<p>Dear Professor <span style="color: rgb(32, 33, 36); font-family: Roboto, Arial, sans-serif; letter-spacing: 0.25px; text-wrap: nowrap;">Iannuzzi</span> ,</p><p>Thank you very much for your detailed and insightful response. Your explanation regarding the poorer initial guess and the potential issues with distorted configurations is very helpful. </p><p>Best regards,</p><p>Xie</p><br /><div class="gmail_quote"><div dir="auto" class="gmail_attr">在2024年6月16日星期日 UTC+8 20:20:37<Marcella Iannuzzi> 写道：<br/></div><blockquote class="gmail_quote" style="margin: 0 0 0 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">Dear Xie,<div><br></div><div>As you already suggested, the reason of the slow or no-convergence of the SCF for configurations extracted from the MD trajectory, is the poorer initial guess, in comparison to what obtained from the extrapolation scheme along the MD.</div><div>This is very clear from the at least 2 order of magnitude difference in the initial value of the "Convergence" at the MD step, in comparison to what obtained in the single point calculation.</div><div>The extracted configurations are probably rather distorted, maybe with almost broken or very elongated bonds, tending to formation of radicals. </div><div>This might require more robust optimisation algorithms. </div><div><br></div><div>By the way, when the SCF becomes unstable and the charge is not anymore conserved, there is no point to continue the simulation, it will never recover.</div><div>This refers to the SCF for step 5200, where after some outer-scf cycles , the energy keeps jumping by several Hartree and the charge on the grid is not zero.</div><div><br></div><div>Best</div><div>Marcella</div><div><br> 35 OT DIIS 0.15E+00 3.6 0.00067106 -1560.8442586833 5.22E-03<br><br> Leaving inner SCF loop after reaching 35 steps.<br> </div><div> Electronic density on regular grids: -724.0022698337 -0.0022698337<br> Core density on regular grids: 723.9999999995 -0.0000000005<br> Total charge density on r-space grids: -0.0022698341<br> Total charge density g-space grids: -0.0022698341<br></div><div>.......</div><div><br></div><div><br></div><div> 35 OT SD 0.15E+00 3.6 0.00066429 -1541.1842316224 -6.68E-02<br><br> Leaving inner SCF loop after reaching 35 steps.<br><br><br> Electronic density on regular grids: -660.6939947665 63.3060052335<br> Core density on regular grids: 723.9999999995 -0.0000000005<br> Total charge density on r-space grids: 63.3060052330<br> Total charge density g-space grids: 63.3060052330<br></div><div><br><br></div><div class="gmail_quote"><div dir="auto" class="gmail_attr">On Sunday, June 16, 2024 at 6:35:23 AM UTC+2 Qiming Xie wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><p>Dear CP2K experts,</p><p>I hope this message finds you well. I am writing to seek your assistance regarding an issue I encountered while using CP2K for molecular dynamics (MD) simulations.<br></p><p>After performing 20,000 steps of MD simulations, I extracted 50 frames and carried out separate SCF calculations to check for convergence. However, I found that not all frames converged as expected. I would appreciate any insights you could provide regarding this issue.</p><ol><li><p>For frames that do converge, the convergence speed during the MD simulations often achieves convergence within one OUTER_SCF step. However, when performing separate SCF calculations, it often takes more than three OUTER_SCF steps to converge. According to the gudieline, the MD process can automatically utilize the wave functions from previous structures to extrapolate the initial guess wave function for the current step, which seems reasonable.</p></li><li><p>However, I have encountered about 10 frames where the separate SCF calculations do not converge. Could this be related to temperature or configuration?</p></li><li><p>Regarding parameter settings, I have tried to keep the SCF and MD calculations as consistent as possible:</p><ul><li>Theoretical method: PBE</li><li>Basis set and pseudopotential: DZVP-MOLOPT-SR-GTH</li><li>Using OT</li><li>EPS_SCF for Inner and Outer SCF: 10E-5</li><li>CUTOFF and REL_CUTOFF: 400, 50 respectively</li><li>MD uses CSVR thermostat with temperature controlled at 300K</li></ul></li></ol><p>I have attached the input and partial output files for your reference.</p><p>Thank you for your time and assistance.</p><p>Best regards,
<br>Xie</p></blockquote></div></blockquote></div>
<p></p>
-- <br />
You received this message because you are subscribed to the Google Groups "cp2k" group.<br />
To unsubscribe from this group and stop receiving emails from it, send an email to <a href="mailto:cp2k+unsubscribe@googlegroups.com">cp2k+unsubscribe@googlegroups.com</a>.<br />
To view this discussion on the web visit <a href="https://groups.google.com/d/msgid/cp2k/6c4fa483-c33d-4174-85f0-cafc6380162fn%40googlegroups.com?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/6c4fa483-c33d-4174-85f0-cafc6380162fn%40googlegroups.com</a>.<br />