Page 1 of 1

convergence

Posted: Tue Jun 02, 2009 3:55 pm
by shurab
I am doing structure relaxation with EDIFFG=-0.01.
Every now and then I get the following message
"aborting loop because EDIFF is reached"
and termination of the structural relaxation
without typical message
"reached required accuracy - stopping structural energy minimisation"
Looks like it is not a normal ending of the atomic relaxation?

convergence

Posted: Wed Jun 10, 2009 12:43 pm
by yuxiao_zihan
I think you could only get this information "reached required accuracy - stopping structural energy minimisation" when you do dynamics. If you do not set NSW in INCAR and dynamics 'T T T' in POSCAR, you will not get this message.