AUTOMATIC VARIANCE REDUCTION FOR 3-DIMENSIONAL MONTE-CARLO SIMULATIONS BY THE LOCAL IMPORTANCE FUNCTION TRANSFORM .2. NUMERICAL RESULTS

The performance of the local importance function transform (LIFT) meth od for several three-dimensional, linearly anisotropic-scattering, one -group, and multigroup transport problems is demonstrated. In these pr oblems, the LIFT method is shown to be more efficient than the AVATAR scheme, which is one of the most efficient variance reduction techniqu es currently available in a production Monte Carlo code. For most of t he problems considered, the LIFT method produces higher figures of mer it than AVATAR even when the LIFT method is used as a black box.