Synthesizing of realistic model textures of mineral raw materials based on quantitative characteristic numbers of the texture as basis for the simulation of comminution processes

• Applicant: Professor Dr.-Ing. Holger Lieberwirth
• Subject Area: Mechanical Process Engineering ; Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing ; Engineering Design, Machine Elements, Product Development ; Mechanical Properties of Metallic Materials and their Microstructural Origins ; Mineralogy, Petrology and Geochemistry
• Term: from 2016 to 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 320596194

The importance of simulation based research increases also in the field of mineral processing technology considering the limited resources for research with machinery on one hand and more and more powerful computer technology at attractive prices on the other hand. Particularly, in the field of comminution technology, known empirical and partly empirical research methods were replaced or supplemented by the specific use of simulation methods. Parallel to the simulation approaches regarding the liberation of valuable minerals, particularly the Discrete-Element-Method has emerged as an accepted tool With regard to the comminution technology, simulations that use so called Bonded Particle models (BPM) within DEM environments seems to be the best choice. However, the number of simulations of comminution behavior is still small and their significance is limited. One reason for this is certainly that so far no sufficiently clear correlations between the parameters of the DEM model on one side and the binding mechanisms of real particles on the other side were found. Furthermore, the literature still shows no adequate method to realistically transfer microstructures of natural mineral raw materials into the simulation environment. However, in order to develop effective processing technologies using simulation tools, the material and mineralogical properties of quantitatively analyzed structures have to be mapped statistically representative by the model. Without these fundamentals, the robustness of the statements derived from the models is limited. The aim of the project is therefore, to realistically map hard rocks and ores on the basis of the quantitative microstructural analysis developed at the Institute of Mineral Processing Machines into a Bonded-Particle-based DEM model. Moreover, the mapping has to involve all determined characteristics of the mineral structure. In addition, the micro hardness and fracture toughness of the minerals have to be considered in order to define the bonding parameters between the discrete elements. By combining a Bonded Particle Model, which represents the grain structure realistically, with strength parameters that are mineral specific, a significantly higher prognostic accuracy of DEM models for hard rocks and ores is expected.

• DFG Programme: Research Grants