Yue-Wen Fang wrote on *Thu May 9 15:40:00 CEST 2013*
> Dear all, > > when we need to make a relaxation of heterojunction (supercell, such as a > system contains 40 atoms), *which one in the following choices is the best > way?* > *A:* full relaxation (volume shape, lattice constant and atomic positions > vary) > *B: * fix the lattice constant > when you don't know the details of your system or it is not a common solid state system you should relax all the parameters. But if you know some little details of your system you can decrease the cost of variable-cell calculations. For instance if your system is a simple cubic with orthogonal vectors you can simply do some SCF calculation and plot E versus lattice constant to have an estimate of your cell parameter. Also it is not bad to use this cell constant as an input in a relaxation (not variable-cell one) process to find the optimized position of atoms. In this way you may be known more about your system and it is easier to start a vc-relax calculation. > any other better relaxation manners is welcome > thank you very much! > You can directly go to doing "vc-relax" and see whether the above procedure is less expensive or the current one ! let us know :) > > -- > > ------------------------------------------------------------------------------------------------------------ > Yue-Wen FANG, PhD candidate > Key Laboratory of Polar Materials and Devices, Ministry of > Education<http://clpm.ecnu.edu.cn/> > East China Normal University <http://www.ecnu.edu.cn/english/> > I will persist until I succeed! > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > http://pwscf.org/pipermail/pw_forum/attachments/20130509/6d87d04d/attachment-0001.html > Best Wishes, m ---------------------------------------- Masoud Nahali SUT masoud.nahali at gmail.com alum.sharif.edu/~m_nahali <http://alum.sharif.edu/%7Em_nahali> -------------- next part -------------- An HTML attachment was scrubbed... URL: http://pwscf.org/pipermail/pw_forum/attachments/20130510/5d0e0b9c/attachment.html
