Thickness and mixed sample effects on areal density measurement with NRTA for particle like debris of melted fuel

Neutron resonance densitometry (NRD) is currently being developed to quantify super nuclear materials in particle-like debris of melted fuel that is generated by a severe accident like the one at the Fukushima Daiichi nuclear power plant. NRD is a combination of neutron resonance transmission analysis (NRTA) and neutron resonance capture analysis (NRCA) or prompt gamma-ray analysis (PGA). Clearly, systematic effects due to the characteristics of samples such as the sample inhomogeneity, presence of impurities, radioactivity and temperature would have a strong impact on the accuracy. To study the uncertainty due to the sample characteristics, NRTA experiments were made at the time-of-flight facility GELINA (Geel Electron LINear Accelerator), with Cu discs with different thickness and Cu discs attached to B4C sample. We show results considering sample thickness effect and mixed sample effect on NRTA measurements. Data are analysed with the resonance shape analysis code REFIT to derive the elemental composition of the sample and the areal density of the main components. In addition the effect of neutron absorbing matrix material will be discussed.