Are you doing a relax calculation? If yes, check the upscale parameter in the doc
-- Lorenzo Paulatto Written on a virtual keyboard with real fingers On Dec 6, 2017 18:28, "David Foster" <[email protected]> wrote: Hi I have set "Conv_Thr" to 1.0D-5 in my calculation. As you see follow, the estimated scf change is less than 10^(-5) but it has not been converged!!! Is there any criteria for SCF convergence in the code which has not been documented in the manual? ====== total cpu time spent up to now is 8119.0 secs total energy = -16045.57783278 Ry Harris-Foulkes estimate = -16045.57783128 Ry estimated scf accuracy < 0.00000207 Ry total magnetization = -0.01 Bohr mag/cell absolute magnetization = 79.80 Bohr mag/cell iteration #164 ecut= 140.00 Ry beta=0.01 Davidson diagonalization with overlap ethr = 3.21E-10, avg # of iterations = 2.0 total cpu time spent up to now is 8138.3 secs total energy = -16045.57783497 Ry Harris-Foulkes estimate = -16045.57783360 Ry estimated scf accuracy < 0.00000201 Ry total magnetization = -0.01 Bohr mag/cell absolute magnetization = 79.80 Bohr mag/cell iteration #165 ecut= 140.00 Ry beta=0.01 Davidson diagonalization with overlap ethr = 3.21E-10, avg # of iterations = 2.0 ================================= _______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
_______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
