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] adjd_v1_0.gz(282 Kbytes)
Manuscript Title: TIMEDEL: a program for the detection and paramaterization of resonances using the time-delay matrix.
Authors: D.T. Stibbe, J. Tennyson
Program title: TIMEDEL
Catalogue identifier: ADJD_v1_0
Distribution format: gz
Journal reference: Comput. Phys. Commun. 114(1998)236
Programming language: Fortran.
Computer: IBM RS6000.
Operating system: AIX 4, DEC UNIX.
RAM: 5M words
Word size: 32
Keywords: Resonances, Scattering, Time-delay matrix, K-matrix, Branching ratios, General purpose, Fit, Molecular physics, Electron, Molecule.
Classification: 4.9, 16.5.

Nature of problem:
TIMEDEL detects and parameterizes resonances found in scattering problems.

Solution method:
K-matrices are read in or calculated on the fly (preferred option). The time-delay [1] is calculated from the K-matrices. The resonances are detected by examining the time-delay experienced by the incident particle and fitted to a characteristic Lorentzian or sum-of-Lorentzian form [2].

The program will only find resonances which have definite maxima in the time-delay. Thus two resonances, very close in energy and width, will be fitted as one resonance. The program also assumes that the effect on the time-delay of more than one resonance is additive. Although this is normally the case, experience has shown this is not always true.

Unusual features:
The program uses a processor-dependent 'double complex' variable definition extension to the F90 standard. TIMEDEL can be used either with a user-supplied file of K-matrices or requires a subroutine (called here GETKMAT) supplied by the user to find the K-matrix of their problem at a particular energy. TIMEDEL also uses NAG library routines: E04FDF, a minimization routine; F04ADF, which solves the matrix equation AS = B; and F02HAF, which finds eigenvectors and values of a matrix. These routines are not supplied here and the code must be compiled with the NAG library [3].

Running time:
2.0 s plus the time taken by the user supplied subroutine to calculate the K-matrices.

[1] F.T. Smith, Phys. Rev. 114 (1960) 349.
[2] D.T. Stibbe and J. Tennyson, J. Phys. B: At. Mol. Opt. Phys. 29 (1996) 4267.
[3] NAG Fortran Library Manual Mark 17 (1996) Numerical Algorithms Group, Oxford, UK.