[QE-users] Query on AIMD simulation with pw.x code in QE and LAMMPS MD simulation.

Jayraj Anadani jayrajanadani at spuvvn.edu
Tue Oct 17 09:40:23 CEST 2023


Hello, QE community!
On my metallic system, which has 128 atoms, I am running AIMD calculations
using the pw.x code. Two pseudopotentials, PAW and USPP, were each
tested individually.
I have the following queries:

   1. Compared to PAW pseudopotential, USPP's convergence occurs *much more
   quickly. why*? (when ecutwfc is 25)
   2. The temperature fluctuation is incredibly small and nearly equal in
   the MD simulation of thousands of atoms in LAMMPS at a specific temperature
   (i.e., 1000K). However, the temperature fluctuation in MD with pw.x (AIMD)
   is significant, so we must set "tolp" which stands for tolerance for
   velocity rescaling. Although the average temperature is kept constant, *how
   closely does this significant fluctuation match the LAMMPS MD simulation?*
   3. For a faster AIMD calculation at a specific temperature, *Can I
   perform the whole simulation with a very low "ecutwfc" and "conv_thr" below
   a minimum cutoff of pseudopotential? *So I can get the next ionic
   position very quickly with a fast scf cycle and at the last MD step I will
   perform a single point energy (SCF) calculation with high "ecutwfc" and
   "conv_thr" to get a good amount of accuracy.

Thank you

regards
JAYRAJ ANADANI
Department of Physics
SPU Research scholar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20231017/5f7a3db8/attachment.html>


More information about the users mailing list