Tag-smass-1

SMASS is the type of ensemble that is determined in the DFTMD of vasp, the specific information given in the manual is as follows:

SMASS =-3 | -2 | -1 | [Real]≥0default:smass =-3 Description:smass controls the velocities during an ab-initio molecular dynamics run. --------------------------------------------------------------------------------Smass=-3For smass=-3 a micro canonical ensemble are simulated (constant energy molecular dynamics). The calculated Hellmann-feynman forces serve as an acceleration acting onto the ions. The total free energy, i.e free electronic energy + Madelung energy of ions + kinetic energy of ions, is conserved.smass=-2For smass=-2 the initial velocities is kept constant. This allows to calculate the energy for a set of different linear dependent positions (for instance frozen phonons, or Dim ERS with varying bond-length). Mind:if Smass=-2 The actual steps taken is Potimx (velocities-read-from-the-poscar-file). To avoid ambiguities, set potim=1.Smass=-1In this case the velocities is scaled each nblock step (starting at the first step i.e. MOD (Nstep,nblock) =1) to the Tempe rature:t=tebeg+ (Teend-tebeg) Xnstep/nsw,where Nstep is the current step (starting from 1). This allows a continuous increase or decrease of the kinetic energy. In the intermediate Period a micro-canonical ensemble is simulated.smass≥0For Smass≥0, a canonical ensemble is simulated using the algorithm of Nosé. The Nosémass controls the frequency of the temperature oscillations during the simulation. [1] [2] [3] for smass=0, a nosé-mass corresponding to period of the time steps would be chosen. The Nosé-mass should is set such that the induced temperature fluctuation show approximately the same frequencies as the T Ypical ' phonon '-frequencies for the specific system. For liquids something like ' phonon '-frequencies might is obtained from the spectrum of the velocity auto-correlation funct Ion. If the ionic frequencies differ by a order of magnitude from the frequencies of the induced temperature fluctuations, Nos Éthermostat and Ionic movement might decouple leading to a non canonical ensemble. The frequency of the approximate temperature fluctuations induced by the Nosé-thermostat was written to the Outcar file.

In the "SMASS≥0" tab, allow decimals to appear, which has a corresponding example in the manual:

If you be a slightly advanced user can also use the damped MD algorithm, which was usually more efficient than the CG One:ibrion =  1  
SMASS  = 0.4          # damped MD  potim  = 0.4         # time step needs to chosen with Carein the case, a too large POT IM'll result in divergence.

The principle of this parameter is explained in detail in section 6.22.5 of the Manual:

If A damping factor is supplied in the Incar file by means of the SMASS tag, a damped second order Equati on of motion was used for the update of the Ionic degrees of freedom:

where SMASS supplies The damping factor, and Potim controls. In fact, a simple velocity Verlet algorithm are used to integrate the equation, the discretised equation reads:

It's immediately recognized, that's equivalent to a simple steepest descent algorithm (of course without line Optimizati ON). Hence, corresponds to maximal damping,corresponds to no damping.The optimal damping factor depends on the Hessian matrix (matrix of the second derivatives of the "the" and respect to T He atomic positions).A Reasonable first guess for is usually 0.4.Mind, we implementation is particular user-friendly, since changing usually does don't require to re-adjust the time St EP (Potim).to choose a optimal time step and damping factor, we recommend the following, step Procedure:first fix (for Insta NCE to 1) and adjust potim. Potim should be chosen as large as possible without getting divergenceThen decrease and keepPotimFixed. IfPotimandSMASSis chosen correctly, the damped molecular dynamics mode usually outperforms the conjugate gradient method by a factor of Both.

corresponds to potim . This equation implies, if the forces is antiparallel to the velocities, the velocities is quenched to zero. Otherwise the velocities was made parallel to the present forces, and they was increased by a amount that's proportional To the forces. Mind : for ibrion=3, a reasonable time step must is supplied by the Potim parameter. Too large time steps would result in divergence, Too small ones would slow down the convergence. The stable time step is usually twice the smallest line minimization step in the conjugate gradient algorithm.

Reference Links:

[1] http://cms.mpi.univie.ac.at/vasp/guide/node118.html

[2] Http://cms.mpi.univie.ac.at/vasp/vasp/Efficient_relaxation_from_unreasonable_starting_guess.html

[3] Http://cms.mpi.univie.ac.at/vasp/vasp/IBRION_3.html

Tag-smass-1