[Pw_forum] Problem in energy conservation in NVE MD using cp.x

[pastore at democritos.it]

Maybe just a few additional comments about this issue may be useful.

Quoting Nicola Marzari <marzari at MIT.EDU>:

>
> CC: Paolo Giannozzi, Carlo Cavazzoni, Roberto Car. Notes below
> also relevant for BO dynamics with pwscf. Please focus on
> the bottom part of the email.
> ...
> I believe the problem is this: you are seeing a transfer of quantity of
> motion between the subset of the ions, and the electronic fictitious
> degrees of freedom. Remember this: the cp lagrangian, when integrated
> correctly (i.e. with a small time step) evolves correctly the full
> dynamical system with all degrees of freedom (physical ionic masses,
> fictitious electronic masses). If there is an electronic gap in your
> system (like here) the two subsets of ions and electrons evolve without
> exchanging energy between each other, and conserving the momentum
> (linear and angular) of the mixed system (ionic + fictitious
> electronic).

Well, the existence of a gap alone is not sufficient to prevent 
transfer of energy from one set of degrees of freedom to another. It is 
also necessary that the gap does not change too much  during the ionic 
evolution (the adiabatic invariant is not the energy of normal modes 
but the ratio energy/frequency,  cfr. formula 28 of the old PRA 28, 
6334 (1991)). I do not think this should be the case for a water 
molecule, if the pseudopotential and the electronic structure are 
reasonable , but it is useful to have in mind such a difference when 
discussing energy transfer issues.

The conserved linear momentum can be evaluated exactly (beware: to the 
electronic term in formula 22 of the above cited PRA,  one  should add 
its complex conjugate, as remarked in Morishita T, Nose S, PRB 59, 
15126 (1999), this last paper was centered on the problem of the 
momentum conservation in CP dynamics.

Periodic boundary conditions prevent angular momentum to be a conserved 
quantity, even for classical simulations, so there is no good reason to 
put constraints on that quantity.

...
> On top of this, the routine that calculates the temperature removes
> the contribution from the motion of the center of mass - it is in
> Modules, so I believe this is also used for pwscf. I believe this
> is never the right thing to do, and maybe this is a good time to
> sort this out.
>

I would not agree with this last point. The physical meaning of 
temperature does not include the center of mass kinetic energy. I can 
remember the very old times  when people using the very first versions 
of CP  code were thinking to have high temperature systems whereas they 
had only  very fast flying clusters!

I would agree with some monitoring of the center of mass motion, but 
without including such a contribution in the ionic physical temperature.

Giorgio

--
Giorgio Pastore - Dipartimento di Fisica Teorica dell' Universita'
                  and CNR-INFM Democritos
                  Strada Costiera 11, 34151 Trieste, Italy
  phone :         + 39 040 2240247   fax: + 39 040 224601
  e-mail:         pastore at ts.infn.it




----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.