scf

Full name: ase2sprkkr.input_parameters.definitions.scf

Description

SCF task input parameters definition

Description of the sections and parameters

SCF - Calculate a self-consistent potential

INPUT PARAMETERS scf contains:
------------------------------
        SECTION CONTROL contains:
    -----------------------------
        DATASET : String                  The custom field for the description of the problem - the output files will have called 'DATASET.<ext>'.
        ADSI : FixedValue(SCF) ≝ SCF      Type of the computation.
        POTFIL : String                   The potential file (see SPRKKR documentation for its format). It isn't necessary to set it, it will be set by the calculator.
        KRWS : Integer ≝ 1  (optional)    If it is 0, RWS is taken from the potential file and scaled. If 1, RWS is calculated by scaling the muffin-tin radii by a common scaling factor. (This setting is forced in the case of FULLPOT.)
        KRMT : AnyOf(0,1,2,3,4,5,6)  (optional)

            Possible values:
              0         RMT is taken from the potential file
              1         RMT = min( x*RWS )
              2         RMT = min( d_ij / 2 )
              3         RMT from atomic charge density (=> KRWS=1)
              4         RMT from atomic Hartree potential (=> KRWS=1)
              5         RMT from total atomic potential (=> KRWS=1)
              6         take average of 3 and 4 (=> KRWS=1)

            It controls how the muffin-tin radii are calculated.

        PRINT : Integer ≝ 0  (optional)   Verbosity of the output (0-5). Do not affect the results in any way, just the amount of the printed output.
        NONMAG : Flag ≝ False             Set this flag, if it is known that the system considered is non-magnetic. This leads to a higher symmetry and a faster calculation.



        SECTION TASK contains:
    --------------------------
        TASK : FixedValue(SCF) ≝ SCF



        SECTION TAU contains:
    -------------------------
        BZINT : AnyOf(POINTS,WEYL) ≝ POINTS

            Possible values:
              POINTS    special points method
              WEYL      Weyl method


            The Weyl method (BZINT=WEYL) is a point sampling method using more or less ran-
            dom points. The number of k-points used for the integration varies quadratically be-
            tween 0.0 and ImE according to the imaginary part of the energy.

            The special point method (BZINT=POINTS) uses a regular k-point grid with NKTAB
            points. It is the standard method and gives a good compromise concerning accuracy
            and efficiency. For BZINT=POINTS the parameter NKTAB will be adjusted to allow a
            regular mesh.


            The mode of BZ-integration used for calculation of the scattering  path operator τ

        NKTAB : Integer ≝ 250  (optional)  Number of points for the special points method
        NKTAB2D : Integer  (optional)     Number of points for the special points method for 2D region of 2D problem
        NKTAB3D : Integer  (optional)     Number of points for the special points method for 3D region of 2D problem
        NKMIN : Integer ≝ 300             Minimal number of k-points used for Weyl integration
        NKMAX : Integer ≝ 500             Maximal number of k-points used for Weyl integration
        KKRMODE : AnyOf(STANDARD-KKR,TB-KKR,LAYER-KKR)  (optional)

        Expert options:
        --------------
            CLUSTER : Flag ≝ False  (optional, expert)  Do cluster type calculation.
            NSHLCLU : Integer  (optional, expert)  Number of atomic shells around the central atom of a cluster
            CLURAD : Real  (optional, expert)  Radius of the cluster in multiples of ALAT.
            IQCNTR : Site  (optional, expert)  The center of the cluster is set at the site position with number IQCNTR of the specified basis.
            ITCNTR : AtomicType  (optional, expert)  The center of the cluster is set at one of the site positions that is occupied by the atomic type ITCNTR.
            NLOUT : Integer ≝ 3  (optional, expert)  The calculated τ -matrix is printed up to lmax=NLOUT.
            MOL : Flag ≝ False  (optional, expert)  Cluster type calculation but for a molecular system. The system is specified as for CLUSTER.



        SECTION ENERGY contains:
    ----------------------------
        GRID : Array(of Integer) ≝ [5]    Type of the grid for the energy-mesh
        NE : Array(of Integer) ≝ [32]     Number of points in energy-mesh
        ImE : Energy (<Real> [Ry|eV]) ≝ 0.0  (optional)
        EMIN : Real  (optional)           The real part of the lowest E-value
        EMINEV : Real  (optional)         EMIN, given in eV with respect to the Fermi level



        SECTION SCF contains:
    -------------------------
        NITER : Integer ≝ 200             Maximal number of iterations of the SCF cycle
        MIX : Real ≝ 0.2                  Mixing parameter
        MIXOP : Real  (optional)
        VXC : AnyOf(VWN,MJW,VBH,PBE,PW92,EV-GGA,BJ,MBJ) ≝ VWN

            Possible values:
              VWN       Vosko, Wilk, Nusair (type: LDA, libxc equivalent: LDA_C_VWN)
              MJW       Janak, Williams, Moruzzigit g (type: LDA, libxc equivalent: -)
              VBH       von Barth, Hedin (type: LDA, libxc equivalent: LDA_C_VBH)
              PBE       Perdew, Burke, Ernzendorfer GGA (type: GGA, libxc equivalent: GGA_X_PBE)
              PW92      Perdew Wang (type: GGA, libxc equivalent: GGA_X_PW91)
              EV-GGA    Engel and Vosko GGA (type: GGA, libxc equivalent: GGA_X_EV93)
              BJ        Becke-Johnson (type: metaGGA, libxc equivalent: MGGA_X_BJ06)
              MBJ       modified Becke-Johnson (type: metaGGA, libxc equivalent: MGGA_X_BJ06)

            parametrisation of the exchange-correlation potential

        ALG : AnyOf(BROYDEN2,TCHEBY) ≝ BROYDEN2

            Possible values:
              BROYDEN2  Broyden’s second method
              TCHEBY    Tchebychev

            Mixing algorithm

        EFGUESS : Real  (optional)        Skip the Fermi energy search in the beginning.
        TOL : Real ≝ 1e-05                Tolerance threshold for the mixing algorithm
        ISTBRY : Integer ≝ 1              Start Broyden after ISTBRY iterations
        FULLPOT : Flag ≝ False            Non-spherical callculation (full-potential) instead of ASA
        ITDEPT : Integer ≝ 40             Iteration depth for Broyden algorithm (length of used history)
        QION : Array(of Real)  (optional)  Guess for the ionic charges Qt for atomic types
        QIONSCL : Real  (optional)        Guess for the ionic charges Qt for atomic types
        MSPIN : Array(of Real)  (optional)  Guess for the magnetic moment μ_{spin,t} for atomic types
        USEVMATT : Flag ≝ False           Set up the starting potential using the original Mattheissconstruction for the potential V instead of the charge density



        SECTION SITES contains:
    ---------------------------
        NL : Array(of Integer) ≝ [3]      Angula momentum cutoff (the first discarded l-space)




    Expert options:
    --------------
                SECTION STRCONST (optional, expert) contains:
        -----------------------------------------------------
            ETA : Real  (optional)        Ewald parameter
            RMAX : Real  (optional)       Convergency radius in real space
            GMAX : Real  (optional)       Convergency radius in reciprocal space

            The calculation of the ~k-dependent KKR structure constant matrix G(~k, E) is controlled by
            three convergence parameters. ETA determines the relative weight of the real and reciprocal
            space lattice sums, that are determined by the convergence radii RMAX and GMAX, respec-
            tively. These convergence parameters have to be optimised anew if the lattice structure, the
            lattice parameter or the energy or ~k-range used is changed. This is done by the program if
            no values are applied via the input file. In some cases, in particular if one works at high
            energies, it might be necessery to set the convergence by hand. For this purpose one can start
            from the values set by kkrgen or kkrscf (see the output file).

                SECTION CPA (optional, expert) contains:
        ------------------------------------------------
            NITER : Integer ≝ 20          Maximum number of CPA iterations
            TOL : Real ≝ 0.0001           Threshold for stopping CPA-cycle

            For a system with substitutional disorder, the CPA is used. The listed variables control the CPA cycle

                SECTION MODE (optional, expert) contains:
        -------------------------------------------------
            MODE : AnyOf(NREL,SREL,SP-SREL)  (optional)

                Possible values:
                  NREL      work in the nonrelativistic mode
                  SREL      work in the scalar-relativistic mode
                  SP-SREL   work in the spin-polarized scalar-relativistic mode

                Using this option you can switch on the spin polarization and relativistic mode. If its not set (or set to FREL), the full relativity mode is used.

            LLOYD : Flag ≝ False          Use LLoyd formula for scattering operator. It can improve the accuracy of the Fermi energy.
            MDIR : Array(of Real of length 3) ≝ [1. 0. 0.]  (optional, add non-default, array)  Common magnetisation direction vector with x, y and z in Cartesian coordinates. The normalisation is arbitrary.
            C : Real ≝ 1.0  (optional, add non-default, array)  Scale the speed of light for a given atom type.
            SOC : Real ≝ 1.0  (optional, add non-default, array)  Scale the strength of the spin-orbit coupling for atom type.

            This section contains options that describe, how to consider relativity and/or spin. If the MODE is not specified otherwise the programs of the SPRKKR-package assume that a magnetic system should be treated in a fully relativistic way. By setting the parameter SP-SREL a slightly faster scalar relativistic calculation can be done instead for a magnetic system.

Module Attributes

input_parameters()

SCF task input parameters definition

Functions

input_parameters()

SCF task input parameters definition