Thank you very much, Prof. Giannozzi.<br>I added nr1 = 32, nr2 = 32, nr3 = 32 in the namecard of SYSTEM, then the vc-relax works well.<br><pre><br>At&nbsp;2011-04-12&nbsp;21:17:11，"Paolo&nbsp;Giannozzi"&nbsp;&lt;giannozz@democritos.it&gt;&nbsp;wrote:

&gt;On&nbsp;Tue,&nbsp;2011-04-12&nbsp;at&nbsp;14:34&nbsp;+0800,&nbsp;GAO&nbsp;Zhe&nbsp;wrote:

&gt;

&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;from&nbsp;sym_rho_init_shell&nbsp;:&nbsp;error&nbsp;#&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;2

&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;lone&nbsp;vector

&gt;

&gt;this&nbsp;is&nbsp;a&nbsp;consequence&nbsp;of&nbsp;non-local&nbsp;interactions&nbsp;between&nbsp;recent

&gt;changes&nbsp;to&nbsp;the&nbsp;way&nbsp;G-vectors&nbsp;are&nbsp;computed&nbsp;and&nbsp;the&nbsp;final&nbsp;scf&nbsp;step,

&gt;performed&nbsp;in&nbsp;variable-cell&nbsp;optimization&nbsp;only,&nbsp;with&nbsp;G-vectors&nbsp;set&nbsp;

&gt;for&nbsp;the&nbsp;final&nbsp;cell&nbsp;(and&nbsp;not&nbsp;for&nbsp;the&nbsp;initial&nbsp;cell&nbsp;as&nbsp;it&nbsp;is&nbsp;done&nbsp;during

&gt;the&nbsp;optimization).&nbsp;Apparently&nbsp;in&nbsp;this&nbsp;last&nbsp;step&nbsp;some&nbsp;G-vectors&nbsp;are

&gt;"lost"&nbsp;and&nbsp;the&nbsp;algorithm&nbsp;that&nbsp;collects&nbsp;G-vector&nbsp;shells&nbsp;complains.

&gt;&nbsp;

&gt;For&nbsp;the&nbsp;time&nbsp;being,&nbsp;waiting&nbsp;for&nbsp;a&nbsp;better&nbsp;solution,&nbsp;increase&nbsp;the

&gt;FFT&nbsp;grid&nbsp;to&nbsp;nr1=32,&nbsp;nr2=32,&nbsp;nr3=32.&nbsp;Also&nbsp;note&nbsp;that&nbsp;you&nbsp;can&nbsp;get&nbsp;the&nbsp;

&gt;final&nbsp;coordinates&nbsp;and&nbsp;lattice&nbsp;parameters&nbsp;from&nbsp;output&nbsp;and&nbsp;perform

&gt;the&nbsp;final&nbsp;scf&nbsp;calculation&nbsp;(as&nbsp;it&nbsp;used&nbsp;to&nbsp;be&nbsp;done&nbsp;some&nbsp;time&nbsp;ago)

&gt;if&nbsp;you&nbsp;want.&nbsp;The&nbsp;last&nbsp;step&nbsp;is&nbsp;done&nbsp;only&nbsp;to&nbsp;make&nbsp;it&nbsp;clear&nbsp;what

&gt;the&nbsp;result&nbsp;will&nbsp;look&nbsp;like&nbsp;with&nbsp;the&nbsp;final&nbsp;cell.

&gt;

&gt;P.

&gt;--&nbsp;

&gt;Paolo&nbsp;Giannozzi,&nbsp;IOM-Democritos&nbsp;and&nbsp;University&nbsp;of&nbsp;Udine,&nbsp;Italy

&gt;

&gt;

&gt;_______________________________________________

&gt;Pw_forum&nbsp;mailing&nbsp;list

&gt;Pw_forum@pwscf.org

&gt;http://www.democritos.it/mailman/listinfo/pw_forum

</pre><br><br><div>--<br>GAO Zhe<br>CMC Lab, MSE, SNU, Seoul, S.Korea<br>        

</div><br><br><span title="neteasefooter"><span id="netease_mail_footer"></span></span>