Imaging multi-materials tightly combined objects by applying grey relational analysis in muon tomography

A novel method developed within Muon Scattering Tomography (MST) to location interface of Multi-Materials Tightly Combined (MMTC) objects. MST has high sensitivity to high-Z materials such as uranium, and is very useful for detecting them in a background of low-Z materials. But commonly used algorithms in MST such as the Point of Closest Approach (PoCA) are difficult to location the interface of MMTC objects, limited its application. In this paper, we investigated Grey Relational Analysis (GRA) based on PoCA to obtain the MMCT interface identification. Then, we applied this novel approach to image the layered structure of the molten pool, which may be of reference value for cooling systems. The results show that 5 days measurement time spans can discriminate the two- and three-layered molten pool models. The further study also shows that this method does not require high position resolution detectors. The uncertainty of interface location in molten pool models under the plastic scintillator detector with a 2cm position resolution is less than 10%. This novel method is practical and provides a new and effective idea for image reconstruction of MMTC objects.