On top of what I said (assuming that your PPs are good): maybe you should also try to increase cutoffs of wavefunctions or charge density. You can first do a convergence test on the forces computed after a simple scf run (remember that forces are harder to converge than total energy).
GS Gabriele Sclauzero wrote: > My "semi-empirical" suggestion is to restart with ion_dynamics='damp' and see > if the > problem persist. Maybe you can do a few steps of damped dynamics and then > restart again > with BFGS. Probably this will not solve your problem, but you can give a try > ;-) > > HTH > > GS > > > PS: you should chose better the subject of your posts next time... > > > Julen Ibanez Azpiroz wrote: >> Hello >> >> I am trying to relax a system with unit cell made of 24 silicon atoms >> and 2 talium atoms, one in each edge. I wish the system to be a surface, >> so the unit cell has vacuum between the talium atoms so that they do not >> "see" each other. I tried to relax it following the Example03 of Quantum >> Espresso which is a surface of "al001". In the example there are two >> methods, I followed the first one, the one of the "bfgs" method. When I >> use Scalar_Relativistic pseudopotentials (for both atoms) it is Ok, >> after some trials the system relaxes. However, when I use Relativistic >> pseudopotentials (for both atoms) the system cannot relax, after many >> trials the system repeats the atomic configuration and stops. I have >> tried both, to let all the atoms move and to let move just a few of the >> surface, keeping the other fixed, but it does not work anyway. The most >> I can get is that the force in every atom is smaller than 0.01Ry/au. >> Both pseudopotentials work allright in simpler models. I would >> appreciate any suggestion. Thank you very much >> >> Julen >> >> UPV/EHU, Fisica de la Matertia Condensada >> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum > -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o
