----- RAHUL SURESH <drrahulsur...@gmail.com> a écrit : >Hi.. Thank you for your reply. My input file is as follows:(...) scf: iscf Eharris(eV) E_KS(eV) FreeEng(eV) dDmax Ef(eV) scf: 1 -557650.0968 20381.7835 20381.6763753.10403******** scf: 2 -4630269.6330 2077870.1836 2077870.1745864.83778******** scf: 3 -2261484.3148 854868.7375 854868.6831***************** scf: 4 -1331002.0702 338981.5615 338981.4669425.38740******** scf: 5 -719651.9273 106868.1468 106868.1251***************** scf: 6 -717147.8071 68971.1766 68971.0943426.43790******** Can I just add MD.UseSaveCG .true. MD.UseSaveXV .true. DM.UseSaveDM .true.
input file without changing anything to the fdf file and re-run? Thank you ====================Dear Rahul:you can do whatever you want, but please take into account thatyour calculation does not converge and produces strongly meaningless results,and this not as it crashed but from the very beginning : scf: iscf Eharris(eV) E_KS(eV) FreeEng(eV) dDmax Ef(eV) scf: 1 -89709.5924 -89842.8329 -89842.9104 1.93966 -6.9260 timer: Routine,Calls,Time,% = IterSCF 1 152.571 78.20 scf: 2 -89677.0881 -89355.7591 -89355.9226 0.51386 -3.5728 scf: 3 -89796.1277 -89389.5132 -89389.7502 0.97916 -2.3163 scf: 4 -105751.2656 -88983.4222 -88983.7173 54.29636 -8.1884 scf: 5 -491486.7932 -40257.9792 -40258.0558259.16731********and in fact your output tells you this in plain words: scf: 999 -779797.2714 78350.0293 78349.9655683.29828******** scf: 1000 -603514.1657 49775.9080 49775.8345***************** SCF_NOT_CONV: SCF did not converge in maximum number of steps. So a good idea might be to fix this, in the first place.Some suggestions:1) Is the structure correct? (Did you look at it?)If yes, could a smaller fragment than 648 atoms be selectedfor a trial calculation to check that it works correctly?2) Did you try a serial calculation? Would it converge? 3) What was the mixing parameter?Etc.Please ask this mailing lest responsibly; this is for asking questions and discussing specific problems,not for depositing millions of lines of output files.Good luckBest regardsAndrei Postnikov
-- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)