Ab-initio MD calculations (smass, temperature, diiffusion)
Posted: Mon May 30, 2011 6:38 pm
Hello,
I am new running MD calculations and I would like to know how I should thermalize my solid. I am running NVT calculations from a GGA relaxation (I copy the CONTCAR of the relaxation to a POSCAR of the NVE calculation). I am fixing the temperature Tfrom zero to 873, in the example below as well as 673, 1073 and 1273 (TEBEG=TEND=desired temperature). MY timestep is 3fs and my SMASS=0. Until now I've integrated 1ps.
Does anybody have experience with this SMASS setting in MD calculations?
How can I now that I have thermalized my solid? my temperature oscillates and my total energy still is going up.
Should I start getting positions once my total energy is constant?
The cell parameters in CONTCAR after one 1ps are exacly the same as in the start.
I analize XDATCAR to obtain MSD.
Surprisingly I do not see diffusion at all for the 673,1073 and 1273 temperatures but for the 873 case I am getting I huuuge diffusion of lithium but also an apreciable diffusion of the rest od species.
Does this make any sence? How real is the temperature in this simulations?
Here is my INCAR.
Plese feel free to make suggestions about it.
May be too many questions...I am sorry I am new with MD
calculations...so any help woul be really apreciated!!!
Thank youuu,
Best,
--Amaia
LiCoO2
ALGO = fast
IWAVPR =12
LREAL = A !Evaluate projection operators in real space
NELMIN = 4 !do a minimum of four electronic steps
BMIX = 2.0 ! mixing parameter
MAXMIX = 50 ! keep dielectric function betwee ionic movements
ISYM = 0 ! switch of symmetry
NSW = 3000 ! number of steps for IOM
NBLOCK = 1
KBLOCK = 1
SMASS = 0 ! microcanonical
POTIM = 3.00 !tima step fs for ion-motion
TEBEG= 873 ! temperature
TEND= 873
PREC = normal
IBRION= 0 # Molecular Dynamics.
ENMAX=600 #cutoff
ISMEAR = 0 #BZ integration method (Bloechl corrections).
SIGMA = 0.1 #smearing width (keep T*S < 1meV/atom).
#PC-FUNCTION
APACO= 10.0 ! Distance for P.C
I am new running MD calculations and I would like to know how I should thermalize my solid. I am running NVT calculations from a GGA relaxation (I copy the CONTCAR of the relaxation to a POSCAR of the NVE calculation). I am fixing the temperature Tfrom zero to 873, in the example below as well as 673, 1073 and 1273 (TEBEG=TEND=desired temperature). MY timestep is 3fs and my SMASS=0. Until now I've integrated 1ps.
Does anybody have experience with this SMASS setting in MD calculations?
How can I now that I have thermalized my solid? my temperature oscillates and my total energy still is going up.
Should I start getting positions once my total energy is constant?
The cell parameters in CONTCAR after one 1ps are exacly the same as in the start.
I analize XDATCAR to obtain MSD.
Surprisingly I do not see diffusion at all for the 673,1073 and 1273 temperatures but for the 873 case I am getting I huuuge diffusion of lithium but also an apreciable diffusion of the rest od species.
Does this make any sence? How real is the temperature in this simulations?
Here is my INCAR.
Plese feel free to make suggestions about it.
May be too many questions...I am sorry I am new with MD
calculations...so any help woul be really apreciated!!!
Thank youuu,
Best,
--Amaia
LiCoO2
ALGO = fast
IWAVPR =12
LREAL = A !Evaluate projection operators in real space
NELMIN = 4 !do a minimum of four electronic steps
BMIX = 2.0 ! mixing parameter
MAXMIX = 50 ! keep dielectric function betwee ionic movements
ISYM = 0 ! switch of symmetry
NSW = 3000 ! number of steps for IOM
NBLOCK = 1
KBLOCK = 1
SMASS = 0 ! microcanonical
POTIM = 3.00 !tima step fs for ion-motion
TEBEG= 873 ! temperature
TEND= 873
PREC = normal
IBRION= 0 # Molecular Dynamics.
ENMAX=600 #cutoff
ISMEAR = 0 #BZ integration method (Bloechl corrections).
SIGMA = 0.1 #smearing width (keep T*S < 1meV/atom).
#PC-FUNCTION
APACO= 10.0 ! Distance for P.C