Your ecutwfc is too low. Check that. and ecutrho should be 8-10 times ecutwfc. The celldm(1) cannot be 1.0. What are the units? check that.<br><br><div class="gmail_quote">On Fri, Aug 27, 2010 at 11:29 AM, joaquin peralta <span dir="ltr"><<a href="mailto:jperaltac@gmail.com">jperaltac@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">Dear pw_forum<br><br>I try to simulate a metallic slab under a electric field. In order to improve the convergence, I checked this web page :<br>
<br><a href="http://blog.sina.com.cn/s/blog_5f15ead20100cw4d.html" target="_blank">http://blog.sina.com.cn/s/blog_5f15ead20100cw4d.html</a><br>

<br>I tried all the suggestion (except band number increase) and the process continue with a very slow convergence, after 283 iterations the ethr=1.2E-9. The important part of the input file is :<br><br>&CONTROL<br>   calculation           = 'scf'<br>


   ...<br>   tefield               = .true.<br>   dipfield              = .true.<br>/<br>&SYSTEM<br>   ibrav                 =  0<br>   celldm(1)             =  1.0<br>   nat                   = 21<br>   ntyp                  = 1<br>


   ecutwfc               = 30<br>   ecutrho               = 500<br>   degauss               = 0.03<br>   occupations           = 'smearing'<br>   smearing              = 'mv'<br>   nspin                 = 2<br>


   starting_magnetization= 0.1<br>   edir                  = 3<br>   emaxpos               = 0.9<br>   eopreg                = 0.2<br>   eamp                  = 0.019447<br>/<br>&ELECTRONS<br>   electron_maxstep      = 300<br>


   mixing_mode           = 'local-TF'<br>   mixing_beta           = 0.1<br>   mixing_ndim           = 10<br>/<br>...<br>K_POINTS (automatic)<br>6 6 6 0 0 0<br>...<br>This process use approx 5Gb of data in the associated files.<br>


<br clear="all">My question are, <br>-If I increase the band number, The space disk will be more?<br>-Do you have suggestion about how to proceed in order to avoid this I/O increase. and improve the convergence or velocity. <br>


-I'm (over/under)estimating some previous value?, k-points, mixing_*, ecut*?<br><br>Thanks in advance<br>Joaquin Peralta<br>Materials Science and Engineering<br>Iowa State University.<br><br>-- <br>----------------------------------------------------<br>


Group of NanoMaterials<br>----------------------------------------------------<br><a href="http://www.gnm.cl" target="_blank">http://www.gnm.cl</a><br>----------------------------------------------------<br>Joaquín Andrés Peralta Camposano<br>


----------------------------------------------------<br><a href="http://www.lpmd.cl/jperalta" target="_blank">http://www.lpmd.cl/jperalta</a><br><br>In a world without frontiers, <br>who needs Gates and Win.<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>
<br></blockquote></div><br><br clear="all"><br>-- <br>Karthik Guda<br>Graduate Student<br>Materials Engineering<br>Purdue University<br><br>