Progress in the Application of Multidimensional Particle Property Distributions: The Separation Function

Authors:
Open Access
Publisher website
Google Scholar
Abstract
In KONA 2022, the fundamentals of two- and multidimensional particle size distributions were introduced. The next question in the field of two- and multidimensional distributions addresses their application to describe a particle process, e.g., agglomeration or separation. A multidimensional separation can be seen as retrieving only particles with a specific set of properties from a multidimensionally distributed system, e.g., retrieving only small particles (below a certain threshold in size) with a compact spherical shape (above a certain threshold in sphericity). The multidimensional separation allows the generation of functional particle systems with specific properties, e.g., semiconducting, optical, or electronic properties, which are required for high-technology applications. Starting from so-called particle-discrete information, i.e., an information vector for each particle containing its compositional, geometrical, and physical properties, it is possible to describe a multidimensional separation in full detail based on various properties. Each particle can be evaluated according to different separation properties, e.g., size, shape, and material composition. With this database, it is possible to define and work with separation functions to describe the multidimensional separation and quantify the separation results. For example, in the two-dimensional case, the median cut size becomes a median cut line, where the probability for a particle to belong to the concentrate is 0.5. Some case studies and examples show different approaches and possibilities to achieve a multidimensional separation in one or several connected process steps.
Funding Information
  • Deutsche Forschungsgemeinschaft (313858373)
  • Federal Ministry of Education and Research (BMBF, Bundesministerium für Bildung und Forschung (03XP0337A)

This publication has 69 references indexed in Scilit: