Elsevier Science Home
Computer Physics Communications Program Library
Full text online from Science Direct
Programs in Physics & Physical Chemistry
CPC Home

[Licence| Download | New Version Template] aanc_v2_0.gz(57 Kbytes)
Manuscript Title: Continuum wavefunction solver for GRASP.
Authors: W.F. Perger, Z. Halabuka, D. Trautmann
Program title: CONTWVG
Catalogue identifier: AANC_v2_0
Distribution format: gz
Journal reference: Comput. Phys. Commun. 76(1993)250
Programming language: Fortran.
Computer: SUN SPARC Station, IBM 4381, IBM RS6000, VAX 850.
Operating system: OS 4.1.1&4.1.2, CMS VM/SP HOP 5.O, AIX 3.1.5, VMS.
RAM: 1600K words
Word size: 8
Keywords: Atomic physics, Racah, Structure, Multiconfiguration, Fractional parentage, Dirac-Fock, Recoupling coefficients, Slater integrals, Complex atoms, Relativistic, Dirac equation, jj-coupling.
Classification: 2.1.

Subprograms used:
Cat Id Title Reference
ABJN_v1_0 GRASP CPC 55(1989)425

Nature of problem:
The relativistic Dirac-Fock equations are set up and solved numerically for continuum wavefunctions within the framework of the GRASP program, K.G. Dyall, et al, CPC 55(1989)425-456.

Solution method:
Relativistic atomic wavefunctions are calculated using a central differences method with deferred corrections within the configuration interaction framework of the GRASP program. The grid spacing at large radial distances (the linear region of the two-piece grid) is automatically determined to provide at least 10 grid points per half cycle of the wavefunction. Lagrange multipliers are automatically calculated without additional input. Reason for the new version This new version is required for obtaining relativistic continuum orbitals with the new version of the bound state program, GRASP, and to add refinements to the original CONTWV program.

Continuum orbitals for atoms ranging from hydrogen to mercury have been calculated, with up to six Lagrange multipliers and energies ranging from zero to 100 atomic units. Cases outside of these limits will likely also succeed. The program automatically sets the spacing of the radial grid to provide for at least 10 grid points per half cycle (of the oscillations in the wavefunction). Consequently, the rapid oscillation of the wavefunction for very large energies will likely result in an error message which instructs the user to increase the number of grid points (which is set by the additional string variable, NPC, in the pre-processor string file).

Unusual features:
The program will calculate a virtual state V**N-1 continuum orbital for a given kappa quantum number and dump that orbital to a file. The program uses a relativistic WKB approximation for normalization of the continuum orbital in the presence of a non-neutral core and it uses a curve-fitting procedure for normalization in the case of a neutral core. The spacing of the radial grid at large radial distances, where the two-piece grid is linear, is automatically calculated to provide at least 10 grid points per half cycle. The procedure for the automatic calculation of the Lagrange multipliers has been improved (relative to CONTWV) to achieve better orthogonality between the orbitals. The program allows the user to choose whether or not to calculate the phase shift, thus providing for better computational speed if the phase shift is not required.

Running time:
The typical running time is 10 seconds to 1 minute on an IBM RS6000, Model 540.