<DIV> </DIV>
<DIV>>Message: 2<BR>>Date: Thu, 15 Oct 2009 15:42:34 +0200<BR>>From: "Lorenzo Paulatto" <paulatto@sissa.it><BR>>Subject: Re: [Pw_forum] total force ascend in the course of VC-relax<BR>>To: "PWSCF Forum" <pw_forum@pwscf.org><BR>>Message-ID: <op.u1ugc8i1a8x26q@paulax><BR>>Content-Type: text/plain; charset=utf-8; format=flowed; delsp=yes<BR>><BR>>In data 15 ottobre 2009 alle ore 15:38:39, Q.J.Wang <wangqj1@126.com> ha <BR>>scritto:<BR>><BR>>> Dear all<BR>>> I did geometry optimization with vc-relax in a supercell. When I <BR>>> increase the concentration of doped atoms ,the value of total force <BR>>> ascend instead of declining.Before I did not increase the concentration <BR>>> of doped atoms ,It converges well .Anyone who encounter the same problem <BR>>> as me ?<BR>><BR>>I sounds normal to me, when you perturb the system you will have to redo <BR>>the relaxation... why do you find it puzzling?</DIV>
<DIV>The problem is when I redo the relaxation ,the value of 'Total force'and the 'P' become large in every bfgs cycle .And at last shows error :</DIV>
<DIV> <EM> from bfgs : error # 1<BR></EM><I> bfgs history already reset at previous step</I></DIV>
<DIV><I>The input file as following :</I></DIV>
<DIV><I> &CONTROL<BR> calculation = 'vc-relax' ,<BR> restart_mode = 'from_scratch' ,<BR> outdir = '/root/work3/' ,<BR> pseudo_dir = '/root/pseudo/' ,<BR> prefix = 'GaAs' ,<BR> nstep = 250 ,<BR> tstress = .true. ,<BR> tprnfor = .true. ,<BR> /<BR> &SYSTEM <BR> nat = 48,<BR> ntyp = 4,<BR> ecutwfc = 30 ,<BR> ecutrho = 300 ,<BR> nosym = .true. ,<BR> occupations = 'smearing' ,<BR> degauss = 0.005 ,<BR> smearing = 'gaussian' ,<BR> nspin = 2 ,<BR> starting_magnetization(1) = 0.5,<BR> starting_magnetization(2) = 0.5,<BR> starting_magnetization(3) = 0.5,<BR> starting_magnetization(4) = 0.5,<BR> lda_plus_u = .false. ,<BR> /<BR> &ELECTRONS<BR> conv_thr = 1.D-7 ,<BR> mixing_mode = 'plain' ,<BR> mixing_beta = 0.3 ,<BR> /<BR> &IONS<BR> ion_dynamics = 'bfgs' ,<BR> phase_space = 'full' ,<BR> pot_extrapolation = 'second_order' ,<BR> wfc_extrapolation = 'second_order' ,<BR> upscale = 100 ,<BR> /<BR> &CELL<BR> cell_dynamics = 'bfgs' ,<BR> /<BR>ATOMIC_SPECIES</I></DIV>
<DIV><EM>Ga ...........</EM></DIV>
<DIV><I>ATOMIC_POSITIONS angstrom </I></DIV>
<DIV><EM>...............</EM></DIV>
<DIV><I> K_POINTS automatic <BR> 3 3 3 1 1 1 <BR></I><I><BR></I></DIV>
<DIV>><BR>><BR>>> The second problem is when I increase the relative distance of the <BR>>> two doped atoms in the supercell and did geometry optimization with <BR>>> vc-relax,and the error turn up as like : from c_bands : error # 1<BR>>> too many bands are not converged<BR>>> So I am puzzled why before I increase relative position of two doped <BR>>> atoms the vc-relax converged and the latter turns up error in the same <BR>>> parameter .<BR>><BR>>It happens with peculiar configurations, the easiest way to get rid of it <BR>>is normally to increase the number of bands.<BR>><BR>>cheers<BR><BR>>-- </DIV>
<DIV>Best regards</DIV>
<DIV> </DIV>
<DIV>Q.J.Wang</DIV>
<DIV> </DIV>
<DIV>XiangTan University </DIV>
<DIV><BR><BR></DIV><br><br><span title="neteasefooter"/><hr/>
<a href="http://allyes.nie.163.com/main/adfclick?db=afanie&bid=1260,614,23&cid=148,4,1&sid=1357&show=ignore&url=http://tx2.163.com/fab.html">看陆川杨幂新片《琴棋书画》,品网易3D国韵网游《天下贰》</a>
</span>