Dear Prof. Iannuzzi,<div><br /></div><div>Thank you so much for the explanation. I further tested on a 2*2*2 cell of the system (starting from scratch), using 1000 for Barostat and 200 for Thermostats coupled to barostat or particles. The Cons Qty still increases after 200 ps. Both the diagonal (starting from ~20 angstrom) and off-diagonal components (starting from 0) drifted away. I also noticed that the system exploded when I constrained the symmetry to orthorhombic. And when I restart the simulation, I got the "CPASSERT failed" from pw/pw_grids.F:2077. Looks like the off-diagonal components of the stress tensor are doing something weird to the simulation...</div><div><br /></div><div>I had some additional questions regarding these observations:</div><div>1. As the Kin and Pot of themostats, barostat and particles fluctuates around some equilibrium values towards the end of the trajectory, why the Cons. Qty still increases? What are the quantities I did not count for the Cons Qty?</div><div>2. Do you have any known system(s) that can be simulated in the NPT_F, for example, in the test suite?</div><div><br /></div><div>Any comments or suggestions are greatly appreciated.</div><div>Thank you!</div><div>Zhuoran</div><div> <img alt="t300_cons.png" width="274px" height="211.49px" src="cid:26fc006e-9961-4c88-a631-538bd94004e8" /><img alt="kin_thermostat.png" width="271px" height="209.504px" src="cid:390ac4bd-2c56-4f69-8c3e-4fb7ab9de54c" /><img alt="cell.png" width="269px" height="207.543px" src="cid:ee87599b-5196-472b-aacc-47c86de74c88" /></div><div><br /></div><div class="gmail_quote"><div dir="auto" class="gmail_attr">在2025年5月30日星期五 UTC-4 04:43:04<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 Zhuoran<div><br></div><div>I would not define those fluctuations in the cell as OK. There is for sure something wrong in the forces.</div><div>The difference between NPT_I and NPT_F comes from the off diagonal elements of the stress tensor. </div><div> I would expect them to be very small for bulk water. </div><div>On the other hand the system you are testing is very small and local fluctuations might have an uncontrolled effect. </div><div>Probably the system and the FF you are using are not suitable to test NPT_F</div><div>Anyway, the fact that the integration is going wrong is already evident from the very beginning of the simulation, the energy drift is from the start not acceptable. </div><div>Regards</div><div>Marcella</div><div><br></div><div class="gmail_quote"><div dir="auto" class="gmail_attr">On Wednesday, May 28, 2025 at 6:44:53 PM UTC+2 <a href data-email-masked rel="nofollow">zlo...@umd.edu</a> 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">Dear Prof. Iannuzzi,<div><br></div><div>Thank you so much for the suggestions. CELL, STRESS, kinetic and potential energy of the barostat all oscillates around some values and I guess that means fine. However I noticed that the kinetic energies of the thermostats keep increases, both for the one applied on the atoms and the one coupled to the barostat. </div><div>I made some tests including </div><div>(a) remove the "TYPE SAME_AS_PARTICLE" so that the TIMECON of the thermostat coupled to the barostat becomes 200 as specified in the input</div><div>(b) from (a), change the TIMECON of the barostat to 2000</div><div>(c) from (b), change the TIMESTEP to 0.2</div><div>and none of them seem to fix the problem. </div><div>I would really appreciate it if you could please provide any additional suggestions.</div><div><br></div><div>Thank you!</div><div>Zhuoran </div><div><br></div><div>The following plots were from the initial simulations but all the tests (a-c) show similar trends. </div><div><img alt="t300_cell.png" width="300px" height="231.787px" src="https://groups.google.com/group/cp2k/attach/1888a510f1e961/t300_cell.png?part=0.2&view=1"><img alt="t300_stress.png" width="297px" height="230.07px" src="https://groups.google.com/group/cp2k/attach/1888a510f1e961/t300_stress.png?part=0.3&view=1"><img alt="t300_e.png" width="290px" height="224.184px" src="https://groups.google.com/group/cp2k/attach/1888a510f1e961/t300_e.png?part=0.1&view=1"></div><div class="gmail_quote"><div dir="auto" class="gmail_attr">On Tuesday, May 27, 2025 at 5:02:27 AM UTC-4 Marcella Iannuzzi 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"><br>Dear Zhuoran<div><br></div><div>Have you also monitored the CELL , the STRESS and the energy and temperature of the barostat?</div><div>When they change too fast, it is difficult for the dynamics to recover. </div><div>In this case you can try to rescale the temperature of the barostat and reduce the time step.</div><div>Regards</div><div>Marcella</div><div><br></div><div class="gmail_quote"><div dir="auto" class="gmail_attr">On Saturday, May 24, 2025 at 10:38:05 AM UTC+2 <a rel="nofollow">zlo...@umd.edu</a> 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">Dear all,<div><br></div><div>I am testing NPT simulations on the <a href="https://www.cp2k.org/exercises:2018_ethz_mmm:h2o_md" rel="nofollow" target="_blank" data-saferedirecturl="https://www.google.com/url?hl=zh-CN&q=https://www.cp2k.org/exercises:2018_ethz_mmm:h2o_md&source=gmail&ust=1748904591409000&usg=AOvVaw3pR9Bndl3J_7QhSGNevVfZ">tutorial water system</a>. I'm including the following changes to the input file in the tutorial example:</div><div><br></div><div>&FORCE_EVAL</div><div> STRESS_TENSOR ANALYTICAL</div><div> ...</div><div>&END FORCE_EVAL<br></div><div>...</div><div>&MOTION<br> &MD<br> ENSEMBLE NPT_I<br> STEPS 200000<br> TIMESTEP 1.0<br> TEMPERATURE 300<br> COMVEL_TOL 1.0E-10<br> &BAROSTAT<br> PRESSURE 1.01325 #[bar]<br> TIMECON 1000<br> VIRIAL XYZ<br> &THERMOSTAT<br> TYPE SAME_AS_PARTICLE<br> &CSVR<br> TIMECON 200<br> &END CSVR<br> &END THERMOSTAT<br> &END BAROSTAT<br> &THERMOSTAT<br> TYPE CSVR<br> REGION MASSIVE<br> &CSVR<br> TIMECON 200<br> &END CSVR<br> &END THERMOSTAT</div><div> ...</div><div> &END MD<br></div><div> ...</div><div>&END MOTION<br></div><div><br></div><div>The Cons Qty in the .ener file converged after about 20 ps. However. when I perform the same simulation from scratch in the NPT-F ensemble (by changing the NPT-I to NPT-F), the Cons Qty kept increasing and exploded after about 80 ps (please see the plot below, so are the kinetic energy and potential energy). I will need the NPT-F in future for a tetragonal covalent organic framework system. Could someone please let me know if I had any wrong or missing parameters? </div><div>And a relevant question, while I set the time constant to be 200 for the CSVR thermostat of the barostat, the corresponding TIMECON in the restart file has a value of ~9.6755. Is that relevant to my question? I'm using the version 2024.2. </div><div><br></div><div>And helps and discussions are greatly appreciated.</div><div>Zhuoran</div><div><img alt="t300.png" width="269px" height="208.094px" src="https://groups.google.com/group/cp2k/attach/1733c25b127436/t300.png?part=0.1&view=1"><br></div></blockquote></div></blockquote></div></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 visit <a href="https://groups.google.com/d/msgid/cp2k/a84851c5-522f-4aae-9985-cdc355a0b23cn%40googlegroups.com?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/a84851c5-522f-4aae-9985-cdc355a0b23cn%40googlegroups.com</a>.<br />