666666667 -0.333333333 -0.250000000 Ca 0.000000000 0.000000000 0.000000000 Ca 0.000000000 0.000000000 0.500000000 K_POINTS automatic 16 16 16 0 0 0 best regart Wang Di -- ----------------------------------------------------------------- *Wang Di** The Crystal Lab,* *AnHui Institute of Optics and Fine Mechanics, *** *Chinese Academy of Sciences,** No350. , Shushanhu Road, Hefei, 230031, China**, E-mail: didi5158 @gmail.ocm** ,* ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/df91893a/attachment.htm From quantumdft at gmail.com Mon Sep 20 10:08:51 2010 From: quantumdft at gmail.com (vega lew) Date: Mon, 20 Sep 2010 16:08:51 +0800 Subject: [Pw_forum] Do you think we could fix certain bond length during the relax calculation In-Reply-To: <4A636B18-C407-4387-A0DA-A6D5AC70EC1D@democritos.it> References: <4C94F735.9070902@gmail.com> <4A636B18-C407-4387-A0DA-A6D5AC70EC1D@democritos.it> Message-ID: <4C971693.5060105@gmail.com> Thank you Paolo, Now i could fix the distance between two atoms. best wishes, vega On 09/19/10 17:03, Paolo Giannozzi wrote: > On Sep 18, 2010, at 19:30 , vega lew wrote: > >> But I found the Q-E have no direct command to control the >> bond length of the system. > there is an implementation of contrained optimization, > using damped MD (not bfgs). See card CONTRAINTS > and file Doc/constraints_HOWTO.tex > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- ================================================ Vega Lew ( weijia liu) Graduate student State Key Laboratory of Materials-oriented Chemical Engineering Nanjing University of Technology, 210009, Nanjing, China *************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, China *************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/7f5a151e/attachment.htm From mohnish.iitk at gmail.com Mon Sep 20 11:35:15 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Mon, 20 Sep 2010 15:05:15 +0530 Subject: [Pw_forum] question about the stress in relax process In-Reply-To: References: Message-ID: Dear Wang, The 'relax' process optimizes the structure within given geometric constraint. It does'nt care for stress. It minimizes force only in the given constraint. If you want to minimize the stress also you have to do 'vc-relax' calculation, It will give you what you want. On Mon, Sep 20, 2010 at 11:37 AM, Wang Di wrote: > Dear all > > I perform the relax calculate to the crystal structure. After 11 step > bfgs, the total force and total scf correction had reach to the criterion, > however, the total stress was so big P=-3443.87. I have no idea about deal > with the case, could anyone give some advices? Thanks in advance. > > the input file > : &CONTROL > calculation = 'scf' , > restart_mode = 'from_scratch' , > wf_collect = .true. , > outdir = '/lustre/AIOFM/gxgu/workLCAF/5/temp/' , > wfcdir = '/lustre/AIOFM/gxgu/workLCAF/5/wftemp/' , > pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , > prefix = 'LCAF' , > etot_conv_thr = 1.0D-7 , > forc_conv_thr = 1.0D-6 , > nstep = 100 , > tstress = .true. , > tprnfor = .true. , > dipfield = .true. , > / > &SYSTEM > ibrav = 4, > celldm(1) = 9.42432151, > celldm(3) = 1.94804220, > nat = 18, > ntyp = 4, > ecutwfc = 30 , > ecutrho = 120 , > nosym = .false. , > / > &ELECTRONS > electron_maxstep = 100, > conv_thr = 1.0D-10 , > / > ATOMIC_SPECIES > Li 6.94100 03-Li.LDA.fhi.UPF > Al 26.98154 13-Al.LDA.fhi.UPF > Ca 40.07800 20-Ca.LDA.fhi.UPF > F 18.99840 09-F.LDA.fhi.UPF > ATOMIC_POSITIONS crystal > Li 0.333333333 0.666666667 0.250000000 > Li -0.333333333 -0.666666667 -0.250000000 > F 0.389997 0.045970 0.148326 > F -0.045970 0.344027 0.148326 > F -0.344027 -0.389997 0.148326 > F -0.045970 -0.389997 0.351674 > F -0.344027 0.045970 0.351674 > F 0.389997 0.344027 0.351674 > F -0.389997 -0.045970 -0.148326 > F 0.045970 -0.344027 -0.148326 > F 0.344027 0.389997 -0.148326 > F 0.045970 0.389997 0.648326 > F 0.344027 -0.045970 0.648326 > F -0.389997 -0.344027 0.648326 > Al 0.666666667 0.333333333 0.250000000 > Al -0.666666667 -0.333333333 -0.250000000 > Ca 0.000000000 0.000000000 0.000000000 > Ca 0.000000000 0.000000000 0.500000000 > K_POINTS automatic > 16 16 16 0 0 0 > > best regart > > Wang Di > > -- > ----------------------------------------------------------------- > *Wang Di** > The Crystal Lab,* > *AnHui Institute of Optics and Fine Mechanics, *** > *Chinese Academy of Sciences,** > No350. , Shushanhu Road, Hefei, 230031, China**, > E-mail: didi5158 @gmail.ocm** ,* > ----------------------------------------------------------------- > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/4257ed30/attachment.htm From didi5158 at gmail.com Mon Sep 20 12:02:16 2010 From: didi5158 at gmail.com (Wang Di) Date: Mon, 20 Sep 2010 18:02:16 +0800 Subject: [Pw_forum] question about the stress in relax process In-Reply-To: References: Message-ID: Dear Mohnish: I have used the 'vc-relax' process optimized the cell parameter, however, the result were seem to deviate from the initial value (experimental value) too much.The aim of the structure optimizes is for calculating the Raman spectra of the structure, I think the bad structure ( not coincide to the experiment) is hard to got the true spectra. So I don't how to deal with the case. According to your experience, what should I do for the next step? Thank you, dear Mohnish, for your kind help. 2010/9/20 mohnish pandey > Dear Wang, > The 'relax' process optimizes the structure within > given geometric constraint. It does'nt care for stress. It minimizes force > only in the given constraint. If you want to minimize the stress also you > have to do 'vc-relax' calculation, It will give you what you want. > > On Mon, Sep 20, 2010 at 11:37 AM, Wang Di wrote: > >> Dear all >> >> I perform the relax calculate to the crystal structure. After 11 step >> bfgs, the total force and total scf correction had reach to the criterion, >> however, the total stress was so big P=-3443.87. I have no idea about deal >> with the case, could anyone give some advices? Thanks in advance. >> >> the input file >> : &CONTROL >> calculation = 'scf' , >> restart_mode = 'from_scratch' , >> wf_collect = .true. , >> outdir = '/lustre/AIOFM/gxgu/workLCAF/5/temp/' , >> wfcdir = '/lustre/AIOFM/gxgu/workLCAF/5/wftemp/' , >> pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , >> prefix = 'LCAF' , >> etot_conv_thr = 1.0D-7 , >> forc_conv_thr = 1.0D-6 , >> nstep = 100 , >> tstress = .true. , >> tprnfor = .true. , >> dipfield = .true. , >> / >> &SYSTEM >> ibrav = 4, >> celldm(1) = 9.42432151, >> celldm(3) = 1.94804220, >> nat = 18, >> ntyp = 4, >> ecutwfc = 30 , >> ecutrho = 120 , >> nosym = .false. , >> / >> &ELECTRONS >> electron_maxstep = 100, >> conv_thr = 1.0D-10 , >> / >> ATOMIC_SPECIES >> Li 6.94100 03-Li.LDA.fhi.UPF >> Al 26.98154 13-Al.LDA.fhi.UPF >> Ca 40.07800 20-Ca.LDA.fhi.UPF >> F 18.99840 09-F.LDA.fhi.UPF >> ATOMIC_POSITIONS crystal >> Li 0.333333333 0.666666667 0.250000000 >> Li -0.333333333 -0.666666667 -0.250000000 >> F 0.389997 0.045970 0.148326 >> F -0.045970 0.344027 0.148326 >> F -0.344027 -0.389997 0.148326 >> F -0.045970 -0.389997 0.351674 >> F -0.344027 0.045970 0.351674 >> F 0.389997 0.344027 0.351674 >> F -0.389997 -0.045970 -0.148326 >> F 0.045970 -0.344027 -0.148326 >> F 0.344027 0.389997 -0.148326 >> F 0.045970 0.389997 0.648326 >> F 0.344027 -0.045970 0.648326 >> F -0.389997 -0.344027 0.648326 >> Al 0.666666667 0.333333333 0.250000000 >> Al -0.666666667 -0.333333333 -0.250000000 >> Ca 0.000000000 0.000000000 0.000000000 >> Ca 0.000000000 0.000000000 0.500000000 >> K_POINTS automatic >> 16 16 16 0 0 0 >> >> best regart >> >> Wang Di >> >> -- >> ----------------------------------------------------------------- >> *Wang Di** >> The Crystal Lab,* >> *AnHui Institute of Optics and Fine Mechanics, *** >> *Chinese Academy of Sciences,** >> No350. , Shushanhu Road, Hefei, 230031, China**, >> E-mail: didi5158 @gmail.ocm** ,* >> ----------------------------------------------------------------- >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> > > > -- > Regards, > MOHNISH, > ----------------------------------------------------------------- > Mohnish Pandey > Y6927262,5th Year dual degree student, > Department of Chemical Engineering, > IIT KANPUR, UP, INDIA > +919235721300 > ----------------------------------------------------------------- > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ----------------------------------------------------------------- *Wang Di** The Crystal Lab,* *AnHui Institute of Optics and Fine Mechanics, *** *Chinese Academy of Sciences,** No350. , Shushanhu Road, Hefei, 230031, China**, E-mail: didi5158 @gmail.ocm** ,* ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/76a50603/attachment-0001.htm From giannozz at democritos.it Mon Sep 20 12:08:37 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 20 Sep 2010 12:08:37 +0200 Subject: [Pw_forum] question about the stress in relax process In-Reply-To: References: Message-ID: <4C9732A5.2020302@democritos.it> Wang Di wrote: > ecutwfc = 30 , > ATOMIC_SPECIES > Li 6.94100 03-Li.LDA.fhi.UPF > Al 26.98154 13-Al.LDA.fhi.UPF > Ca 40.07800 20-Ca.LDA.fhi.UPF > F 18.99840 09-F.LDA.fhi.UPF 30 Ry is too low for norm-conserving F. Also, you should use a PP for Li (maybe also for Ca) with nonlinear core corrections P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From mohnish.iitk at gmail.com Mon Sep 20 12:09:24 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Mon, 20 Sep 2010 15:39:24 +0530 Subject: [Pw_forum] question about the stress in relax process In-Reply-To: References: Message-ID: Dear Wang, Are you sure that this is your input file? Because the calculation flag says it is "scf" ..Which will not definitely do the relax calculation.... On Mon, Sep 20, 2010 at 3:32 PM, Wang Di wrote: > Dear Mohnish: > I have used the 'vc-relax' process optimized the cell parameter, however, > the result were seem to deviate from the initial value (experimental value) > too much.The aim of the structure optimizes is for calculating the Raman > spectra of the structure, I think the bad structure ( not coincide to the > experiment) is hard to got the true spectra. So I don't how to deal with the > case. According to your experience, what should I do for the next step? > > Thank you, dear Mohnish, for your kind help. > > > > 2010/9/20 mohnish pandey > > Dear Wang, >> The 'relax' process optimizes the structure within >> given geometric constraint. It does'nt care for stress. It minimizes force >> only in the given constraint. If you want to minimize the stress also you >> have to do 'vc-relax' calculation, It will give you what you want. >> >> On Mon, Sep 20, 2010 at 11:37 AM, Wang Di wrote: >> >>> Dear all >>> >>> I perform the relax calculate to the crystal structure. After 11 step >>> bfgs, the total force and total scf correction had reach to the criterion, >>> however, the total stress was so big P=-3443.87. I have no idea about deal >>> with the case, could anyone give some advices? Thanks in advance. >>> >>> the input file >>> : &CONTROL >>> calculation = 'scf' , >>> restart_mode = 'from_scratch' , >>> wf_collect = .true. , >>> outdir = '/lustre/AIOFM/gxgu/workLCAF/5/temp/' , >>> wfcdir = '/lustre/AIOFM/gxgu/workLCAF/5/wftemp/' , >>> pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , >>> prefix = 'LCAF' , >>> etot_conv_thr = 1.0D-7 , >>> forc_conv_thr = 1.0D-6 , >>> nstep = 100 , >>> tstress = .true. , >>> tprnfor = .true. , >>> dipfield = .true. , >>> / >>> &SYSTEM >>> ibrav = 4, >>> celldm(1) = 9.42432151, >>> celldm(3) = 1.94804220, >>> nat = 18, >>> ntyp = 4, >>> ecutwfc = 30 , >>> ecutrho = 120 , >>> nosym = .false. , >>> / >>> &ELECTRONS >>> electron_maxstep = 100, >>> conv_thr = 1.0D-10 , >>> / >>> ATOMIC_SPECIES >>> Li 6.94100 03-Li.LDA.fhi.UPF >>> Al 26.98154 13-Al.LDA.fhi.UPF >>> Ca 40.07800 20-Ca.LDA.fhi.UPF >>> F 18.99840 09-F.LDA.fhi.UPF >>> ATOMIC_POSITIONS crystal >>> Li 0.333333333 0.666666667 0.250000000 >>> Li -0.333333333 -0.666666667 -0.250000000 >>> F 0.389997 0.045970 0.148326 >>> F -0.045970 0.344027 0.148326 >>> F -0.344027 -0.389997 0.148326 >>> F -0.045970 -0.389997 0.351674 >>> F -0.344027 0.045970 0.351674 >>> F 0.389997 0.344027 0.351674 >>> F -0.389997 -0.045970 -0.148326 >>> F 0.045970 -0.344027 -0.148326 >>> F 0.344027 0.389997 -0.148326 >>> F 0.045970 0.389997 0.648326 >>> F 0.344027 -0.045970 0.648326 >>> F -0.389997 -0.344027 0.648326 >>> Al 0.666666667 0.333333333 0.250000000 >>> Al -0.666666667 -0.333333333 -0.250000000 >>> Ca 0.000000000 0.000000000 0.000000000 >>> Ca 0.000000000 0.000000000 0.500000000 >>> K_POINTS automatic >>> 16 16 16 0 0 0 >>> >>> best regart >>> >>> Wang Di >>> >>> -- >>> ----------------------------------------------------------------- >>> *Wang Di** >>> The Crystal Lab,* >>> *AnHui Institute of Optics and Fine Mechanics, *** >>> *Chinese Academy of Sciences,** >>> No350. , Shushanhu Road, Hefei, 230031, China**, >>> E-mail: didi5158 @gmail.ocm** ,* >>> ----------------------------------------------------------------- >>> >>> _______________________________________________ >>> Pw_forum mailing list >>> Pw_forum at pwscf.org >>> http://www.democritos.it/mailman/listinfo/pw_forum >>> >>> >> >> >> -- >> Regards, >> MOHNISH, >> ----------------------------------------------------------------- >> Mohnish Pandey >> Y6927262,5th Year dual degree student, >> Department of Chemical Engineering, >> IIT KANPUR, UP, INDIA >> +919235721300 >> ----------------------------------------------------------------- >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> > > > -- > ----------------------------------------------------------------- > *Wang Di** > The Crystal Lab,* > *AnHui Institute of Optics and Fine Mechanics, *** > *Chinese Academy of Sciences,** > No350. , Shushanhu Road, Hefei, 230031, China**, > E-mail: didi5158 @gmail.ocm** ,* > ----------------------------------------------------------------- > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/6f30a75e/attachment.htm From nkxirainbow at gmail.com Mon Sep 20 12:12:38 2010 From: nkxirainbow at gmail.com (xirainbow) Date: Mon, 20 Sep 2010 18:12:38 +0800 Subject: [Pw_forum] question about the stress in relax process In-Reply-To: References: Message-ID: Dear Wang: All these parameters in the blow article may influence the relaxed structure. *?Designing meaningful density functional theory calculations in materials science?a prime.?* * * *** * On Mon, Sep 20, 2010 at 6:02 PM, Wang Di wrote: > Dear Mohnish: > I have used the 'vc-relax' process optimized the cell parameter, however, > the result were seem to deviate from the initial value (experimental value) > too much.The aim of the structure optimizes is for calculating the Raman > spectra of the structure, I think the bad structure ( not coincide to the > experiment) is hard to got the true spectra. So I don't how to deal with the > case. According to your experience, what should I do for the next step? > > Thank you, dear Mohnish, for your kind help. > > > > 2010/9/20 mohnish pandey > > Dear Wang, >> The 'relax' process optimizes the structure within >> given geometric constraint. It does'nt care for stress. It minimizes force >> only in the given constraint. If you want to minimize the stress also you >> have to do 'vc-relax' calculation, It will give you what you want. >> >> On Mon, Sep 20, 2010 at 11:37 AM, Wang Di wrote: >> >>> Dear all >>> >>> I perform the relax calculate to the crystal structure. After 11 step >>> bfgs, the total force and total scf correction had reach to the criterion, >>> however, the total stress was so big P=-3443.87. I have no idea about deal >>> with the case, could anyone give some advices? Thanks in advance. >>> >>> the input file >>> : &CONTROL >>> calculation = 'scf' , >>> restart_mode = 'from_scratch' , >>> wf_collect = .true. , >>> outdir = '/lustre/AIOFM/gxgu/workLCAF/5/temp/' , >>> wfcdir = '/lustre/AIOFM/gxgu/workLCAF/5/wftemp/' , >>> pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , >>> prefix = 'LCAF' , >>> etot_conv_thr = 1.0D-7 , >>> forc_conv_thr = 1.0D-6 , >>> nstep = 100 , >>> tstress = .true. , >>> tprnfor = .true. , >>> dipfield = .true. , >>> / >>> &SYSTEM >>> ibrav = 4, >>> celldm(1) = 9.42432151, >>> celldm(3) = 1.94804220, >>> nat = 18, >>> ntyp = 4, >>> ecutwfc = 30 , >>> ecutrho = 120 , >>> nosym = .false. , >>> / >>> &ELECTRONS >>> electron_maxstep = 100, >>> conv_thr = 1.0D-10 , >>> / >>> ATOMIC_SPECIES >>> Li 6.94100 03-Li.LDA.fhi.UPF >>> Al 26.98154 13-Al.LDA.fhi.UPF >>> Ca 40.07800 20-Ca.LDA.fhi.UPF >>> F 18.99840 09-F.LDA.fhi.UPF >>> ATOMIC_POSITIONS crystal >>> Li 0.333333333 0.666666667 0.250000000 >>> Li -0.333333333 -0.666666667 -0.250000000 >>> F 0.389997 0.045970 0.148326 >>> F -0.045970 0.344027 0.148326 >>> F -0.344027 -0.389997 0.148326 >>> F -0.045970 -0.389997 0.351674 >>> F -0.344027 0.045970 0.351674 >>> F 0.389997 0.344027 0.351674 >>> F -0.389997 -0.045970 -0.148326 >>> F 0.045970 -0.344027 -0.148326 >>> F 0.344027 0.389997 -0.148326 >>> F 0.045970 0.389997 0.648326 >>> F 0.344027 -0.045970 0.648326 >>> F -0.389997 -0.344027 0.648326 >>> Al 0.666666667 0.333333333 0.250000000 >>> Al -0.666666667 -0.333333333 -0.250000000 >>> Ca 0.000000000 0.000000000 0.000000000 >>> Ca 0.000000000 0.000000000 0.500000000 >>> K_POINTS automatic >>> 16 16 16 0 0 0 >>> >>> best regart >>> >>> Wang Di >>> >>> -- >>> ----------------------------------------------------------------- >>> *Wang Di** >>> The Crystal Lab,* >>> *AnHui Institute of Optics and Fine Mechanics, *** >>> *Chinese Academy of Sciences,** >>> No350. , Shushanhu Road, Hefei, 230031, China**, >>> E-mail: didi5158 @gmail.ocm** ,* >>> ----------------------------------------------------------------- >>> >>> _______________________________________________ >>> Pw_forum mailing list >>> Pw_forum at pwscf.org >>> http://www.democritos.it/mailman/listinfo/pw_forum >>> >>> >> >> >> -- >> Regards, >> MOHNISH, >> ----------------------------------------------------------------- >> Mohnish Pandey >> Y6927262,5th Year dual degree student, >> Department of Chemical Engineering, >> IIT KANPUR, UP, INDIA >> +919235721300 >> ----------------------------------------------------------------- >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> > > > -- > ----------------------------------------------------------------- > *Wang Di** > The Crystal Lab,* > *AnHui Institute of Optics and Fine Mechanics, *** > *Chinese Academy of Sciences,** > No350. , Shushanhu Road, Hefei, 230031, China**, > E-mail: didi5158 @gmail.ocm** ,* > ----------------------------------------------------------------- > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ____________________________________ Hui Wang School of physics, Fudan University, Shanghai, China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/8b778c8f/attachment-0001.htm From didi5158 at gmail.com Mon Sep 20 12:14:07 2010 From: didi5158 at gmail.com (Wang Di) Date: Mon, 20 Sep 2010 18:14:07 +0800 Subject: [Pw_forum] question about the stress in relax process In-Reply-To: References: Message-ID: I am very sorry that the input file was wrong, the relax input file is the similar to the file. &CONTROL calculation = 'relax' , restart_mode = 'from_scratch' , wf_collect = .true. , outdir = '/lustre/AIOFM/gxgu/workLCAF/6/temp/' , wfcdir = '/lustre/AIOFM/gxgu/workLCAF/6/wftemp/' , pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , prefix = 'LCAF' , etot_conv_thr = 1.0D-6 , forc_conv_thr = 1.0D-4 , nstep = 100 , tstress = .true. , tprnfor = .true. , dipfield = .true. , / &SYSTEM ibrav = 4, celldm(1) = 9.4264, celldm(3) = 1.9287, nat = 18, ntyp = 4, ecutwfc = 30 , ecutrho = 120 , nosym = .false. , / &ELECTRONS electron_maxstep = 100, conv_thr = 1.0D-8 , / &IONs ion_dynamics = 'bfgs' , / ATOMIC_SPECIES Li 6.94100 03-Li.LDA.fhi.UPF Al 26.98154 13-Al.LDA.fhi.UPF Ca 40.07800 20-Ca.LDA.fhi.UPF F 18.99840 09-F.LDA.fhi.UPF ATOMIC_POSITIONS crystal Li 0.333333333 0.666666667 0.250000000 Li -0.333333333 -0.666666667 -0.250000000 F 0.375499994 0.031899999 0.143800005 F -0.031899999 0.343599993 0.143800005 F -0.343599993 -0.375499993 0.143800005 F -0.031899999 -0.375499993 0.356199994 F -0.343599993 0.031899999 0.356199994 F 0.375499993 0.343599993 0.356199994 F -0.375499993 -0.031899998 -0.143800005 F 0.031899999 -0.343599993 -0.143800005 F 0.343599993 0.375499993 -0.143800005 F 0.031899999 0.375499993 0.643800005 F 0.343599993 -0.031899999 0.643800005 F -0.375499993 -0.343599993 0.643800005 Al 0.666666667 0.333333333 0.250000000 Al -0.666666667 -0.333333333 -0.250000000 Ca 0.000000000 0.000000000 0.000000000 Ca 0.000000000 0.000000000 0.500000000 K_POINTS automatic 16 16 16 0 0 0 2010/9/20 mohnish pandey > Dear Wang, > Are you sure that this is your input file? Because the > calculation flag says it is "scf" ..Which will not definitely do the relax > calculation.... > > > On Mon, Sep 20, 2010 at 3:32 PM, Wang Di wrote: > >> Dear Mohnish: >> I have used the 'vc-relax' process optimized the cell parameter, however, >> the result were seem to deviate from the initial value (experimental value) >> too much.The aim of the structure optimizes is for calculating the Raman >> spectra of the structure, I think the bad structure ( not coincide to the >> experiment) is hard to got the true spectra. So I don't how to deal with the >> case. According to your experience, what should I do for the next step? >> >> Thank you, dear Mohnish, for your kind help. >> >> >> >> 2010/9/20 mohnish pandey >> >> Dear Wang, >>> The 'relax' process optimizes the structure within >>> given geometric constraint. It does'nt care for stress. It minimizes force >>> only in the given constraint. If you want to minimize the stress also you >>> have to do 'vc-relax' calculation, It will give you what you want. >>> >>> On Mon, Sep 20, 2010 at 11:37 AM, Wang Di wrote: >>> >>>> Dear all >>>> >>>> I perform the relax calculate to the crystal structure. After 11 step >>>> bfgs, the total force and total scf correction had reach to the criterion, >>>> however, the total stress was so big P=-3443.87. I have no idea about deal >>>> with the case, could anyone give some advices? Thanks in advance. >>>> >>>> the input file >>>> : &CONTROL >>>> calculation = 'scf' , >>>> restart_mode = 'from_scratch' , >>>> wf_collect = .true. , >>>> outdir = '/lustre/AIOFM/gxgu/workLCAF/5/temp/' , >>>> wfcdir = '/lustre/AIOFM/gxgu/workLCAF/5/wftemp/' , >>>> pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , >>>> prefix = 'LCAF' , >>>> etot_conv_thr = 1.0D-7 , >>>> forc_conv_thr = 1.0D-6 , >>>> nstep = 100 , >>>> tstress = .true. , >>>> tprnfor = .true. , >>>> dipfield = .true. , >>>> / >>>> &SYSTEM >>>> ibrav = 4, >>>> celldm(1) = 9.42432151, >>>> celldm(3) = 1.94804220, >>>> nat = 18, >>>> ntyp = 4, >>>> ecutwfc = 30 , >>>> ecutrho = 120 , >>>> nosym = .false. , >>>> / >>>> &ELECTRONS >>>> electron_maxstep = 100, >>>> conv_thr = 1.0D-10 , >>>> / >>>> ATOMIC_SPECIES >>>> Li 6.94100 03-Li.LDA.fhi.UPF >>>> Al 26.98154 13-Al.LDA.fhi.UPF >>>> Ca 40.07800 20-Ca.LDA.fhi.UPF >>>> F 18.99840 09-F.LDA.fhi.UPF >>>> ATOMIC_POSITIONS crystal >>>> Li 0.333333333 0.666666667 0.250000000 >>>> Li -0.333333333 -0.666666667 -0.250000000 >>>> F 0.389997 0.045970 0.148326 >>>> F -0.045970 0.344027 0.148326 >>>> F -0.344027 -0.389997 0.148326 >>>> F -0.045970 -0.389997 0.351674 >>>> F -0.344027 0.045970 0.351674 >>>> F 0.389997 0.344027 0.351674 >>>> F -0.389997 -0.045970 -0.148326 >>>> F 0.045970 -0.344027 -0.148326 >>>> F 0.344027 0.389997 -0.148326 >>>> F 0.045970 0.389997 0.648326 >>>> F 0.344027 -0.045970 0.648326 >>>> F -0.389997 -0.344027 0.648326 >>>> Al 0.666666667 0.333333333 0.250000000 >>>> Al -0.666666667 -0.333333333 -0.250000000 >>>> Ca 0.000000000 0.000000000 0.000000000 >>>> Ca 0.000000000 0.000000000 0.500000000 >>>> K_POINTS automatic >>>> 16 16 16 0 0 0 >>>> >>>> best regart >>>> >>>> Wang Di >>>> >>>> -- >>>> ----------------------------------------------------------------- >>>> *Wang Di** >>>> The Crystal Lab,* >>>> *AnHui Institute of Optics and Fine Mechanics, *** >>>> *Chinese Academy of Sciences,** >>>> No350. , Shushanhu Road, Hefei, 230031, China**, >>>> E-mail: didi5158 @gmail.ocm** ,* >>>> ----------------------------------------------------------------- >>>> >>>> _______________________________________________ >>>> Pw_forum mailing list >>>> Pw_forum at pwscf.org >>>> http://www.democritos.it/mailman/listinfo/pw_forum >>>> >>>> >>> >>> >>> -- >>> Regards, >>> MOHNISH, >>> ----------------------------------------------------------------- >>> Mohnish Pandey >>> Y6927262,5th Year dual degree student, >>> Department of Chemical Engineering, >>> IIT KANPUR, UP, INDIA >>> +919235721300 >>> ----------------------------------------------------------------- >>> >>> _______________________________________________ >>> Pw_forum mailing list >>> Pw_forum at pwscf.org >>> http://www.democritos.it/mailman/listinfo/pw_forum >>> >>> >> >> >> -- >> ----------------------------------------------------------------- >> *Wang Di** >> The Crystal Lab,* >> *AnHui Institute of Optics and Fine Mechanics, *** >> *Chinese Academy of Sciences,** >> No350. , Shushanhu Road, Hefei, 230031, China**, >> E-mail: didi5158 @gmail.ocm** ,* >> ----------------------------------------------------------------- >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> > > > -- > Regards, > MOHNISH, > ----------------------------------------------------------------- > Mohnish Pandey > Y6927262,5th Year dual degree student, > Department of Chemical Engineering, > IIT KANPUR, UP, INDIA > +919235721300 > ----------------------------------------------------------------- > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ----------------------------------------------------------------- *Wang Di** The Crystal Lab,* *AnHui Institute of Optics and Fine Mechanics, *** *Chinese Academy of Sciences,** No350. , Shushanhu Road, Hefei, 230031, China**, E-mail: didi5158 @gmail.ocm** ,* ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/350a198d/attachment.htm From didi5158 at gmail.com Mon Sep 20 12:18:20 2010 From: didi5158 at gmail.com (Wang Di) Date: Mon, 20 Sep 2010 18:18:20 +0800 Subject: [Pw_forum] question about the stress in relax process In-Reply-To: References: Message-ID: *Thanks for your advices! * > ----------------------------------------------------------------- >> *Wang Di** >> The Crystal Lab,* >> *AnHui Institute of Optics and Fine Mechanics, *** >> *Chinese Academy of Sciences,** >> No350. , Shushanhu Road, Hefei, 230031, China**, >> E-mail: didi5158 @gmail.ocm** ,* >> ----------------------------------------------------------------- >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> > > > -- > ____________________________________ > Hui Wang > School of physics, Fudan University, Shanghai, China > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ----------------------------------------------------------------- *Wang Di** The Crystal Lab,* *AnHui Institute of Optics and Fine Mechanics, *** *Chinese Academy of Sciences,** No350. , Shushanhu Road, Hefei, 230031, China**, E-mail: didi5158 @gmail.ocm** ,* ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/3521bd1d/attachment-0001.htm From lfhuang at theory.issp.ac.cn Mon Sep 20 11:59:53 2010 From: lfhuang at theory.issp.ac.cn (=?utf-8?B?bGZodWFuZw==?=) Date: Mon, 20 Sep 2010 17:59:53 +0800 Subject: [Pw_forum] =?utf-8?q?question_about_the_stress_in_relax_process?= Message-ID: <20100920095954.20178.qmail@ms.hfcas.ac.cn> Dear Wang Di: Are you using Norm-Conserving pseudopotentials? If you are, please increase the ecutwfc to be about 50 Ry or larger, and the density of the k mesh should be high enough (by tests). This is because the accuracy of the calculation of stress is much dependent on the completeness of the plane wave basis set. In addition, the forc_conv_thr=1.0d-4 already be great (usually 1.0d-3 is used), smaller thresholds (e.g. 1.0d-6) won't make any difference and only waste much of your invaluable time. And I am very curious about why did you use "dipfield = .true." for your 3D-periodic crystalline system? PS: we are working at the same island (science island), but two institutes, calculating in the same super-computer. I am very glad to meet you on the same Forum :) Best Wishes! Yours Sincerely L. F. Huang > Date: Mon, 20 Sep 2010 14:07:30 +0800 > From: Wang Di > Subject: [Pw_forum] question about the stress in relax process > To: PWSCF Forum > Message-ID: > > Content-Type: text/plain; charset="iso-8859-1" > > Dear all > > I perform the relax calculate to the crystal structure. After 11 step bfgs, > the total force and total scf correction had reach to the criterion, > however, the total stress was so big P=-3443.87. I have no idea about deal > with the case, could anyone give some advices? Thanks in advance. > > the input file > : &CONTROL > calculation = 'scf' , > restart_mode = 'from_scratch' , > wf_collect = .true. , > outdir = '/lustre/AIOFM/gxgu/workLCAF/5/temp/' , > wfcdir = '/lustre/AIOFM/gxgu/workLCAF/5/wftemp/' , > pseudo_dir = '/home/AIOFM/gxgu/work/NCPP/' , > prefix = 'LCAF' , > etot_conv_thr = 1.0D-7 , > forc_conv_thr = 1.0D-6 , > nstep = 100 , > tstress = .true. , > tprnfor = .true. , > dipfield = .true. , > / > &SYSTEM > ibrav = 4, > celldm(1) = 9.42432151, > celldm(3) = 1.94804220, > nat = 18, > ntyp = 4, > ecutwfc = 30 , > ecutrho = 120 , > nosym = .false. , > / > &ELECTRONS > electron_maxstep = 100, > conv_thr = 1.0D-10 , > / > ATOMIC_SPECIES > Li 6.94100 03-Li.LDA.fhi.UPF > Al 26.98154 13-Al.LDA.fhi.UPF > Ca 40.07800 20-Ca.LDA.fhi.UPF > F 18.99840 09-F.LDA.fhi.UPF > ATOMIC_POSITIONS crystal > Li 0.333333333 0.666666667 0.250000000 > Li -0.333333333 -0.666666667 -0.250000000 > F 0.389997 0.045970 0.148326 > F -0.045970 0.344027 0.148326 > F -0.344027 -0.389997 0.148326 > F -0.045970 -0.389997 0.351674 > F -0.344027 0.045970 0.351674 > F 0.389997 0.344027 0.351674 > F -0.389997 -0.045970 -0.148326 > F 0.045970 -0.344027 -0.148326 > F 0.344027 0.389997 -0.148326 > F 0.045970 0.389997 0.648326 > F 0.344027 -0.045970 0.648326 > F -0.389997 -0.344027 0.648326 > Al 0.666666667 0.333333333 0.250000000 > Al -0.666666667 -0.333333333 -0.250000000 > Ca 0.000000000 0.000000000 0.000000000 > Ca 0.000000000 0.000000000 0.500000000 > K_POINTS automatic > 16 16 16 0 0 0 > > best regart > > Wang Di > > -- > ----------------------------------------------------------------- > *Wang Di** > The Crystal Lab,* > *AnHui Institute of Optics and Fine Mechanics, *** > *Chinese Academy of Sciences,** > No350. , Shushanhu Road, Hefei, 230031, China**, > E-mail: didi5158 @gmail.ocm** ,* > ----------------------------------------------------------------- > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/df91893a/attachment-0001.htm > > ------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > End of Pw_forum Digest, Vol 39, Issue 31 > **************************************** > ------ ====================================================================== L.F.Huang(???) DFT and phonon physics ====================================================================== Add: Research Laboratory for Computational Materials Sciences, Instutue of Solid State Physics,the Chinese Academy of Sciences, P.O.Box 1129, Hefei 230031, P.R.China Tel: 86-551-5591464-326(office) Fax: 86-551-5591434 Our group: http://theory.issp.ac.cn ====================================================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100920/7b0adea6/attachment.htm From quantumdft at gmail.com Mon Sep 20 18:48:58 2010 From: quantumdft at gmail.com (vega lew) Date: Tue, 21 Sep 2010 00:48:58 +0800 Subject: [Pw_forum] Do you think we could fix certain bond length during the relax calculation In-Reply-To: <4A636B18-C407-4387-A0DA-A6D5AC70EC1D@democritos.it> References: <4C94F735.9070902@gmail.com> <4A636B18-C407-4387-A0DA-A6D5AC70EC1D@democritos.it> Message-ID: <4C97907A.1000208@gmail.com> Dear Paolo, I found when I specified BFGS method, the 'relax' calculation also could run smoothly. Do you think the BFGS method rather than Damped MD is OK for the 'relax' calculation with 'distance' constrains? best wishes, vega On 09/19/10 17:03, Paolo Giannozzi wrote: > On Sep 18, 2010, at 19:30 , vega lew wrote: > >> But I found the Q-E have no direct command to control the >> bond length of the system. > there is an implementation of contrained optimization, > using damped MD (not bfgs). See card CONTRAINTS > and file Doc/constraints_HOWTO.tex > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- ================================================ Vega Lew ( weijia liu) Graduate student State Key Laboratory of Materials-oriented Chemical Engineering Nanjing University of Technology, 210009, Nanjing, China *************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, China *************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100921/10707a83/attachment.htm From modaresi.mohsen at gmail.com Tue Sep 21 09:41:52 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Tue, 21 Sep 2010 12:11:52 +0430 Subject: [Pw_forum] Young's Modulus Message-ID: Dear all, Can Q. E calculate Young's Modulus? Thanks for your reply -- Mohsen Modarresi, PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. Phone +98-9133452131 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100921/42f966f7/attachment.htm From baroni at sissa.it Tue Sep 21 13:39:33 2010 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 21 Sep 2010 13:39:33 +0200 Subject: [Pw_forum] Young's Modulus In-Reply-To: References: Message-ID: <97B125B7-A57F-40A2-ADE9-C17B17ADCF3D@sissa.it> Mohsen: I think you can answer by yourself to your question. Answer this question first: 1) How is the Young's modulus defined? Ask then yourself the following question: Can the ingredients entering the answer to question #1 be calculated by QE? Please, revert to us with the answer to the above chain of questions. SB On Sep 21, 2010, at 9:41 AM, mohsen modaresi wrote: > Dear all, > Can Q. E calculate Young's Modulus? > Thanks for your reply > > -- > Mohsen Modarresi, > PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. > Phone +98-9133452131 > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100921/8ac169ef/attachment.htm From somesh.kb at gmail.com Tue Sep 21 13:57:15 2010 From: somesh.kb at gmail.com (Somesh Kumar Bhattacharya) Date: Tue, 21 Sep 2010 13:57:15 +0200 Subject: [Pw_forum] prob with cp-wf Message-ID: Dear All, I am trying to calculate the wannier function with CP code using QE-4.2.1 My system has 64 H2O molecules and an excess electron. The input file is : &SYSTEM ibrav= 0, celldm(1) = 1.0, nat= 192, ntyp= 2, nspin = 1, ecutwfc =80.0, ecfixed = 68.0, q2sigma = 8.0, nbnd=513, tot_charge = -1, / &ELECTRONS emass = 200.d0, emass_cutoff = 2.5d0, orthogonalization = 'ortho', ortho_eps = 5.d-8, ortho_max = 28, electron_dynamics = 'damp' / &IONS ion_dynamics = 'none' ion_radius(1) = 0.8d0, ion_radius(2) = 0.8d0, / &WANNIER adapt = .true., nsteps = 100, / ATOMIC_SPECIES O 15.9990 O.blyp-mt.UPF H 1.0008 H.blyp-vbc.UPF ATOMIC_POSITIONS (angstrom) O 0.10050E+02 0.76609E+01 0.53241E+01 1 1 1 1 O 0.14681E+02 0.43002E+01 0.26406E+01 1 1 1 1 O 0.16186E+02 0.10293E+02 0.10190E+02 1 1 1 1 O 0.76912E+01 0.12696E+02 0.58596E+01 1 1 1 1 O 0.10611E+02 0.11355E+02 0.23346E+01 1 1 1 1 O 0.72755E+01 0.49979E+01 0.60073E+01 1 1 1 1 O 0.60692E+01 0.11413E+02 0.76575E+01 1 1 1 1 O 0.58212E+01 0.35473E+01 0.26764E+01 1 1 1 1 O 0.16143E+02 0.52818E+01 0.10493E+02 1 1 1 1 O 0.11928E+02 0.40990E+01 0.62768E+01 1 1 1 1 O 0.59251E+01 0.11427E+02 0.10525E+02 1 1 1 1 O 0.62330E+01 0.15141E+01 0.10873E+02 1 1 1 1 O 0.13516E+02 0.39298E+01 0.10963E+02 1 1 1 1 O 0.75307E+01 0.10319E+02 -0.21023E+00 1 1 1 1 O 0.61103E+01 0.26737E+01 0.53330E+01 1 1 1 1 O 0.39421E+01 0.54188E+01 0.23858E+01 1 1 1 1 O 0.98251E+01 0.13416E+02 0.72376E+01 1 1 1 1 O 0.11775E+02 0.13821E+02 0.27902E+01 1 1 1 1 O 0.40894E+01 0.68185E+01 -0.50521E+00 1 1 1 1 O 0.70352E+01 0.77315E+01 -0.67776E+00 1 1 1 1 O 0.13032E+02 0.11912E+02 0.54367E+01 1 1 1 1 O 0.79946E+01 0.83715E+01 0.37856E+01 1 1 1 1 O 0.94996E+01 0.31317E+01 0.88059E+01 1 1 1 1 O 0.10342E+02 0.30701E+01 0.11860E+01 1 1 1 1 O 0.13662E+02 0.13606E+02 0.11005E+02 1 1 1 1 O 0.38246E+01 0.74449E+01 0.39497E+01 1 1 1 1 O 0.16573E+02 0.12982E+02 0.70078E+01 1 1 1 1 O 0.14325E+02 0.16642E+01 0.27419E+01 1 1 1 1 O 0.12619E+02 0.44514E+01 0.74222E+00 1 1 1 1 O 0.12217E+02 0.78936E+01 0.38662E+01 1 1 1 1 O 0.38544E+01 0.10165E+02 0.33184E+01 1 1 1 1 O 0.55168E+01 0.13913E+02 0.98810E+00 1 1 1 1 O 0.14958E+02 0.10728E+02 0.14855E+01 1 1 1 1 O 0.14063E+02 0.82235E+01 0.75328E+01 1 1 1 1 O 0.96928E+01 0.50255E+01 0.49634E+01 1 1 1 1 O 0.11058E+02 0.12569E+02 0.10845E+02 1 1 1 1 O 0.10019E+02 0.84837E+01 0.80330E+01 1 1 1 1 O 0.15473E+02 0.12464E+02 0.46456E+01 1 1 1 1 O 0.49611E+01 0.35401E+01 0.93536E+01 1 1 1 1 O 0.10158E+02 0.10677E+02 -0.16527E+00 1 1 1 1 O 0.73662E+01 0.46998E+01 0.86410E+01 1 1 1 1 O 0.15119E+02 0.89701E+01 -0.56502E+00 1 1 1 1 O 0.11612E+02 0.14173E+02 0.54365E+01 1 1 1 1 O 0.12117E+02 0.74237E+01 0.93207E+01 1 1 1 1 O 0.11520E+02 0.69213E+01 0.13692E+01 1 1 1 1 O 0.11904E+02 0.46856E+01 0.89255E+01 1 1 1 1 O 0.12867E+02 0.10408E+02 0.32626E+01 1 1 1 1 O 0.15391E+02 0.13102E+02 0.54147E+00 1 1 1 1 O 0.81972E+01 0.11085E+02 0.35952E+01 1 1 1 1 O 0.13291E+02 0.10721E+02 0.76854E+01 1 1 1 1 O 0.92727E+01 0.20574E+01 0.11342E+02 1 1 1 1 O 0.61288E+01 0.11462E+02 0.18765E+01 1 1 1 1 O 0.12226E+02 0.85331E+01 -0.69902E+00 1 1 1 1 O 0.86513E+01 0.70391E+01 0.15026E+01 1 1 1 1 O 0.57127E+01 0.86340E+01 0.78603E+01 1 1 1 1 O 0.84309E+01 0.44819E+01 0.24723E+01 1 1 1 1 O 0.14508E+02 0.43435E+01 0.54487E+01 1 1 1 1 O 0.14382E+02 0.70165E+01 0.51478E+01 1 1 1 1 O 0.60352E+01 0.73393E+01 0.53135E+01 1 1 1 1 O 0.38800E+01 0.28723E+01 0.69386E+01 1 1 1 1 O 0.76488E+01 0.73864E+01 0.91298E+01 1 1 1 1 O 0.10756E+02 0.11160E+02 0.84779E+01 1 1 1 1 O 0.16357E+02 0.79793E+01 0.87477E+01 1 1 1 1 O 0.14930E+02 0.12304E+02 0.90595E+01 1 1 1 1 H 0.92869E+01 0.80362E+01 0.47862E+01 1 1 1 1 H 0.11409E+02 0.78848E+01 0.44717E+01 1 1 1 1 H 0.99546E+01 0.66663E+01 0.52326E+01 1 1 1 1 H 0.10054E+02 0.82394E+01 0.70652E+01 1 1 1 1 H 0.15576E+02 0.48351E+01 0.24255E+01 1 1 1 1 H 0.13314E+02 0.43563E+01 0.14654E+01 1 1 1 1 H 0.14527E+02 0.44368E+01 0.36066E+01 1 1 1 1 H 0.14541E+02 0.26524E+01 0.27434E+01 1 1 1 1 H 0.15402E+02 0.96043E+01 0.11348E+02 1 1 1 1 H 0.50370E+01 0.10996E+02 0.10569E+02 1 1 1 1 H 0.16268E+02 0.94998E+01 0.95968E+01 1 1 1 1 H 0.15699E+02 0.11052E+02 0.97368E+01 1 1 1 1 H 0.85189E+01 0.12976E+02 0.63556E+01 1 1 1 1 H 0.79306E+01 0.12146E+02 0.50636E+01 1 1 1 1 H 0.67294E+01 0.11848E+02 0.70324E+01 1 1 1 1 H 0.67894E+01 0.14390E+02 0.55014E+01 1 1 1 1 H 0.11426E+02 0.12910E+02 0.25605E+01 1 1 1 1 H 0.10357E+02 0.10938E+02 0.78748E+00 1 1 1 1 H 0.11369E+02 0.10824E+02 0.27242E+01 1 1 1 1 H 0.97614E+01 0.11271E+02 0.28637E+01 1 1 1 1 H 0.88101E+01 0.49465E+01 0.54490E+01 1 1 1 1 H 0.72651E+01 0.49616E+01 0.70094E+01 1 1 1 1 H 0.67752E+01 0.58133E+01 0.56959E+01 1 1 1 1 H 0.66659E+01 0.34677E+01 0.55799E+01 1 1 1 1 H 0.62833E+01 0.11559E+02 0.86118E+01 1 1 1 1 H 0.45358E+01 0.12282E+02 0.72811E+01 1 1 1 1 H 0.58247E+01 0.96207E+01 0.77476E+01 1 1 1 1 H 0.74973E+01 0.41139E+01 0.25379E+01 1 1 1 1 H 0.51133E+01 0.42967E+01 0.25845E+01 1 1 1 1 H 0.57914E+01 0.31861E+01 0.36022E+01 1 1 1 1 H 0.54639E+01 0.21426E+01 0.15556E+01 1 1 1 1 H 0.16520E+02 0.60752E+01 0.11283E+02 1 1 1 1 H 0.15251E+02 0.48439E+01 0.10646E+02 1 1 1 1 H 0.39571E+01 0.44434E+01 0.10134E+02 1 1 1 1 H 0.16401E+02 0.71265E+01 0.92266E+01 1 1 1 1 H 0.11150E+02 0.45127E+01 0.58040E+01 1 1 1 1 H 0.11934E+02 0.43354E+01 0.72499E+01 1 1 1 1 H 0.13599E+02 0.41417E+01 0.57977E+01 1 1 1 1 H 0.11698E+02 0.25024E+01 0.57457E+01 1 1 1 1 H 0.65437E+01 0.10979E+02 0.11180E+02 1 1 1 1 H 0.62370E+01 0.13118E+02 0.10708E+02 1 1 1 1 H 0.72059E+01 0.16370E+01 0.11002E+02 1 1 1 1 H 0.58739E+01 0.14096E+02 0.85167E-01 1 1 1 1 H 0.53187E+01 0.27474E+01 0.98564E+01 1 1 1 1 H 0.13229E+02 0.41775E+01 -0.77408E+00 1 1 1 1 H 0.13551E+02 0.29344E+01 0.10952E+02 1 1 1 1 H 0.12529E+02 0.43453E+01 0.96318E+01 1 1 1 1 H 0.85426E+01 0.10454E+02 -0.20510E+00 1 1 1 1 H 0.71455E+01 0.87188E+01 -0.50977E+00 1 1 1 1 H 0.70889E+01 0.10691E+02 0.60263E+00 1 1 1 1 H 0.46458E+01 0.28817E+01 0.62852E+01 1 1 1 1 H 0.40182E+01 0.57681E+01 0.14764E+01 1 1 1 1 H 0.38960E+01 0.66285E+01 0.32547E+01 1 1 1 1 H 0.95539E+01 0.23147E+01 0.82241E+01 1 1 1 1 H 0.10489E+02 0.13665E+02 0.65203E+01 1 1 1 1 H 0.10141E+02 0.12578E+02 0.76933E+01 1 1 1 1 H 0.10698E+02 0.23546E+01 0.17885E+01 1 1 1 1 H 0.12579E+02 0.13178E+01 0.27062E+01 1 1 1 1 H 0.11514E+02 0.14172E+02 0.44369E+01 1 1 1 1 H 0.50299E+01 0.71540E+01 -0.53941E+00 1 1 1 1 H 0.16274E+02 0.77607E+01 -0.50468E+00 1 1 1 1 H 0.81302E+01 0.72294E+01 0.67181E+00 1 1 1 1 H 0.72421E+01 0.75867E+01 0.11032E+02 1 1 1 1 H 0.14554E+02 0.12212E+02 0.49822E+01 1 1 1 1 H 0.12522E+02 0.12771E+02 0.55111E+01 1 1 1 1 H 0.12977E+02 0.10989E+02 0.40789E+01 1 1 1 1 H 0.13069E+02 0.11428E+02 0.63226E+01 1 1 1 1 H 0.81050E+01 0.10097E+02 0.37381E+01 1 1 1 1 H 0.81547E+01 0.79005E+01 0.29180E+01 1 1 1 1 H 0.72104E+01 0.79837E+01 0.42850E+01 1 1 1 1 H 0.94063E+01 0.26410E+01 0.10552E+02 1 1 1 1 H 0.10345E+02 0.36559E+01 0.87763E+01 1 1 1 1 H 0.81383E+01 0.40646E+01 0.87539E+01 1 1 1 1 H 0.11798E+02 0.38945E+01 0.92349E+00 1 1 1 1 H 0.96193E+01 0.35809E+01 0.16635E+01 1 1 1 1 H 0.96635E+01 0.24733E+01 -0.45115E+00 1 1 1 1 H 0.12738E+02 0.13219E+02 0.11038E+02 1 1 1 1 H 0.14284E+02 0.13378E+02 -0.81055E+00 1 1 1 1 H 0.14491E+02 0.12930E+02 0.97192E+01 1 1 1 1 H 0.38965E+01 0.91982E+01 0.36044E+01 1 1 1 1 H 0.15856E+02 0.74852E+01 0.43660E+01 1 1 1 1 H 0.51353E+01 0.73381E+01 0.48137E+01 1 1 1 1 H 0.16544E+02 0.20901E+01 0.69498E+01 1 1 1 1 H 0.16223E+02 0.12770E+02 0.60809E+01 1 1 1 1 H 0.15500E+02 0.12731E+02 0.83546E+01 1 1 1 1 H 0.15291E+02 0.13196E+02 0.39835E+01 1 1 1 1 H 0.14840E+02 0.13877E+02 0.18832E+01 1 1 1 1 H 0.11912E+02 0.60290E+01 0.11531E+01 1 1 1 1 H 0.11819E+02 0.72121E+01 0.22820E+01 1 1 1 1 H 0.12480E+02 0.88403E+01 0.36626E+01 1 1 1 1 H 0.13564E+02 0.72374E+01 0.45975E+01 1 1 1 1 H 0.15699E+02 0.10526E+02 0.21594E+01 1 1 1 1 H 0.16399E+02 0.10978E+02 0.39319E+01 1 1 1 1 H 0.53648E+01 0.11011E+02 0.23101E+01 1 1 1 1 H 0.34596E+01 0.13170E+02 0.60397E+00 1 1 1 1 H 0.58772E+01 0.12396E+02 0.15901E+01 1 1 1 1 H 0.14990E+02 0.10027E+02 0.77363E+00 1 1 1 1 H 0.13592E+02 0.10568E+02 0.25894E+01 1 1 1 1 H 0.15281E+02 0.12136E+02 0.86419E+00 1 1 1 1 H 0.12891E+02 0.77016E+01 0.87494E+01 1 1 1 1 H 0.13612E+02 0.97653E+01 0.76083E+01 1 1 1 1 H 0.14111E+02 0.77803E+01 0.66252E+01 1 1 1 1 H 0.14969E+02 0.81163E+01 0.79451E+01 1 1 1 1 H 0.94604E+01 0.47722E+01 0.40302E+01 1 1 1 1 H 0.10437E+02 0.11415E+02 0.11883E+02 1 1 1 1 H 0.10411E+02 0.13330E+02 0.10986E+02 1 1 1 1 H 0.10895E+02 0.12106E+02 0.99767E+01 1 1 1 1 H 0.10795E+02 0.80582E+01 0.85021E+01 1 1 1 1 H 0.85166E+01 0.77592E+01 0.87930E+01 1 1 1 1 H 0.10273E+02 0.10294E+02 0.83887E+01 1 1 1 1 H 0.65202E+01 0.43007E+01 0.89727E+01 1 1 1 1 H 0.45505E+01 0.32707E+01 0.84554E+01 1 1 1 1 H 0.11578E+02 0.92806E+01 -0.63868E+00 1 1 1 1 H 0.76180E+01 0.63999E+01 0.89681E+01 1 1 1 1 H 0.14153E+02 0.87817E+01 -0.71366E+00 1 1 1 1 H 0.12173E+02 0.77953E+01 0.10245E+02 1 1 1 1 H 0.11858E+02 0.56699E+01 0.90758E+01 1 1 1 1 H 0.11987E+02 0.79213E+01 0.63115E-01 1 1 1 1 H 0.96337E+01 0.70726E+01 0.13330E+01 1 1 1 1 H 0.74597E+01 0.11328E+02 0.29631E+01 1 1 1 1 H 0.11701E+02 0.10960E+02 0.81966E+01 1 1 1 1 H 0.13916E+02 0.11279E+02 0.82366E+01 1 1 1 1 H 0.83870E+01 0.53998E+01 0.20864E+01 1 1 1 1 H 0.64813E+01 0.82363E+01 0.83833E+01 1 1 1 1 H 0.59144E+01 0.78362E+01 0.61630E+01 1 1 1 1 H 0.43681E+01 0.80285E+01 0.83954E+01 1 1 1 1 H 0.14453E+02 0.60275E+01 0.52899E+01 1 1 1 1 H 0.15223E+02 0.38965E+01 0.59663E+01 1 1 1 1 CELL_PARAMETERS 24.347851385876538 0.36920984085018538 -0.16930057214376279 0.37628734578661066 23.766046075564496 3.59168252649787120E-002 -0.17599055343917119 3.60113432262023428E-002 23.89319163700252 I am trying to calculate the spread of the wannier functions. However, I am facing a prob with segmentation fault. The code starts but stops in the first cp step. Can anybody provide a solution for this? regards Somesh -- Somesh Kr. Bhattacharya Post Doctoral Fellow Room No. 263, Leonardo Building, The Abdus Salam International Centre for Theoretical Physics Strada Costiera, 11 I-34014 Trieste Italy Phone: +39-040-2240399 http://portal.ictp.it/cmsp/members/postdoctoral-fellows/somesh-kumar-bhattacharya/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100921/3275b2f2/attachment-0001.htm From giannozz at democritos.it Tue Sep 21 14:36:22 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 21 Sep 2010 14:36:22 +0200 Subject: [Pw_forum] prob with cp-wf In-Reply-To: References: Message-ID: <4C98A6C6.6020900@democritos.it> Somesh Kumar Bhattacharya wrote: > The input file is : incomplete (&control namelist is missing)... > Can anybody provide a solution for this? ...so nobody can provide any solution. As explained a diverging number of times, "segmentation fault" means nothing but "something very wrong happened". There is NO WAY anybody can say anything sensible about a "segmentation fault" without additional information. First and most important piece of information: * Is the problem reproducible with different compilers / libraries / op.sys / hardware / input options / examples / phases of the moon? P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Tue Sep 21 14:39:04 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 21 Sep 2010 14:39:04 +0200 Subject: [Pw_forum] Do you think we could fix certain bond length during the relax calculation In-Reply-To: <4C97907A.1000208@gmail.com> References: <4C94F735.9070902@gmail.com> <4A636B18-C407-4387-A0DA-A6D5AC70EC1D@democritos.it> <4C97907A.1000208@gmail.com> Message-ID: <4C98A768.2000007@democritos.it> vega lew wrote: > I found when I specified BFGS method, the 'relax' calculation also could > run smoothly. Do you think the BFGS method rather than Damped MD is OK > for the 'relax' calculation with 'distance' constrains? are constraints implemented in bfgs? P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From eyvaz_isaev at yahoo.com Tue Sep 21 14:44:58 2010 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 21 Sep 2010 05:44:58 -0700 (PDT) Subject: [Pw_forum] Young's Modulus In-Reply-To: References: Message-ID: <90545.24036.qm@web65707.mail.ac4.yahoo.com> ________________________________ From: mohsen modaresi >Can Q. E calculate Young's Modulus? No, no ab initio program can calculate E directly. You have to calculate elastic constants and then you can easily get the Young modulus. See papers: Mehl et al., PhysRev.B 41, 10311, 1990 and Mehl et al, in Intermetallic Compounds: Principles and Practice, vol.1, Principles, J.H. Wesbrook and R.L. Fleischer, eds, John Wiley and Sons (London, 1995), ch.9 (pp. 195-210). Search cst-www.nrl.navy.mil/users/mehl/papers/cij452.ps.gz via google. Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100921/7f31c034/attachment.htm From somesh.kb at gmail.com Tue Sep 21 15:09:19 2010 From: somesh.kb at gmail.com (Somesh Kumar Bhattacharya) Date: Tue, 21 Sep 2010 15:09:19 +0200 Subject: [Pw_forum] prob with cp-wf In-Reply-To: <4C98A6C6.6020900@democritos.it> References: <4C98A6C6.6020900@democritos.it> Message-ID: Hi Paolo, Sorry for the input file : The complete one is below : &CONTROL calculation = 'cp-wf', title = 'Amorphous Ice and EE', prefix='amor-ice-ee' ndr = 50, ndw = 51, restart_mode = 'from_scratch' nstep = 1000, iprint = 100, isave = 100, tprnfor = .TRUE., dt = 4.d0 etot_conv_thr = 1.d-5, pseudo_dir= '/gpfs/scratch/userictp/ictpcm16/pseudo/' max_seconds = 21000, / &SYSTEM ibrav= 0, celldm(1) = 1.0, nat= 192, ntyp= 2, nspin = 1, ecutwfc =80.0, ecfixed = 68.0, q2sigma = 8.0, nbnd=513, tot_charge = -1, / &ELECTRONS emass = 200.d0, emass_cutoff = 2.5d0, orthogonalization = 'ortho', ortho_eps = 5.d-8, ortho_max = 28, electron_dynamics = 'damp' / &IONS ion_dynamics = 'none' ion_radius(1) = 0.8d0, ion_radius(2) = 0.8d0, / &WANNIER adapt = .true., nsteps = 100, / ATOMIC_SPECIES O 15.9990 O.blyp-mt.UPF H 1.0008 H.blyp-vbc.UPF ATOMIC_POSITIONS (angstrom) O 0.10050E+02 0.76609E+01 0.53241E+01 1 1 1 1 O 0.14681E+02 0.43002E+01 0.26406E+01 1 1 1 1 O 0.16186E+02 0.10293E+02 0.10190E+02 1 1 1 1 O 0.76912E+01 0.12696E+02 0.58596E+01 1 1 1 1 O 0.10611E+02 0.11355E+02 0.23346E+01 1 1 1 1 O 0.72755E+01 0.49979E+01 0.60073E+01 1 1 1 1 O 0.60692E+01 0.11413E+02 0.76575E+01 1 1 1 1 O 0.58212E+01 0.35473E+01 0.26764E+01 1 1 1 1 O 0.16143E+02 0.52818E+01 0.10493E+02 1 1 1 1 O 0.11928E+02 0.40990E+01 0.62768E+01 1 1 1 1 O 0.59251E+01 0.11427E+02 0.10525E+02 1 1 1 1 O 0.62330E+01 0.15141E+01 0.10873E+02 1 1 1 1 O 0.13516E+02 0.39298E+01 0.10963E+02 1 1 1 1 O 0.75307E+01 0.10319E+02 -0.21023E+00 1 1 1 1 O 0.61103E+01 0.26737E+01 0.53330E+01 1 1 1 1 O 0.39421E+01 0.54188E+01 0.23858E+01 1 1 1 1 O 0.98251E+01 0.13416E+02 0.72376E+01 1 1 1 1 O 0.11775E+02 0.13821E+02 0.27902E+01 1 1 1 1 O 0.40894E+01 0.68185E+01 -0.50521E+00 1 1 1 1 O 0.70352E+01 0.77315E+01 -0.67776E+00 1 1 1 1 O 0.13032E+02 0.11912E+02 0.54367E+01 1 1 1 1 O 0.79946E+01 0.83715E+01 0.37856E+01 1 1 1 1 O 0.94996E+01 0.31317E+01 0.88059E+01 1 1 1 1 O 0.10342E+02 0.30701E+01 0.11860E+01 1 1 1 1 O 0.13662E+02 0.13606E+02 0.11005E+02 1 1 1 1 O 0.38246E+01 0.74449E+01 0.39497E+01 1 1 1 1 O 0.16573E+02 0.12982E+02 0.70078E+01 1 1 1 1 O 0.14325E+02 0.16642E+01 0.27419E+01 1 1 1 1 O 0.12619E+02 0.44514E+01 0.74222E+00 1 1 1 1 O 0.12217E+02 0.78936E+01 0.38662E+01 1 1 1 1 O 0.38544E+01 0.10165E+02 0.33184E+01 1 1 1 1 O 0.55168E+01 0.13913E+02 0.98810E+00 1 1 1 1 O 0.14958E+02 0.10728E+02 0.14855E+01 1 1 1 1 O 0.14063E+02 0.82235E+01 0.75328E+01 1 1 1 1 O 0.96928E+01 0.50255E+01 0.49634E+01 1 1 1 1 O 0.11058E+02 0.12569E+02 0.10845E+02 1 1 1 1 O 0.10019E+02 0.84837E+01 0.80330E+01 1 1 1 1 O 0.15473E+02 0.12464E+02 0.46456E+01 1 1 1 1 O 0.49611E+01 0.35401E+01 0.93536E+01 1 1 1 1 O 0.10158E+02 0.10677E+02 -0.16527E+00 1 1 1 1 O 0.73662E+01 0.46998E+01 0.86410E+01 1 1 1 1 O 0.15119E+02 0.89701E+01 -0.56502E+00 1 1 1 1 O 0.11612E+02 0.14173E+02 0.54365E+01 1 1 1 1 O 0.12117E+02 0.74237E+01 0.93207E+01 1 1 1 1 O 0.11520E+02 0.69213E+01 0.13692E+01 1 1 1 1 O 0.11904E+02 0.46856E+01 0.89255E+01 1 1 1 1 O 0.12867E+02 0.10408E+02 0.32626E+01 1 1 1 1 O 0.15391E+02 0.13102E+02 0.54147E+00 1 1 1 1 O 0.81972E+01 0.11085E+02 0.35952E+01 1 1 1 1 O 0.13291E+02 0.10721E+02 0.76854E+01 1 1 1 1 O 0.92727E+01 0.20574E+01 0.11342E+02 1 1 1 1 O 0.61288E+01 0.11462E+02 0.18765E+01 1 1 1 1 O 0.12226E+02 0.85331E+01 -0.69902E+00 1 1 1 1 O 0.86513E+01 0.70391E+01 0.15026E+01 1 1 1 1 O 0.57127E+01 0.86340E+01 0.78603E+01 1 1 1 1 O 0.84309E+01 0.44819E+01 0.24723E+01 1 1 1 1 O 0.14508E+02 0.43435E+01 0.54487E+01 1 1 1 1 O 0.14382E+02 0.70165E+01 0.51478E+01 1 1 1 1 O 0.60352E+01 0.73393E+01 0.53135E+01 1 1 1 1 O 0.38800E+01 0.28723E+01 0.69386E+01 1 1 1 1 O 0.76488E+01 0.73864E+01 0.91298E+01 1 1 1 1 O 0.10756E+02 0.11160E+02 0.84779E+01 1 1 1 1 O 0.16357E+02 0.79793E+01 0.87477E+01 1 1 1 1 O 0.14930E+02 0.12304E+02 0.90595E+01 1 1 1 1 H 0.92869E+01 0.80362E+01 0.47862E+01 1 1 1 1 H 0.11409E+02 0.78848E+01 0.44717E+01 1 1 1 1 H 0.99546E+01 0.66663E+01 0.52326E+01 1 1 1 1 H 0.10054E+02 0.82394E+01 0.70652E+01 1 1 1 1 H 0.15576E+02 0.48351E+01 0.24255E+01 1 1 1 1 H 0.13314E+02 0.43563E+01 0.14654E+01 1 1 1 1 H 0.14527E+02 0.44368E+01 0.36066E+01 1 1 1 1 H 0.14541E+02 0.26524E+01 0.27434E+01 1 1 1 1 H 0.15402E+02 0.96043E+01 0.11348E+02 1 1 1 1 H 0.50370E+01 0.10996E+02 0.10569E+02 1 1 1 1 H 0.16268E+02 0.94998E+01 0.95968E+01 1 1 1 1 H 0.15699E+02 0.11052E+02 0.97368E+01 1 1 1 1 H 0.85189E+01 0.12976E+02 0.63556E+01 1 1 1 1 H 0.79306E+01 0.12146E+02 0.50636E+01 1 1 1 1 H 0.67294E+01 0.11848E+02 0.70324E+01 1 1 1 1 H 0.67894E+01 0.14390E+02 0.55014E+01 1 1 1 1 H 0.11426E+02 0.12910E+02 0.25605E+01 1 1 1 1 H 0.10357E+02 0.10938E+02 0.78748E+00 1 1 1 1 H 0.11369E+02 0.10824E+02 0.27242E+01 1 1 1 1 H 0.97614E+01 0.11271E+02 0.28637E+01 1 1 1 1 H 0.88101E+01 0.49465E+01 0.54490E+01 1 1 1 1 H 0.72651E+01 0.49616E+01 0.70094E+01 1 1 1 1 H 0.67752E+01 0.58133E+01 0.56959E+01 1 1 1 1 H 0.66659E+01 0.34677E+01 0.55799E+01 1 1 1 1 H 0.62833E+01 0.11559E+02 0.86118E+01 1 1 1 1 H 0.45358E+01 0.12282E+02 0.72811E+01 1 1 1 1 H 0.58247E+01 0.96207E+01 0.77476E+01 1 1 1 1 H 0.74973E+01 0.41139E+01 0.25379E+01 1 1 1 1 H 0.51133E+01 0.42967E+01 0.25845E+01 1 1 1 1 H 0.57914E+01 0.31861E+01 0.36022E+01 1 1 1 1 H 0.54639E+01 0.21426E+01 0.15556E+01 1 1 1 1 H 0.16520E+02 0.60752E+01 0.11283E+02 1 1 1 1 H 0.15251E+02 0.48439E+01 0.10646E+02 1 1 1 1 H 0.39571E+01 0.44434E+01 0.10134E+02 1 1 1 1 H 0.16401E+02 0.71265E+01 0.92266E+01 1 1 1 1 H 0.11150E+02 0.45127E+01 0.58040E+01 1 1 1 1 H 0.11934E+02 0.43354E+01 0.72499E+01 1 1 1 1 H 0.13599E+02 0.41417E+01 0.57977E+01 1 1 1 1 H 0.11698E+02 0.25024E+01 0.57457E+01 1 1 1 1 H 0.65437E+01 0.10979E+02 0.11180E+02 1 1 1 1 H 0.62370E+01 0.13118E+02 0.10708E+02 1 1 1 1 H 0.72059E+01 0.16370E+01 0.11002E+02 1 1 1 1 H 0.58739E+01 0.14096E+02 0.85167E-01 1 1 1 1 H 0.53187E+01 0.27474E+01 0.98564E+01 1 1 1 1 H 0.13229E+02 0.41775E+01 -0.77408E+00 1 1 1 1 H 0.13551E+02 0.29344E+01 0.10952E+02 1 1 1 1 H 0.12529E+02 0.43453E+01 0.96318E+01 1 1 1 1 H 0.85426E+01 0.10454E+02 -0.20510E+00 1 1 1 1 H 0.71455E+01 0.87188E+01 -0.50977E+00 1 1 1 1 H 0.70889E+01 0.10691E+02 0.60263E+00 1 1 1 1 H 0.46458E+01 0.28817E+01 0.62852E+01 1 1 1 1 H 0.40182E+01 0.57681E+01 0.14764E+01 1 1 1 1 H 0.38960E+01 0.66285E+01 0.32547E+01 1 1 1 1 H 0.95539E+01 0.23147E+01 0.82241E+01 1 1 1 1 H 0.10489E+02 0.13665E+02 0.65203E+01 1 1 1 1 H 0.10141E+02 0.12578E+02 0.76933E+01 1 1 1 1 H 0.10698E+02 0.23546E+01 0.17885E+01 1 1 1 1 H 0.12579E+02 0.13178E+01 0.27062E+01 1 1 1 1 H 0.11514E+02 0.14172E+02 0.44369E+01 1 1 1 1 H 0.50299E+01 0.71540E+01 -0.53941E+00 1 1 1 1 H 0.16274E+02 0.77607E+01 -0.50468E+00 1 1 1 1 H 0.81302E+01 0.72294E+01 0.67181E+00 1 1 1 1 H 0.72421E+01 0.75867E+01 0.11032E+02 1 1 1 1 H 0.14554E+02 0.12212E+02 0.49822E+01 1 1 1 1 H 0.12522E+02 0.12771E+02 0.55111E+01 1 1 1 1 H 0.12977E+02 0.10989E+02 0.40789E+01 1 1 1 1 H 0.13069E+02 0.11428E+02 0.63226E+01 1 1 1 1 H 0.81050E+01 0.10097E+02 0.37381E+01 1 1 1 1 H 0.81547E+01 0.79005E+01 0.29180E+01 1 1 1 1 H 0.72104E+01 0.79837E+01 0.42850E+01 1 1 1 1 H 0.94063E+01 0.26410E+01 0.10552E+02 1 1 1 1 H 0.10345E+02 0.36559E+01 0.87763E+01 1 1 1 1 H 0.81383E+01 0.40646E+01 0.87539E+01 1 1 1 1 H 0.11798E+02 0.38945E+01 0.92349E+00 1 1 1 1 H 0.96193E+01 0.35809E+01 0.16635E+01 1 1 1 1 H 0.96635E+01 0.24733E+01 -0.45115E+00 1 1 1 1 H 0.12738E+02 0.13219E+02 0.11038E+02 1 1 1 1 H 0.14284E+02 0.13378E+02 -0.81055E+00 1 1 1 1 H 0.14491E+02 0.12930E+02 0.97192E+01 1 1 1 1 H 0.38965E+01 0.91982E+01 0.36044E+01 1 1 1 1 H 0.15856E+02 0.74852E+01 0.43660E+01 1 1 1 1 H 0.51353E+01 0.73381E+01 0.48137E+01 1 1 1 1 H 0.16544E+02 0.20901E+01 0.69498E+01 1 1 1 1 H 0.16223E+02 0.12770E+02 0.60809E+01 1 1 1 1 H 0.15500E+02 0.12731E+02 0.83546E+01 1 1 1 1 H 0.15291E+02 0.13196E+02 0.39835E+01 1 1 1 1 H 0.14840E+02 0.13877E+02 0.18832E+01 1 1 1 1 H 0.11912E+02 0.60290E+01 0.11531E+01 1 1 1 1 H 0.11819E+02 0.72121E+01 0.22820E+01 1 1 1 1 H 0.12480E+02 0.88403E+01 0.36626E+01 1 1 1 1 H 0.13564E+02 0.72374E+01 0.45975E+01 1 1 1 1 H 0.15699E+02 0.10526E+02 0.21594E+01 1 1 1 1 H 0.16399E+02 0.10978E+02 0.39319E+01 1 1 1 1 H 0.53648E+01 0.11011E+02 0.23101E+01 1 1 1 1 H 0.34596E+01 0.13170E+02 0.60397E+00 1 1 1 1 H 0.58772E+01 0.12396E+02 0.15901E+01 1 1 1 1 H 0.14990E+02 0.10027E+02 0.77363E+00 1 1 1 1 H 0.13592E+02 0.10568E+02 0.25894E+01 1 1 1 1 H 0.15281E+02 0.12136E+02 0.86419E+00 1 1 1 1 H 0.12891E+02 0.77016E+01 0.87494E+01 1 1 1 1 H 0.13612E+02 0.97653E+01 0.76083E+01 1 1 1 1 H 0.14111E+02 0.77803E+01 0.66252E+01 1 1 1 1 H 0.14969E+02 0.81163E+01 0.79451E+01 1 1 1 1 H 0.94604E+01 0.47722E+01 0.40302E+01 1 1 1 1 H 0.10437E+02 0.11415E+02 0.11883E+02 1 1 1 1 H 0.10411E+02 0.13330E+02 0.10986E+02 1 1 1 1 H 0.10895E+02 0.12106E+02 0.99767E+01 1 1 1 1 H 0.10795E+02 0.80582E+01 0.85021E+01 1 1 1 1 H 0.85166E+01 0.77592E+01 0.87930E+01 1 1 1 1 H 0.10273E+02 0.10294E+02 0.83887E+01 1 1 1 1 H 0.65202E+01 0.43007E+01 0.89727E+01 1 1 1 1 H 0.45505E+01 0.32707E+01 0.84554E+01 1 1 1 1 H 0.11578E+02 0.92806E+01 -0.63868E+00 1 1 1 1 H 0.76180E+01 0.63999E+01 0.89681E+01 1 1 1 1 H 0.14153E+02 0.87817E+01 -0.71366E+00 1 1 1 1 H 0.12173E+02 0.77953E+01 0.10245E+02 1 1 1 1 H 0.11858E+02 0.56699E+01 0.90758E+01 1 1 1 1 H 0.11987E+02 0.79213E+01 0.63115E-01 1 1 1 1 H 0.96337E+01 0.70726E+01 0.13330E+01 1 1 1 1 H 0.74597E+01 0.11328E+02 0.29631E+01 1 1 1 1 H 0.11701E+02 0.10960E+02 0.81966E+01 1 1 1 1 H 0.13916E+02 0.11279E+02 0.82366E+01 1 1 1 1 H 0.83870E+01 0.53998E+01 0.20864E+01 1 1 1 1 H 0.64813E+01 0.82363E+01 0.83833E+01 1 1 1 1 H 0.59144E+01 0.78362E+01 0.61630E+01 1 1 1 1 H 0.43681E+01 0.80285E+01 0.83954E+01 1 1 1 1 H 0.14453E+02 0.60275E+01 0.52899E+01 1 1 1 1 H 0.15223E+02 0.38965E+01 0.59663E+01 1 1 1 1 CELL_PARAMETERS 24.347851385876538 0.36920984085018538 -0.16930057214376279 0.37628734578661066 23.766046075564496 3.59168252649787120E-002 -0.17599055343917119 3.60113432262023428E-002 23.893191637002520 The calculation fails on cineca machines. I don't have any other machine to test it. Secondly, I can run a similar calculation of a single H2O molecule in serial on a desktop. However, when I try to do the same single molecule H2O on a parallel version, it gives the same segmentation error. I checked the document and the code. We need to have less number of processors than the wannier functions to be calculated. So for a single molecule ( with nspin =2 ), when there are 8 wannier functions to be calculated ( 4 for up and 4 for down spin) , the parallel version fails while a serial version gives correct results. While doing the parallel calculation I am keeping the number of processors < 8, then I have the segmentation fault error. For number of processors > 8, I have the error : task # 10 from cp-wf : error # 1 Number of Processors is greater than the number of states Hope I am able to explain the problem properly. Waiting for your comment. Somesh On Tue, Sep 21, 2010 at 2:36 PM, Paolo Giannozzi wrote: > Somesh Kumar Bhattacharya wrote: > > > The input file is : > > incomplete (&control namelist is missing)... > > > Can anybody provide a solution for this? > > ...so nobody can provide any solution. > > As explained a diverging number of times, "segmentation fault" > means nothing but "something very wrong happened". > There is NO WAY anybody can say anything sensible about a > "segmentation fault" without additional information. > First and most important piece of information: > * Is the problem reproducible with different compilers / > libraries / op.sys / hardware / input options / examples / > phases of the moon? > > P. > -- > Paolo Giannozzi, Democritos and University of Udine, Italy > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Somesh Kr. Bhattacharya Post Doctoral Fellow Room No. 263, Leonardo Building, The Abdus Salam International Centre for Theoretical Physics Strada Costiera, 11 I-34014 Trieste Italy Phone: +39-040-2240399 http://portal.ictp.it/cmsp/members/postdoctoral-fellows/somesh-kumar-bhattacharya/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100921/a21ac910/attachment-0001.htm From quantumdft at gmail.com Tue Sep 21 15:34:04 2010 From: quantumdft at gmail.com (vega lew) Date: Tue, 21 Sep 2010 21:34:04 +0800 Subject: [Pw_forum] Do you think we could fix certain bond length during the relax calculation In-Reply-To: <4C98A768.2000007@democritos.it> References: <4C94F735.9070902@gmail.com> <4A636B18-C407-4387-A0DA-A6D5AC70EC1D@democritos.it> <4C97907A.1000208@gmail.com> <4C98A768.2000007@democritos.it> Message-ID: <4C98B44C.9000909@gmail.com> I am not sure. But when I specify the 'ion_dynamics = 'bfgs' ,', the relax calculation could run smoothly. And the output file reported bfgs steps as common relax calculation. best, vega On 09/21/10 20:39, Paolo Giannozzi wrote: > vega lew wrote: > >> I found when I specified BFGS method, the 'relax' calculation also could >> run smoothly. Do you think the BFGS method rather than Damped MD is OK >> for the 'relax' calculation with 'distance' constrains? > are constraints implemented in bfgs? > > P. -- ================================================ Vega Lew ( weijia liu) Graduate student State Key Laboratory of Materials-oriented Chemical Engineering Nanjing University of Technology, 210009, Nanjing, China *************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, China *************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100921/5a463b3d/attachment.htm From giannozz at democritos.it Tue Sep 21 15:37:37 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 21 Sep 2010 15:37:37 +0200 Subject: [Pw_forum] Do you think we could fix certain bond length during the relax calculation In-Reply-To: <4C98B44C.9000909@gmail.com> References: <4C94F735.9070902@gmail.com> <4A636B18-C407-4387-A0DA-A6D5AC70EC1D@democritos.it> <4C97907A.1000208@gmail.com> <4C98A768.2000007@democritos.it> <4C98B44C.9000909@gmail.com> Message-ID: <4C98B521.4040402@democritos.it> vega lew wrote: > I am not sure. But when I specify the 'ion_dynamics = 'bfgs' ,', the > relax calculation could run smoothly. of course it does, but does it enforce constraints? -- Paolo Giannozzi, Democritos and University of Udine, Italy From Michael.Mehl at nrl.navy.mil Tue Sep 21 15:39:02 2010 From: Michael.Mehl at nrl.navy.mil (Mike Mehl) Date: Tue, 21 Sep 2010 09:39:02 -0400 Subject: [Pw_forum] Young's Modulus In-Reply-To: <90545.24036.qm@web65707.mail.ac4.yahoo.com> References: <90545.24036.qm@web65707.mail.ac4.yahoo.com> Message-ID: <4C98B576.2070406@nrl.navy.mil> We've changed the link slightly. It's now: http://cst-www.nrl.navy.mil/users/mehl/papers/cij453.pdf On 09/21/2010 08:44 AM, Eyvaz Isaev wrote: > > ------------------------------------------------------------------------ > *From:* mohsen modaresi > See papers: Mehl et al., PhysRev.B 41, 10311, 1990 and Mehl et al, in > Intermetallic Compounds: Principles and Practice, vol.1, Principles, > J.H. Wesbrook and R.L. Fleischer, eds, John Wiley and Sons (London, > 1995), ch.9 (pp. 195-210). -- Michael J. Mehl Head, Center for Computational Materials Science Naval Research Laboratory Code 6390 Washington DC From giannozz at democritos.it Tue Sep 21 16:31:48 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 21 Sep 2010 16:31:48 +0200 Subject: [Pw_forum] prob with cp-wf In-Reply-To: References: <4C98A6C6.6020900@democritos.it> Message-ID: <4C98C1D4.7050805@democritos.it> Somesh Kumar Bhattacharya wrote: > The calculation fails on cineca machines. I don't have any other machine > to test it. it's too big to be run on a desktop pc. You should first of all find out if the memory is sufficient. > Secondly, I can run a similar calculation of a single H2O molecule in > serial on a desktop. However, when I try to do the same single molecule > H2O on a parallel version, it gives the same segmentation error. how do you know it is "the same"? There are 1001 reasons that can lead to a segmentation fault, insufficient memory being the most obvious. Anyway: please provide a small test that can be run on a desktop PC P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From somesh.kb at gmail.com Tue Sep 21 21:55:14 2010 From: somesh.kb at gmail.com (Somesh Kumar Bhattacharya) Date: Tue, 21 Sep 2010 21:55:14 +0200 Subject: [Pw_forum] prob with cp-wf In-Reply-To: <4C98C1D4.7050805@democritos.it> References: <4C98A6C6.6020900@democritos.it> <4C98C1D4.7050805@democritos.it> Message-ID: Below is the input file for H2O molecule that works for a serial case : &CONTROL calculation = "cp-wf", restart_mode = "from_scratch", nstep = 1000, iprint = 100, isave = 100, dt = 4.D0, ndr = 50, ndw = 51, etot_conv_thr = 1.D-16, ekin_conv_thr = 1.D-16, prefix = "h2o_mol", pseudo_dir = '/home/somesh/QE/pseudo/' outdir = '.', / &SYSTEM ibrav = 1, celldm(1) = 20.0, nat = 3, ntyp = 2, ecutwfc = 25.D0, nspin = 2, tot_magnetization = 0 ! nr1b = 10, nr2b = 10, nr3b = 10, / &ELECTRONS emass = 350.D0, emass_cutoff = 3.D0, ortho_eps = 5.D-8, ortho_max = 250, electron_dynamics = "damp", electron_damping = 0.15, / &IONS ion_dynamics = "none", / &WANNIER adapt = .true., nsteps = 100, / ATOMIC_SPECIES O 16.D0 O.blyp-mt.UPF H 2.D0 H.blyp-vbc.UPF ATOMIC_POSITIONS (bohr) O 10.0000 10.0000 10.000 H 11.7325 9.6757 10.000 H 9.6757 11.7325 10.000 When I tried to run it in parallel, it fails. As I described in my last message, I was giving number of processors < number of wannier functions. Regards Somesh P.S. I never tried to run the 64 molecule case on a desktop. On Tue, Sep 21, 2010 at 4:31 PM, Paolo Giannozzi wrote: > Somesh Kumar Bhattacharya wrote: > > > The calculation fails on cineca machines. I don't have any other machine > > to test it. > > it's too big to be run on a desktop pc. You should first of all > find out if the memory is sufficient. > > > Secondly, I can run a similar calculation of a single H2O molecule in > > serial on a desktop. However, when I try to do the same single molecule > > H2O on a parallel version, it gives the same segmentation error. > > how do you know it is "the same"? There are 1001 reasons that can > lead to a segmentation fault, insufficient memory being the most > obvious. > > Anyway: please provide a small test that can be run on a desktop PC > > P. > -- > Paolo Giannozzi, Democritos and University of Udine, Italy > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Somesh Kr. Bhattacharya Post Doctoral Fellow Room No. 263, Leonardo Building, The Abdus Salam International Centre for Theoretical Physics Strada Costiera, 11 I-34014 Trieste Italy Phone: +39-040-2240399 http://portal.ictp.it/cmsp/members/postdoctoral-fellows/somesh-kumar-bhattacharya/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100921/b76da516/attachment.htm From zdw2000 at gmail.com Wed Sep 22 07:04:07 2010 From: zdw2000 at gmail.com (Wei Zhou) Date: Wed, 22 Sep 2010 13:04:07 +0800 Subject: [Pw_forum] about the installation pwscf2.1.5 Message-ID: I use the PGI7.0 to install pwscf 2.1.5 , it always show the errors as following; > make pw test -d bin || mkdir bin ( cd Modules; make all ) make[1]: Entering directory `/home/user15/projects/pwwork/PW/espresso-2.1.5/Modules' /lib/cpp -P -traditional -D__LINUX -D__PGI -D__FFTW -D__USE_INTERNAL_FFTW -I/home/user15/projects/pwwork/PW/espresso-2.1.5/include -I./ atom.f90 atom.F90 pgf90 -fast -r8 -I/home/user15/projects/pwwork/PW/espresso-2.1.5/Modules -I/home/user15/projects/pwwork/PW/espresso-2.1.5/PW -I/home/user15/projects/pwwork/PW/espresso-2.1.5/PH -c atom.F90 -o atom.o PGF90-F-0004-Unable to open MODULE file kinds.mod (atom.F90: 14) PGF90/x86-64 Linux 7.0-2: compilation aborted make[1]: *** [atom.o] Error 2 make[1]: Leaving directory `/home/user15/projects/pwwork/PW/espresso-2.1.5/Modules' make: *** [mods] Error 2 doses anybody can solve this problem? the following is the make.sys. # make.sys. Generated from make.sys.in by configure. CC = mpicc CCFLAGS = -fast $(DFLAGS) $(IFLAGS) # See include/defs.h.README for a list of precompilation options # (possible arguments to -D or -U) and their meaning DFLAGS = -D__LINUX64 -D__PGI -D__MPI -D__PARA -D__FFTW -D__USE_INTERNAL_FFTW FDFLAGS = $(DFLAGS) IFLAGS = -I. -I../include -I../Modules -I../PW -I../PH CPP = cpp CPPFLAGS = -P -traditional $(DFLAGS) $(IFLAGS) F77 = mpif77 F90 = mpif90 FFLAGS = -fast -r8 $(IFLAGS) F77FLAGS = $(FFLAGS) F90FLAGS = $(FFLAGS) $(FDFLAGS) F77FLAGS_NOOPT = -O0 LD = mpif90 LDFLAGS = $(LIBOBJS) $(LIBS) LIBOBJS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a # LIBS must contain the location of all needed external libraries LIBS = -llapack -lblas # MYLIB can be one of the following (depending on LIBS): # blas : compile the local copy of blas routines # lapack : compile the local copy of lapack routines # blas_and_lapack : all of the above - use this for a quick test # or if you don't have an optimized blas/lapack library # lapack_ibm : compile only lapack routines not present in IBM ESSL # use this together with IBM ESSL # lapack_t3e : compile only lapack routines not present in T3E scilib # use this together with T3E scilib # lapack_mkl : compile only lapack routines not present in Intel MKL # use this together with Intel MKL MYLIB = AR = ar ARFLAGS = ruv RANLIB = echo -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/79c0d053/attachment.htm From giannozz at democritos.it Wed Sep 22 08:41:50 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 22 Sep 2010 08:41:50 +0200 Subject: [Pw_forum] about the installation pwscf2.1.5 In-Reply-To: References: Message-ID: On Sep 22, 2010, at 7:04 , Wei Zhou wrote: > I use the PGI7.0 to install pwscf 2.1.5 version 2.1.5 was released 5 years ago! P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at democritos.it Wed Sep 22 08:58:03 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 22 Sep 2010 08:58:03 +0200 Subject: [Pw_forum] prob with cp-wf In-Reply-To: References: <4C98A6C6.6020900@democritos.it> <4C98C1D4.7050805@democritos.it> Message-ID: <2E920475-5860-443E-928F-210FA4688CB4@democritos.it> On Sep 21, 2010, at 21:55 , Somesh Kumar Bhattacharya wrote: > Below is the input file for H2O molecule that works for a serial case it works for me in the parallel case as well P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From sclauzer at sissa.it Wed Sep 22 10:07:51 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Wed, 22 Sep 2010 10:07:51 +0200 Subject: [Pw_forum] dosup and pdosup in pdos_tot file In-Reply-To: References: Message-ID: <4C99B957.3040007@sissa.it> On 09/12/2010 08:11 PM, B. Koa wrote: > Dear PWScf Users > > It seems that "dosup" and "pdosup" which are in the pdos_tot file must > be equal and really they are equal in many cases. But in some cases > they differ a very bit with each other. I think that the origin of > this very small difference is only the numerical procedure and not any > physical one. Am I correct ? I don't think your statement is correct. DOS (density of states) and PDOS (projected DOS) are not the same quantity (otherwise why should one give them distinct names?). Please spend some time in searching their respective definitions (on textbooks, in the forum archives, or even looking at the code itself). Regards, GS > > Many Thanks > B. Koa > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/85635e33/attachment.htm From modaresi.mohsen at gmail.com Wed Sep 22 15:45:14 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Wed, 22 Sep 2010 17:15:14 +0330 Subject: [Pw_forum] Young's Modulus In-Reply-To: <4C98B576.2070406@nrl.navy.mil> References: <90545.24036.qm@web65707.mail.ac4.yahoo.com> <4C98B576.2070406@nrl.navy.mil> Message-ID: Dear Stefano, Eyvaz and Mike, Thanks for your reply. I follow this procedure and get Young's Modulus for CNT (3,3): 1) relax the structure. 2) compressed the structure in Z direction (parallel to the axis of nanotube) and calculate "Force" between atoms and then calculate the Young's Modulus by using it's definition ( E=(F/A)/((L-L0)/L0) ). The result has "good" agreement with reported one. But one of my friend work with Abinit. Abinit can calculate elastic constant in a systematic way. Dr. Eyvaz Isaev thanks for the link. On Tue, Sep 21, 2010 at 5:09 PM, Mike Mehl wrote: > We've changed the link slightly. It's now: > > http://cst-www.nrl.navy.mil/users/mehl/papers/cij453.pdf > > On 09/21/2010 08:44 AM, Eyvaz Isaev wrote: > > > > ------------------------------------------------------------------------ > > *From:* mohsen modaresi > > > See papers: Mehl et al., PhysRev.B 41, 10311, 1990 and Mehl et al, in > > Intermetallic Compounds: Principles and Practice, vol.1, Principles, > > J.H. Wesbrook and R.L. Fleischer, eds, John Wiley and Sons (London, > > 1995), ch.9 (pp. 195-210). > > -- > Michael J. Mehl > Head, Center for Computational Materials Science > Naval Research Laboratory Code 6390 > Washington DC > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Mohsen Modarresi, PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. Phone +98-9133452131 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/55717c3e/attachment.htm From elbuesta at icqmail.com Wed Sep 22 16:10:08 2010 From: elbuesta at icqmail.com (elbuesta at icqmail.com) Date: Wed, 22 Sep 2010 10:10:08 -0400 Subject: [Pw_forum] PP with semi-core states Message-ID: <8CD2876FF67CC25-8A8-1C9C@Webmail-d121.sysops.aol.com> Dear users, We are tying to build a PP for Cu with 19 electrons in the valence using the ld1 code. We would like to know: 1) Is there an input example using semi-core states where two of the wavefunctions that we have to pseudize have the same angular momenta, like 3s and 4s? 1) Could we generate with ld1 a single-projector PP for the last case? Thanks, Luca Sementa and Fabio Negreiros IPCF - CNR, Pisa - Italia -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/6a27314b/attachment.htm From quantumdft at gmail.com Wed Sep 22 17:10:34 2010 From: quantumdft at gmail.com (vega lew) Date: Wed, 22 Sep 2010 23:10:34 +0800 Subject: [Pw_forum] Do you think we could fix certain bond length during the relax calculation In-Reply-To: <4C98B521.4040402@democritos.it> References: <4C94F735.9070902@gmail.com> <4A636B18-C407-4387-A0DA-A6D5AC70EC1D@democritos.it> <4C97907A.1000208@gmail.com> <4C98A768.2000007@democritos.it> <4C98B44C.9000909@gmail.com> <4C98B521.4040402@democritos.it> Message-ID: <4C9A1C6A.6070807@gmail.com> Dear Paolo, Well, I find the constrains did not take effect when using bfgs method. I think at least a warning should be added to the program, so that we could see something is wrong from the output. best, vega On 09/21/10 21:37, Paolo Giannozzi wrote: > vega lew wrote: > >> I am not sure. But when I specify the 'ion_dynamics = 'bfgs' ,', the >> relax calculation could run smoothly. > of course it does, but does it enforce constraints? > -- ================================================ Vega Lew ( weijia liu) Graduate student State Key Laboratory of Materials-oriented Chemical Engineering Nanjing University of Technology, 210009, Nanjing, China *************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, China *************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/0f3aa51c/attachment.htm From baroni at sissa.it Wed Sep 22 18:16:38 2010 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 22 Sep 2010 18:16:38 +0200 Subject: [Pw_forum] Young's Modulus In-Reply-To: References: <90545.24036.qm@web65707.mail.ac4.yahoo.com> <4C98B576.2070406@nrl.navy.mil> Message-ID: <5B97D492-70AF-49B8-A7A8-44E3D60C9D1F@sissa.it> The application of linear response theory to the calculation of elastic constants implemented in abinit was introduced by Giannozzi, Testa, and myself soon after our paper on linear-response theory, back in 1987 (PRL 59m p. 2662), and rediscovered a few years ago in the abinit group. We never implemented our LRT approach because we always thought that, beside elegance, LRT had few advantages with respect to the brute-force calculation of the elastic constants. We may be wrong, and if anybody volunteers to implement this old idea of ours, I would be delighted to help. In any case, I would be curious to know if the "systematic" calculation of elastic constants in abinit is any faster than what you can get "by hands" either using QE. Cheers - SB On Sep 22, 2010, at 3:45 PM, mohsen modaresi wrote: > Dear Stefano, Eyvaz and Mike, > Thanks for your reply. I follow this procedure and get Young's Modulus for CNT (3,3): > 1) relax the structure. > 2) compressed the structure in Z direction (parallel to the axis of nanotube) and calculate "Force" between atoms and then calculate the Young's Modulus by using it's definition ( E=(F/A)/((L-L0)/L0) ). The result has "good" agreement with reported one. > But one of my friend work with Abinit. Abinit can calculate elastic constant in a systematic way. > Dr. Eyvaz Isaev thanks for the link. > > > > > On Tue, Sep 21, 2010 at 5:09 PM, Mike Mehl wrote: > We've changed the link slightly. It's now: > > http://cst-www.nrl.navy.mil/users/mehl/papers/cij453.pdf > > On 09/21/2010 08:44 AM, Eyvaz Isaev wrote: > > > > ------------------------------------------------------------------------ > > *From:* mohsen modaresi > > > See papers: Mehl et al., PhysRev.B 41, 10311, 1990 and Mehl et al, in > > Intermetallic Compounds: Principles and Practice, vol.1, Principles, > > J.H. Wesbrook and R.L. Fleischer, eds, John Wiley and Sons (London, > > 1995), ch.9 (pp. 195-210). > > -- > Michael J. Mehl > Head, Center for Computational Materials Science > Naval Research Laboratory Code 6390 > Washington DC > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > > -- > Mohsen Modarresi, > PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. > Phone +98-9133452131 > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/bbb4bbaa/attachment-0001.htm From eyvaz_isaev at yahoo.com Wed Sep 22 19:14:23 2010 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 22 Sep 2010 10:14:23 -0700 (PDT) Subject: [Pw_forum] Young's Modulus In-Reply-To: References: <90545.24036.qm@web65707.mail.ac4.yahoo.com> <4C98B576.2070406@nrl.navy.mil> Message-ID: <611374.25444.qm@web65710.mail.ac4.yahoo.com> Hi, Yes, you are right about Abinit for elastic constants calculations. The main question is which method (Abinit or via total energy) is easier and less time-consuming. Bests, Eyvaz.------------------------------------------------------------------- Prof. Eyvaz Isaev, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com ________________________________ From: mohsen modaresi To: PWSCF Forum Sent: Wed, September 22, 2010 5:45:14 PM Subject: Re: [Pw_forum] Young's Modulus Dear Stefano, Eyvaz and Mike, Thanks for your reply. I follow this procedure and get Young's Modulus for CNT (3,3): 1) relax the structure. 2) compressed the structure in Z direction (parallel to the axis of nanotube) and calculate "Force" between atoms and then calculate the Young's Modulus by using it's definition ( E=(F/A)/((L-L0)/L0) ). The result has "good" agreement with reported one. But one of my friend work with Abinit. Abinit can calculate elastic constant in a systematic way. Dr. Eyvaz Isaev thanks for the link. On Tue, Sep 21, 2010 at 5:09 PM, Mike Mehl wrote: We've changed the link slightly. It's now: > >http://cst-www.nrl.navy.mil/users/mehl/papers/cij453.pdf> > >On 09/21/2010 08:44 AM, Eyvaz Isaev wrote: >> >> ------------------------------------------------------------------------ >> *From:* mohsen modaresi > > >> See papers: Mehl et al., PhysRev.B 41, 10311, 1990 and Mehl et al, in >> Intermetallic Compounds: Principles and Practice, vol.1, Principles, >> J.H. Wesbrook and R.L. Fleischer, eds, John Wiley and Sons (London, >> 1995), ch.9 (pp. 195-210). > >-- >Michael J. Mehl >Head, Center for Computational Materials Science >Naval Research Laboratory Code 6390 >Washington DC > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > -- Mohsen Modarresi, PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. Phone +98-9133452131 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/4e48d581/attachment.htm From metokal at gmail.com Wed Sep 22 19:30:07 2010 From: metokal at gmail.com (Mehmet Topsakal) Date: Wed, 22 Sep 2010 20:30:07 +0300 Subject: [Pw_forum] Young's Modulus In-Reply-To: <611374.25444.qm@web65710.mail.ac4.yahoo.com> References: <90545.24036.qm@web65707.mail.ac4.yahoo.com> <4C98B576.2070406@nrl.navy.mil> <611374.25444.qm@web65710.mail.ac4.yahoo.com> Message-ID: Dear Eyvaz, I think you omitted one more thing to mention. Brute-force method also allows one to know what he is doing. On Wed, Sep 22, 2010 at 8:14 PM, Eyvaz Isaev wrote: > Hi, > > Yes, you are right about Abinit for elastic constants calculations. The > main question is which method (Abinit or via total energy) is easier > and less time-consuming. > > Bests, > Eyvaz. > ------------------------------------------------------------------- > Prof. Eyvaz Isaev, > Department of Physics, Chemistry, and Biology (IFM), Linkoping University, > Sweden > Theoretical Physics Department, Moscow State Institute of Steel & Alloys, > Russia, > isaev at ifm.liu.se, eyvaz_isaev at yahoo.com > > > ------------------------------ > *From:* mohsen modaresi > *To:* PWSCF Forum > *Sent:* Wed, September 22, 2010 5:45:14 PM > *Subject:* Re: [Pw_forum] Young's Modulus > > Dear Stefano, Eyvaz and Mike, > Thanks for your reply. I follow this procedure and get Young's Modulus for > CNT (3,3): > 1) relax the structure. > 2) compressed the structure in Z direction (parallel to the axis of > nanotube) and calculate "Force" between atoms and then calculate the Young's > Modulus by using it's definition ( E=(F/A)/((L-L0)/L0) ). The result has > "good" agreement with reported one. > But one of my friend work with Abinit. Abinit can calculate elastic > constant in a systematic way. > Dr. Eyvaz Isaev thanks for the link. > > > > > On Tue, Sep 21, 2010 at 5:09 PM, Mike Mehl wrote: > >> We've changed the link slightly. It's now: >> >> http://cst-www.nrl.navy.mil/users/mehl/papers/cij453.pdf> >> >> On 09/21/2010 08:44 AM, Eyvaz Isaev wrote: >> > >> > ------------------------------------------------------------------------ >> > *From:* mohsen modaresi >> >> > See papers: Mehl et al., PhysRev.B 41, 10311, 1990 and Mehl et al, in >> > Intermetallic Compounds: Principles and Practice, vol.1, Principles, >> > J.H. Wesbrook and R.L. Fleischer, eds, John Wiley and Sons (London, >> > 1995), ch.9 (pp. 195-210). >> >> -- >> Michael J. Mehl >> Head, Center for Computational Materials Science >> Naval Research Laboratory Code 6390 >> Washington DC >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> > > > > -- > Mohsen Modarresi, > PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. > Phone +98-9133452131 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Mehmet Topsakal (Ph.D. Student) UNAM-Institute of Materials Science and Nanotechnology. Bilkent University. 06800 Bilkent, Ankara/T?rkiye Tel: 0090 312 290 3527 ; Fax: 0090 312 266 4365 UNAM-web : www.nano.org.tr -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/948334fc/attachment.htm From baroni at sissa.it Wed Sep 22 19:43:15 2010 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 22 Sep 2010 19:43:15 +0200 Subject: [Pw_forum] Young's Modulus In-Reply-To: References: <90545.24036.qm@web65707.mail.ac4.yahoo.com> <4C98B576.2070406@nrl.navy.mil> <611374.25444.qm@web65710.mail.ac4.yahoo.com> Message-ID: <00D5E238-5ADF-4253-99A6-F65CB7805C4F@sissa.it> For that, also LRT akkows one to know what one is doing ... ;-) SB On Sep 22, 2010, at 7:30 PM, Mehmet Topsakal wrote: > Dear Eyvaz, > > I think you omitted one more thing to mention. Brute-force method also allows one to know what he is doing. > > > > On Wed, Sep 22, 2010 at 8:14 PM, Eyvaz Isaev wrote: > Hi, > > Yes, you are right about Abinit for elastic constants calculations. The main question is which method (Abinit or via total energy) is easier > and less time-consuming. > > Bests, > Eyvaz. > ------------------------------------------------------------------- > Prof. Eyvaz Isaev, > Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden > Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, > isaev at ifm.liu.se, eyvaz_isaev at yahoo.com > > > From: mohsen modaresi > To: PWSCF Forum > Sent: Wed, September 22, 2010 5:45:14 PM > Subject: Re: [Pw_forum] Young's Modulus > > Dear Stefano, Eyvaz and Mike, > Thanks for your reply. I follow this procedure and get Young's Modulus for CNT (3,3): > 1) relax the structure. > 2) compressed the structure in Z direction (parallel to the axis of nanotube) and calculate "Force" between atoms and then calculate the Young's Modulus by using it's definition ( E=(F/A)/((L-L0)/L0) ). The result has "good" agreement with reported one. > But one of my friend work with Abinit. Abinit can calculate elastic constant in a systematic way. > Dr. Eyvaz Isaev thanks for the link. > > > > > On Tue, Sep 21, 2010 at 5:09 PM, Mike Mehl wrote: > We've changed the link slightly. It's now: > > http://cst-www.nrl.navy.mil/users/mehl/papers/cij453.pdf> > > On 09/21/2010 08:44 AM, Eyvaz Isaev wrote: > > > > ------------------------------------------------------------------------ > > *From:* mohsen modaresi > > > See papers: Mehl et al., PhysRev.B 41, 10311, 1990 and Mehl et al, in > > Intermetallic Compounds: Principles and Practice, vol.1, Principles, > > J.H. Wesbrook and R.L. Fleischer, eds, John Wiley and Sons (London, > > 1995), ch.9 (pp. 195-210). > > -- > Michael J. Mehl > Head, Center for Computational Materials Science > Naval Research Laboratory Code 6390 > Washington DC > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > > -- > Mohsen Modarresi, > PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. > Phone +98-9133452131 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > -- > > Mehmet Topsakal (Ph.D. Student) > UNAM-Institute of Materials Science and Nanotechnology. > Bilkent University. 06800 Bilkent, Ankara/T?rkiye > Tel: 0090 312 290 3527 ; Fax: 0090 312 266 4365 > UNAM-web : www.nano.org.tr > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/d2968487/attachment-0001.htm From miguel.martinez at ehu.es Wed Sep 22 20:39:44 2010 From: miguel.martinez at ehu.es (Miguel =?ISO-8859-1?B?TWFydO1uZXo=?=) Date: Wed, 22 Sep 2010 19:39:44 +0100 Subject: [Pw_forum] PP with semi-core states In-Reply-To: <8CD2876FF67CC25-8A8-1C9C@Webmail-d121.sysops.aol.com> References: <8CD2876FF67CC25-8A8-1C9C@Webmail-d121.sysops.aol.com> Message-ID: <20100922193944.1810b894@mark> Dear Luca, There are a few examples of that in the standard package. The paw library has three such cases: N,a Li and Pd, available at: espresso-4.2.1/atomic_doc/paw_library/input Note that, as far as I know, the Li-1s example doesn't currently work and that the Pd isn't strictly speaking an example with semicore electrons. As another example, I submitted a 9-electron Cs pseudo input to the mailing list not so long ago. For your second question, I suppose you could do so, but you'll probably lose a lot of accuracy modelling two states separated by a few eV from a single reference energy. A glance at the logarithmic derivatives should give you a hint. Finally, don't despair if you find generating a pseudo with semicore electrons difficult. It *is* difficult. Regards, Miguel On Wed, 22 Sep 2010 10:10:08 -0400 elbuesta at icqmail.com wrote: > > Dear users, > > We are tying to build a PP for Cu with 19 electrons in the valence > using the ld1 code. We would like to know: > > 1) Is there an input example using semi-core states where two of the > wavefunctions that we have to pseudize have the same angular momenta, > like 3s and 4s? 1) Could we generate with ld1 a single-projector PP > for the last case? > > Thanks, > Luca Sementa and Fabio Negreiros > IPCF - CNR, Pisa - Italia > > -- ---------------------------------------- Dr. Miguel Martinez Canales Department of Physics & Astronomy University College London Gower Street, London, WC1E 6BT (UK) Fax: +44 207 679 0595 Tlf: +44 207 678 3476 ---------------------------------------- "If you have an apple and I have an apple and we exchange these apples then you and I will still each have one apple. But if you have an idea and I have an idea and we exchange these ideas, then each of us will have two ideas." George Bernard Shaw From barfi.koa at gmail.com Wed Sep 22 23:01:00 2010 From: barfi.koa at gmail.com (B. Koa) Date: Thu, 23 Sep 2010 00:31:00 +0330 Subject: [Pw_forum] dosup and pdosup in pdos_tot file In-Reply-To: References: Message-ID: Dear PWScf Users and Gabriele As I have read in the tutorial_magn.pdf which prepared by Paolo Giannozzi they are too close to each other but not exactly the same as each other and I don't understand the meaning of dosup and pdosup in the pdos_tot file well; but generally I know the definition of DOS and PDOS. Also some values of dosup and pdosup in the pdos_tot file have negative sign ! (at higher energies ) but there are not such negative sign of dosup in the *.dos file. What do the negative values mean? Truly Yours Koa me : > > Dear PWScf Users > > > > It seems that "dosup" and "pdosup" which are in the pdos_tot file must > > be equal and really they are equal in many cases. But in some cases > > they differ a very bit with each other. I think that the origin of > > this very small difference is only the numerical procedure and not any > > physical one. Am I correct ? > > Gabriele : > I don't think your statement is correct. DOS (density of states) and > PDOS (projected DOS) are not the same quantity (otherwise why should one > give them distinct names?). Please spend some time in searching their > respective definitions (on textbooks, in the forum archives, or even > looking at the code itself). > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100923/5078c601/attachment.htm From eariel99 at gmail.com Wed Sep 22 23:28:41 2010 From: eariel99 at gmail.com (Eduardo Ariel Menendez Proupin) Date: Wed, 22 Sep 2010 17:28:41 -0400 Subject: [Pw_forum] Young's Modulus Message-ID: Hi, I think we always know what WE do, but not all of what computers do :-(. Mohsen, did you relaxed the atoms after the compression? Why did you calculate force and not stress. Have you used a finite nanotube and computed the forces upon the atoms of the edge? Finally, how is defined A in E=(F/A)/((L-L0)/L0) for the case of a single nonotube? Best regards Eduardo Eduardo Menendez Departamento de Fisica Facultad de Ciencias Universidad de Chile Phone: (56)(2)9787439 URL: http://fisica.ciencias.uchile.cl/~emenendez Let's pray for the 33 trapped miners! Four months to rescue > ---------- Mensaje reenviado ---------- > From: Stefano Baroni > To: PWSCF Forum > Date: Wed, 22 Sep 2010 19:43:15 +0200 > Subject: Re: [Pw_forum] Young's Modulus > For that, also LRT akkows one to know what one is doing ... ;-) > SB > > On Sep 22, 2010, at 7:30 PM, Mehmet Topsakal wrote: > > Dear Eyvaz, > > I think you omitted one more thing to mention. Brute-force method also > allows one to know what he is doing. > > > > On Wed, Sep 22, 2010 at 8:14 PM, Eyvaz Isaev wrote: > >> Hi, >> >> Yes, you are right about Abinit for elastic constants calculations. The >> main question is which method (Abinit or via total energy) is easier >> and less time-consuming. >> >> Bests, >> Eyvaz. >> ------------------------------------------------------------------- >> Prof. Eyvaz Isaev, >> Department of Physics, Chemistry, and Biology (IFM), Linkoping University, >> Sweden >> Theoretical Physics Department, Moscow State Institute of Steel & Alloys, >> Russia, >> isaev at ifm.liu.se, eyvaz_isaev at yahoo.com >> > >> -- . -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100922/23ecc42f/attachment.htm From modaresi.mohsen at gmail.com Wed Sep 22 23:51:15 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Thu, 23 Sep 2010 01:21:15 +0330 Subject: [Pw_forum] Young's Modulus In-Reply-To: References: Message-ID: Dear Eduardo Ariel Menendez Proupin, At first i relaxed nano-tube and obtain the position of carbon atoms in the ground state. Then i decrease the tube length. A in the relation is the area of nano-tube (A=pi*R*R which R is the radius of tube). F is the atomic force in the Z axis which can be calculated by using Q.E. I think it is better to calculate total energy in some configuration and use (F=dE/dz, where E is the total energy in different configuration). Is there any problem in this procedure? On Thu, Sep 23, 2010 at 12:58 AM, Eduardo Ariel Menendez Proupin < eariel99 at gmail.com> wrote: > Hi, > I think we always know what WE do, but not all of what computers do :-(. > > Mohsen, did you relaxed the atoms after the compression? Why did you > calculate force and not stress. Have you used a finite nanotube and computed > the forces upon the atoms of the edge? Finally, how is defined A in E=(F/A)/((L-L0)/L0) > for the case of a single nonotube? > Best regards > Eduardo > > Eduardo Menendez > Departamento de Fisica > Facultad de Ciencias > Universidad de Chile > Phone: (56)(2)9787439 > URL: http://fisica.ciencias.uchile.cl/~emenendez > > Let's pray for the 33 trapped miners! Four months to rescue > > > >> ---------- Mensaje reenviado ---------- >> >> From: Stefano Baroni >> To: PWSCF Forum >> Date: Wed, 22 Sep 2010 19:43:15 +0200 >> >> Subject: Re: [Pw_forum] Young's Modulus >> For that, also LRT akkows one to know what one is doing ... ;-) >> SB >> >> On Sep 22, 2010, at 7:30 PM, Mehmet Topsakal wrote: >> >> Dear Eyvaz, >> >> I think you omitted one more thing to mention. Brute-force method also >> allows one to know what he is doing. >> >> >> >> On Wed, Sep 22, 2010 at 8:14 PM, Eyvaz Isaev >> wrote: >> >>> Hi, >>> >>> Yes, you are right about Abinit for elastic constants calculations. The >>> main question is which method (Abinit or via total energy) is easier >>> and less time-consuming. >>> >>> Bests, >>> Eyvaz. >>> ------------------------------------------------------------------- >>> Prof. Eyvaz Isaev, >>> Department of Physics, Chemistry, and Biology (IFM), Linkoping >>> University, Sweden >>> Theoretical Physics Department, Moscow State Institute of Steel & Alloys, >>> Russia, >>> isaev at ifm.liu.se, eyvaz_isaev at yahoo.com >>> >> >>> > -- > > > . > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Mohsen Modarresi, PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. Phone +98-9133452131 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100923/d7060f3b/attachment.htm From modaresi.mohsen at gmail.com Thu Sep 23 00:46:45 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Thu, 23 Sep 2010 02:16:45 +0330 Subject: [Pw_forum] Young's Modulus In-Reply-To: References: Message-ID: Dear Eduardo Ariel Menendez Proupin, I forget to answer other questions. If we relaxed the tube after compression there is no net force on carbon atoms so I did not relax the system after compression. *Have you used a finite nanotube and computed the forces upon the atoms of the edge?*.....> A finite nano-tube? I consider a unit cell (which can make an infinite tube ) and calculate force between atoms. Is there any problem? Mohsen Modarresi On Thu, Sep 23, 2010 at 1:21 AM, mohsen modaresi wrote: > Dear Eduardo Ariel Menendez Proupin, > At first i relaxed nano-tube and obtain the position of carbon atoms in the > ground state. Then i decrease the tube length. A in the relation is the area > of nano-tube (A=pi*R*R which R is the radius of tube). F is the atomic force > in the Z axis which can be calculated by using Q.E. > I think it is better to calculate total energy in some configuration and > use (F=dE/dz, where E is the total energy in different configuration). > Is there any problem in this procedure? > > > > On Thu, Sep 23, 2010 at 12:58 AM, Eduardo Ariel Menendez Proupin < > eariel99 at gmail.com> wrote: > >> Hi, >> I think we always know what WE do, but not all of what computers do :-(. >> >> Mohsen, did you relaxed the atoms after the compression? Why did you >> calculate force and not stress. Have you used a finite nanotube and computed >> the forces upon the atoms of the edge? Finally, how is defined A in E=(F/A)/((L-L0)/L0) >> for the case of a single nonotube? >> Best regards >> Eduardo >> >> Eduardo Menendez >> Departamento de Fisica >> Facultad de Ciencias >> Universidad de Chile >> Phone: (56)(2)9787439 >> URL: http://fisica.ciencias.uchile.cl/~emenendez >> >> Let's pray for the 33 trapped miners! Four months to rescue >> >> >> >>> ---------- Mensaje reenviado ---------- >>> >>> From: Stefano Baroni >>> To: PWSCF Forum >>> Date: Wed, 22 Sep 2010 19:43:15 +0200 >>> >>> Subject: Re: [Pw_forum] Young's Modulus >>> For that, also LRT akkows one to know what one is doing ... ;-) >>> SB >>> >>> On Sep 22, 2010, at 7:30 PM, Mehmet Topsakal wrote: >>> >>> Dear Eyvaz, >>> >>> I think you omitted one more thing to mention. Brute-force method also >>> allows one to know what he is doing. >>> >>> >>> >>> On Wed, Sep 22, 2010 at 8:14 PM, Eyvaz Isaev >>> wrote: >>> >>>> Hi, >>>> >>>> Yes, you are right about Abinit for elastic constants calculations. The >>>> main question is which method (Abinit or via total energy) is easier >>>> and less time-consuming. >>>> >>>> Bests, >>>> Eyvaz. >>>> ------------------------------------------------------------------- >>>> Prof. Eyvaz Isaev, >>>> Department of Physics, Chemistry, and Biology (IFM), Linkoping >>>> University, Sweden >>>> Theoretical Physics Department, Moscow State Institute of Steel & >>>> Alloys, Russia, >>>> isaev at ifm.liu.se, eyvaz_isaev at yahoo.com >>>> >>> >>>> >> -- >> >> >> . >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> > > > -- > Mohsen Modarresi, > PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. > Phone +98-9133452131 > > -- Mohsen Modarresi, PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. Phone +98-9133452131 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100923/ab8c3061/attachment-0001.htm From weboffice1 at 126.com Thu Sep 23 04:23:34 2010 From: weboffice1 at 126.com (web) Date: Thu, 23 Sep 2010 10:23:34 +0800 (CST) Subject: [Pw_forum] maillist Message-ID: <465b29a3.1266.12b3c6727ea.Coremail.weboffice1@126.com> -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100923/0b82151c/attachment.htm From sclauzer at sissa.it Thu Sep 23 10:34:23 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Thu, 23 Sep 2010 10:34:23 +0200 Subject: [Pw_forum] dosup and pdosup in pdos_tot file In-Reply-To: References: Message-ID: <4C9B110F.8060803@sissa.it> Dear Koa (please specify your affiliation), On 09/22/2010 11:01 PM, B. Koa wrote: > > Dear PWScf Users and Gabriele > > As I have read in the tutorial_magn.pdf which prepared by Paolo > Giannozzi they are too close to each other but not exactly the same as > each other and I don't understand the meaning of dosup and pdosup in > the pdos_tot file well; but generally I know the definition of DOS and > PDOS. So it should be easy for you to understand: dosup is the DOS for up-spin states, while pdosup is the PDOS for up-spin states. Since you know the definitions of DOS and PDOS, you must be able to understand why they are not the same and why they can be almost equal in some energy range (the agreement is usually better at lower energies). > Also some values of dosup and pdosup in the pdos_tot file have > negative sign ! (at higher energies ) but there are not such negative > sign of dosup in the *.dos file. What do the negative values mean? Negative values in the PDOS (and in the DOS as well) may appear if you have used a smearing function like the Methfessel-Paxton one. For plotting the DOS it is usually more convenient to adopt the gaussian smearing (ngauss=0), even if you used a different one in the scf/nscf calculations. For computing DOS integrals, instead, you should use the same. HTH GS > > Truly Yours > Koa > > me : > > Dear PWScf Users > > > > It seems that "dosup" and "pdosup" which are in the pdos_tot > file must > > be equal and really they are equal in many cases. But in some cases > > they differ a very bit with each other. I think that the origin of > > this very small difference is only the numerical procedure and > not any > > physical one. Am I correct ? > > Gabriele : > > I don't think your statement is correct. DOS (density of states) and > PDOS (projected DOS) are not the same quantity (otherwise why > should one > give them distinct names?). Please spend some time in searching their > respective definitions (on textbooks, in the forum archives, or even > looking at the code itself). > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100923/6c6c8310/attachment.htm From arphysis at gmail.com Thu Sep 23 10:55:21 2010 From: arphysis at gmail.com (Artem R. Khabibullin) Date: Thu, 23 Sep 2010 12:55:21 +0400 Subject: [Pw_forum] Parallel calculations with NVIDIA CUDA cards. Message-ID: Dear users, I have got the problem with parallel calculations on PC with four video cards NVIDIA CUDA, under Ubuntu 9.04. Unfortunately, I don't know the steps of preparing for working with parallel calculations with Espresso used NVIDIA CUDA cards as the additional processors. And does Espresso work with NVIDIA CUDA cards? Anybody know how to do it? Thanks, for answers. with best regards Artem R. Khabibullin, Kazan Federal University, Faculty of Physics and Mathematics Education. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100923/6a902e9c/attachment.htm From Silvia.Bahmann at physik.tu-freiberg.de Thu Sep 23 10:46:55 2010 From: Silvia.Bahmann at physik.tu-freiberg.de (Silvia Bahmann) Date: Thu, 23 Sep 2010 10:46:55 +0200 Subject: [Pw_forum] Availability of phonon and raman calculations Message-ID: <201009231046.56677.Silvia.Bahmann@physik.tu-freiberg.de> Hi all, maybe there's a better mailing list to ask this, but I wanted to enquire if there are recent developments in the code that would allow to do: phonons with LDA+U raman for spin-polarized systems raman with US-PP raman with LDA+U in the near future. Thanks, Silvia -- Dipl.-Nat. Silvia Bahmann TU Bergakademie Freiberg Institut f?r Theoretische Physik Leipziger Stra?e 23, 09596 Freiberg Tel.: 03731 / 39 2591 Email: silvia.bahmann at physik.tu-freiberg.de From arphysis at gmail.com Thu Sep 23 11:17:56 2010 From: arphysis at gmail.com (Artem R. Khabibullin) Date: Thu, 23 Sep 2010 13:17:56 +0400 Subject: [Pw_forum] Parallel calculations with NVIDIA CUDA cards. In-Reply-To: References: Message-ID: Dear users, I have got the problem with parallel calculations on PC with four video cards NVIDIA CUDA, under Ubuntu 9.04. Unfortunately, I don't know the steps of preparing for working with parallel calculations with Espresso used NVIDIA CUDA cards as the additional processors. And does Espresso work with NVIDIA CUDA cards? Anybody know how to do it? Thanks, for answers. with best regards Artem R. Khabibullin, Kazan Federal University, Faculty of Physics and Mathematics Education. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100923/112a75b0/attachment.htm From eariel99 at gmail.com Thu Sep 23 11:26:27 2010 From: eariel99 at gmail.com (Eduardo Ariel Menendez Proupin) Date: Thu, 23 Sep 2010 05:26:27 -0400 Subject: [Pw_forum] Young's Modulus Message-ID: Dear Mohsen Let me comment on your procedure >At first i relaxed nano-tube and obtain the position of carbon atoms in the >ground state. Then i decrease the tube length. A in the relation is the area of >nano-tube (A=pi*R*R which R is the radius of tube). F is the atomic force in >the Z axis which can be calculated by using Q.E. >I think it is better to calculate total energy in some configuration and use >(F=dE/dz, where E is the total energy in different configuration). >Is there any problem in this procedure? First, are you calculating the Young modulus of an isolated nanotube, or of a material composed of (3,3) nanotubes perfectly oriented and with a well defined density ? The Young modulus is a concept that is well suited for a homogeneous material, e.g., a bulk of nanotubes. It will depend on the density and orientations of the nanotubes, but will be indendent of the transverse area (the force and the energy of deformation is prportional to the area). In a bulk of nanotubes perfectly oriented, the relevant area A would be the transverse area of the unit cell, and the size and shape of the unit cell depends on the density of nanotubes, or must be optimized by a mimimum energy criterium. If you are studying the Young modulus of a single isolated nanotube, then note that the Young modulus is well defined only if everybody use the same definition of A. Hence, be sure of considering the same definition as the other results that you compare. Second, be sure to use a unit cell sufficiently wide, so that the preriodic replicas of the nanotube are far enough and the results do not depend on the transversal dimensions of the unit cell. >If we relaxed the tube after compression there is no net force on carbon >atoms so I did not relax the system after compression. *>Have you used a finite nanotube and computed the forces upon the >atoms of the edge?*.....> A finite nano-tube? I consider a unit cell (which can >make an infinite tube ) and calculate force between atoms. >Is there any problem? Yes. The state with the atoms not relaxed is not a state of equilibrium, and according to Boltzmann law, if its energy difference with the relaxed state is larger than kT (0.026 eV at 300 K), this state may happen only once in the lifetime of the universe. Then, the property you calculate may have nothing to do with a real situation. Hence, relax the atoms. If you relax, of course there will be no net force upon the atoms, but the tube have stress. Use tstress=.true. and use the appropriate component of the stress tensor as F/A. Remember to do it for various transverse sizes of the simulation cell and be sure that the stress is independent of the cell size, or extrapolate. Of course, also must be independent of the cutoffs and smearing parameters (degauss). Alternatively, you may use the second derivative of the total energy to obtain the Young modulus. It is healthy to use both methods to check that the results are fine. The numbers will not be exactly equal due to numerical reasons, and you can systematically improve the agreement increasing the cutoffs and the number of kpoints. When I asked about a finite nanotube, I was thinking in an alternative way to do it. Use a large cubic cell, with a finite nanotube inside, made a constrained relaxation fixing the z coordinate (I assume that z is the nanotube axis) of the edge atoms, relax the other atoms, and sum the forces acting upon the atoms at one edge (that must be the negative of the total force upon the other edge). With that you can obtain the Young modulus of a finite nanotube. If you want it for the infinite nanotube you must do it with several lengths and extrapolate to infinite length. Of course this is much more expensive (really brute force) if your goal is the infinite nanotube, but is is what you should do if your interest is in short nanotubes. Best regards Eduardo -- Eduardo Menendez Departamento de Fisica Facultad de Ciencias Universidad de Chile Phone: (56)(2)9787439 URL: http://fisica.ciencias.uchile.cl/~emenendez Let's pray for the 33 trapped miners! Four months to rescue. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100923/e5fe19fd/attachment.htm From degironc at sissa.it Thu Sep 23 13:24:14 2010 From: degironc at sissa.it (Stefano de Gironcoli) Date: Thu, 23 Sep 2010 13:24:14 +0200 Subject: [Pw_forum] Parallel calculations with NVIDIA CUDA cards. In-Reply-To: References: Message-ID: <4C9B38DE.50604@sissa.it> Dear Artem R. Khabibullin there is some work going on, notably--to my knowledge--at ICHEC (Ireland's High-Performance Computing Centre,http://www.ichec.ie/), on porting Quantum ESPRESSO to hybrid CPU-GPU systems. I don't know the current status of the effort . I forwarded your message to someone there that may be better informed. best, Stefano de Gironcoli Artem R. Khabibullin wrote: > Dear users, > > I have got the problem with parallel calculations on PC with four video > cards NVIDIA CUDA, under Ubuntu 9.04. Unfortunately, I don't know the steps > of preparing for working with parallel calculations with Espresso used > NVIDIA CUDA cards as the additional processors. And does Espresso work with > NVIDIA CUDA cards? Anybody know how to do it? Thanks, for answers. > > with best regards > Artem R. Khabibullin, > Kazan Federal University, > Faculty of Physics and Mathematics Education. > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From modaresi.mohsen at gmail.com Thu Sep 23 23:06:26 2010 From: modaresi.mohsen at gmail.com (mohsen modaresi) Date: Fri, 24 Sep 2010 00:36:26 +0330 Subject: [Pw_forum] Young's Modulus In-Reply-To: References: Message-ID: Dear Eduardo Ariel Menendez Proupin, Thanks for your useful comments. Mohsen On Thu, Sep 23, 2010 at 12:56 PM, Eduardo Ariel Menendez Proupin < eariel99 at gmail.com> wrote: > Dear Mohsen > > Let me comment on your procedure > > >At first i relaxed nano-tube and obtain the position of carbon atoms in > the >ground state. Then i decrease the tube length. A in the relation is the > area of >nano-tube (A=pi*R*R which R is the radius of tube). F is the atomic > force in >the Z axis which can be calculated by using Q.E. > >I think it is better to calculate total energy in some configuration and > use >(F=dE/dz, where E is the total energy in different configuration). > >Is there any problem in this procedure? > > First, are you calculating the Young modulus of an isolated nanotube, or of > a material composed of (3,3) nanotubes perfectly oriented and with a well > defined density ? > > The Young modulus is a concept that is well suited for a homogeneous > material, e.g., a bulk of nanotubes. It will depend on the density and > orientations of the nanotubes, but will be indendent of the transverse area > (the force and the energy of deformation is prportional to the area). In a > bulk of nanotubes perfectly oriented, the relevant area A would be the > transverse area of the unit cell, and the size and shape of the unit cell > depends on the density of nanotubes, or must be optimized by a mimimum > energy criterium. If you are studying the Young modulus of a single isolated > nanotube, then note that the Young modulus is well defined only if everybody > use the same definition of A. Hence, be sure of considering the same > definition as the other results that you compare. Second, be sure to use a > unit cell sufficiently wide, so that the preriodic replicas of the nanotube > are far enough and the results do not depend on the transversal dimensions > of the unit cell. > > > >If we relaxed the tube after compression there is no net force on carbon > >atoms so I did not relax the system after compression. > *>Have you used a finite nanotube and computed the forces upon the >atoms > of the edge?*.....> A finite nano-tube? I consider a unit cell (which can > >make an infinite tube ) and calculate force between atoms. > >Is there any problem? > Yes. The state with the atoms not relaxed is not a state of equilibrium, > and according to Boltzmann law, if its energy difference with the relaxed > state is larger than kT (0.026 eV at 300 K), this state may happen only once > in the lifetime of the universe. Then, the property you calculate may have > nothing to do with a real situation. Hence, relax the atoms. > > If you relax, of course there will be no net force upon the atoms, but the > tube have stress. Use tstress=.true. and use the appropriate component of > the stress tensor as F/A. Remember to do it for various transverse sizes of > the simulation cell and be sure that the stress is independent of the cell > size, or extrapolate. Of course, also must be independent of the cutoffs and > smearing parameters (degauss). Alternatively, you may use the second > derivative of the total energy to obtain the Young modulus. It is healthy to > use both methods to check that the results are fine. The numbers will not be > exactly equal due to numerical reasons, and you can systematically improve > the agreement increasing the cutoffs and the number of kpoints. > > > When I asked about a finite nanotube, I was thinking in an alternative way > to do it. Use a large cubic cell, with a finite nanotube inside, made a > constrained relaxation fixing the z coordinate (I assume that z is the > nanotube axis) of the edge atoms, relax the other atoms, and sum the forces > acting upon the atoms at one edge (that must be the negative of the total > force upon the other edge). With that you can obtain the Young modulus of a > finite nanotube. If you want it for the infinite nanotube you must do it > with several lengths and extrapolate to infinite length. Of course this is > much more expensive (really brute force) if your goal is the infinite > nanotube, but is is what you should do if your interest is in short > nanotubes. > > Best regards > Eduardo > > -- > > > Eduardo Menendez > Departamento de Fisica > Facultad de Ciencias > Universidad de Chile > Phone: (56)(2)9787439 > URL: http://fisica.ciencias.uchile.cl/~emenendez > > Let's pray for the 33 trapped miners! Four months to rescue. > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Mohsen Modarresi, PhD student of Solid State Physics, Ferdowsi University of Mashhad, Iran. Phone +98-9133452131 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100924/9a44948a/attachment.htm From Matthieu.Verstraete at ulg.ac.be Fri Sep 24 10:50:33 2010 From: Matthieu.Verstraete at ulg.ac.be (matthieu verstraete) Date: Fri, 24 Sep 2010 10:50:33 +0200 Subject: [Pw_forum] d3 errors in file reading Message-ID: Hello everyone, I am trying to run d3 for a metal, and get a variety of error messages which are not helpful. Does anyone know if it works properly in 4.2.1? For the parallel executable I get >>> from d3_readin : error # 1 reading title stopping ... <<< which someone on the forum attributed to d3 using the wrong structures for input file reading (Paolo Giannozzi was recommending going back to v.1.x, which I will try eventually). For the sequential executable the code gets further along, lists the q-points and irreps, but then stops with: >>> from davcio : error # 25 error while reading from file stopping ... <<< I have _no_ idea what is going on. The phonons and ground state are fine, I have calculated all the drho* files I need for each q-point (right now I am just trying gamma), I am using norm-conserving psp and I have looked through the code for some indication of what error 25 is in davcio (unit number?), but to no avail. Does anyone know which file is problematic? Running on an IBM machine with xlf 12.1 (magerit in Spain) Thanks! Matthieu -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Prof. Matthieu Verstraete Universite de Li?ge Institut de Physique, Bat. B5, 3/7 All?e du 6 aout, 17 B- 4000 Sart Tilman, Li?ge Belgium From barfi.koa at gmail.com Fri Sep 24 11:31:20 2010 From: barfi.koa at gmail.com (B. Koa) Date: Fri, 24 Sep 2010 13:01:20 +0330 Subject: [Pw_forum] dosup and pdosup in pdos_tot file Message-ID: Dear Gabriele Many thanks for your answer. I replaced ngauss=-1 by ngauss=0 and now I don't see any negative value. Also, occupations = 'tetrahedra' and degauss=0.03 have been used for the nscf calculation. Truly Yours B. Koa Azad University (Saveh Branch) Iran > > > > Dear Koa (please specify your affiliation), > > > On 09/22/2010 11:01 PM, B. Koa wrote: > > > > Dear PWScf Users and Gabriele > > > > As I have read in the tutorial_magn.pdf which prepared by Paolo > > Giannozzi they are too close to each other but not exactly the same as > > each other and I don't understand the meaning of dosup and pdosup in > > the pdos_tot file well; but generally I know the definition of DOS and > > PDOS. > > So it should be easy for you to understand: dosup is the DOS for up-spin > states, while pdosup is the PDOS for up-spin states. Since you know the > definitions of DOS and PDOS, you must be able to understand why they are > not the same and why they can be almost equal in some energy range (the > agreement is usually better at lower energies). > > > > Also some values of dosup and pdosup in the pdos_tot file have > > negative sign ! (at higher energies ) but there are not such negative > > sign of dosup in the *.dos file. What do the negative values mean? > > Negative values in the PDOS (and in the DOS as well) may appear if you > have used a smearing function like the Methfessel-Paxton one. For > plotting the DOS it is usually more convenient to adopt the gaussian > smearing (ngauss=0), even if you used a different one in the scf/nscf > calculations. For computing DOS integrals, instead, you should use the > same. > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100924/4b1a9a0c/attachment.htm From sclauzer at sissa.it Fri Sep 24 12:06:41 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Fri, 24 Sep 2010 12:06:41 +0200 Subject: [Pw_forum] dosup and pdosup in pdos_tot file In-Reply-To: References: Message-ID: <1A75B1D9-FC69-4AFB-8780-3F963C7F0A0A@sissa.it> Il giorno 24/set/2010, alle ore 11.31, B. Koa ha scritto: > > Dear Gabriele > > > Many thanks for your answer. I replaced ngauss=-1 by ngauss=0 and now I don't see any negative value. Also, occupations = 'tetrahedra' > and degauss=0.03 have been used for the nscf calculation. The tetrahedron method is implemented in dos.x, but not in projwfc.x In order to use the tetrahedra in the calculation of the DOS you must NOT specify degauss in the namelist. The ngauss variable is ignored used and there should be no negative densities. Please read the documentation. HTH GS > > Truly Yours > B. Koa > Azad University (Saveh Branch) > Iran > > > > > > Dear Koa (please specify your affiliation), > > > On 09/22/2010 11:01 PM, B. Koa wrote: > > > > Dear PWScf Users and Gabriele > > > > As I have read in the tutorial_magn.pdf which prepared by Paolo > > Giannozzi they are too close to each other but not exactly the same as > > each other and I don't understand the meaning of dosup and pdosup in > > the pdos_tot file well; but generally I know the definition of DOS and > > PDOS. > > So it should be easy for you to understand: dosup is the DOS for up-spin > states, while pdosup is the PDOS for up-spin states. Since you know the > definitions of DOS and PDOS, you must be able to understand why they are > not the same and why they can be almost equal in some energy range (the > agreement is usually better at lower energies). > > > > Also some values of dosup and pdosup in the pdos_tot file have > > negative sign ! (at higher energies ) but there are not such negative > > sign of dosup in the *.dos file. What do the negative values mean? > > Negative values in the PDOS (and in the DOS as well) may appear if you > have used a smearing function like the Methfessel-Paxton one. For > plotting the DOS it is usually more convenient to adopt the gaussian > smearing (ngauss=0), even if you used a different one in the scf/nscf > calculations. For computing DOS integrals, instead, you should use the same. > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum ? Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100924/cb7509a4/attachment.htm From giannozz at democritos.it Fri Sep 24 18:48:12 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 24 Sep 2010 18:48:12 +0200 Subject: [Pw_forum] d3 errors in file reading In-Reply-To: References: Message-ID: On Sep 24, 2010, at 10:50 , matthieu verstraete wrote: > (Paolo Giannozzi was recommending going back to v.1.x, I beg your pardon? when did I recommand anything like this? The first error you get means that there is nothing in the input data. Try "-inp input-file" For the second: try example 14 first P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From jperaltac at gmail.com Fri Sep 24 22:51:55 2010 From: jperaltac at gmail.com (joaquin peralta) Date: Fri, 24 Sep 2010 15:51:55 -0500 Subject: [Pw_forum] Improve convergence, slab and electric field. In-Reply-To: <1283504718.4010.63.camel@walk.ijs.si> References: <1283164236.2883.42.camel@walk.ijs.si> <1283237963.27836.13.camel@walk.ijs.si> <1283504718.4010.63.camel@walk.ijs.si> Message-ID: After many time, finally I ran some simulations (cluster problems), the 'jumps' in the energy convergence remain, but reducing the k-points and cutoffs, was more fast. The more fast convergence was (and without many jumps) : -increasing the mixing_beta from 0.1 to 0.3 -NO use local-TF in the mixing_mode -NO change the mixing_ndim -Reduce the vaccum. -Reduce cutoff energy. When I only change the mixing_mode and mixing_beta from this previous data (even reduce from 0.1 to 0.09) the convergence problem remain and increase. I hope that with this, the problem disappear. Probably always was the vaccum, and the default values (except mixing_beta) was fine. After how many SCF cycles these "fluctuations" appear? Usually each 10 steps. Related with 0.1 of mixing_beta?? And thank you very much for all the answers, and any other suggestion is welcome. Best Regards Joaquin On Fri, Sep 3, 2010 at 4:05 AM, Tone Kokalj wrote: > On Thu, 2010-09-02 at 08:38 -0500, joaquin peralta wrote: > > Thanks, i obtain more fast convergence for 1V/A, but ... > > > > If I use the previous results from 1V/A in order to calculate an > > increase in the Field i obtain fluctuation in the energy. > > > > total energy = -261.46819272 Ry > > total energy = -261.46815798 Ry > > total energy = -261.46816720 Ry > > total energy = -261.46260676 Ry > > total energy = -261.37365516 Ry > > total energy = -261.46820680 Ry > > total energy = -261.46821935 Ry > > total energy = -261.46807352 Ry > > total energy = -261.46825528 Ry > > total energy = -261.46820900 Ry > > total energy = -261.46803348 Ry > > > > for my case, is recommendable reduce the precision by default of the > > energy convergence from 1D-4 to 1D-3? > > To what energy threshold are you referring: conv_thr (in &electrons) or > etot_conv_thr (in &control). According to the value of 1d-3/1d-4 I would > assume the latter; note however that the latter has nothing to do with > the SCF convergence, as it is an optimization threshold criteria for > geometry optimizer. > > If you use 1d-4 for the conv_thr, then the value is much too high, it > should be 1d-6 or, even better, lower than that. > > After how many SCF cycles these "fluctuations" appear? > > You may try to reduce mixing_beta to 5%, ... > > The following considerations will speed-up your calculations, hence you > will be able to do more SCF convergence tests: from the previous emails > of yours, I notice that you use over 50 Angs of vacuum, which is HUGE (I > would assume that about 25 Angs should be sufficient; in this case there > will be more than 10 Angs spacing from the slab to the dipole-layer). I > also noticed that ecutwfc/ecutrho of 30/500 ry is a very large number > for Aluminum. > > Regards, Tone > -- > Anton Kokalj > J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia > (tel: +386-1-477-3523 // fax:+386-1-477-3822) > > Please, if possible, avoid sending me Word or PowerPoint attachments. > See: http://www.gnu.org/philosophy/no-word-attachments.html > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- ---------------------------------------------------- Group of NanoMaterials ---------------------------------------------------- http://www.gnm.cl ---------------------------------------------------- Joaqu?n Andr?s Peralta Camposano ---------------------------------------------------- http://www.lpmd.cl/jperalta In a world without frontiers, who needs Gates and Win. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100924/132c0d10/attachment-0001.htm From alex.msilva at uol.com.br Fri Sep 24 23:01:19 2010 From: alex.msilva at uol.com.br (alex.msilva) Date: Fri, 24 Sep 2010 18:01:19 -0300 Subject: [Pw_forum] compilation error Message-ID: <4c9d119f76fac_efb36a3e7c16c@weasel28.tmail> An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100924/fac0186d/attachment.htm From alex.msilva at uol.com.br Sat Sep 25 01:58:46 2010 From: alex.msilva at uol.com.br (alex.msilva) Date: Fri, 24 Sep 2010 20:58:46 -0300 Subject: [Pw_forum] compilation error -- solved In-Reply-To: <4c9d119f76fac_efb36a3e7c16c@weasel28.tmail> References: <4c9d119f76fac_efb36a3e7c16c@weasel28.tmail> Message-ID: <4c9d3b3620c7d_45d636a3e7c18f@weasel28.tmail> An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100924/a4e3772b/attachment.htm From Matthieu.Verstraete at ulg.ac.be Sat Sep 25 02:49:28 2010 From: Matthieu.Verstraete at ulg.ac.be (matthieu verstraete) Date: Sat, 25 Sep 2010 02:49:28 +0200 Subject: [Pw_forum] d3 errors in file reading In-Reply-To: References: Message-ID: Hello Paolo, thanks for answering On Fri, Sep 24, 2010 at 6:48 PM, Paolo Giannozzi wrote: > > On Sep 24, 2010, at 10:50 , matthieu verstraete wrote: > >> (Paolo Giannozzi was recommending going back to v.1.x, > > I beg your pardon? when did I recommand anything like > this? this was in an old post in the pw_forum archives. I found but few more recent posts about d3, so imagined it might still be appropriate. This is obviously not the desirable path. http://www.democritos.it/pipermail/pw_forum/2003-June/000339.html > The first error you get means that there is nothing in the > input data. Try "-inp input-file" in the parallel case this is what I do (mandatory for all q-e executables in IBM batch jobs). In sequential (when input reading _does_ work) I just use d3.x < i.d3 pw.x and ph.x have no problems with the -inp directives. Don't know why d3 does. > For the second: try example 14 first I did this of course. On my linux box it works fine. On IBM it's much more pain because the parallel executable can not be launched from the command line. I don't expect qualitative differences. My input is directly copied from example14 - one possible difference is that it's a metal (antimony) instead of silicon, but I still haven't managed to find out what the davcio error is due to. cheers Matthieu > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Prof. Matthieu Verstraete Universite de Li?ge Institut de Physique, Bat. B5, 3/7 All?e du 6 aout, 17 B- 4000 Sart Tilman, Li?ge Belgium Associate member European Theoretical Spectroscopy Facility (ETSF http://www.etsf.eu) Phone : +32 4 366 90 17 Fax?? : +32 4 366 36 29 Mail : matthieu.verstraete at ulg.ac.be ? ? ? ? ? matthieu.jean.verstraete at gmail.com From hgm20090101 at 163.com Sat Sep 25 05:43:08 2010 From: hgm20090101 at 163.com (hgm20090101) Date: Sat, 25 Sep 2010 11:43:08 +0800 (CST) Subject: [Pw_forum] pwgui: cannot load tcl library Message-ID: <426683c7.53ba.12b46fcb7d9.Coremail.hgm20090101@163.com> Dear all, Using PWgui under linux, it appears problem as the following, would you give me hints on how to deal with it. Your faithfully g.m He ---------check the tcl lib -------- $ ls /usr/lib/tk8.4/../libtk8.4.so /usr/lib/tk8.4/../libtk8.4.so $ ldd /usr/lib/tk8.4/../libtk8.4.so linux-gate.so.1 => (0xffffe000) libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0xf7e37000) libdl.so.2 => /lib/libdl.so.2 (0xf7e33000) libm.so.6 => /lib/tls/libm.so.6 (0xf7e10000) libc.so.6 => /lib/tls/libc.so.6 (0xf7ce7000) /lib/ld-linux.so.2 (0x56555000) ---------------below is the pwgui results---------------- ================================================== This is PWgui version: 4.0cvs -------------------------------------------------- PWgui: using the system default "tclsh" interpreter PWGUI : /export/home/lsp/hgm/hh/PWgui-4.0cvs GUIB engine : /export/home/lsp/hgm/hh/PWgui-4.0cvs/lib/Guib-0.4 couldn't load file "/usr/lib/tk8.4/../libtk8.4.so": /usr/lib/tk8.4/../libtk8.4.so: cannot open shared object file: No such file or directory while executing "load /usr/lib/tk8.4/../libtk8.4.so Tk" ("package ifneeded" script) invoked from within "package require Tk " (file "/export/home/lsp/hgm/hh/PWgui-4.0cvs/lib/Guib-0.4/init.tcl" line 9) invoked from within "source /export/home/lsp/hgm/hh/PWgui-4.0cvs/lib/Guib-0.4/init.tcl" ("package ifneeded" script) invoked from within "package require Guib 0.4" (file "/export/home/lsp/hgm/hh/PWgui-4.0cvs/init.tcl" line 5) invoked from within "source [file join $env(PWGUI) init.tcl]" (file "/export/home/lsp/hgm/hh/PWgui-4.0cvs/pwgui.tcl" line 49) ??????????????24??????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100925/abdaa197/attachment.htm From ivan.girotto at ichec.ie Sat Sep 25 06:39:00 2010 From: ivan.girotto at ichec.ie (Ivan Girotto) Date: Sat, 25 Sep 2010 05:39:00 +0100 Subject: [Pw_forum] Parallel calculations with NVIDIA CUDA cards. In-Reply-To: <4C9B38DE.50604@sissa.it> References: <4C9B38DE.50604@sissa.it> Message-ID: <4C9D7CE4.3030502@ichec.ie> Dear Artem R. Khabibullin, I'm personally working on the porting of PWscf code on NVIDA GPGPU. My work it's still in progress and the code has never been released so I don't think you can run any QE codes available from the web on your GPUs architecture. We can discuss the possibility to let you have at the code if you are interested. Please, contact me on my personal mail address (ivan.girotto at ichec.ie) as I guess this topic is out of interest to this mailing list yet. Best, Ivan On 23/09/2010 12:24, Stefano de Gironcoli wrote: > Dear Artem R. Khabibullin > there is some work going on, notably--to my knowledge--at ICHEC > (Ireland's High-Performance Computing Centre,http://www.ichec.ie/), on > porting Quantum ESPRESSO to hybrid CPU-GPU systems. > I don't know the current status of the effort . I forwarded your message > to someone there that may be better informed. > best, > Stefano de Gironcoli > > > Artem R. Khabibullin wrote: >> Dear users, >> >> I have got the problem with parallel calculations on PC with four video >> cards NVIDIA CUDA, under Ubuntu 9.04. Unfortunately, I don't know the steps >> of preparing for working with parallel calculations with Espresso used >> NVIDIA CUDA cards as the additional processors. And does Espresso work with >> NVIDIA CUDA cards? Anybody know how to do it? Thanks, for answers. >> >> with best regards >> Artem R. Khabibullin, >> Kazan Federal University, >> Faculty of Physics and Mathematics Education. >> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- Girotto Ivan - ivan.girotto at ichec.ie ICHEC - Computational Group - http://www.ichec.ie The Tower - 7th floor Trinity Technology& Enterprise Campus Grand Canal Quay - Dublin 2 - Ireland +353-1-5241608 ex. 32 (ph) / +353-1-7645845 (fax) From giannozz at democritos.it Sat Sep 25 16:11:19 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sat, 25 Sep 2010 16:11:19 +0200 Subject: [Pw_forum] d3 errors in file reading In-Reply-To: References: Message-ID: <513A18AF-BE20-44D4-A458-2B57D3DA2B29@democritos.it> On Sep 25, 2010, at 2:49 , matthieu verstraete wrote: > http://www.democritos.it/pipermail/pw_forum/2003-June/000339.html that was 7 years ago! > pw.x and ph.x have no problems with the -inp directives. Don't know > why d3 does. oh well. Because support for -inp it is not there. Quick-and dirty fix: in D3/d3_readin.f90, add CALL input_from_file ( ) between IF ( ionode ) THEN and ! ! Read the first line of the input file ! READ (5, '(a)', err = 100, iostat = ios) title Before starting an enquiry on the "davcio error" (seomthing that should never happen) It would be useful to know whether the problem shows up on other machines as well P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From hgm20090101 at 163.com Sun Sep 26 11:10:04 2010 From: hgm20090101 at 163.com (hgm20090101) Date: Sun, 26 Sep 2010 17:10:04 +0800 (CST) Subject: [Pw_forum] compile on ia64-hp-hpux11.23 Message-ID: <12fa3d2.17eac.12b4d4e64d1.Coremail.hgm20090101@163.com> Dear all, If the espresso has been compiled success on the below system, please tell me some hints on how to compile it. The pw.x that I compiled cannot perform output correctly. For example, when runing example01, though it indicates " Writing output data file silicon.save", only the charge file has been written, and ended abnormal by " MPI Application rank 0 killed before MPI_Finalize() with signal 11". I also encounter problem in iotk. For example, when running test.x in iotk/ , it cannot run normally which ends with " Memory fault(coredump)". ====The system is ========= hp unix ia64-hp-hpux11.23 mlib for lapack and blas mpif90 for mpi ================= Your faithfully, g. m. he --------run on Si of example01 --------- Program PWSCF v.4.1.3 starts ... Today is 26Sep2010 at 16:58: 0 Parallel version (MPI) Number of processors in use: 1 For Norm-Conserving or Ultrasoft (Vanderbilt) Pseudopotentials or PAW Current dimensions of program pwscf are: Max number of different atomic species (ntypx) = 10 Max number of k-points (npk) = 40000 Max angular momentum in pseudopotentials (lmaxx) = 3 Waiting for input... Subspace diagonalization in iterative solution of the eigenvalue problem: Too few procs for parallel algorithm we need at least 4 procs per pool a serial algorithm will be used Planes per process (thick) : nr3 = 20 npp = 20 ncplane = 400 Proc/ planes cols G planes cols G columns G Pool (dense grid) (smooth grid) (wavefct grid) 1 20 253 2733 20 253 2733 85 531 bravais-lattice index = 2 lattice parameter (a_0) = 10.2000 a.u. unit-cell volume = 265.3020 (a.u.)^3 number of atoms/cell = 2 number of atomic types = 1 number of electrons = 8.00 number of Kohn-Sham states= 4 kinetic-energy cutoff = 18.0000 Ry charge density cutoff = 72.0000 Ry convergence threshold = 1.0E-08 mixing beta = 0.7000 number of iterations used = 8 plain mixing Exchange-correlation = SLA PZ NOGX NOGC (1100) celldm(1)= 10.200000 celldm(2)= 0.000000 celldm(3)= 0.000000 celldm(4)= 0.000000 celldm(5)= 0.000000 celldm(6)= 0.000000 crystal axes: (cart. coord. in units of a_0) a(1) = ( -0.500000 0.000000 0.500000 ) a(2) = ( 0.000000 0.500000 0.500000 ) a(3) = ( -0.500000 0.500000 0.000000 ) reciprocal axes: (cart. coord. in units 2 pi/a_0) b(1) = ( -1.000000 -1.000000 1.000000 ) b(2) = ( 1.000000 1.000000 1.000000 ) b(3) = ( -1.000000 1.000000 -1.000000 ) PseudoPot. # 1 for Si read from file Si.pz-vbc.UPF Pseudo is Norm-conserving, Zval = 4.0 Generated by new atomic code, or converted to UPF format Using radial grid of 431 points, 2 beta functions with: l(1) = 0 l(2) = 1 atomic species valence mass pseudopotential Si 4.00 28.08600 Si( 1.00) 48 Sym.Ops. (with inversion) Cartesian axes site n. atom positions (a_0 units) 1 Si tau( 1) = ( 0.0000000 0.0000000 0.0000000 ) 2 Si tau( 2) = ( 0.2500000 0.2500000 0.2500000 ) number of k points= 10 cart. coord. in units 2pi/a_0 k( 1) = ( 0.1250000 0.1250000 0.1250000), wk = 0.0625000 k( 2) = ( 0.1250000 0.1250000 0.3750000), wk = 0.1875000 k( 3) = ( 0.1250000 0.1250000 0.6250000), wk = 0.1875000 k( 4) = ( 0.1250000 0.1250000 0.8750000), wk = 0.1875000 k( 5) = ( 0.1250000 0.3750000 0.3750000), wk = 0.1875000 k( 6) = ( 0.1250000 0.3750000 0.6250000), wk = 0.3750000 k( 7) = ( 0.1250000 0.3750000 0.8750000), wk = 0.3750000 k( 8) = ( 0.1250000 0.6250000 0.6250000), wk = 0.1875000 k( 9) = ( 0.3750000 0.3750000 0.3750000), wk = 0.0625000 k( 10) = ( 0.3750000 0.3750000 0.6250000), wk = 0.1875000 G cutoff = 189.7462 ( 2733 G-vectors) FFT grid: ( 20, 20, 20) Largest allocated arrays est. size (Mb) dimensions Kohn-Sham Wavefunctions 0.02 Mb ( 350, 4) NL pseudopotentials 0.04 Mb ( 350, 8) Each V/rho on FFT grid 0.12 Mb ( 8000) Each G-vector array 0.02 Mb ( 2733) G-vector shells 0.00 Mb ( 65) Largest temporary arrays est. size (Mb) dimensions Auxiliary wavefunctions 0.09 Mb ( 350, 16) Each subspace H/S matrix 0.00 Mb ( 16, 16) Each matrix 0.00 Mb ( 8, 4) Arrays for rho mixing 0.98 Mb ( 8000, 8) Initial potential from superposition of free atoms starting charge 7.99901, renormalised to 8.00000 Starting wfc are 8 atomic wfcs total cpu time spent up to now is 0.15 secs per-process dynamical memory: 2.5 Mb Self-consistent Calculation iteration # 1 ecut= 18.00 Ry beta=0.70 Davidson diagonalization with overlap ethr = 1.00E-02, avg # of iterations = 2.0 Threshold (ethr) on eigenvalues was too large: Diagonalizing with lowered threshold Davidson diagonalization with overlap ethr = 7.75E-04, avg # of iterations = 1.0 total cpu time spent up to now is 0.46 secs total energy = -15.84097415 Ry Harris-Foulkes estimate = -15.86197052 Ry estimated scf accuracy < 0.06141563 Ry iteration # 2 ecut= 18.00 Ry beta=0.70 Davidson diagonalization with overlap ethr = 7.68E-04, avg # of iterations = 1.0 total cpu time spent up to now is 0.60 secs total energy = -15.84406636 Ry Harris-Foulkes estimate = -15.84437081 Ry estimated scf accuracy < 0.00214295 Ry iteration # 3 ecut= 18.00 Ry beta=0.70 Davidson diagonalization with overlap ethr = 2.68E-05, avg # of iterations = 2.5 total cpu time spent up to now is 0.77 secs total energy = -15.84451020 Ry Harris-Foulkes estimate = -15.84454237 Ry estimated scf accuracy < 0.00007086 Ry iteration # 4 ecut= 18.00 Ry beta=0.70 Davidson diagonalization with overlap ethr = 8.86E-07, avg # of iterations = 2.1 total cpu time spent up to now is 0.96 secs total energy = -15.84452620 Ry Harris-Foulkes estimate = -15.84452929 Ry estimated scf accuracy < 0.00000682 Ry iteration # 5 ecut= 18.00 Ry beta=0.70 Davidson diagonalization with overlap ethr = 8.52E-08, avg # of iterations = 2.0 total cpu time spent up to now is 1.14 secs total energy = -15.84452724 Ry Harris-Foulkes estimate = -15.84452726 Ry estimated scf accuracy < 0.00000006 Ry iteration # 6 ecut= 18.00 Ry beta=0.70 Davidson diagonalization with overlap ethr = 7.18E-10, avg # of iterations = 2.7 total cpu time spent up to now is 1.34 secs End of self-consistent calculation k = 0.1250 0.1250 0.1250 ( 335 PWs) bands (ev): -5.6039 4.6467 5.9568 5.9568 k = 0.1250 0.1250 0.3750 ( 338 PWs) bands (ev): -5.0584 3.0175 4.9012 4.9909 k = 0.1250 0.1250 0.6250 ( 337 PWs) bands (ev): -3.9883 1.3106 3.5165 3.9919 k = 0.1250 0.1250 0.8750 ( 343 PWs) bands (ev): -2.4615 -0.5936 2.7226 3.5069 k = 0.1250 0.3750 0.3750 ( 341 PWs) bands (ev): -4.5395 1.5909 3.8905 5.4636 k = 0.1250 0.3750 0.6250 ( 340 PWs) bands (ev): -3.5491 0.3750 2.8565 4.2745 k = 0.1250 0.3750 0.8750 ( 347 PWs) bands (ev): -2.2719 -0.7033 2.0783 3.2106 k = 0.1250 0.6250 0.6250 ( 344 PWs) bands (ev): -2.8220 -0.4390 2.1614 4.3230 k = 0.3750 0.3750 0.3750 ( 350 PWs) bands (ev): -4.0849 0.2304 5.1432 5.1432 k = 0.3750 0.3750 0.6250 ( 343 PWs) bands (ev): -3.3347 -0.5842 3.9340 4.6556 ! total energy = -15.84452726 Ry Harris-Foulkes estimate = -15.84452726 Ry estimated scf accuracy < 8.8E-10 Ry The total energy is the sum of the following terms: one-electron contribution = 4.79352695 Ry hartree contribution = 1.07664132 Ry xc contribution = -4.81493686 Ry ewald contribution = -16.89975867 Ry convergence has been achieved in 6 iterations Forces acting on atoms (Ry/au): atom 1 type 1 force = 0.00000000 0.00000000 0.00000000 atom 2 type 1 force = 0.00000000 0.00000000 0.00000000 Total force = 0.000000 Total SCF correction = 0.000000 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -10.23 -0.00006958 0.00000000 -0.00000000 -10.23 0.00 -0.00 -0.00000000 -0.00006958 -0.00000000 -0.00 -10.23 -0.00 -0.00000000 -0.00000000 -0.00006958 -0.00 -0.00 -10.23 Writing output data file silicon.save MPI Application rank 0 killed before MPI_Finalize() with signal 11 -----------------end of run on Si in example01-------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100926/83000099/attachment-0001.htm From lfhuang at theory.issp.ac.cn Sun Sep 26 11:31:24 2010 From: lfhuang at theory.issp.ac.cn (=?utf-8?B?bGZodWFuZw==?=) Date: Sun, 26 Sep 2010 17:31:24 +0800 Subject: [Pw_forum] =?utf-8?q?compile_on_ia64-hp-hpux11=2E23?= Message-ID: <20100926093124.31325.qmail@ms.hfcas.ac.cn> Dear G. M. He: Maybe your problem is due to the usage of the compiler mpif90, because one year ago I encountered a problem seemly the same as yours (I'm sorry for that I have forgotten whether it exactly is). I used the intel fortran compiler and MKL for instead, then the problem was solved. Best Wishes! Yours Sincerely L. F. Huang > ====The system is ========= > hp unix > ia64-hp-hpux11.23 > mlib for lapack and blas > mpif90 for mpi > ================= > Your faithfully, > g. m. he > Writing output data file silicon.save > MPI Application rank 0 killed before MPI_Finalize() with signal 11 ------ ====================================================================== L.F.Huang(???) DFT and phonon physics ====================================================================== Add: Research Laboratory for Computational Materials Sciences, Instutue of Solid State Physics,the Chinese Academy of Sciences, P.O.Box 1129, Hefei 230031, P.R.China Tel: 86-551-5591464-326(office) Fax: 86-551-5591434 Our group: http://theory.issp.ac.cn ====================================================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100926/ca841a3c/attachment.htm From baroni at sissa.it Sun Sep 26 18:09:30 2010 From: baroni at sissa.it (Stefano Baroni) Date: Sun, 26 Sep 2010 18:09:30 +0200 Subject: [Pw_forum] Parallel calculations with NVIDIA CUDA cards. In-Reply-To: <4C9D7CE4.3030502@ichec.ie> References: <4C9B38DE.50604@sissa.it> <4C9D7CE4.3030502@ichec.ie> Message-ID: On Sep 25, 2010, at 6:39 AM, Ivan Girotto wrote: thank you, Ivan > > > Dear Artem R. Khabibullin, > > I'm personally working on the porting of PWscf code on NVIDA GPGPU. > My work it's still in progress and the code has never been released so I don't think you can run any QE codes available from the web on your GPUs architecture. > > We can discuss the possibility to let you have at the code if you are interested. > Please, contact me on my personal mail address (ivan.girotto at ichec.ie) as I guess this topic is out of interest to this mailing list yet. It is not. If you want to report on your progresses (you do not have to ;-), I believe many people here would be most interested to learn what's goingon. Cheers - SB --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100926/7f8b2c19/attachment.htm From mayankaditya at gmail.com Mon Sep 27 09:33:04 2010 From: mayankaditya at gmail.com (mayank gupta) Date: Mon, 27 Sep 2010 13:03:04 +0530 Subject: [Pw_forum] energy cutoff for oxides Message-ID: Dear all pwscf users I have some problem regarding ecutoff in scf run. I did a series of calculation for oxides like lithium oxide silver oxide and copper oxide. To get the energy convergence of the order of meV, the ecuttoff parameters is beyond 100Ry in all the above cases even with upps. Is it ok? Mayank kumar gupta Contact No- 9869834437 ? ? ? ? ? ? ? ? ? 8097400037 ? ? ? ? ? ? ? ? ?8080458227 From Lorenzo.Paulatto at impmc.upmc.fr Mon Sep 27 10:00:39 2010 From: Lorenzo.Paulatto at impmc.upmc.fr (Lorenzo Paulatto) Date: Mon, 27 Sep 2010 10:00:39 +0200 Subject: [Pw_forum] energy cutoff for oxides In-Reply-To: References: Message-ID: Dear Mayak, the cutoff energy required for convergence does not depend on the chemical composition but on the pseudopotentials chosen. If you have use norm-conserving pseudos for Oxygen, and/or pseudopotentials for Lithium with semi-core (1s) electrons than 100Ry is reasonable. To be sure, you should tell us which pseudopotential you've used. best regards In data 27 settembre 2010 alle ore 09:33:04, mayank gupta ha scritto: > Dear all pwscf users > > I have some problem regarding ecutoff in scf run. I did a series of > calculation for oxides like lithium oxide silver oxide and copper > oxide. To get the energy convergence of the order of meV, the ecuttoff > parameters is beyond 100Ry in all the above cases even with upps. Is > it ok? > > > Mayank kumar gupta > Contact No- 9869834437 > 8097400037 > 8080458227 > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- Lorenzo Paulatto post-doc @ IMPMC/UPMC - Universit? Paris 6 phone: +33 (0)1 44 27 74 89 www: http://www-int.impmc.upmc.fr/~paulatto/ previously (take note of the change!): phd student @ SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 www: http://people.sissa.it/~paulatto/ From arphysis at gmail.com Mon Sep 27 10:49:58 2010 From: arphysis at gmail.com (Artem R. Khabibullin) Date: Mon, 27 Sep 2010 12:49:58 +0400 Subject: [Pw_forum] Parallel calculations with NVIDIA CUDA cards. Message-ID: Dear colleagues, thanks to all for your useful recommendations and answers. Artem R. Khabibullin, Kazan Federal University, Faculty of Physics and Mathematics Education. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100927/74be6cf0/attachment.htm From mohnish.iitk at gmail.com Mon Sep 27 12:12:14 2010 From: mohnish.iitk at gmail.com (mohnish pandey) Date: Mon, 27 Sep 2010 15:42:14 +0530 Subject: [Pw_forum] energy cutoff for oxides In-Reply-To: References: Message-ID: Dear Mayank! Are you sure you are relaxing the bulk system for convergence test? For convergence test you have to vc-relax calculation not 'scf'. 'SCF' calculation will not convergence stress and forces in the system so your energy will show the behavior that you are getting and by the way as Dr. Lorenzo said you should specify the type of pseudopotential you are using. And mention you affiliation also not only phone numbers ;) On Mon, Sep 27, 2010 at 1:03 PM, mayank gupta wrote: > Dear all pwscf users > > I have some problem regarding ecutoff in scf run. I did a series of > calculation for oxides like lithium oxide silver oxide and copper > oxide. To get the energy convergence of the order of meV, the ecuttoff > parameters is beyond 100Ry in all the above cases even with upps. Is > it ok? > > > Mayank kumar gupta > Contact No- 9869834437 > 8097400037 > 8080458227 > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Regards, MOHNISH, ----------------------------------------------------------------- Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA +919235721300 ----------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100927/31b4322f/attachment.htm From Matthieu.Verstraete at ulg.ac.be Mon Sep 27 15:25:17 2010 From: Matthieu.Verstraete at ulg.ac.be (matthieu verstraete) Date: Mon, 27 Sep 2010 15:25:17 +0200 Subject: [Pw_forum] d3 errors in file reading In-Reply-To: <513A18AF-BE20-44D4-A458-2B57D3DA2B29@democritos.it> References: <513A18AF-BE20-44D4-A458-2B57D3DA2B29@democritos.it> Message-ID: Brilliant, thanks Paolo. The parallel executable with the added input_from_file appears to be working. For the moment with 1 processor - is d3 fully parallelized? Can I run it with more? The davcio error was for the sequential executable, which is either incompatible with the parallel generated files, or just buggy. Ciao Matthieu On Sat, Sep 25, 2010 at 4:11 PM, Paolo Giannozzi wrote: > > On Sep 25, 2010, at 2:49 , matthieu verstraete wrote: > >> http://www.democritos.it/pipermail/pw_forum/2003-June/000339.html > > that was 7 years ago! > >> pw.x and ph.x have no problems with the -inp directives. Don't know >> why d3 does. > > oh well. Because support for -inp it is not there. Quick-and dirty > fix: in D3/d3_readin.f90, > add > > ? ? ?CALL input_from_file ( ) > > between > > ? IF ( ionode ) THEN > > and > ? ? ?! > ? ? ?! ? ?Read the first line of the input file > ? ? ?! > ? ? ?READ (5, '(a)', err = 100, iostat = ios) title > > Before starting an enquiry on the "davcio error" (seomthing that > should never happen) > It would be useful to know whether the problem shows up on other > machines as well > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Prof. Matthieu Verstraete Universite de Li?ge Institut de Physique, Bat. B5, 3/7 All?e du 6 aout, 17 B- 4000 Sart Tilman, Li?ge Belgium Associate member European Theoretical Spectroscopy Facility (ETSF http://www.etsf.eu) Phone : +32 4 366 90 17 Fax?? : +32 4 366 36 29 Mail : matthieu.verstraete at ulg.ac.be ? ? ? ? ? matthieu.jean.verstraete at gmail.com From 1009ukumar at gmail.com Tue Sep 28 10:19:01 2010 From: 1009ukumar at gmail.com (sonu kumar) Date: Tue, 28 Sep 2010 13:49:01 +0530 Subject: [Pw_forum] energy cutoff for oxides Message-ID: Dear mohnish pandey, for convergene tests in bulk, both "relax"(for atomic coordinates) and "vc-relax"(for lattice parameters with some what high values of ecutwfc and kpts than converged values)) options have to be used. please correct me if i am wrong? ! Are you sure you are relaxing the bulk system for ! convergence test? For convergence test you have to *vc-relax calculation*not ! 'scf'. 'SCF' calculation will not convergence stress and forces in the ! system so your energy will show the behavior that you are getting and by the! ! way as Dr. Lorenzo said you should specify the type of pseudopotential you ! are using. And mention you affiliation also not only phone numbers ;) -- Sonu Kumar Phd Student Physics Department Indian Institute of Technology Delhi-110016, India web:-http://www.iitd.ac.in/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100928/63db090b/attachment.htm From ihsanas at yahoo.com Tue Sep 28 14:43:07 2010 From: ihsanas at yahoo.com (Ihsan Erikat) Date: Tue, 28 Sep 2010 05:43:07 -0700 (PDT) Subject: [Pw_forum] Hessian matrix Message-ID: <844940.23942.qm@web52305.mail.re2.yahoo.com> Dear all after performing bfgs relaxation where can I find Hessian matrix . I try to search in the outdirectory (tmp) at prefix.hess file but I can`t find it, where I can find this matrix . best regards Ihsan Erikat Jordan University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100928/803ab8fb/attachment.htm From baroni at sissa.it Tue Sep 28 15:16:12 2010 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 28 Sep 2010 15:16:12 +0200 Subject: [Pw_forum] Hessian matrix In-Reply-To: <844940.23942.qm@web52305.mail.re2.yahoo.com> References: <844940.23942.qm@web52305.mail.re2.yahoo.com> Message-ID: The Hessian matrix is the matrix of interatomic force constants, calculated by ph.x using density-functional perturbation theory - SB On Sep 28, 2010, at 2:43 PM, Ihsan Erikat wrote: > Dear all > after performing bfgs relaxation where can I find Hessian matrix . I try to search in the outdirectory (tmp) at prefix.hess file but I can`t find it, where I can find this matrix . > best regards > Ihsan Erikat > Jordan University > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100928/684fe99c/attachment.htm From elie.moujaes at hotmail.co.uk Tue Sep 28 22:41:52 2010 From: elie.moujaes at hotmail.co.uk (Elie Moujaes) Date: Tue, 28 Sep 2010 21:41:52 +0100 Subject: [Pw_forum] Bilayer graphene with an electric field applied Message-ID: Dear all, I I am trying to get the band structure of graphene under the effect of an electric field. The problem is that the result at the end is very messy and jiggly. I repeated the calculations many times but I still get the same result. I first performed a scf calculation without the electric field followed by a scf calculation with an electric field and then the band calculation. SCF with no electric field: control prefix='bi elgraphene', calculation='scf', restart_mode='from_scratch', lelfield = .true., pseudo_dir = '/espresso-4.2/pseudo/', outdir='/tmp/results_MOUJAES/' / &system ibrav= 4, celldm(1) =4.608737, celldm(3)=5.330410, nat=4, ntyp= 1, ecutwfc = 110.D0,occupations='smearing', smearing='methfessel-paxton', degauss=0.01 / &electrons conv_thr=1.D-10, mixing_mode='plain' efield_cart(1) = 0.0, efield_cart(2) = 0.0, efield_cart(3) = 0.0, startingwfc = 'random' / ATOMIC_SPECIES C 12.0107 C.pz-vbc.UPF ATOMIC_POSITIONS crystal C 0.000000 0.000000 0.000000 C 0.000000 0.000000 0.257692 C 0.333333 -0.33333 0.000000 C -0.333333 0.33333 0.257692 K_POINTS automatic 38 38 1 0 0 0 SCF with electric field &control prefix='bi elgraphene', calculation='scf', restart_mode='from_scratch', lelfield = .true., pseudo_dir = '/espresso-4.2/pseudo/', outdir='/tmp/results_MOUJAES/' / &system ibrav= 4, celldm(1) =4.608737, celldm(3)=5.330410, nat=4, ntyp= 1, ecutwfc = 110.D0,occupations='smearing',smearing='methfessel-paxton', degauss=0.01 / &electrons conv_thr=1.D-10, mixing_mode='plain' efield_cart(1) = 0.0, efield_cart(2) = 0.0, efield_cart(3) = 0.001, startingwfc = 'random' / ATOMIC_SPECIES C 12.0107 C.pz-vbc.UPF ATOMIC_POSITIONS crystal C 0.000000 0.000000 0.000000 C 0.000000 0.000000 0.257692 C 0.333333 -0.33333 0.000000 C -0.333333 0.33333 0.257692 K_POINTS automatic 38 38 1 0 0 0 I noticed that part of the bands calculation output had a problem with the convergence of one of thh eigenavalues as some of the input looked like: per-process dynamical memory: 30.2 Mb Band Structure Calculation Davidson diagonalization with overlap c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged Please can anyone suggest me of what could have gone wrong... Thanks Elie Moujaes University of Nottingham NG7 2RD UK -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100928/68755545/attachment-0001.htm From sclauzer at sissa.it Wed Sep 29 09:55:03 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Wed, 29 Sep 2010 09:55:03 +0200 Subject: [Pw_forum] Bilayer graphene with an electric field applied In-Reply-To: References: Message-ID: <4CA2F0D7.1040205@sissa.it> Dear Elje, since you are dealing with a slab geometry and you want to apply the field along the direction perpendicular to the surface, why don't you use the sawtooth potential (tefield=.TRUE. instead of lefield=.TRUE.) to generate the E field? On 09/28/2010 10:41 PM, Elie Moujaes wrote: > Dear all, > I > I am trying to get the band structure of graphene under the effect of > an electric field. The problem is that the result at the end is very > messy and jiggly. I repeated the calculations many times but I still > get the same result. I first performed a scf calculation without the > electric field followed by a scf calculation with an electric field > and then the band calculation. > Do you have some evidence that your "messy and jiggly" band structure is wrong? Are you sure that you are visualizing it in the correct way? HTH GS > > SCF with electric field > > &control > > prefix='bi elgraphene', > > calculation='scf', > > restart_mode='from_scratch', > > lelfield = .true., > > pseudo_dir = '/espresso-4.2/pseudo/', > > outdir='/tmp/results_MOUJAES/' > > / > > &system > > ibrav= 4, celldm(1) =4.608737, celldm(3)=5.330410, nat=4, ntyp= 1, > > ecutwfc = 110.D0,occupations='smearing',smearing='methfessel-paxton', > degauss=0.01 > > / > > &electrons > > conv_thr=1.D-10, > > mixing_mode='plain' > > efield_cart(1) = 0.0, > > efield_cart(2) = 0.0, > > efield_cart(3) = 0.001, > > startingwfc = 'random' > > / > > ATOMIC_SPECIES > > C 12.0107 C.pz-vbc.UPF > > ATOMIC_POSITIONS crystal > > C 0.000000 0.000000 0.000000 > > C 0.000000 0.000000 0.257692 > > C 0.333333 -0.33333 0.000000 > > C -0.333333 0.33333 0.257692 > > K_POINTS automatic > > 38 38 1 0 0 0 > > > I noticed that part of the bands calculation output had a problem with > the convergence of one of thh eigenavalues as some of the input looked > like: > > per-process dynamical memory: 30.2 Mb > > Band Structure Calculation > Davidson diagonalization with overlap > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > > Please can anyone suggest me of what could have gone wrong... > > Thanks > > Elie Moujaes > > University of Nottingham > > NG7 2RD > > UK > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100929/2ee513a6/attachment.htm From giannozz at democritos.it Wed Sep 29 14:02:27 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 29 Sep 2010 14:02:27 +0200 Subject: [Pw_forum] d3 errors in file reading In-Reply-To: References: <513A18AF-BE20-44D4-A458-2B57D3DA2B29@democritos.it> Message-ID: <4CA32AD3.3080606@democritos.it> matthieu verstraete wrote: > - is d3 fully parallelized? Can I run it with more? you can, but I think you need to run d3 on the same number of processors as for ph.x P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Wed Sep 29 14:10:33 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 29 Sep 2010 14:10:33 +0200 Subject: [Pw_forum] Availability of phonon and raman calculations In-Reply-To: <201009231046.56677.Silvia.Bahmann@physik.tu-freiberg.de> References: <201009231046.56677.Silvia.Bahmann@physik.tu-freiberg.de> Message-ID: <4CA32CB9.1090109@democritos.it> Silvia Bahmann wrote: > Hi all, > > maybe there's a better mailing list to ask this q-e-developers at qe-forge.org, only for technical questions to developer, but this is also ok > phonons with LDA+U missing in action: last seen in http://www.democritos.it/pipermail/pw_forum/2009-March/011784.html > raman for spin-polarized systems this shouldn't be difficult > raman with US-PP > raman with LDA+U these are rather tough in my opinion. I am not aware of anybody working on this P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From mirnezhad.mm at gmail.com Wed Sep 29 14:20:17 2010 From: mirnezhad.mm at gmail.com (Mahdi Mirnezhad) Date: Wed, 29 Sep 2010 05:20:17 -0700 Subject: [Pw_forum] A Quastion about stress Message-ID: Dear all, I am a new user of esperesso. I want to calculate the stress of a graphene like structure. What are the axises in the stress tensor labaled ? How can we exert a shear force? Thanks for your answers Mahdi Mirnezad, Guilan University From giannozz at democritos.it Wed Sep 29 14:27:19 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 29 Sep 2010 14:27:19 +0200 Subject: [Pw_forum] A Quastion about stress In-Reply-To: References: Message-ID: <4CA330A7.4060904@democritos.it> Mahdi Mirnezhad wrote: > What are the axises in the stress tensor labaled ? cartesian: x,y,z P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From metokal at gmail.com Wed Sep 29 16:54:32 2010 From: metokal at gmail.com (Mehmet Topsakal) Date: Wed, 29 Sep 2010 17:54:32 +0300 Subject: [Pw_forum] weird phonon frequencies for 3X 3X 1 KPOINTS Message-ID: Dear all, For graphene structure, I noticed that I get completely wrong phonon frequencies when I use KPOINTS such as 9 9 1, 12 12 1, 15 15 1 For example: K_POINTS {automatic} 11 11 1 0 0 0 gives realistic frequencies as in 11.phonon.out file BUT K_POINTS {automatic} 12 12 1 0 0 0 gives "omega( 1) = -149.592762 [THz] = -4989.910586 [cm-1]" as in 12.phonon.out What should be the reason ???? Thanks. -- Mehmet Topsakal (Ph.D. Student) UNAM-Institute of Materials Science and Nanotechnology. Bilkent University. 06800 Bilkent, Ankara/T?rkiye Tel: 0090 312 290 3527 ; Fax: 0090 312 266 4365 UNAM-web : www.nano.org.tr -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e6cd9669/attachment-0001.htm -------------- next part -------------- Other files are available at http://db.tt/xK2Jtu1 NOTE: DropBox is an EXCELLENT service. You can get 2.25 Gb online storage with this link http://www.dropbox.com/referrals/NTczODUzMDY5 -------------- next part -------------- A non-text attachment was scrubbed... Name: 11.scf.in Type: application/octet-stream Size: 501 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e6cd9669/attachment-0008.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: 11.scf.out Type: application/octet-stream Size: 12825 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e6cd9669/attachment-0009.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: 11.phonon.in Type: application/octet-stream Size: 147 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e6cd9669/attachment-0010.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: 11.phonon.out Type: application/octet-stream Size: 14663 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e6cd9669/attachment-0011.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: 12.scf.in Type: application/octet-stream Size: 501 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e6cd9669/attachment-0012.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: 12.scf.out Type: application/octet-stream Size: 13362 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e6cd9669/attachment-0013.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: 12.phonon.in Type: application/octet-stream Size: 147 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e6cd9669/attachment-0014.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: 12.phonon.out Type: application/octet-stream Size: 15287 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e6cd9669/attachment-0015.obj From masoudnahali at live.com Wed Sep 29 16:59:06 2010 From: masoudnahali at live.com (Masoud Nahali) Date: Wed, 29 Sep 2010 18:29:06 +0330 Subject: [Pw_forum] relaxation and TS Message-ID: Dear PWscf Users Through the relaxation of a oxygen molecule on a surface by QE4.2, it dissociated and adsorbed on the surface. The scf energy of each step has decreased in compared with the previous step until the relaxation completed. Now, can one say that there is not any TS for O2 dissociation here ? Or is the reaction spontaneous ? If there was a TS in this procedure (dissociation and adsorption of O2) would the optimization algorithm of QE4.2 stop the work or continue ? Any help would be *appreciate*. Many Thanks ** ** *-----------------* *Masoud Nahali* *PhD Student of Physical Chemistry Sharif University of Technology * -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100929/431b9443/attachment.htm From daijiayu at nudt.edu.cn Wed Sep 29 16:46:27 2010 From: daijiayu at nudt.edu.cn (jiayudai) Date: Wed, 29 Sep 2010 22:46:27 +0800 Subject: [Pw_forum] negative values in XANES Message-ID: Dear developers and users, I am doing the calculation for the XANES of CHx system. There is a problem that the absorption is always negative, locating at about -100 eV. According to my understanding, the absorption should be positive. Besides, the fermi energy in my systems is very high, about tens of eV (high pressure). Therefore, if i used the cutoff of fermi energy, the absorption should be very small. Is it reasonable? Or there is some problems in the calculations? Thanks in advance. Jiayu ---------------- ------------------------------------------- Jiayu Dai Department of Physics National University of Defense Technology, Changsha, 410073, P R China ----------------------------------------- From quantumdft at gmail.com Wed Sep 29 17:32:12 2010 From: quantumdft at gmail.com (vega lew) Date: Wed, 29 Sep 2010 23:32:12 +0800 Subject: [Pw_forum] relaxation and TS In-Reply-To: References: Message-ID: <4CA35BFC.9090408@gmail.com> Through the relaxation of a oxygen molecule on a surface by QE4.2, it dissociated and adsorbed on the surface. The scf energy of each step has decreased in compared with the previous step until the relaxation completed. Now, can one say that there is not any TS for O2 dissociation here ? Probably not. The potential energy surface might be not very corrugated and the energy barrier for the dissociation might be not very large, so that the dissociation of O2 could happen in a relax calculation. I think there might be a very minute TS in the dissociation process. Or is the reaction spontaneous ? If the total energy of system goes down continuously and the adsorption energy of dissociated state is lower than the one of molecular state, you could say the reaction is favorable. Whether it is spontaneous, it is depend on how large of the energy barrier. If there was a TS in this procedure (dissociation and adsorption of O2) would the optimization algorithm of QE4.2 stop the work or continue ? If there is only one bond broken in the process you considered, you could try constrained optimization technique. For you information, please refer to B.Hammer (Hammer, B.; Jacobsen, K. W.; Norskov, J. K. Phys. Rev. Lett. 1992, 69, 1971) or PJ Hu's work. (Alavi, A.; Hu, P.; Deutsch, T.; Silvestrelli, P. L.; Hutter, J. Phys. Rev. Lett. 1998, 80, 3650). Or recent developed method by ZP Liu (J. Chem. Theory Comput. 2010, 6, 1136--1144 ) hope helps, best vega On 09/29/10 22:59, Masoud Nahali wrote: > > Dear PWscf Users > > Through the relaxation of a oxygen molecule on a surface by QE4.2, it > dissociated and adsorbed on the surface. The scf energy of each step > has decreased in compared with the previous step until the relaxation > completed. Now, can one say that there is not any TS for O2 > dissociation here ? Or is the reaction spontaneous ? If there was a TS > in this procedure (dissociation and adsorption of O2) would the > optimization algorithm of QE4.2 stop the work or continue ? Any help > would be *appreciate*. Many Thanks > /-----------------/ > /Masoud Nahali/ > //PhD Student of Physical Chemistry > Sharif University of Technology/ > > / > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- ================================================ Vega Lew ( weijia liu) Graduate student State Key Laboratory of Materials-oriented Chemical Engineering Nanjing University of Technology, 210009, Nanjing, China *************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, China *************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100929/a969af5a/attachment.htm From elie.moujaes at hotmail.co.uk Wed Sep 29 18:36:02 2010 From: elie.moujaes at hotmail.co.uk (Elie Moujaes) Date: Wed, 29 Sep 2010 17:36:02 +0100 Subject: [Pw_forum] Bilayer graphene with an electric field applied In-Reply-To: <4CA2F0D7.1040205@sissa.it> References: , <4CA2F0D7.1040205@sissa.it> Message-ID: Thanks very much for your help. I will use the sawtooth potential and see what happens. I will have another look on the band structure as well because I have done the DOS calculations and the graph seems pretty ok.. Regards Elie Date: Wed, 29 Sep 2010 09:55:03 +0200 From: sclauzer at sissa.it To: pw_forum at pwscf.org Subject: Re: [Pw_forum] Bilayer graphene with an electric field applied Dear Elje, since you are dealing with a slab geometry and you want to apply the field along the direction perpendicular to the surface, why don't you use the sawtooth potential (tefield=.TRUE. instead of lefield=.TRUE.) to generate the E field? On 09/28/2010 10:41 PM, Elie Moujaes wrote: Dear all, I I am trying to get the band structure of graphene under the effect of an electric field. The problem is that the result at the end is very messy and jiggly. I repeated the calculations many times but I still get the same result. I first performed a scf calculation without the electric field followed by a scf calculation with an electric field and then the band calculation. Do you have some evidence that your "messy and jiggly" band structure is wrong? Are you sure that you are visualizing it in the correct way? HTH GS SCF with electric field &control prefix='bi elgraphene', calculation='scf', restart_mode='from_scratch', lelfield = .true., pseudo_dir = '/espresso-4.2/pseudo/', outdir='/tmp/results_MOUJAES/' / &system ibrav= 4, celldm(1) =4.608737, celldm(3)=5.330410, nat=4, ntyp= 1, ecutwfc = 110.D0,occupations='smearing',smearing='methfessel-paxton', degauss=0.01 / &electrons conv_thr=1.D-10, mixing_mode='plain' efield_cart(1) = 0.0, efield_cart(2) = 0.0, efield_cart(3) = 0.001, startingwfc = 'random' / ATOMIC_SPECIES C 12.0107 C.pz-vbc.UPF ATOMIC_POSITIONS crystal C 0.000000 0.000000 0.000000 C 0.000000 0.000000 0.257692 C 0.333333 -0.33333 0.000000 C -0.333333 0.33333 0.257692 K_POINTS automatic 38 38 1 0 0 0 I noticed that part of the bands calculation output had a problem with the convergence of one of thh eigenavalues as some of the input looked like: per-process dynamical memory: 30.2 Mb Band Structure Calculation Davidson diagonalization with overlap c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged Please can anyone suggest me of what could have gone wrong... Thanks Elie Moujaes University of Nottingham NG7 2RD UK _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100929/a9193dee/attachment-0001.htm From eariel99 at gmail.com Wed Sep 29 18:39:26 2010 From: eariel99 at gmail.com (Eduardo Ariel Menendez Proupin) Date: Wed, 29 Sep 2010 12:39:26 -0400 Subject: [Pw_forum] A Quastion about stress Message-ID: > How can we exert a shear force? A short list of readings Introduction to Solid State Physics, C. Kittel, Theory of Elasticity, Landau &Lifshits For simulation with Quantum ESPRESSO, you define a strain and the code calculates stress. You must obtain a relaxed structure (calculation='vc-relax'), then you apply deformations to the lattice vectors, but relaxing the atom positions (calculation='relax') One simple example of how to deform the lattice vectors is explained in J. Appl. Phys. 99, 103504 (2006). See also Phys Rev B 1995; 51: 17431. An explanation on the theory of ab initio stress calculations is in Nielsen OH, Richard M. First?Principles Calculation of Stress. Phys Rev Lett 1983;50: 697, and more detailed in PRB 32, 3780 (1985) Good luck -- Eduardo Menendez Departamento de Fisica Facultad de Ciencias Universidad de Chile Phone: (56)(2)9787439 URL: http://fisica.ciencias.uchile.cl/~emenendez Let's pray for the 33 trapped miners! Four months to rescue. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100929/e2c674c7/attachment.htm From Jennifer.Wohlwend.ctr at wpafb.af.mil Wed Sep 29 20:46:15 2010 From: Jennifer.Wohlwend.ctr at wpafb.af.mil (Wohlwend, Jennifer L CTR USAF AFMC AFRL/RXBT) Date: Wed, 29 Sep 2010 14:46:15 -0400 Subject: [Pw_forum] symmetries Message-ID: <9AC197D8D0788140BC98A478FB3852A8471716@VFOHMLMC11.Enterprise.afmc.ds.af.mil> Is ltaucry=.true. still a resolution to the No Symmetry ! issue? I've tried putting it in various places in the input and now get: task # 12 from read_namelists : error # 19 reading namelist control Which I did not get previously so I'm thinking it is b/c of the ltaucry tag? A bit of background: I'm running P21/c structure with coordinates given in angstroms; input is attached. <<...>> Jennifer L. Wohlwend, Ph.D. Thermal Sciences and Materials Branch Materials and Manufacturing Directorate Air Force Research Laboratory Ctr: Universal Technology Corporation -------------- next part -------------- A non-text attachment was scrubbed... Name: LiAlH4.in Type: application/octet-stream Size: 2739 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/f5ef8814/attachment.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 5738 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/f5ef8814/attachment.bin From davide.ceresoli at materials.ox.ac.uk Wed Sep 29 21:49:29 2010 From: davide.ceresoli at materials.ox.ac.uk (Davide Ceresoli) Date: Wed, 29 Sep 2010 20:49:29 +0100 Subject: [Pw_forum] symmetries In-Reply-To: <9AC197D8D0788140BC98A478FB3852A8471716@VFOHMLMC11.Enterprise.afmc.ds.af.mil> References: <9AC197D8D0788140BC98A478FB3852A8471716@VFOHMLMC11.Enterprise.afmc.ds.af.mil> Message-ID: <4CA39849.1000703@materials.ox.ac.uk> On 09/29/2010 07:46 PM, Wohlwend, Jennifer L CTR USAF AFMC AFRL/RXBT wrote: > Is ltaucry=.true. still a resolution to the No Symmetry ! issue? I've tried > putting it in various places in the input and now get: > > task # 12 > > from read_namelists : error # 19 > > reading namelist control > > Which I did not get previously so I'm thinking it is b/c of the ltaucry tag? Dear Jennifer, I believe ltaucry was present in an old version of the code. To specify positions in fractional coordinates, use: ATOMIC_POSITIONS crystal Davide From Jennifer.Wohlwend.ctr at wpafb.af.mil Wed Sep 29 21:58:01 2010 From: Jennifer.Wohlwend.ctr at wpafb.af.mil (Wohlwend, Jennifer L CTR USAF AFMC AFRL/RXBT) Date: Wed, 29 Sep 2010 15:58:01 -0400 Subject: [Pw_forum] symmetries Message-ID: <9AC197D8D0788140BC98A478FB3852A847175B@VFOHMLMC11.Enterprise.afmc.ds.af.mil> >On 09/29/2010 07:46 PM, Wohlwend, Jennifer L CTR USAF AFMC AFRL/RXBT wrote: >> Is ltaucry=.true. still a resolution to the No Symmetry ! issue? I've tried >> putting it in various places in the input and now get: >> >> task # 12 >> >> from read_namelists : error # 19 >> >> reading namelist control >> >> Which I did not get previously so I'm thinking it is b/c of the ltaucry tag? >Dear Jennifer, > I believe ltaucry was present in an old version of the code. >To specify positions in fractional coordinates, use: >ATOMIC_POSITIONS crystal > >Davide Right, but mine's in angstrom so I'm using ATOMIC_POSITIONS angstrom but I'm not sure why I'm getting no symmetry, I guess that was my real question. Thanks, Jen Jennifer L. Wohlwend, Ph.D. Thermal Sciences and Materials Branch Materials and Manufacturing Directorate Air Force Research Laboratory Ctr: Universal Technology Corporation Email: jennifer.wohlwend.ctr at wpafb.af.mil -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100929/d033d6ff/attachment-0001.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 5738 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100929/d033d6ff/attachment-0001.bin From giannozz at democritos.it Wed Sep 29 22:07:54 2010 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 29 Sep 2010 22:07:54 +0200 Subject: [Pw_forum] symmetries In-Reply-To: <9AC197D8D0788140BC98A478FB3852A847175B@VFOHMLMC11.Enterprise.afmc.ds.af.mil> References: <9AC197D8D0788140BC98A478FB3852A847175B@VFOHMLMC11.Enterprise.afmc.ds.af.mil> Message-ID: <7722C2F7-005C-47A9-8CD3-68B3D82238D1@democritos.it> On Sep 29, 2010, at 21:58 , Wohlwend, Jennifer L CTR USAF AFMC AFRL/ RXBT wrote: > Right, but mine?s in angstrom so I?m using ATOMIC_POSITIONS > angstrom but I?m not sure why I?m getting no symmetry > because there isn't any. Really. Please see here: http://www.quantum-espresso.org/user_guide/ node50.html#SECTION0001110190000000000000 and here: http://www.quantum-espresso.org/user_guide/node57.html P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From matteo.calandra at impmc.jussieu.fr Thu Sep 30 09:13:15 2010 From: matteo.calandra at impmc.jussieu.fr (matteo calandra) Date: Thu, 30 Sep 2010 09:13:15 +0200 Subject: [Pw_forum] negative values in XANES (jiayudai) Message-ID: <20100930091315.70871lr5ma0pqf7v@www1.impmc.upmc.fr> Dear Jiayu, I assume you're using XSPECTRA to calculate Xanes K-edge. Maybe you can better specify what does it means "negative absorption". If this means that the value on the XANES intensity is negative then you must have some problem somewhere as the cross section implemented in XSPECTRA is positive definite. However, from your e-mail it is not very clear if the absorption is positive definite (as you say) or if the energy scale (the x axis in the plot) is negative but not the X-ray absorption intensity (y-axis in the plot). If y is positive and x is negative, then you simply forgot to specify the variable ef_r in the XSPECTRA input file. This variable tells you the zero energy of the scale. From the XSPECTRA input file: ef_r real(DP) DEFAULT=0.0 Fermi energy in Rydberg. This value combined with the option cut_occupied_states can be used to exclude the occupied states in a smooth way from the final plot. I hope it helps, M. > Dear developers and users, > > I am doing the calculation for the XANES of CHx system. There is a > problem that the absorption is always negative, locating at about > -100 eV. According to my >understanding, the absorption should be > positive. Besides, the fermi energy in >my systems is very high, > about tens of eV (high pressure). Therefore, if i used the cutoff of > fermi energy, the absorption should be very small. Is it > >reasonable? Or there is some problems in the calculations? > > Thanks in advance. > > Jiayu ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From baroni at sissa.it Thu Sep 30 10:21:37 2010 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 30 Sep 2010 10:21:37 +0200 Subject: [Pw_forum] relaxation and TS In-Reply-To: <4CA35BFC.9090408@gmail.com> References: <4CA35BFC.9090408@gmail.com> Message-ID: <2B6F3476-5E4E-4828-B54B-6F39114BFBC6@sissa.it> On Sep 29, 2010, at 5:32 PM, vega lew wrote: > Through the relaxation of a oxygen molecule on a surface by QE4.2, it dissociated and adsorbed on the surface. The scf energy of each step has decreased in compared with the previous step until the relaxation completed. Now, can one say that there is not any TS for O2 dissociation here ? > >> Probably not. The potential energy surface might be not very corrugated and the energy barrier for the dissociation might be not very large, so that the dissociation of O2 could happen in a relax calculation. I think there might be a very minute TS in the dissociation process. here, I beg to differ. If a state, say B, can be reached from another state, say A, following a path along which the energy always decreases, this means *by definition* that there is *no* energy barrier between A and B. I admit that if the energy barrier is so "minute" as to be smaller that the numerical accuracy of the minimization algorithm, it can escape proper detection. (notice that this can be said of most, if not all, numerical computations of any quantity) > > Or is the reaction spontaneous ? It is spontaneous >> If the total energy of system goes down continuously and the adsorption energy of dissociated state is lower than the one of molecular state, you could say the reaction is favorable. Whether it is spontaneous, it is depend on how large of the energy barrier. in the reported conditions, the energy barrier is zero >> If there was a TS in this procedure (dissociation and adsorption of O2) would the optimization algorithm of QE4.2 stop the work or continue ? In the presence of an energy barrier, the system would stay stuck in the starting configuration Stefano > If there is only one bond broken in the process you considered, you could try constrained optimization technique. For you information, please refer to B.Hammer (Hammer, B.; Jacobsen, K. W.; Norskov, J. K. Phys. Rev. Lett. 1992, 69, 1971) or PJ Hu's work. (Alavi, A.; Hu, P.; Deutsch, T.; Silvestrelli, P. L.; Hutter, J. Phys. Rev. Lett. 1998, 80, 3650). Or recent developed method by ZP Liu (J. Chem. Theory Comput. 2010, 6, 1136?1144 ) If I have not misunderstood some important point, no constrained minimization or other advanced techniques are needed here. Stefano B --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/f7b226a1/attachment.htm From baroni at sissa.it Thu Sep 30 10:24:16 2010 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 30 Sep 2010 10:24:16 +0200 Subject: [Pw_forum] negative values in XANES (jiayudai) In-Reply-To: <20100930091315.70871lr5ma0pqf7v@www1.impmc.upmc.fr> References: <20100930091315.70871lr5ma0pqf7v@www1.impmc.upmc.fr> Message-ID: Matteo, thank you for taking the time to participate in the discussion. In the future, please hit "reply" when answering a post, rather than starting a new thread. Have all a good day - Stefano B On Sep 30, 2010, at 9:13 AM, matteo calandra wrote: > Dear Jiayu, > > I assume you're using XSPECTRA to calculate Xanes K-edge. > Maybe you can better specify what does it means "negative > absorption". If this means that the value on the XANES intensity is negative > then you must have some problem somewhere as the cross section > implemented in XSPECTRA is positive definite. > > However, from your e-mail it is not very clear if the > absorption is positive definite (as you say) or if the energy scale > (the x axis in the plot) > is negative but not the X-ray absorption intensity (y-axis in the plot). > If y is positive and x is negative, then you simply forgot > to specify the variable > ef_r in the XSPECTRA input file. This variable tells you the zero > energy of the scale. From the XSPECTRA input file: > > ef_r real(DP) DEFAULT=0.0 > Fermi energy in Rydberg. This value combined > with the option cut_occupied_states can be > used to exclude the occupied states in a > smooth way from the final plot. > > I hope it helps, > > M. > > >> Dear developers and users, >> >> I am doing the calculation for the XANES of CHx system. There is a >> problem that the absorption is always negative, locating at about >> -100 eV. According to my >understanding, the absorption should be >> positive. Besides, the fermi energy in >my systems is very high, >> about tens of eV (high pressure). Therefore, if i used the cutoff of >> fermi energy, the absorption should be very small. Is it >>> reasonable? Or there is some problems in the calculations? >> >> Thanks in advance. >> >> Jiayu > > ---------------------------------------------------------------- > This message was sent using IMP, the Internet Messaging Program. > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/c1998bf7/attachment-0001.htm From mayankaditya at gmail.com Thu Sep 30 13:53:20 2010 From: mayankaditya at gmail.com (mayank gupta) Date: Thu, 30 Sep 2010 17:23:20 +0530 Subject: [Pw_forum] parallel machine Message-ID: Dear all Quantum Espresso users I am facing problem during phonon calculation in parallel machine for silver oxide, while the same calculation with same parameters has been successfully executed in single PC. PW.X run is successfully run in the parallel machine, but in ph.x run during electric field calculation k band not converges. I used 16 processors with 4 pools. Thanks -- Mayank kumar gupta From quantumdft at gmail.com Thu Sep 30 14:01:14 2010 From: quantumdft at gmail.com (vega lew) Date: Thu, 30 Sep 2010 20:01:14 +0800 Subject: [Pw_forum] relaxation and TS In-Reply-To: <2B6F3476-5E4E-4828-B54B-6F39114BFBC6@sissa.it> References: <4CA35BFC.9090408@gmail.com> <2B6F3476-5E4E-4828-B54B-6F39114BFBC6@sissa.it> Message-ID: <4CA47C0A.5040500@gmail.com> Dear sir, Thank you for your detailed explanation. My understanding on the TS is improved. best vega On 09/30/10 16:21, Stefano Baroni wrote: > > On Sep 29, 2010, at 5:32 PM, vega lew wrote: > >> Through the relaxation of a oxygen molecule on a surface by QE4.2, it >> dissociated and adsorbed on the surface. The scf energy of each step >> has decreased in compared with the previous step until the relaxation >> completed. Now, can one say that there is not any TS for O2 >> dissociation here ? >> >>> Probably not. The potential energy surface might be not very >>> corrugated and the energy barrier for the dissociation might be not >>> very large, so that the dissociation of O2 could happen in a relax >>> calculation. I think there might be a very minute TS in the >>> dissociation process. > > here, I beg to differ. If a state, say B, can be reached from another > state, say A, following a path along which the energy always > decreases, this means *by definition* that there is *no* energy > barrier between A and B. I admit that if the energy barrier is so > "minute" as to be smaller that the numerical accuracy of the > minimization algorithm, it can escape proper detection. (notice that > this can be said of most, if not all, numerical computations of any > quantity) > >> >> Or is the reaction spontaneous ? > > It is spontaneous > >>> If the total energy of system goes down continuously and the >>> adsorption energy of dissociated state is lower than the one of >>> molecular state, you could say the reaction is favorable. Whether it >>> is spontaneous, it is depend on how large of the energy barrier. > > in the reported conditions, the energy barrier is zero > >>> If there was a TS in this procedure (dissociation and adsorption of >>> O2) would the optimization algorithm of QE4.2 stop the work or >>> continue ? > > In the presence of an energy barrier, the system would stay stuck in > the starting configuration > > Stefano > >> If there is only one bond broken in the process you considered, you >> could try constrained optimization technique. For you information, >> please refer to B.Hammer (Hammer, B.; Jacobsen, K. W.; Norskov, J. K. >> Phys. Rev. Lett. 1992, 69, 1971) or PJ Hu's work. (Alavi, A.; Hu, P.; >> Deutsch, T.; Silvestrelli, P. L.; Hutter, J. Phys. Rev. Lett. 1998, >> 80, 3650). Or recent developed method by ZP Liu (J. Chem. Theory >> Comput. 2010, 6, 1136?1144 ) > > If I have not misunderstood some important point, no constrained > minimization or other advanced techniques are needed here. > > Stefano B > > --- > Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste > http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / > stefanobaroni (skype) > > La morale est une logique de l'action comme la logique est une morale > de la pens?e - Jean Piaget > > Please, if possible, don't send me MS Word or PowerPoint attachments > Why? See: http://www.gnu.org/philosophy/no-word-attachments.html > > > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- ================================================ Vega Lew ( weijia liu) Graduate student State Key Laboratory of Materials-oriented Chemical Engineering Nanjing University of Technology, 210009, Nanjing, China *************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, China *************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/64c2d3b6/attachment.htm From daijiayu at nudt.edu.cn Thu Sep 30 13:54:51 2010 From: daijiayu at nudt.edu.cn (jiayudai) Date: Thu, 30 Sep 2010 19:54:51 +0800 Subject: [Pw_forum] negative values in XANES Message-ID: Dear matteo, Thanks for your reply to me. The "negative" means the negative position in x-axis. And i added the "ef_r" in the input file, and the position of the absorption is reasonable now. However, when the pressure is high (high density), the fermi energy should be also very high. Even i added the "ef_r", the absorption after cutting the contribution of occupied states is only a isolated peak. In the same case but with lower density, there is not this problem. And the analysis of density of states for the system tells me that there should be absorption at the position of high energy. In this case, the absorption before occ_cut and after occ_cut is very different (the intensity is very weak, about 10-8). So, i can not find physcal reason to explain it. Is there also some problems in my calculation? By the way, i changed some parameters of &occ_cut& part, but the results were the same. Thanks again, and best wishes. Jiayu Subject: Re: [Pw_forum] negative values in XANES (jiayudai) To: PWSCF Forum Message-ID: Content-Type: text/plain; charset="iso-8859-1" Matteo, thank you for taking the time to participate in the discussion. In the future, please hit "reply" when answering a post, rather than starting a new thread. Have all a good day - Stefano B On Sep 30, 2010, at 9:13 AM, matteo calandra wrote: > Dear Jiayu, > > I assume you're using XSPECTRA to calculate Xanes K-edge. > Maybe you can better specify what does it means "negative > absorption". If this means that the value on the XANES intensity is negative > then you must have some problem somewhere as the cross section > implemented in XSPECTRA is positive definite. > > However, from your e-mail it is not very clear if the > absorption is positive definite (as you say) or if the energy scale > (the x axis in the plot) > is negative but not the X-ray absorption intensity (y-axis in the plot). > If y is positive and x is negative, then you simply forgot > to specify the variable > ef_r in the XSPECTRA input file. This variable tells you the zero > energy of the scale. From the XSPECTRA input file: > > ef_r real(DP) DEFAULT=0.0 > Fermi energy in Rydberg. This value combined > with the option cut_occupied_states can be > used to exclude the occupied states in a > smooth way from the final plot. > > I hope it helps, > > M. > > >> Dear developers and users, >> >> I am doing the calculation for the XANES of CHx system. There is a >> problem that the absorption is always negative, locating at about >> -100 eV. According to my >understanding, the absorption should be >> positive. Besides, the fermi energy in >my systems is very high, >> about tens of eV (high pressure). Therefore, if i used the cutoff of >> fermi energy, the absorption should be very small. Is it >>> reasonable? Or there is some problems in the calculations? >> >> Thanks in advance. >> >> Jiayu > > ---------------------------------------------------------------- > This message was sent using IMP, the Internet Messaging Program. > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/c1998bf7/attachment.htm ------------------------------ _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum End of Pw_forum Digest, Vol 39, Issue 57 **************************************** From archygu at gmail.com Thu Sep 30 15:03:06 2010 From: archygu at gmail.com (archygu) Date: Thu, 30 Sep 2010 21:03:06 +0800 Subject: [Pw_forum] the keyword in the pw.x Message-ID: <201009302103032657359@gmail.com> Dear QE Users, As a beginner of QE, I get problem in understanding some keywords in PW.x, it will be very kind of you to give me a more detailed explanation: nbnd: acturally I do not know how to use this parameter , sometimes, a lot of levels appear in the valence bands, but only several in the conduction band. so I want to know how can I accurately set the parameter, then can obtain the energy levels properly. nosym: if it is true, that means the symmetry in the structure will be ignored? nraise: how can we define one "nraise" step, it is the same as one time step if we set nraise=1; many thanks for you help. 2010-09-30 archygu,NTU -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/2ef733ed/attachment-0001.htm From elie.moujaes at hotmail.co.uk Thu Sep 30 15:30:01 2010 From: elie.moujaes at hotmail.co.uk (Elie Moujaes) Date: Thu, 30 Sep 2010 14:30:01 +0100 Subject: [Pw_forum] Bilayer graphene with an electric field applied In-Reply-To: <4CA2F0D7.1040205@sissa.it> References: , <4CA2F0D7.1040205@sissa.it> Message-ID: Dear Gabriele, I have tried the tefield but i am getting the same messy results again.. The thing is there are no problems in the DOS calculations; only in the band structure. I have increased the E field in the new calculation to 0.1 instead of the 0.001 old value. I am attaching a very small part of the results as a tif file because of the attachment limitations in the forum. When you have time, please let me know what you think... Thank you Elie Date: Wed, 29 Sep 2010 09:55:03 +0200 From: sclauzer at sissa.it To: pw_forum at pwscf.org Subject: Re: [Pw_forum] Bilayer graphene with an electric field applied Dear Elje, since you are dealing with a slab geometry and you want to apply the field along the direction perpendicular to the surface, why don't you use the sawtooth potential (tefield=.TRUE. instead of lefield=.TRUE.) to generate the E field? On 09/28/2010 10:41 PM, Elie Moujaes wrote: Dear all, I I am trying to get the band structure of graphene under the effect of an electric field. The problem is that the result at the end is very messy and jiggly. I repeated the calculations many times but I still get the same result. I first performed a scf calculation without the electric field followed by a scf calculation with an electric field and then the band calculation. Do you have some evidence that your "messy and jiggly" band structure is wrong? Are you sure that you are visualizing it in the correct way? HTH GS SCF with electric field &control prefix='bi elgraphene', calculation='scf', restart_mode='from_scratch', lelfield = .true., pseudo_dir = '/espresso-4.2/pseudo/', outdir='/tmp/results_MOUJAES/' / &system ibrav= 4, celldm(1) =4.608737, celldm(3)=5.330410, nat=4, ntyp= 1, ecutwfc = 110.D0,occupations='smearing',smearing='methfessel-paxton', degauss=0.01 / &electrons conv_thr=1.D-10, mixing_mode='plain' efield_cart(1) = 0.0, efield_cart(2) = 0.0, efield_cart(3) = 0.001, startingwfc = 'random' / ATOMIC_SPECIES C 12.0107 C.pz-vbc.UPF ATOMIC_POSITIONS crystal C 0.000000 0.000000 0.000000 C 0.000000 0.000000 0.257692 C 0.333333 -0.33333 0.000000 C -0.333333 0.33333 0.257692 K_POINTS automatic 38 38 1 0 0 0 I noticed that part of the bands calculation output had a problem with the convergence of one of thh eigenavalues as some of the input looked like: per-process dynamical memory: 30.2 Mb Band Structure Calculation Davidson diagonalization with overlap c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged Please can anyone suggest me of what could have gone wrong... Thanks Elie Moujaes University of Nottingham NG7 2RD UK _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/c21ec792/attachment-0001.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: bigraphel.tif Type: image/tiff Size: 102582 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100930/c21ec792/attachment-0001.tif From sclauzer at sissa.it Thu Sep 30 15:45:17 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Thu, 30 Sep 2010 15:45:17 +0200 Subject: [Pw_forum] Bilayer graphene with an electric field applied In-Reply-To: References: , <4CA2F0D7.1040205@sissa.it> Message-ID: <4CA4946D.5010007@sissa.it> On 09/30/2010 03:30 PM, Elie Moujaes wrote: > Dear Gabriele, > > I have tried the tefield but i am getting the same messy results > again.. The thing is there are no problems in the DOS calculations; > only in the band structure. I have increased the E field in the new > calculation to 0.1 instead of the 0.001 old value. I am attaching a > very small part of the results as a tif file because of the attachment > limitations in the forum. When you have time, please let me know what > you think... > > Thank you > > Elie > Sorry, your picture is ununderstandable. Please try to upload the file somewhere or convert the image in another format. Please also explain us the procedure you followed to compute and display the band structure. GS -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/f3f92504/attachment.htm From sclauzer at sissa.it Thu Sep 30 15:47:39 2010 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Thu, 30 Sep 2010 15:47:39 +0200 Subject: [Pw_forum] weird phonon frequencies for 3X 3X 1 KPOINTS In-Reply-To: References: Message-ID: <4CA494FB.9040506@sissa.it> Dear Mehmet, graphene has zero gap, hence you probably need to include smearing of the occupations in the scf calculations. HTH GS On 09/29/2010 04:54 PM, Mehmet Topsakal wrote: > Dear all, > > For graphene structure, I noticed that I get completely wrong phonon > frequencies when I use KPOINTS such as 9 9 1, 12 12 1, 15 15 1 > > For example: > > K_POINTS {automatic} > 11 11 1 0 0 0 > > gives realistic frequencies as in 11.phonon.out file > > BUT > > K_POINTS {automatic} > 12 12 1 0 0 0 > > gives "omega( 1) = -149.592762 [THz] = -4989.910586 [cm-1]" as > in 12.phonon.out > > > What should be the reason ???? > > Thanks. > > -- > > Mehmet Topsakal (Ph.D. Student) > UNAM-Institute of Materials Science and Nanotechnology. > Bilkent University. 06800 Bilkent, Ankara/T?rkiye > Tel: 0090 312 290 3527 ; Fax: 0090 312 266 4365 > UNAM-web : www.nano.org.tr > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Gabriele Sclauzero, EPFL SB ITP CSEA PH H2 462, Station 3, CH-1015 Lausanne -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/2d8b617d/attachment.htm From masoudnahali at live.com Thu Sep 30 15:49:49 2010 From: masoudnahali at live.com (Masoud Nahali) Date: Thu, 30 Sep 2010 17:19:49 +0330 Subject: [Pw_forum] relaxation and TS Message-ID: Dear Prof. Stefano Baroni and Vega Lew Many many thanks for your detailed answers and helpful comments. This is scf energy of each relaxation steps : -411.0652023864 -411.1576447124 -411.1877929276 -411.1989785058 -411.1027361193** -411.2006610355 -411.2078087316 -411.2087637085 -411.2094335080 -411.2112728535 -411.2137042660 -411.2145098860 -411.2163180058 -411.2200044391 -411.2186337506** -411.2238781353 -411.2272606876 -411.2335499997 -411.2410057850 -411.2486666714 -411.2432716037** -411.2545344840 -411.2632404207 -411.2686094326 -411.2697596541 -411.2688830229** -411.2723156669 -411.2728025269 -411.2731592912 -411.2735337540 -411.2741229900 -411.2749931725 -411.2755105410 -411.2761617487 -411.2768678588 -411.2772028256 -411.2775503379 -411.2785154793 -411.2823072885 -411.2298751795** -411.2825004743 -411.2845206434 -411.2875009929 -411.2801470356** -411.2954982819 -411.3096061836 -411.3466223605 -411.367570164 -411.3724319076 -411.3795692495 -411.3812435563 -411.3823959522 -411.3829995766 -411.3833540501 -411.3835954004 -411.3839027669 -411.3841960285 -411.3843908495 -411.3844986818 -411.3845647753 -411.3846244537 -411.3846644112 -411.3846945470 -411.3847346519 -411.3848191008 -411.3848652595 -411.3849367730 -411.3849811533 -411.3851262954 -411.3534987061** -411.3850020755 -411.3851753523 -411.3852665074 -411.3853358233 -411.3855056618 -411.3858383834 -411.3860833880 -411.3866016967 -411.3874393558 -411.3878819548 -411.3881741927 -411.3884732905 -411.3886017423 -411.3886998952 -411.3887918295 -411.3888586159 -411.3888992755 -411.3889304888 -411.3889622530 -411.3889906826 -411.3890146960 -411.3890447538 -411.3890806506 -411.3891138607 -411.3891306978 -411.3891510321 -411.3891853920 -411.3892253795 -411.3892680354 -411.3893152439 -411.3893766070 -411.3894890010 -411.3895516220 -411.3896539375 -411.3896995617 -411.3897936720 -411.3898406152 -411.3898796419 -411.3899004418 ** shows the step in which the energy rises. I don't think that these points indicate a TS or something like that since one may see such points through any optimization procedure; am I correct ? Also, here is the input : &CONTROL calculation = "relax", pseudo_dir = "/home/soft/qe4.2/espresso-4.2/pseudo", outdir = "/home/AB/tmp", restart_mode = 'from_scratch' etot_conv_thr= 1.0D-4, forc_conv_thr= 1.0D-3, nstep=120 / &SYSTEM ibrav = 4, a = 4.864990484, b = 4.864990484, c = 16.418933296, cosab = -0.5, cosac = 1.0, cosbc = 1.0, nat = 26, ntyp = 3, ecutwfc = 40.D0, ecutrho = 480.D0, occupations = 'smearing' smearing ='mp', degauss = 0.03, nspin = 2, starting_magnetization(1)= 0.01, starting_magnetization(2)= 0.01, starting_magnetization(3)= 0.01, london=.true., nosym=.true. / &ELECTRONS conv_thr = 1.D-6, mixing_beta = 0.7D0, diagonalization = "david", / &IONS ion_dynamics="bfgs" / ATOMIC_SPECIES C 12.0107 C.pbe-rrkjus.UPF Ni 58.69 Ni.pbe-nd-rrkjus.UPF O 15.9994 O.pbe-rrkjus.UPF ATOMIC_POSITIONS {angstrom} C 0.000000000 0.000000000 0.000000000 0 0 0 C 2.461165257 -0.001796837 0.002770193 C -1.232804188 2.130698110 0.002678921 C 1.230011567 2.131364695 0.003376901 C -0.003333507 1.422214727 0.001510708 C -1.234100323 3.553382228 0.001755289 C 2.458567713 1.421054385 0.003703007 C 1.228614904 3.554143930 0.002513562 C -0.235163108 0.160826352 3.247131267 C 2.244119337 0.161060766 3.248352597 C -1.457317029 2.298122349 3.233987380 C 1.004488573 2.308122685 3.246477338 C 1.003490453 0.877294779 3.260744638 C -0.228726740 3.007907781 3.234369538 C 3.466400555 0.879138405 3.235278922 C 2.237111888 3.008093688 3.233133671 C 1.014869884 -0.620629524 6.400327421 C 3.476740887 -0.544001537 6.494151711 C -0.280354238 1.625208172 6.378065136 C 2.312052832 1.626645989 6.400900301 Ni 1.023326058 0.872911705 5.337347244 O 1.023326000 0.872911000 8.700000000 O 2.012052000 1.326645000 8.700000000 C -0.206008752 3.013580161 6.506783380 C 3.478548719 0.866737695 6.505990211 C 2.235793764 3.015530163 6.513155404 K_POINTS {automatic} 4 4 1 1 1 1 *-----------------* *Masoud Nahali* *PhD Student of Physical Chemistry Sharif University of Technology * Stefano Baroni wrote : > > On Sep 29, 2010, at 5:32 PM, vega lew wrote: > > > Through the relaxation of a oxygen molecule on a surface by QE4.2, it > dissociated and adsorbed on the surface. The scf energy of each step has > decreased in compared with the previous step until the relaxation completed. > Now, can one say that there is not any TS for O2 dissociation here ? > > > >> Probably not. The potential energy surface might be not very corrugated > and the energy barrier for the dissociation might be not very large, so that > the dissociation of O2 could happen in a relax calculation. I think there > might be a very minute TS in the dissociation process. > > here, I beg to differ. If a state, say B, can be reached from another > state, say A, following a path along which the energy always decreases, this > means *by definition* that there is *no* energy barrier between A and B. I > admit that if the energy barrier is so "minute" as to be smaller that the > numerical accuracy of the minimization algorithm, it can escape proper > detection. (notice that this can be said of most, if not all, numerical > computations of any quantity) > > > > > Or is the reaction spontaneous ? > > It is spontaneous > > >> If the total energy of system goes down continuously and the adsorption > energy of dissociated state is lower than the one of molecular state, you > could say the reaction is favorable. Whether it is spontaneous, it is depend > on how large of the energy barrier. > > in the reported conditions, the energy barrier is zero > > >> If there was a TS in this procedure (dissociation and adsorption of O2) > would the optimization algorithm of QE4.2 stop the work or continue ? > > In the presence of an energy barrier, the system would stay stuck in the > starting configuration > > Stefano > > > If there is only one bond broken in the process you considered, you could > try constrained optimization technique. For you information, please refer to > B.Hammer (Hammer, B.; Jacobsen, K. W.; Norskov, J. K. Phys. Rev. Lett. 1992, > 69, 1971) or PJ Hu's work. (Alavi, A.; Hu, P.; Deutsch, T.; Silvestrelli, P. > L.; Hutter, J. Phys. Rev. Lett. 1998, 80, 3650). Or recent developed method > by ZP Liu (J. Chem. Theory Comput. 2010, 6, 1136?1144 ) > > If I have not misunderstood some important point, no constrained > minimization or other advanced techniques are needed here. > > Stefano B > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/fd2cd4da/attachment.htm From matteo.calandra at impmc.jussieu.fr Thu Sep 30 15:50:01 2010 From: matteo.calandra at impmc.jussieu.fr (matteo calandra) Date: Thu, 30 Sep 2010 15:50:01 +0200 Subject: [Pw_forum] negative values in XANES (jiayudai) In-Reply-To: References: Message-ID: <20100930155001.10885gvznlnufc7d@www1.impmc.upmc.fr> Dear Jiayudai, it is a bit difficult to understand what do you mean without having a picture. Maybe you can try the following for the high pressure case. You take a GIPAW pseudopotential compatible with XSPECTRA without core hole and do again the calculation without cutting occupied states but inserting your ef_r value. Then you compare your results with the corresponding projected density of states calcolated with projwfc. For corresponding I mean corresponding to the choice of your polarization (and k-vector if you are calculating the quadrupolar part). They should be the same (or very similar as probably you don't have all the same parameters in the two calculations, number of bands, broadening...) except for a prefactor along the y axis. So you can check if you are not making a mistake somewhere. All the best, M. > > Message: 3 > Date: Thu, 30 Sep 2010 19:54:51 +0800 > From: "jiayudai" > Subject: Re: [Pw_forum] negative values in XANES > To: pw_forum at pwscf.org > Message-ID: > Content-Type: text/plain; charset="gb2312" > > Dear matteo, > > Thanks for your reply to me. The "negative" means the negative > position in x-axis. And i added the "ef_r" in the input file, and > the position of the absorption is reasonable now. However, when the > pressure is high (high density), the fermi energy should be also > very high. Even i added the "ef_r", the absorption after cutting the > contribution of occupied states is only a isolated peak. In the same > case but with lower density, there is not this problem. And the > analysis of density of states for the system tells me that there > should be absorption at the position of high energy. In this case, > the absorption before occ_cut and after occ_cut is very different > (the intensity is very weak, about 10-8). So, i can not find physcal > reason to explain it. Is there also some problems in my calculation? > > By the way, i changed some parameters of &occ_cut& part, but the > results were the same. > > Thanks again, and best wishes. > > Jiayu > > > Subject: Re: [Pw_forum] negative values in XANES (jiayudai) > To: PWSCF Forum > > Message-ID: > > Content-Type: text/plain; charset="iso-8859-1" > > Matteo, thank you for taking the time to participate in the > discussion. In the future, please hit "reply" when answering a post, > rather than starting a new thread. Have all a good day - Stefano B > > On Sep 30, 2010, at 9:13 AM, matteo calandra wrote: > >> Dear Jiayu, >> >> I assume you're using XSPECTRA to calculate Xanes K-edge. >> Maybe you can better specify what does it means "negative >> absorption". If this means that the value on the XANES intensity is negative >> then you must have some problem somewhere as the cross section >> implemented in XSPECTRA is positive definite. >> >> However, from your e-mail it is not very clear if the >> absorption is positive definite (as you say) or if the energy scale >> (the x axis in the plot) >> is negative but not the X-ray absorption intensity (y-axis in the plot). >> If y is positive and x is negative, then you simply forgot >> to specify the variable >> ef_r in the XSPECTRA input file. This variable tells you the zero >> energy of the scale. From the XSPECTRA input file: >> >> ef_r real(DP) >> DEFAULT=0.0 >> Fermi energy in Rydberg. This value combined >> with the option cut_occupied_states can be >> used to exclude the occupied states in a >> smooth way from the final plot. >> >> I hope it helps, >> >> M. >> >> >>> Dear developers and users, >>> >>> I am doing the calculation for the XANES of CHx system. There is a >>> problem that the absorption is always negative, locating at about >>> -100 eV. According to my >understanding, the absorption should be >>> positive. Besides, the fermi energy in >my systems is very high, >>> about tens of eV (high pressure). Therefore, if i used the cutoff of >>> fermi energy, the absorption should be very small. Is it >>>> reasonable? Or there is some problems in the calculations? >>> >>> Thanks in advance. >>> >>> Jiayu >> >> ---------------------------------------------------------------- >> This message was sent using IMP, the Internet Messaging Program. >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum > > --- > Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste > http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / > stefanobaroni (skype) > > La morale est une logique de l'action comme la logique est une > morale de la pens?e - Jean Piaget > > Please, if possible, don't send me MS Word or PowerPoint attachments > Why? See: http://www.gnu.org/philosophy/no-word-attachments.html > > > > > > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > http://www.democritos.it/pipermail/pw_forum/attachments/20100930/c1998bf7/attachment.htm > > ------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > End of Pw_forum Digest, Vol 39, Issue 57 > **************************************** > > > > ------------------------------ > > Message: 4 > Date: Thu, 30 Sep 2010 21:03:06 +0800 > From: "archygu" > Subject: [Pw_forum] the keyword in the pw.x > To: pw_forum > Message-ID: <201009302103032657359 at gmail.com> > Content-Type: text/plain; charset="gb2312" > > Dear QE Users, > > As a beginner of QE, I get problem in understanding some keywords in > PW.x, it will be very kind of you to give me a more > > detailed explanation: > > nbnd: acturally I do not know how to use this parameter , sometimes, > a lot of levels appear in the valence bands, but only several in the > conduction band. > so I want to know how can I accurately set the parameter, > then can obtain the energy levels properly. > > nosym: if it is true, that means the symmetry in the structure will > be ignored? > > > > nraise: how can we define one "nraise" step, it is the same as one > time step if we set nraise=1; > > many thanks for you help. > > > 2010-09-30 > > > > archygu,NTU > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > http://www.democritos.it/pipermail/pw_forum/attachments/20100930/2ef733ed/attachment.htm > > ------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > End of Pw_forum Digest, Vol 39, Issue 58 > **************************************** > ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From chakraa at tcd.ie Thu Sep 30 15:50:29 2010 From: chakraa at tcd.ie (Aurab) Date: Thu, 30 Sep 2010 14:50:29 +0100 Subject: [Pw_forum] Calculating g-tensor using gipaw Message-ID: Dear QE users and developers, I started to do the g-tensor calculation of ZnO using gipaw and got stuck with the error. from init_gipaw_1 : error # 1 impossible value for nrc I am using my own pseudopotentials with gipaw enabled. I looked into the source and found that the program init_gipaw_1.f90 (Version espresso-4.2.1) has it in line 133. Could you please instruct me how to get past this and what is the correct way to calculate g-tensor using gipaw. The gipaw input is given below, inputgipaw job = 'g_tensor' prefix = 'ZnO' tmp_dir = './' isolve = 0 iverbosity = 1 file_reconstruction(1) = 'Zn.pbe-gipaw.recon' file_reconstruction(2) = 'O.pbe-gipaw.recon' q_gipaw = 0.01 / Regards, Aurab D. Chakrabarty Trinity College Dublin *=*=*=*=*=*=*=*=*=*=*=*=*=*=* -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/6bc4364e/attachment.htm From Ari.P.Seitsonen at iki.fi Thu Sep 30 15:58:08 2010 From: Ari.P.Seitsonen at iki.fi (Ari P Seitsonen) Date: Thu, 30 Sep 2010 15:58:08 +0200 (CEST) Subject: [Pw_forum] Calculating g-tensor using gipaw In-Reply-To: References: Message-ID: Dear Aurab, One possibility is that the core radii read from the GIPAW file are zero; those are automatically written into to UPF file by the atomic/ld1 code, but for some reason that is not completely clear to me some of them are sometimes zero (occurs only for empty states). One way to check if this is your case you can try #> awk '/ Dear QE users and developers, > I started to do the g-tensor calculation of ZnO using gipaw and got stuck > with the error. > > ???? from init_gipaw_1 : error #???????? 1 > ???? impossible value for nrc > > I am using my own pseudopotentials with gipaw enabled. I looked into the > source and found that the program init_gipaw_1.f90 (Version espresso-4.2.1) > has it in line 133. Could you please instruct me how to get past this and > what is the correct way to calculate g-tensor using gipaw. > The gipaw input is given below, > > inputgipaw > ?job = 'g_tensor' > ?prefix = 'ZnO' > ?tmp_dir = './' > ?isolve = 0 > ?iverbosity = 1 > ?file_reconstruction(1) = 'Zn.pbe-gipaw.recon' > ?file_reconstruction(2) = 'O.pbe-gipaw.recon' > ?q_gipaw = 0.01 > / > > Regards, > > Aurab D. Chakrabarty > Trinity College Dublin > *=*=*=*=*=*=*=*=*=*=*=*=*=*=* > > From Lorenzo.Paulatto at impmc.upmc.fr Thu Sep 30 17:46:39 2010 From: Lorenzo.Paulatto at impmc.upmc.fr (Lorenzo Paulatto) Date: Thu, 30 Sep 2010 17:46:39 +0200 Subject: [Pw_forum] Bilayer graphene with an electric field applied In-Reply-To: References: <4CA2F0D7.1040205@sissa.it> Message-ID: In data 30 settembre 2010 alle ore 15:30:01, Elie Moujaes ha scritto: > The thing is there are no problems in the DOS calculations; only in the > band structure. Are you sure you did a proper band plot along a path in the Brillouin zone (like example01) and not just plot of the band energies of the regular grid of k-points? regards -- Lorenzo Paulatto post-doc @ IMPMC/UPMC - Universit? Paris 6 phone: +33 (0)1 44 27 74 89 www: http://www-int.impmc.upmc.fr/~paulatto/ previously (take note of the change!): phd student @ SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 www: http://people.sissa.it/~paulatto/ From metokal at gmail.com Thu Sep 30 20:47:57 2010 From: metokal at gmail.com (Mehmet Topsakal) Date: Thu, 30 Sep 2010 21:47:57 +0300 Subject: [Pw_forum] [SOLVED ?] Re: weird phonon frequencies for 3X 3X 1 KPOINTS In-Reply-To: <4CA494FB.9040506@sissa.it> References: <4CA494FB.9040506@sissa.it> Message-ID: Dear Gabriele, Thank you very much for your suggestion. After including some smearing, -5000 frequencies disappeared. But this still cannot explain the 3X 3X 1 situation. For this reason, I plotted the positions of scf KPOINTS (see the attachments) and realized that the corner of the Brillouin zone, which is also known as K-special-point, is NOT included for KPOINTS other than 3X 3X 1. For graphene, K-special-point also corresponds to the Dirac-Point where zero-band-gap occurs. As a result, problem should be related with K-special-point. BUT. Still I cannot understand why inclusion of a point in BZ corresponds to such a HUGE difference in phonon frequencies. Maybe someone in this list gives a better explanation for that.... All of the files are available at http://db.tt/xK2Jtu1 Ragards. On Thu, Sep 30, 2010 at 4:47 PM, Gabriele Sclauzero wrote: > Dear Mehmet, > > graphene has zero gap, hence you probably need to include smearing of > the occupations in the scf calculations. > > > HTH > > GS > > > On 09/29/2010 04:54 PM, Mehmet Topsakal wrote: > > Dear all, > > For graphene structure, I noticed that I get completely wrong phonon > frequencies when I use KPOINTS such as 9 9 1, 12 12 1, 15 15 1 > > For example: > > K_POINTS {automatic} > 11 11 1 0 0 0 > > gives realistic frequencies as in 11.phonon.out file > > BUT > > K_POINTS {automatic} > 12 12 1 0 0 0 > > gives "omega( 1) = -149.592762 [THz] = -4989.910586 [cm-1]" as > in 12.phonon.out > > > What should be the reason ???? > > Thanks. > > -- > > Mehmet Topsakal (Ph.D. Student) > UNAM-Institute of Materials Science and Nanotechnology. > Bilkent University. 06800 Bilkent, Ankara/T?rkiye > Tel: 0090 312 290 3527 ; Fax: 0090 312 266 4365 > UNAM-web : www.nano.org.tr > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.orghttp://www.democritos.it/mailman/listinfo/pw_forum > > > > -- > > Gabriele Sclauzero, EPFL SB ITP CSEA > PH H2 462, Station 3, CH-1015 Lausanne > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Mehmet Topsakal (Ph.D. Student) UNAM-Institute of Materials Science and Nanotechnology. Bilkent University. 06800 Bilkent, Ankara/T?rkiye Tel: 0090 312 290 3527 ; Fax: 0090 312 266 4365 UNAM-web : www.nano.org.tr -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100930/3d39d54a/attachment-0001.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: BZ.png Type: image/png Size: 98121 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100930/3d39d54a/attachment-0001.png -------------- next part -------------- A non-text attachment was scrubbed... Name: draw_bz.m Type: text/x-objcsrc Size: 9206 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100930/3d39d54a/attachment-0001.bin From snd2mra at yahoo.com Thu Sep 30 21:40:14 2010 From: snd2mra at yahoo.com (Madan Mithra. L. M) Date: Thu, 30 Sep 2010 12:40:14 -0700 (PDT) Subject: [Pw_forum] EXX on 1D supercells fail Message-ID: <38919.63978.qm@web53308.mail.re2.yahoo.com> Dear All, I compiled the latest version of pwscf (4.2.1) on IBM Power5 machine (xlf compiler) as well as on Dell Cluster (Intel ifort 11.1.056). In both cases, calculation including EXX (PBE0) fails on 1D supercell. However, the same job ran without dumping the core with pwscf 4.1.3. Any suggestion ? I would like to use the latest version as it supports restarting the EXX calculation. Thank you - Madan ---------------------------------------- Madan Mithra .L.M Senior Research Fellow Dept.of Physics Indian Institute of Science Bangalore - 560 012 INDIA ......................................... e-mail : mit at physics.iisc.ernet.in : snd2mra at yahoo.com Phone : 91-80-22932313 (Lab) FAX : (080) 360 2602 -----------------------------------------