lattice constant optimize

Problems running VASP: crashes, internal errors, "wrong" results.


Moderators: Global Moderator, Moderator

Post Reply
Message
Author
huijiany
Newbie
Newbie
Posts: 1
Joined: Thu Nov 27, 2008 2:03 am

lattice constant optimize

#1 Post by huijiany » Fri Feb 27, 2009 12:46 am

when optimize the lattice constant, I just want to make C axis
change, and keep a and b axis unchanged. How can I do?
Last edited by huijiany on Fri Feb 27, 2009 12:46 am, edited 1 time in total.

forsdan
Sr. Member
Sr. Member
Posts: 339
Joined: Mon Apr 24, 2006 9:07 am
License Nr.: 173
Location: Gothenburg, Sweden

lattice constant optimize

#2 Post by forsdan » Fri Feb 27, 2009 7:38 am

There is no automatic way to do this in VASP, since there is no support in the current version for this kind of constraint relaxation.

The recommended thing that you can do is however to write a small script that generate your input files with different c-axis where you keep the a,b axis unchanged, and then run VASP for each input-set under constant volume and constant shape.


Best regards,
/Dan
Last edited by forsdan on Fri Feb 27, 2009 7:38 am, edited 1 time in total.

mrm
Newbie
Newbie
Posts: 20
Joined: Mon Jul 07, 2008 2:24 am

lattice constant optimize

#3 Post by mrm » Sat Feb 28, 2009 2:58 am

Hello,

I think it is possible to use "Selective Dynamics" in the POSCAR file and choose explicitly (by saying T or F for each x, y, z coordinate of each atom) if a degree of freedom can change. I believe that you can constrain the degrees of freedom to the c-axis in this way. Please check VASP manual to make sure about this. I hope this helps (and I hope I am not wrong). :-)
Last edited by mrm on Sat Feb 28, 2009 2:58 am, edited 1 time in total.

forsdan
Sr. Member
Sr. Member
Posts: 339
Joined: Mon Apr 24, 2006 9:07 am
License Nr.: 173
Location: Gothenburg, Sweden

lattice constant optimize

#4 Post by forsdan » Sat Feb 28, 2009 10:38 am

Dear mrm,

The 'Selective Dynamics' tag only locks the direct coordintes of the atoms and not the cartesian coordinates. This implies that you can't use this tag to minimize the c-axis with respect to an arbitrary constraint on the magnitude of the a,b lattice vectors.

cheers,
/Dan

<span class='smallblacktext'>[ Edited Sat Feb 28 2009, 11:40AM ]</span>
Last edited by forsdan on Sat Feb 28, 2009 10:38 am, edited 1 time in total.

Post Reply