Preliminary results for simply sparse as a preconditioner to SIM

Simply sparse is a new method for compressing matrices that occur in integral equation computer programs such as NEC2. The resulting sparse matrix is used in this investigation as a preconditioner to the sparse iterative method (SIM) in an attempt to improve the spectral properties of the coefficient matrix. The memory requirements become unacceptably high for the SIM iterative scheme when using an unmodified simply sparse matrix. Reducing the density of the simply sparse matrix improves the memory requirements while maintaining good solution times. This is, however, structure specific since, for example, the Hawk aircraft problem does not converge while the helicopter problem converges in almost half the time taken by SuperNEC with a matrix density of 1.62% before factorization and 2.35% after factorization.