On 31 January 2013 08:32, Wei-Bing Zhang <span dir="ltr"><<a href="mailto:weibingzhangavh@gmail.com" target="_blank">weibingzhangavh@gmail.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
(i.e vc-relax directly and<br>
atomic relaxation with different lattice parameters).<br></blockquote><div><br></div><div>Dear Wei,</div><div>please triple-check that you are doing everything correctly: </div><div>* that after the relax you are putting back the final parameters in the pw input</div>
<div>* that in the SCF calculation that precedes the phonon one, forces and stress are actually zero</div><div><br></div><div>I would also suggest to NOT play to input parameters: you are using input_dft and exx_div_threatement both of which do not apply in your case. Finally, ecutwfc = 65 is incredibly high for a Si ultrasoft pseudopotential. Try to read something about this subject i.e. here <<a href="http://www.stanford.edu/~hkulik/www/Tutorials/Entries/2012/4/17_Convergence_101.html">http://www.stanford.edu/~hkulik/www/Tutorials/Entries/2012/4/17_Convergence_101.html</a>> (the "dual" is ecutrho/ecutwfc)</div>
<div><br></div><div>bests</div></div><br clear="all"><div><br></div>-- <br><div>Dr. Lorenzo Paulatto </div><div>IdR @ IMPMC -- CNRS & Université Paris 6</div><div>phone: +33 (0)1 44275 084 / skype: paulatz</div><div>www: <a href="http://www-int.impmc.upmc.fr/~paulatto/" target="_blank">http://www-int.impmc.upmc.fr/~paulatto/</a></div>
<div>mail: 23-24/4é16 Boîte courrier 115, 4 place Jussieu 75252 Paris Cédex 05</div>