questions on AIMD input file

Questions and discussions regarding the use of Qbox
Forum rules
You must be a registered user to post in this forum. Registered users may also post new topics if they consider that their subject does not correspond to any topic already present on the forum.
Post Reply
luke419
Posts: 16
Joined: Tue Feb 27, 2024 2:50 am

questions on AIMD input file

Post by luke419 »

Dear Francois,

Hello?
First many thanks for your helps.
I'd like to ask some questions on AIMD input file of qbox shown below.

set nrowmax 72
load gs.xml
set xc PBE
set wf_dyn PSDA
set ecutprec 4
set atoms_dyn MD
set dt 10
set thermostat BDP
set th_temp 400
set th_time 10000
run 10 5
save md.xml

1. it seems that load gs.xml means the output of gs job.
Should we always perform gs first before AIMD? Can we replace "load gs.xml" with "h2o64.i" as such in gs.i?

2. If possible, how can we make an input file to perform pre-equilibration, equilibration and analysis works subsequently?
I'd like to raise the temperature gradually from 10 to 300 K for prequilibration and save the trajectory only in the analysis process.
What command should we use to save the trajectory of atoms?

3. Would you let me know where we can download the potential files for qbox?

4. If there is an utility to convert a cif file to the system coordinate file of qbox (e.g., h2o64.i), would you let me know?


Thank you for your answers in advance.

Best regards,
Young
fgygi
Site Admin
Posts: 167
Joined: Tue Jun 17, 2008 7:03 pm

Re: questions on AIMD input file

Post by fgygi »

Hello Young,

It is necessary to first compute the electronic ground state before you can start an MD simulation. First you need to define the initial positions (and optionally velocities) of all atoms, and define the unit cell. This can be done using the following input file:
h2o64.i
H2O64 sample from PBE400
(17.63 KiB) Downloaded 741 times
The file h2o64.i uses the pseudopotential files D_ONCV_PBE-1.2.xml and O_ONCV_PBE-1.2.xml which can be downloaded from http://www.quantum-simulation.org/poten ... _oncv/xml/ .
The following input file gs.i can be used to compute the ground state and save it in the restart file gs.xml:

Code: Select all

h2o64.i
fold_in_ws
set ecut 65
set xc PBE
set wf_dyn JD
set scf_tol 1.e-8
run 0 30 10
save gs.xml
Once the ground state has been computed, you can start an MD simulation using the following input file md.i:

Code: Select all

load gs.xml
set xc PBE
set wf_dyn PSDA
set atoms_dyn MD
set dt 20
set thermostat BDP
set th_temp 400
rseed
reset_vcm
run 100 5
save md.xml
The file md.xml is saved at the end of the simulation and can be used to continue the simulation using the following input file md2.i:

Code: Select all

load md.xml
set xc PBE
set wf_dyn PSDA
set atoms_dyn MD
set dt 20
set thermostat BDP
set th_temp 400
rseed
reset_vcm
run 100 5
save md2.xml
Note the following:
  • The velocities in the file h2o64.i are such that the temperature of the sample is around 400 K at the beginning of the MD simulation.
  • The time step is 20 (a.u.) but can be changed to 20.670681 without any additional computational cost.
  • The file D_ONCV_PBE-1.2.xml defines deuterium with a mass of 2.014.
  • You can perform long MD simulations by running multiple simulations that save and load restart files. Typically we have used simulations of 2000 steps of 20 (a.u.) which amounts to approximately 1 ps. Depending on how long you can run jobs on your computer you will have to adjust the number of steps of a simulation.
  • The rseed command is used in each run to seed the random number generator that is used by the thermostat. The seed depends on the time at which the simulation starts, which ensures randomness.
  • The reset_vcm command is used before each run to reset the velocity of the center of mass to zero. This is done to avoid the "flying ice cube" problem in which the center of mass acquires a finite velocity due to an accumulation of numerical errors (S. C. Harvey et al, J. Comput. Chem. 19, 726 (1998)).
  • The Qbox distribution does not include a conversion program taking cif files as input. However, if you can generate an xyz file of the coordinates, you can then used the xyz2qbox.py script to generate a Qbox input file, which needs minor editing to define species.
Best,
Francois
Post Reply