The last decade has seen tremendous advancements in computer technology, not only in computational speed but also in memory and storage resources. Calculations which twenty years ago could only be done on a main-frame computer, can now easily be performed on many personal computers. This book, Computational Atomic Structure, was written partly in response to this change.
Our book differs from other books on atomic theory or quantum mechanics in that it deals not only with theory but also the computational aspects of actually determining an atomic property. An important goal of this book is to review the atomic structure theory needed for the computation of atomic properties using programs from the MCHF ATOMIC STRUCTURE PACKAGE that are available from a number of ftp sites. Such calculations can be at a simple, Hartree-Fock level or at a more advance, multiconfiguration Hartree-Fock level. Another goal is to explain the computational procedures that are used. Included in the book are sample calculations performed in an interactive mode, to show what input is expected from the user and sections of screen output. The latter serves two purposes - it gives the reader an idea of the information that can be obtained and allows the person installing the system to confirm proper execution of the code. In the interest of space, only selected output is provided. Complete output will be provided with the code where examples too complex for a book will be presented. Finally, it was anticipated that some users may wish to modify the programs. To encourage such changes, we deal briefly with the computational aspects and outline some of the numerical procedures.
In the MCHF approach, the first step is always to find an approximate wave function for bound states of an atomic system. An energy criterion is used for this purpose and several chapters are devoted to the study of factors affecting energy level structures. Once a wave function has been determined, a number of atomic properties can be computed. The chapter on transition probabilities is the ``grand finale'' that ties everything together. However, the book does not stop there. It includes a brief excursion into the theory of continuum processes, just enough to establish a bridge between these two areas.
With the book and the code it should be possible to explore many atomic properties. An experimental approach is recommended - some of these experiments will be more accurate than others. The goal should always be an understanding of the important effects. However, the present code should not be viewed as a ``state-of-the-art'' code. It's development started in the mid 1960's and much could be done to improve it. Indeed, for large-scale calculations, the authors have already modified this version for sparse-matrix methods and dynamic memory allocation. The resulting code, however is not standard whereas the present version, with at most minor modifications, will compile on both PC's and workstations. It is, in a sense, a small case version where more intermediate data can be printed, making it more suited as a learning tool. Missing entirely is a graphical interface partly because of a lack of a standard. Such an interface could be a valuable contribution.
The code will be available from several sites:
These programs have been tested on many platforms and applied to many atomic systems. But as research software, they cannot be guaranteed for any particular purpose. Neither the authors nor the publisher offer any warranties, nor do they accept any liabilities with respect to the programs and their application.
Your comments and suggestions are welcome. Please send electronic mail to Charlotte.F.Fischer@Vanderbilt.edu or brage@kurslab.fysik.lu.se.
