<div dir="ltr"><div><div><div><div><br></div>Dear Dmitry,<br><br></div>what you describe is probably true for all the components of the forces, not just the Hubbard one. The reason is that the Hellmann-Feynman theorem, used to compute the forces printed in the output, only applies to the total energy, not to its separate components. In fact, the electronic wave functions used in the calculation of the matrix elements are eigenstates of the whole Hamiltonian, not of its pieces.<br></div>Of course, the code still compute the forces "piece by piece". So each of those term is "wrong" (in the sense it does not equal the derivative of the corresponding term of the energy) but when you sum them up together, these errors cancel each other and you get the right force.<br><br><br></div><div>Regards,<br><br>Matteo<br></div><div><div><br><br><br><div><div><div class="gmail_extra"><br><div class="gmail_quote">On Thu, Oct 23, 2014 at 10:24 AM, Dmitry Novoselov <span dir="ltr"><<a href="mailto:dnovoselov@gmail.com" target="_blank">dnovoselov@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div>Dear all,<br><br></div>I have performed the set of <span>LSDA</span>+U calculations to determine the Hubbard forces acting on Ni atom in a <span lang="en"><span>well-known <span>NiO</span>.<br></span></span><div><div></div><div>For this purpose I was displacing one Ni atom in the x-direction up to 0.1 angstroms with 0.025 angstroms step.</div><div><br></div><div>How we know a force may be evaluate like:<br><div style="margin-left:40px">$F_{\alpha i} = -\frac{\partial E}{\partial \tau_{\alpha i}}$.<br></div></div><div>That allows us to calculate a force by <span lang="en"><span>taking</span> a<span> numerical derivative of the energy </span></span><span lang="en">with respect to the displacement $</span><span lang="en">\tau_{\alpha i}$ by least square approximation for example.</span></div><div><span lang="en"><br></span></div><div><span lang="en">If I make it for the total energy (see total_energy.<span>eps</span>) I get a good agreement between analytical </span>(x-component for the displaced Ni atom) and numerical value of the total force (see total_force.<span>eps</span>).</div><div><span lang="en">But if I repeat it for the Hubbard energy (see <span>hubbard</span>_energy.<span>eps</span>) I get some discrepancy expressed in the mismatch between analytical (x-component for the displaced Ni atom) and numerical value of the Hubbard force (</span>see <span>hubbard</span>_force.<span>eps</span>) with -0.5 factor (see expected_<span>hubbard</span>_force.<span>eps</span>).</div><div><br></div><div>What can be the reason for this discrepancy?<br></div><div><br></div><div>Thank you!</div><div><br></div><div>P.S.</div><div>The values of the energy and forces (x-component for the displaced Ni atom) obtained during the <span>LSDA</span>+U calculation respect to the displacement of one Ni atom in the x-direction are contained in the attached file result.<span>dat</span>. <span class="HOEnZb"><font color="#888888"><br></font></span></div><span class="HOEnZb"><font color="#888888"><div><br></div><div>-- <br><div dir="ltr"><p><b><i><span style="border-collapse:collapse;color:rgb(51,51,51);font-family:arial,sans-serif;font-size:13px;font-style:normal;font-weight:normal"></span></i></b></p><div><i><span style="border-collapse:collapse;color:rgb(51,51,51);font-family:arial,sans-serif;font-size:13px"><i>Best regards,</i></span></i><br></div><div><i><span style="border-collapse:collapse"><i style="color:rgb(51,51,51);font-family:arial,sans-serif;font-size:13px">Dr. <span>Dmitry</span> <span>Novoselov</span></i><br><br><font color="#333333" face="arial, sans-serif">Institute for Metal Physics,</font></span></i></div><div><i><span style="border-collapse:collapse"><font color="#333333" face="arial, sans-serif">Yekaterinburg, Russia</font></span></i></div><p></p></div>
</div></font></span></div></div>
<br>_______________________________________________<br>
Pw_forum mailing list<br>
<a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
<a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br></blockquote></div><br></div></div></div></div></div></div>