search

ferchault / APDFT

APDFT calculates quantumchemical results for many molecules at once.
https://ferchault.github.io/APDFT/
GNU General Public License v3.0
14 stars 4 forks source link

BOMD in Profess for Si8 #239

Closed michasahre closed 4 years ago

michasahre commented 4 years ago

see /home/misa/projects/APDFT-CPMD/data/singlepoint_PROFESS and /home/misa/projects/APDFT-CPMD/data/BOMD as well as http://dx.doi.org/10.1063/1.5023926 and PHYSICAL REVIEW B 81, 045206 (2010)

michasahre commented 4 years ago

Optimize ion positions (What if not CD structure anymore?) Do NPT simulation, which settings? Heat up, how? Consecutive MD simulations at increasing temperatures?

michasahre commented 4 years ago

1) Lattice optimization with cubic cell conserved. temp

ferchault commented 4 years ago

image

michasahre commented 4 years ago

NPT: go to 1 atm, room temperature by heating up.

Steps of 50 K, ~0.5 ps (till volume is constant and temperature at desired value) at target temperature ~1000 steps constant

michasahre commented 4 years ago

qmas

Temperature drops to zero or goes crazy but does not converge to 50K. cell is not cubic anymore, is this because random velocities in one direction larger than in the others -> expansion in this direction this effect could be reduced by going to a larger cell?

michasahre commented 4 years ago
michasahre commented 4 years ago

todo

michasahre commented 4 years ago

~- look at new data~ ~- add pressure~ ~- add fractional coordinates: axis label, legend~ ~- repeat for all simulations at 10e-5 and at the once for 1e-7 and 1e-7~ ~- send to Guido~

michasahre commented 4 years ago

QMAS5e4_cubic

michasahre commented 4 years ago

Make unit cell with 64 atoms run MD just for fun: /home/misa/projects/APDFT-CPMD/data/BOMD/Si64_hc_NPT/QMAS_5e4 get access to fast computer, then run it there

michasahre commented 4 years ago