Proceedings of the second international seminar on the generator coordinate method for nuclear bound states and reactions

In the method of the complete diagonalization one solves exactly (for a given effective in the action) the eigenvalue problem within the N-particle model space spanned by all Slater determinants of chosen single-particle functions. In principle one can obtain the same solution also with the GCM if an appropriate generator function (x,a) and a sufficient number of GC's are chosen. However, in practice one intends to apply the GCM with few GC's only and a generator function which is easy to handle. Since those functions span only a subspace of the N-particle model space there arises the question for the quality of such an approximation in comparison with the complete shell model diagonalization even if the effective interaction is not further adapted to account for this truncation. Recently in two publications 1, 2 This idea has been tested on the nuclei 160, 20Ne, 14.Ne (for which there exist results of the complete diagonalization). The results are encouraging. In all cases the appropriate GC's have been guessed by analyzing a few experimental data, and by interpreting results of comparatively simple calculations.