C. Kloss1, M. Schilling2, C. Goniva1
1Johannes Kepler University/AT, 2BASF Ludwigshafen/DE
The second proposed mini-symposium focuses on the modelling of particulate flow by means of particle-based methods and their coupling to Computational Fluid Dynamics (CFD).
In contrast to continuum approach based models, which consider the bulk of particles as an artificial continuum, particle-based methods picture the motion of each particle (or a set of representative particles) individually, with a detailed handling for collisions. Many such methods have been developed over the years, most prominently the Discrete Element Method (DEM) originally proposed Cundall and Strack in 1979. Typically, a soft-particle or time-driven approach is used, where trajectories are integrated using differential momentum as it can capture the physics of static and slowly moving packings. Thus, the main advantage over currently available continuum models is that both solid-like and fluid-like state of granular media can be pictured.
Using the Discrete Element method in conjunction with Computational Fluid Dynamics (CFD), the so called CFD-DEM Method, fluid-granular systems can be calculated. Hereby particle motion is resolved by DEM and gas flow by CFD solving a modified set of Navier Stokes Equations and a set of dedicated sub-models. Models for physical phenomena such as inter-phase momentum, turbulence and heat exchange as well as numerical aspects such as Lagrangian-Eulerian field mapping are subject of current research. These modelling efforts are very demanding as non-sphericity or cohesion are to be accounted for. Nevertheless provided a sound model framework is used this method spans and enormous range of applications and application fields, such as bulk solids handling in the minerals and mining industry, metallurgical, pharmaceutical or agricultural processes, pneumatic conveying, cyclone separation, hopper flow, die filling, fluidized beds, granulation processes, risers or similar equipment.
Lately, the DEM and CFD-DEM methods come more and more into the focus of engineers and researchers. Yet these methods require high CPU effort making them a typical application for high performance computing (HPC). Still, most large-scale industrial applications cannot be simulated on original scale using the DEM approach. For this reason, coarse-graining schemes for the DEM approach are being eagerly discussed in recent literature.
As a summary, this mini-symposium will cover four main aspects:
This mini-symposium will provide a platform for discussion on these or related topics. Potential participants are encouraged to present their theoretical research as well as applications and/or experimental validations.