Project Details
Projekt Print View

Agglomeration and cluster formation of microscacle particles and droplets in highly loaded turbulent gas flows

Subject Area Fluid Mechanics
Term from 2012 to 2020
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 215541820
 
Final Report Year 2019

Final Report Abstract

The objective of the present project was to take the simulation possibilities of disperse turbulent multiphase flows with high mass loadings a decisive step forward. The basis for this should be the Euler-Lagrange approach which allows to predict the continuous flow based on high-fidelity eddyresolving technique such as the LES technique. Even more essential is that the particulate phase can be described based on first principles. Thus, starting from a four-way coupled Euler-Lagrange methodology with a deterministic collision model, the task of the project was to find, evaluate and incorporate appropriate modeling and simulation strategies for two physical processes, which dictate the size distribution of the disperse phase, i.e., agglomeration and breakup. The first phase of the project was devoted to agglomeration processes, considering both the agglomeration of dry electrostatically neutral particles and the coalescence of surface-tension dominated droplets. For the agglomeration process two different models were derived, a momentum-based model and an energy-based model. In the former particle agglomeration is considered by evaluating particle-particle collisions in the framework of the hard-sphere model taking the cohesive van-der-Waals forces acting between the collision partners into account. The latter relies on an energy balance of the collision partners including friction. Both agglomeration models were successfully validated based on theoretical results. Afterwards, a comparative study of both models with applications to turbulent particle-laden flows in a downward directed vertical channel flow was carried out. Furthermore, the momentum-based model formed the starting point to derive an enhanced particle–wall adhesion model for dry rigid particles (deposition), which takes the physics of the problem more realistically into account than previous models. Similar to the other models a thorough validation based on a variety of test cases and appropriate reference data was carried out. Another final outcome of the first phase was that the entire methodology was extended towards surface-tension dominated droplets. For this purpose, a deterministic composite collision outcome mode for droplets was developed and validated considering the four relevant regimes of binary collisions, i.e., bouncing, fast coalescence, reflexive and stretching separation. The second phase of the project was devoted to breakup processes. Here the breakup provoked by three different fluid-induced forces (turbulence, drag and rotation) and by collisions with walls or other agglomerates/particles was taken into account. For the former individual models were derived for each mechanism under the challenging constraints to incorporate as much as possible the physics of the underlying problem while keeping the computational effort at a reasonable level in order to describe multiphase flows with a high mass loading and thus a huge number of particles/agglomerates. The models developed are fulfilling these conditions. They include distinctive conditions for each breakup mechanism and deliver characteristic properties of the arising fragments (size, density and velocity) to be tracked through the flow field after breakage. Besides other test cases the modeling strategy was extensively evaluated based on the practical application of the turbulent flow inside a disperser. Investigating two different flow cases and three different powders allowed to extensively analyze the performance of the derived models confirming the excellent predictive quality of the simulation approach. In order to derive appropriate data-driven models for the impact of agglomerates at walls (or with other agglomerates), extensive DEM simulations were carried out to characterize the phenomenon and to evaluate the influence of all relevant parameters on the outcome of breakage. Based on the results a new dimensionless number was proposed, which allows to reasonably provide a unified description of the results generated under a wide variety of conditions. In addition, a detailed investigation revealed that in some cases (re-)agglomeration plays a dominant role leading to a deviation from the expected behavior. Accordingly, relationships to describe the breakage behavior and the resulting particle size distribution were proposed and fitted to the data least influenced by (re-)agglomeration since the latter phenomenon is separately accounted for in the present modeling framework. Moreover, additional modeling work is in progress to provide a proper description of the post-impact velocities of the fragments in order to complete the work package by incorporating the models in the Euler-Lagrange approach. In summary, the project was very successful in the derivation, validation and integration of various models for the agglomeration and breakup phenomena highly relevant for reliable simulations of disperse turbulent multiphase flows with high mass loadings.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung