Project 3.2

Multiscale and reactive granular flows


This project aims to use numerical simulations to optimize operation and reduce emissions of metallurgical facilities, which feature reactive granular flows on large and small scales. The blast furnace (BF), MIDREX and HYFOR (see also Project 2.2) processes for ironmaking, and the rotary kiln for refractory material production, exemplify such processes.

The numerical description of these processes needs to combine effects varying in scale by multiple orders of magnitude. Therefore, this project considers models ranging from detailed single particle descriptions, Euler-Lagrange methods (reduced order particle models, CFD-DEM coupling, MPPIC), coarse graining, and Euler-Euler methods. Existing and new models are utilized, improved, newly developed and integrated into different frameworks.

The alternative reducing agent (ARA) reactor, investigated in P1.1, and a novel macro-TGA developed as part of this project, serve to increase the accuracy and fidelity of the numerical models. They provide the validation data for the thermochemical conversion kinetics that are the basis of the numerical simulations. Also process data from industrial settings will be used for validating simulation models.

Objectives and Motivation

  • Evaluation and optimization of the injection process of sustainable ARAs by means of large-scale and particle resolved modeling
  • Improving the detailed particle model by including swelling, shrinking, softening, and melting
  • Using ARA reactor experiments to improve numerical ARA conversion rate modeling
  • Obtaining the defining raceway characteristics from detailed simulations in form of boundary conditions for full scale blast furnace simulations to reduce computational cost
  • Creating a comprehensive iron ore reduction model
  • Determining coarse-grained material parameters from fine-grained simulations containing complex processes like fragmentation, mixing, or segregation
  • Extension of popular coarse-graining methods developed for simple, idealized conditions to more realistic settings (e.g. strong polydispersity, cohesive powders, etc.)
  • Deriving a comprehensive rheological model for cohesive granular materials with broad particle size distributions and sub-grid models accounting for small geometry features
  • Application of these methods to the BF, MIDREX and HYFOR processes
  • Numerical and experimental investigation of the calcination and sintering process of refractory materials under realistic industrial conditions
  • Implementation and design of a macro-TGA suitable to study industrial-grade materials
  • Utilizing particle models and experiments as basis towards CCU in calcination processes


  • Literature- and experiment-driven model development
  • Detailed single particle CFD simulations
  • CFD-aided conversion rate extraction from experimental ARA reactor
  • Various CFD and DEM based methods for modeling particulate flows, e.g., Euler-Lagrange (CFD-DEM, MPPIC) and Euler-Euler
  • Fine-grained simulations providing detailed information about complex conditions, and generalized simple coarse-graining
  • Time-extrapolation techniques to study dense particle beds over the duration of hours at relatively little numerical costs
  • Particle based methods to model agglomeration
  • Literature research, engineering and construction of a macro-TGA
  • Systematic experimental evaluation using DoE (design of experiment) methods

Results and Application

The detailed particle model for ARA conversion, developed in the precious funding period, is extended to predict particle behavior, including shrinkage, swelling, softening, and melting. The model accuracy is improved by CFD-aided experimental data evaluation using the novel test rig for ARA conversion developed in P1.1. In-depth understanding of the thermo-chemical conversion process of metallurgical reducing agents (solid ARAs, coke, gaseous, and liquid agents) under blast furnace conditions is obtained.

The description of the BF, MIDREX, and HYFOR processes is improved. Existing simulation techniques are improved, and new simulation techniques are created in order to carry out numerical experiments in real scale over process-relevant durations. The systemic inclusion of thermochemical and physical effects, such as carburization, particle fragmentation, segregation, mixing, rheology, and heterogeneity allows for qualitatively realistic results.

The role of fines in process irregularities within moving particle beds with respect to fines transport and deposition is studied numerically. Insights into irregular behavior are gained, leading to the formulation of rules and guidelines to prevent such events. Numerical analysis identifies the impact of gas velocity, agglomerates, and reactor geometry on local irregularities and process efficiency.

A macro-TGA setup for the study of calcination and sintering of industrial input materials is designed and built. A detailed particle model is developed for the calcination and sintering of refractory materials based on results extracted from the macro-TGA. A simplified calcination-sintering model is then derived from the detailed model and experiments, complementing investigations in Area 2related to carbonization, methanation, and the simulation of green furnaces.