ISIF, box and ions relaxarion

Message

pmignon · #1 Post by **pmignon** » Thu Jun 14, 2007 7:39 am

Hello,

I am tying to relax both the box and ions but keeping the same shape for the box. By using ISIF=3 the shape is also changed. So I do a volume relax(without shape ISIF=7) + ions relax(ISIF=2) + volume relax (without shape) +ions relax + ... until convergence
Do you see/know other possibilities than the multiple step that i am doing to do so ? and keeping the same shape ?
thank you,
pierre.

#2 Post by **admin** » Fri Jun 15, 2007 11:57 am

I am afraid this will be the only way to do it. If the ions are not kept in their internal positions by symmetry, ISIF 7 and 2 runs will be necessary. As soon as the geometry is pre-converged, you could do a few steps of 7 and 2, respectively, before you switch the relaxation type.

nab · #3 Post by **nab** » Mon Sep 07, 2009 10:36 pm

Hello,

I have an orthorhombic cell and so I am also performing isif=7 and 2 runs. What is the stopping criterion for isif=7 calculations? I thought it would be the pressure, but my calculations continue even after the external pressure = 0 kB.

Thank you,
Nicole

#4 Post by **admin** » Mon Oct 12, 2009 1:14 pm

if your run continues, some atoms are not in their equilibrium positions yet (please have a look whether the forces on the atoms as written in OUTCAR match the EDIFFG criterium) eventually, you will have to do another run of ionic relaxations.