I am optimizing geometry of a slab with vacuum. The cutoff energy is 400eV with LDA exchange-correlation functionals.
Firstly, I got the convergence structure without applying dipole correction( EDIFF = 5.D-4 and EDIFFG = -5.D-2).
Secondly, I set LDIPOL = .TRUE. and IDIPOL = 3. Whereas, the forces became very large and the structure could note reach convergence.
I am sure that my structure is correct because I have compared mine to others' for many times.
![Smiley :)](./images/smilies/smiley.gif)
regards!
Yuewen