[QE-users] [SUSPECT ATTACHMENT REMOVED] vc-relax changes the lattice constants too much

[Xiaowei Zhang]

Hi experts,
         I came across a problem when I was doing variable-cell relaxations for the metal Gallium. 
1)	I started the relaxations using the experimental structure.
2)	I chose the PBE xc functional and the Dojo-ONCV pseudopotential. 
3)	As the attached log files show, after several ionic steps, the b axis is shorten by near 0.8 angstrom and at the same time, the c axis is lengthen by near 0.6 angstrom.  
4)	I’ve tried to increase the ecutwfc to 130 Ry, but it didn’t improve at all.
5)	I tried to use the ultrasoft pseudopotential, but things were not getting better.
6)	 I also used VASP to do the same thing and VASP gives the optimized lattice constants very similar to the experimental ones.
I’m very confused. Look forward to any suggestions and helps!

Thanks!

Bet wishes,
Xiaowei

------------------------------
Ph.D. Student Xiao-Wei Zhang
ICQM, School of Physics, Peking University
Email: willzxw at pku.edu.cn
------------------------------


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20200720/89f21fe7/attachment.html>