Analyzing Iron Dust Bunsen Flames using Numerical Simulations

T. Hazenberg, D. Braig, M. A. Fedoryk, J. Mich, F. P. Hagen, S. R. Harth, B. Stelzner, A. Scholtissek, D. Trimis, C. Hasse · 2025

Read Analyzing Iron Dust Bunsen Flames using Numerical Simulations by T. Hazenberg, D. Braig, M. A. Fedoryk, J. Mich, F. P. Hagen, S. R. Harth, B. Stelzner, A. Scholtissek, D. Trimis, C. Hasse (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-20692.txt.

Introduction

Thijs Hazenberg and co-authors simulate a flame fed by iron dust in this 2025 paper. Iron dust can burn like a fuel, and a Bunsen-style flame is a standard set-up to measure how fast that burning front moves. Getting that speed right in a model is key for future iron fuel systems.

Their three-dimensional model was checked against lab results, and the burning speed drawn from it matched the experiments better than any prior model. A comparison with a simpler one-dimensional model showed why the real flame differs. Flow from the side mixes into the burnt zone and raises oxygen in the reaction layer, which lifts the speed. Stretch and curve of the flame front also play a part.

Catalog record

CREATORT. Hazenberg, D. Braig, M. A. Fedoryk, J. Mich, F. P. Hagen, S. R. Harth, B. Stelzner, A. Scholtissek, D. Trimis, C. Hasse
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE49 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-20692

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Analyzing Iron Dust Bunsen Flames using Numerical Simulations T. Hazenberg, D. Braig, M. A. Fedoryk, J. Mich, F. P. Hagen, S. R. Harth, B. Stelzner, A. Scholtissek, D. Trimis, C. Hasse Abstract This article presents numerical simulations of an iron dust Bunsen flame. The results are validated against experimental results. The burning velocity is extracted from the 3D simulation results, as in the experiments. The agreement of the burning velocity between the model and experiment is the best to date for iron dust flames. A comparison is performed between 3D and 1D simulations to improve our understanding of how the 3D Bunsen flame deviates from an ideal 1D flame. This comparison reveals that the co-flow mixes with the post-flame zone, increasing the oxygen concentration in the reaction layer, which increases the burning velocity. Moreover, the analysis also reveals that stretch and curvature affect the burning velocity. These results are valuable for the future development of experimental setups aimed at measuring the burning velocity. Journal: XXX 1 Introduction Moving away from fossil fuels to renewable energy sources provides several challenges. One of these is introducing intermittent renewable energy sources, i.e., solar and wind, to the electricity grid while maintaining stable operation. Renewable chemical energy carriers then become a requirement for temporary energy storage. These can account for load fluctuations or transport energy from regions with high to low renewable energy potential. An interesting category of carbon-free energy carriers, that have received increased attention lately, is metals, e.g., [ 1 , 2 , 3 ] . Metal powders can be used to operate a carbon-free energy cycle. First, the metal powder is combusted, releasing energy, and the combustion product are metal oxides. The metal oxides are captured and reduced with green hydrogen (energy storage), returning the metal oxide to metal and closing the cycle. A promising candidate for constructing such a metal-based energy cycle is iron, which is the focus of this work. Compared to other solid fuels, like coal and biomass, research on iron powders for energy storage and supply is still in its infancy. A lot of pioneering work on iron (and other metal fuels for heat and power supply) has been performed at McGill University in Canada, e.g., [ 4 ] . Energy from iron powder can be rapidly oxidized with air in iron dust flames, releasing the stored energy as heat. However, the operation of iron dust flames is non-trivial since several fundamental differences exist between such (non-volatile) solid fuel flames, and conventional gas flames as pointed out by Goroshin et al. [ 4 ] . Among others, the particle temperature can significantly exceed the gas temperature; fuel and oxidizer concentration relate non-trivially as the fuel does not diffuse, and the flame structure becomes distorted due to the particles‘ thermal inertia. Consequently, the burning velocity and temperature respond differently to the fuel-oxygen equivalence ratio than gas flames [ 5 ] . In the past, iron dust flames have been established experimentally in various configurations, such as counterflow flames [ 6 ] , methane/iron hybrid Bunsen flames [ 7 ] , conventional Bunsen-type flames [ 8 ] , spherical flames [ 9 , 10 ] , methane/iron hybrid flat flame burners [ 11 ] , lifted flames [ 12 ] , and top-fired swirled tornado flames [ 13 ] . Even though several stable flames could be operated in the lab, important flame characteristics, such as the flame stabilization mechanisms, are not thoroughly understood and warrant further research. Parallel to the experimental work, significant progress has been made in the model development of metal dust flames. In one of the earlier numerical works by Soo et al. [ 14 ] , they developed a single-particle model and used this to investigate the transient propagation of a hypothetical dust flame. Later, Hazenberg and van Oijen [ 5 ] extended this model with temperature-dependent transport properties and based the particle model properties on that of iron; they than simulated the first 1D iron flames. Several model variations and studies have been performed with this model as a base. Van Gool et al. [ 15 ] extended the model with temperature-dependent thermodynamic properties for iron and considered oxidation to the highest oxidation state ( \text{Fe}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}}\text{O}{\vphantom{\text{X}}}_{\smash[t]{\text{3}}} ). The utilized thermodynamic table was later improved by Mich et al. [ 16 ] by accounting for non-ideal mixing enthalpy and entropy to capture the “reactive cooling” after peak particle temperature. The original model and variants (inspired by [ 17 , 15 , 18 , 19 ] ) were used by Ravi et al. [ 20 , 21 ] and Mich et al. [ 22 ] to investigate the influence of particle size distributions. The first to explore 3D effects on a 1D flame was van Gool et al. [ 23 ] , who studied the impact of flame stretch using a 1D approximation to the counterflow setup. Until now, only two numerical studies of laminar flame propagation have been performed in 3D: 1) Wen et al. [ 24 ] they studied the counterflow setup of McRae et al. [ 6 ] , and 2) Ramaekers et al. [ 25 ] they studied the impact of radiation on the tube flames of Tang et al. [ 26 , 27 ] . Rather differently, but not less useful, Vance et al. [ 28 , 29 , 30 ] have studied flame propagation using boundary layer resolved simulations, providing useful insights into the discrete regime [ 31 , 32 , 33 ] . In this article, we will present 3D numerical simulations of the experimental Bunsen flame of Fedoryk et al. [ 8 ] . To this end, a point particle method in OpenFOAM® is utilized. The particle model is that by Hazenberg and van Oijen [ 5 ] , with the thermodynamic properties and Stefan flow of Mich et al. [ 22 ] . A particle size distribution is utilized in the experiments, which we approximate with a log-hyperbolic distribution function [ 34 ] . Our goal is twofold: First and foremost, we would like to validate our numerical model against experiments, and second, we would like to identify if the Bunsen flame can, locally, be approximated as a 1D flame. This article is split into several sections: first, we provide a short description of the experimental setup; second, the numerical model is introduced; third, the results are presented; and finally, we give an outlook and conclusions. 2 Experimental Bunsen flame setup The experimental setup utilizes iron powder of purity greater than 99.5% (PMCtec GmbH, type YTF-HY2) as a fuel, which is contained in a cylindrical tube and shifted upwards by a piston connected to a stepper motor as shown in Fig. 1 . The dispersion of the iron powder occurs in a so-called air-knife seeder in a small gap of approximately 30\text{\,}\mathrm{\SIUnitSymbolMicro m} height with high gas velocity. Mass flow controllers control the air mass flow in the seeder/burner and co-flow stream. Figure 1: Sketch of the experimental configuration presented by Fedoryk et al. [ 8 ] . Several of the relevant parts of the setup are indicated. The computational domain, which extends 8\text{\,}\mathrm{cm} above the combustion tube and co-flow outlet, is marked in red. A

Read the complete text