In my opinion, the stricter the criterion, the more ecut and more K points sampling.<br><br>
<div class="gmail_quote">2009/9/2 Huiqun Zhou <span dir="ltr"><<a href="mailto:hqzhou@nju.edu.cn">hqzhou@nju.edu.cn</a>></span><br>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">Pablo,<br><br>Usually, setting conv_thr = 1d-8 is sufficient, please make sure if you<br>need such strict convergence criterion.<br>
<font color="#888888"><br><br>Huiqun Zhou<br>@Earth Sciences, Nanjing University, China<br></font>
<div>
<div></div>
<div class="h5"><br><br>----- Original Message -----<br>From: "Pablo Aguado" <<a href="mailto:paguado@gmail.com">paguado@gmail.com</a>><br>To: <<a href="mailto:pw_forum@pwscf.org">pw_forum@pwscf.org</a>><br>
Sent: Tuesday, September 01, 2009 6:38 PM<br>Subject: [Pw_forum] Relaxation doesn't converge<br><br><br>> Dear all,<br>><br>> I'm doing some tests on the structural optimization with pw. I'm<br>> currently testing with the tetragonal structure of PbTiO3, using the<br>
> following input:<br>><br>> &control<br>> calculation = 'vc-relax'<br>> restart_mode = 'restart'<br>> pseudo_dir = '/****/espresso-4.1/pseudo/'<br>> outdir = '/***/tmp'<br>
> forc_conv_thr = 1d-3<br>> nstep = 50<br>> /<br>> &system<br>> ibrav=6<br>> celldm(1)=7.3699<br>> celldm(3)=1.04<br>> nat=5<br>> ntyp=3<br>> nbnd=28<br>> ecutwfc=30.0<br>
> occupations = 'fixed'<br>> /<br>> &electrons<br>> conv_thr = 1d-12,<br>> mixing_beta=0.4,<br>> /<br>> &ions<br>> ion_dynamics = 'bfgs'<br>> bfgs_ndim = 5<br>
> /<br>> &cell<br>> cell_dynamics = 'damp-pr'<br>> press_conv_thr = 1.0d0<br>> cell_dofree = 'z'<br>> /<br>> ATOMIC_SPECIES<br>> Pb 207.2 Pb.pz-d-van.UPF<br>> Ti 47.867 Ti.pz-sp-van_ak.UPF<br>
> O 15.9994 O.pz-van_ak.UPF<br>> ATOMIC_POSITIONS crystal<br>> Pb 0.000 0.000 0.040 0 0 1<br>> Ti 0.500 0.500 0.520 0 0 1<br>> O 0.000 0.500 0.500 0 0 1<br>> O 0.500 0.500 0.000 0 0 1<br>
> O 0.500 0.000 0.500 0 0 1<br>> K_POINTS automatic<br>> 6 6 6 1 1 1<br>><br>> The relaxation is taking forever so I've checked the output and I've<br>> found several steps where everything seems to be converged (notice I'm<br>
> only relaxing out-of-plane lattice vector) but the code keeps running.<br>> An example:<br>><br>> - Energy difference = 4e-7<br>> - Forces acting on atoms (Ry/au):<br>><br>> atom 1 type 1 force = 0.00000000 0.00000000 0.00085679<br>
> atom 2 type 2 force = 0.00000000 0.00000000 -0.00001779<br>> atom 3 type 3 force = 0.00000000 0.00000000 -0.00024617<br>> atom 4 type 3 force = 0.00000000 0.00000000 -0.00034665<br>
> atom 5 type 3 force = 0.00000000 0.00000000 -0.00024617<br>><br>> Total force = 0.000988 Total SCF correction = 0.000002<br>><br>> total stress (Ry/bohr**3) (kbar)<br>
> P= -0.84<br>> -0.00001150 0.00000000 0.00000000 -1.69 0.00 0.00<br>> 0.00000000 -0.00001150 0.00000000 0.00 -1.69 0.00<br>> 0.00000000 0.00000000 0.00000591 0.00 0.00 0.87<br>
><br>> Any idea of what's going on?<br>><br>> Thank you very much<br>><br>> Pablo<br>> _______________________________________________<br>> Pw_forum mailing list<br>> <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
> <a href="http://www.democritos.it/mailman/listinfo/pw_forum" target="_blank">http://www.democritos.it/mailman/listinfo/pw_forum</a><br>><br><br>_______________________________________________<br>Pw_forum mailing list<br>
<a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br><a href="http://www.democritos.it/mailman/listinfo/pw_forum" target="_blank">http://www.democritos.it/mailman/listinfo/pw_forum</a><br></div></div></blockquote>
</div><br><br clear="all">
<div></div><br>-- <br>Y. C. Cheng<br>Department of Physics<br>Nanjing University<br>Nanjing 210093<br>P. R. China<br>Tel: 86-25-83592907<br>Email: <a href="mailto:yccheng.nju@gmail.com">yccheng.nju@gmail.com</a><br>