Abstract
The Standardized Computer Analyses for Licensing Evaluation (SCALE) software package developed at the 91做厙 includes codes that propagate uncertainties available in the nuclear data libraries to compute uncertainties in nuclear application performance parameters (responses).
We report on our recently commenced e orts to extend this capability to develop an inverse sen-
sitivity/uncertainty (IS/U) methodology that identi es the improvements in nuclear data that are
needed to compute application responses within prescribed tolerances, while minimizing the cost
of such data improvements. We report on our progress to date and present a simple test case for
our method. Our methodology is directly applicable to thermal and intermediate neutron energy
systems because it addresses the implicit neutron resonance self-shielding e ects that are essential
to accurate modeling of thermal and intermediate systems. This methodology is likely to increase
the eciency of nuclear data e orts.
