[QE-users] NVT variable cell dynamics - protocol and issues

[Manuel Pérez Escribano]

Dear users,

I am quite new to QuantumEspresso. In the past weeks, I managed to run 
some BOMD NVT simulations, for which I previously relaxed the electronic 
structure and perform a NVE thermalization. However, I am now interested 
in repeating the procedure but allowing the crystal cell to move. I 
first relaxed the electronic structure without moving the atoms, nor the 
cell:

&ELECTRONS
   emass = 500,
   emass_cutoff = 3.0,
   orthogonalization = 'ortho',
   electron_dynamics = 'damp',
   electron_damping = 0.1
/

&IONS
   ion_dynamics = 'none',
   ion_temperature = 'not_controlled',
   tranp(1)= .true.
   tranp(2)= .true.
   tranp(3)= .true.
   amprp(1)= 0.35
   amprp(2)= 0.35
   amprp(3)= 0.35
/

&CELL
   cell_parameters='default'
   cell_dynamics='none'
/

Then I tried to perform a NVE trajectory allowing both the cell and the 
atoms to move:


&ELECTRONS
   emass = 1000,
   emass_cutoff = 3.0,
   orthogonalization = 'ortho',
   ortho_max=1000,
   electron_dynamics = 'verlet',
/

&IONS
   ion_dynamics = 'verlet',
   ion_temperature = 'not_controlled',
/

&CELL
   cell_dynamics='pr',
   cell_velocities='zero'
/

Restarting from the electronic relaxation. However, I get stuck with a 
orthogonalization problem and I am not sure whether my protocol is not 
valid and maybe I should do either NVE on the ions and the cell 
separately firstly, and then allow both to move. Can you please suggest 
how to proceed?

Manuel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20230413/4781985c/attachment.html>