ROPT

From VASP Wiki
Revision as of 16:16, 1 February 2011 by Mmars (talk | contribs)

ROPT = [real array], one entry for each species on the POTCAR file 

Default: ROPT = 0.67 if LREAL=On and PREC=Low
= 1.0 if LREAL=On and PREC=Med
= 1.5 if LREAL=On and PREC=High
= 0.01 if LREAL=Auto and PREC=Low
= 0.002 if LREAL=Auto and PREC=Med
= 1E-4 if LREAL=Auto and PREC=High

Description: ROPT determines how precise the projectors are represented in real space.


Depending on their value, VASP interprets ROPT entries in one of two ways:

  • ROPT ≥ 0.1: "conventional"-mode.
Sets the number of real space points within the cutoff sphere for a particular species to approximately 1000×ROPT. So for instance the line
ROPT = 0.7 1.5
will set the number of real space points within the cutoff sphere for the first species to approximately 700, and that for the second species to 1500.
  • ROPT < 0.1: "precision"-mode.
In this case the real-space operators will be optimized for an accuracy of approximately ROPT eV/atom. So, for instance specifying the following line in the INCAR file
ROPT = 1E-3 1E-3
Tells VASP to optimize the real-space projector operators for species 1 and 2 for an accuracy of approximately 1 meV/atom (10-3). The "precision" mode works both for LREAL=On and LREAL=Auto (but to maintain compatibility with older VASP versions it is only selected if LREAL=Auto is specified in the INCAR file).

The "precision" and "conventional" modes may be intermixed, i.e., it is possible to specify

ROPT = 0.7 1E-3

in which case the number of real space points within the cutoff sphere for the first species will be approximately 700, whereas the real space projector functions for the second species are optimized for an accuracy of approximately 1 meV. We recommend to use the "precision" mode with a target accuracy of around 10-3 eV/atom if your VASP version supports this mode.


Optimization of the real space projectors

maximal supplied Q-value         = 12.85
optimization between [QCUT,QGAM] = [ 4.75, 9.51] = [ 6.33, 25.32] Ry
Optimized for a Real-space Cutoff    2.30 Angstroem

  l   X(QCUT)   X(cont)   X(QGAM)  max X(q)  W(q)/X(q)  e(spline)

  0    9.518     9.484     -.004   18.582    .11E-03     .16E-06
  0   -2.149    -2.145      .001    3.059    .17E-03     .25E-06
  1    8.957     8.942      .003    9.950    .14E-03     .34E-06
  1    1.870     1.870      .001    1.837    .95E-03     .51E-06
  2    3.874     3.866      .000    4.764    .15E-03     .68E-07

QCUT and QGAM are parameters chosen for the generation of this particular PAW dataset (or ultrasoft pseudopotential). The most important information is given in the column W(q)/X(q) (respectively the column W(low)/X(q) for LREAL=Auto). The values in these columns must be as small as possible. If these values are too large, increase the ROPT from the default value. As a rule of thumb the maximum allowed value in this column is 10-3 for PREC=Med. (For PREC=Low errors might be around 10-2 and for PREC=High errors should be smaller than 10-4). If W(q)/X(q) is larger than 10-2 the errors introduced by the real space projections can be substantial. In this case ROPT must be specified in the INCAR file to avoid incorrect results. If the new "precision" mode is used in VASP.4.4 (ROPT<0.1) the code automatically selects the real-space cutoff so that the required precision is reached.

Related Tags and Sections

LREAL


Contents