Rational Mapping (RAM) of In-Core Data

Abstract

The paper describes and demonstrates a unique processing of in-core flux detector data, such that the detailed in-core power distribution can be derived with great accuracy by combining a specially "smoothed-out" set of in-core data with neutron diffusion theory.RAM is designed in such a way that erratic detector signals are recognized very efficiently and can be eliminated from the experimental data set: this is achie1Ted by modal expansion of the difference between theoretical fluxes and experimental fluxes at the detector sites.Sensitivity studies have shown that RAM is quite stable, does not absorb the "wild" detector errors in the mapping procedure and results in mapped fluxes with errors about three times smaller than would be obtained by direct interpolation of detector readings.