Zusammenhänge zwischen sichtbaren Eigenschaften und der chemischen Zusammensetzung von Shredderschrottpartikeln

These master¿s thesis addresses the question of whether visually detectable characteristics of magnetic shredder scrap particles allow conclusions to be drawn about their chemical composition. The aim of the thesis was to investigate the suitability of visual properties for the quantitative separation of material fractions and for the targeted enrichment of depletion of certain alloying elements ¿ with the overarching goal of improving the quality of secondary raw materials. To this end, data from a comprehensive database of shredder scrap particles was analysed using MATLAB. Visual characteristics such as material type, shape and condition were systematically evaluated and correlated with chemical element contents. The analysis confirms that copper, non-metallic and heterogeneous composite materials in particular are to be considered quality-reducing components. Analysis of the particle composition showed that copper particles accounted for only about 8 % of the total particles but were characterised by high copper content (> 68 %) and low iron content. The non-metallic fractions comprised a variety of materials, including wood, glass, textiles, minerals, plastics and composite materials, the latter often representing combinations of different types of materials and predominantly originating from electrical engineering. The metallic fractions showed high material recycling potential and could largely be classified as high-quality fractions. Clear content classes enable the identification of homogeneous subfractions. Fractions with a high iron content (> 80 %) and low accompanying elements were rated as high quality, while lower iron contents were associated with increased concentrations of zinc, nickel and copper. The analysis of specific component types, in particular screws and nuts, showed a clear correlation between the alloying elements nickel and zinc and the respective material composition. This demonstrated that visual characteristics of certain component geometries allow conclusions to be drawn about chemical compositions, even if the accuracy of such visual classifications remains limited. At the same time, the limitations of purely visual classification became apparent, especially in the case of chemically variable feature groups. The results thus underline the potential of visual sorting approaches, while at the same time pointing to the need of complementary analytical methods to ensure precise and reliable material separation.