[QE-users] What is the difference between 'relax' and 'vc-relax' when cell_dofree is set to '2Dxy' ?

[连云龙]

Dear QE users,




I am trying to understand the difference between 'relax' and 'vc-relax' when cell_dofree is set to '2Dxy' in the case of 2D structure relaxation.




According to the help files, the 'relax' mode only moves the ions and the 'vc-relax' mode move both the ions and the cell parameters. In the 2D case, is it exactly the same? 




Moreover, how to see that in the source code? I found in ./Modules/read_namelists.f90 that for the 'relax' mode, only ion_dynamics = 'bfgs' has been set, while for the 'vc-relax' mode, both cell_dynamics and ion_dynamics are set to 'bfgs'. Meanwhile, the iforceh matrix serves as a mask to control the behaviour of the relaxation of ion positions. But how does this matrix affect the relaxation of the cell parameters for the 'vc-relax' mode ?




One more related question : is there a setting for the relaxation "all but cell parameter c" with 'vc-relax' in the 2D case? I mean, can I set a mode in which all parameters can change but the cell parameter c is fixed (which is usually set to a large value)?




Thank you very much for any hints !




Best regards,

Yunlong LIAN




ORCID : https://orcid.org/0000-0002-9474-2181

CSNS, Dongguan Neutron Science Center
Dongguan, Guangdong, P. R. China

连云龙
中国散裂中子源 中子科学部 
A1-629 (内线 303)
中国广东省东莞市
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20201130/f5887776/attachment.html>