Constant-pressure molecular dynamics in VASP.5.3.2

Message

pblanko · #1 Post by **pblanko** » Thu Sep 20, 2012 5:02 am

In the release notes of VASP.5.3.2 it is written:

2012-05-30
Constant-pressure molecular dynamics should now work.

However, there are no additional information on how to use constant pressure MD. The manual is not updated to cover the recent changes either.

Does anyone knows how to enable constant pressure MD in the new code?

zhangyg · #2 Post by **zhangyg** » Thu Sep 20, 2012 10:54 am

I have the same question. Hope the manual will be updated sooner so that we can use the NPT ensemble.

#3 Post by **admin** » Fri Sep 21, 2012 5:25 pm

At the moment we provide only very basic instructions on the use of our NpT molecular dynamics. Currently, Parrinello-Rahman NpT dynamics with Langevin thermostat is available

1.) compile VASP with precompiler flag "tbdyn" (see also http://cms.mpi.univie.ac.at/vasp/vasp/A ... iques.html

2.) in INCAR, define usual MD related flags (in praticular IBRION=0, POTIM, TEBEG) plus add he following flags:

ISIF=3 #variable cell shape and volume
MDALGO = 3
PSTRESS = [scalar]

LANGEVIN_GAMMA = [real array] # friction coeficients (in ps^-1) for atomic degrees of freedom, one number for each species defined in POSCAR must be provided

LANGEVIN_GAMMA_L= [real] #friction coeficient (in ps^-1) for lattice degrees of freedom (one number)

PMASS=[real] . # mass (in amu) for lattice degrees of freedom (one number)

The optimal values for the friction coefficients are system dependent - we don't have easy rule to determine the optimal values (typical values are from interval between 0 ps^-1 (no thermostating) and 100 ps^-1 ) - please check the vast literature on Langevin dynamics (good starting point is certainly the book of Allen and Tildesley: "Computer simulations of liquids"). As to the value for mass of lattice degrees of freedom (PMASS) - this is again very much system dependent, the optimal value is a compromise between two opposing factors: too large value leads to very slow motion of lattice degrees of freedom (and hence the sampling is inefficient) while too small POMASS (without adjusting POTIM to very small value) could cause numerical problems. In our experience, reasonable values of POMASS are from interval 1 and 10 - but once again we recommend to check literature and make careful tests for your particular system of interest before you perform "production run".
[ Edited Tue Sep 25 2012, 01:51PM ]
[ Edited Tue Sep 25 2012, 01:56PM ]
[ Edited Tue Sep 25 2012, 03:35PM ]
[ Edited Fri Sep 28 2012, 10:03AM ]

Vahid Askarpour · #4 Post by **Vahid Askarpour** » Thu Sep 27, 2012 7:50 pm

Dear Admin,

You mentioned that PSTRESS =[real array]. So for a target pressure of 300GPa, I set
PSTRESS=3000 3000 3000 0 0 0
but in the output, I get

internal error: XML_INCAR called with a vector, use XML_INCAR_V instead PSTRESS

Am I defining PSTRESS correctly?

Thanks,
Vahid

zhangyg · #5 Post by **zhangyg** » Fri Sep 28, 2012 3:57 am

Dear Sir:
(1)In the program, PSTRESS is defined as a scalar, not an array.
(2)After several MD steps, our MD cell becomes non cuboid. How to keep the MD cell's shape fixed.

zhangyg · #6 Post by **zhangyg** » Fri Sep 28, 2012 6:02 am

Hopefully there is an option for ISIF that allows relax ion, fix shape, and change volume.

#7 Post by **admin** » Fri Sep 28, 2012 10:09 am

@vahid Thank you, corrected.
@zhangyg No, ISIF=4 for fixed shape of the cell is not implemented yet.
Note that PREC=low and low value of ENCUT are not recommended for this kind of simulations as the quality of computed stress tensor (playing role of forces in equations of motion) is poor... In any case, you should make sure that the Pulay stress is negligible for your setting
http://cms.mpi.univie.ac.at/vasp/vasp/V ... tress.html

[ Edited Fri Sep 28 2012, 10:10AM ]

pblanko · #8 Post by **pblanko** » Sun Sep 30, 2012 4:59 am

After running 11196 MD steps using NPT the simulation crashed with this error:
ERROR FEXCP: supplied Exchange-correletion table
is too small, maximal index : 14257

Any ideas on how to solve?
I am using PREC=HIGH as recommended. PSTRESS=10 !1GPa

#9 Post by **admin** » Mon Oct 01, 2012 11:22 am

This appears at unreasonable geometry (perhaps your POTIM value is too large). Check FEXCP-contributions in forum.

felixvasp · #10 Post by **felixvasp** » Mon Oct 01, 2012 11:41 am

Hi, All
The following is from REPORT file:

MD step No. 1
========================================

>Damping parameter for lattice DOF:
PMASS = 6.00000
LATTICE_CONSTRAINTS = T T T (X,Y,Z) = 9 degree(s) of freedom

t_b> 1000.000 988.669 710.853 985.451

>Energies
E_tot E_pot E_kin E_const EPS ES
e_b> -0.12110244E+04 -0.12440159E+04 0.32991452E+02 0.00000000E+00 0.00000000E+00 0.00000000E+00

RANDOM_SEED = 1451059310 1744574122

What those temperatures stand for? Among them, I know 1000 K is the target and initial Temp.

[quote author=0, POTIM, TEBEG) plus add he following flags:

ISIF=3 #variable cell shape and volume
MDALGO = 3
PSTRESS = [scalar]

LANGEVIN_GAMMA = [real array] # friction coeficients (in ps^-1) for atomic degrees of freedom, one number for each species defined in POSCAR must be provided

LANGEVIN_GAMMA_L= [real] #friction coeficient (in ps^-1) for lattice degrees of freedom (one number)

PMASS=[real] . # mass (in amu) for lattice degrees of freedom (one number)

The optimal values for the friction coefficients are system dependent - we don't have easy rule to determine the optimal values (typical values are from interval between 0 ps^-1 (no thermostating) and 100 ps^-1 ) - please check the vast literature on Langevin dynamics (good starting point is certainly the book of Allen and Tildesley: "Computer simulations of liquids"). As to the value for mass of lattice degrees of freedom (PMASS) - this is again very much system dependent, the optimal value is a compromise between two opposing factors: too large value leads to very slow motion of lattice degrees of freedom (and hence the sampling is inefficient) while too small POMASS (without adjusting POTIM to very small value) could cause numerical problems. In our experience, reasonable values of POMASS are from interval 1 and 10 - but once again we recommend to check literature and make careful tests for your particular system of interest before you perform "production run".
[/quote][/color]

[ Edited Tue Oct 02 2012, 10:26PM ]

felixvasp · #11 Post by **felixvasp** » Tue Oct 02, 2012 9:24 pm

Hi, everyone,
Does anyone know whether VASP.5.3.2 can do constrained NPT?Thanks.
[ Edited Tue Oct 02 2012, 09:25PM ]

#12 Post by **admin** » Wed Oct 03, 2012 11:10 am

Constraints are in VASP.5.3.2 not implemented yet.

pblanko · #13 Post by **pblanko** » Tue Oct 09, 2012 3:33 am

[quote="admin"]This appears at unreasonable geometry (perhaps your POTIM value is too large). Check FEXCP-contributions in forum.[/quote]

I restarted the MD simulation from the geometry at the crash point (CONTCAR). To my surprise it can continue the MD simulation without a complaint. The geometry and the POTIM used does not seem to be the responsible for the crash.

pblanko · #14 Post by **pblanko** » Tue Oct 09, 2012 3:35 am

[quote author========================================

>Damping parameter for lattice DOF:
PMASS = 6.00000
LATTICE_CONSTRAINTS = T T T (X,Y,Z) = 9 degree(s) of freedom

t_b> 1000.000 988.669 710.853 985.451

>Energies
E_tot E_pot E_kin E_const EPS ES
e_b> -0.12110244E+04 -0.12440159E+04 0.32991452E+02 0.00000000E+00 0.00000000E+00 0.00000000E+00

RANDOM_SEED = 1451059310 1744574122

What those temperatures stand for? Among them, I know 1000 K is the target and initial Temp. [/quote]

felixvasp · #15 Post by **felixvasp** » Sat Oct 13, 2012 12:10 am

Thanks for explanations. pblanko.