<div dir="ltr"><div>Hi all!</div><div><br></div><div>I am exploring single molecule AIMD (H2 in this case) for different purposes. The main problem that I face is temperature/rotational modes. I try to run a AIMD using NVE and NVT ensembles, but there are some problems.</div><div>For an NVE ensemble, the system starts rotating with a conserved angular moment, <span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-">even if I set equal to zero the total angular moment... This <span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-">disrupt the linear movement of H2 bond</span></span></span></span> oscillation. The rotation couples with the bond and generates a fake two frequency oscillator. If, after the AIMD, I correct the trajectory to a zero total angular moment for all frames (simple for diatomic homonuclear molecule, but not so easy for polyatomic molecules), the oscillator <span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-">recovers the single frequency motion.<br></span></span></div><div><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-">For an NVT  ensemble, the thermalization is so hard for few atoms... and the system is very bad behaved.</span></span></div><div><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-"><br></span></span></div><div><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-">Now I try an NVE ensemble with constrain MD, fixing the Y and Z position of both atoms, and put the bond along the X direction.</span></span></div><div><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-"><br></span></span></div><div><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-">  &CONSTRAINT<br>    &FIXED_ATOMS<br>      COMPONENTS_TO_FIX      YZ<br>      LIST                   1 2<br>    &END FIXED_ATOMS<br>  &END CONSTRAINT</span></span></div><div><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-"><br></span></span></div><div><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-">I think with it, I <span class="gmail-gt-baf-term-text"><span class="gmail-gt-baf-cell gmail-gt-baf-word-clickable">force the MD to evolve only over the X axis... but the initial step gives me a NaN temperature and kinetic energy,</span></span></span></span><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-"> no matter what initial temperature I use. I attach both files.</span></span></div><div><br></div><div>The output file prints:</div><div> Calculation of degrees of freedom<br>                                                      Number of atoms:         2<br>                                 Number of Intramolecular constraints:         4<br>                                 Number of Intermolecular constraints:         0<br>                                  Invariants(translation + rotations):         5<br>                                                   Degrees of freedom:        -3<br><br><br> Restraints Information<br>                                  Number of Intramolecular restraints:         0<br>                                  Number of Intermolecular restraints:         0<br> ************************** Velocities initialization **************************<br> Initial Temperature                                                       NaN K<br> COM velocity:                       NaN                 NaN                 NaN<br> COM position:            5.669178398657      5.669178398657      5.669178398657<br> Angular velocity:                   NaN                 NaN                 NaN<br> *******************************************************************************</div><div><br></div><div>It seems, the degrees of freedom is: 3N - constrains - invariants...</div><div>Is there a way to force the MD only along one axis to avoid rotation?</div><div><br></div><div>Regards!<br></div><div><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-"></span></span><span class="gmail-tlid-translation-gender-indicator gmail-translation-gender-indicator"></span><span class="gmail-tlid-trans-verified-button gmail-trans-verified-button"></span><span class="gmail-tlid-translation gmail-translation" lang="en"><span title="" class="gmail-"><span class="gmail-gt-baf-term-text"><span class="gmail-gt-baf-cell gmail-gt-baf-word-clickable"></span></span></span></span></div></div>