From particles via decives to autonomous processes

Hermann Nirschl, Karlsruhe Institute of Technology (KIT), Germany

In recent years huge progress has been made in the field of simulation of the particle properties and processes. The simulations, often based on CFD or DEM methods, can help to predict the particle behaviour in machines and devices. But the number of particles in the simulations is still limited and very often the description of the interaction between the single particles is still a huge challenge. Therefore experimental methods have been developed to get information about the real particle or better bulk behaviour and to use them in lower order models. Also AI methods can help here to reduce the experimental efforts and further to predict the material properties directly from inline measurements. This makes it possible to realize autonomous processes by means of a closed control loop, where the process models and the insitu measurement technology are coupled with a process control strategy. It is clear that reduced order models, abstracted from complex simulations or lower order models, are necessary to have a ‚faster as real time‘ performance which helps to predict the process. Also insitu characterization devices will become more and more important for the realization of a ‚model predictive control‘ strategy. This not only allows an automatic optimization of the target variables, but also helps to ensure a high resource efficiency according to raw materials and energy consumption. The presentation explains the basics of autonomous processes and their implementation in devices for particle manufacturing and handling.

Karlsruhe Institute of Technology (KIT)
Institute for Mechanical Process Engineering
Straße am Forum 8
76131 Karlsruhe, Germany