Predictive Calculation of Hetero-Agglomeration Processes in Multi-Substance Suspensions

• Applicant: Dr.-Ing. Frank Rhein
• Subject Area: Mechanical Process Engineering
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 520278921

Agglomeration is a unit operation of mechanical process engineering and therefore relevant for almost all processes in particle technology. Targeted or selective agglomeration in multi-substance suspensions has a wide range of applications. A current example is the microstructure in lithium-ion batteries, which is adjusted by hetero-agglomeration and significantly influences the resulting product properties. Only limited analytical capabilities exist to determine relevant properties, such as the material-specific agglomerate composition. This motivates the search for predictive computational methods that provide insight into the micro-processes and thus grant access to process information that cannot be measured directly. Generally, population balance equations are used for this purpose, which allow the calculation of macroscopic agglomeration processes. A reliable determination of the agglomeration and breakage rates, the so-called kernels, is essential. Especially in multi-substance suspensions this is challenging and up to now hardly described, since all interactions between all occurring components have to be considered and their description is complicated by locally heterogeneous surface properties. The proposed project follows a multi-scale approach in which different methods complement each other: CFD-DEM simulations on macro- and micro-scale provide the necessary kernels under incorporation of experimental data. The analyses are initially limited to the single- and two-substance system. The variation of relevant process parameters creates a broad data set, which is extended and transformed by machine learning methods into a model capable of interpolation. This is essential for real-time integration into the population balance equations and at the same time realizes the necessary multi-scale coupling. Subsequently, the focus lies on criteria and models for transferring the kernels from the two- to the arbitrary multi-substance system. Agglomeration experiments in the multi-substance system as well as detailed analyses of the produced hetero-agglomerates allow the validation of the predictive calculation. Finally, the established procedure is transferred to technically relevant material systems that have not been investigated so far, in order to test its generalization capability. At the end of the project, a systematic methodology exists with which hetero-agglomeration processes in arbitrary multi-substance suspensions can be calculated from scratch, i.e. without prior knowledge. This methodology offers broad application potential for almost all areas of particle technology and helps to make the often complex micro processes more transparent and controllable.

• DFG Programme: Research Grants
• Co-Investigator: Professor Dr.-Ing. Hermann Nirschl