Dear Payam, what you report is consistent with a too high threshold for the self-consistency loop (ie forces are not computed accurately enough): try reducing conv_thr (say to 1d-8) or use upscale.
Also have a look to the total energy: is that decreasing, while forces are fluctuating? Then, the system is still relaxing towards its minimum even though forces are small. HTH, Guido PS Notice that it occurs sometimes that the system relaxes in one direction, reducing the force quite rapidly. Then the algorithm finds new possibilities to further reduce the total energy, yet the forces are small. An example of that might be when you put a strongly deformed molecule close to a surface, above an adsorption site which is not the optimal one. First, the molecule relaxes more or less where you placed it, then it finds its way towards another adsorption site (if symmetry allows it to do so). Il 03/15/2012 04:25 PM, Payam Norouzzadeh ha scritto: > Dear QE users > > I usually look at total force to check if my structure is being relaxed > or not.Most of times it reduces to a certain level (~0.003) continuously > and > then starts to fluctuate. I mean it reverses the trend of reducing total > force.Simply total force begins to grow and sometimes never gets smaller. > The question is that how can I push QE to reduce total force > continuously when I am trying to relax a structure?Why total force > fluctuates? > > Best regards,Payam Norouzzadeh > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- Guido Fratesi Dipartimento di Scienza dei Materiali Universita` degli Studi di Milano-Bicocca via Cozzi 53, 20125 Milano, Italy
