A comparison of mechanistic models for the combustion of iron microparticles and their application to polydisperse iron-air suspensions

Abstract

Metals can serve as carbon-free energy carriers, e.g., in innovative metal-metal oxide cycles as proposed by Bergthorson (Prog. Energy Combust. Sci., 2018). For this purpose, iron powder is a suitable candidate since it can be oxidized with air, exhibits a high energy density, is non-toxic and abundant. Nevertheless, the combustion of iron powder in air is challenging especially with respect to flame stabilization which depends on the particle size distribution and the morphology of the iron microparticles among other factors. Models for the prediction of reaction front speed in iron-air suspensions can contribute to overcoming this challenge. To this end, three different models for iron particle oxidation are integrated into a laminar flame solver for simulating such reaction fronts. The scientific objective of this work is to elucidate the influence of polydispersity of the iron particles on the reaction front speed, which is still not satisfactorily understood. In a systematic approach, cases with successively increasing complexity are considered: The investigation starts with single particle combustion and then proceeds with iron-air suspensions prescribing binary particle size distributions (PSDs), parametrized generic PSDs, and an exemplary PSD measured for a real iron powder sample. The simulations show that, dependent on the kinetic particle model, the particles’ thermal inertia, and the PSD, particles undergo thermochemical conversion in a sequential manner according to their size and every particle fraction exhibits an individual combustion environment (surrounding gas phase temperature and oxygen concentration). The local environment can be leaner or richer than the overall iron-to-air ratio would suggest and can be very different from single particle experiments. The contribution of individual particle fractions to the overall reaction front speed depends on its ranking within the PSD. The study further demonstrates, that although the three particle models show good agreement for single particle combustion, they lead to very different reaction front speeds. This is due to the different ignition behavior predicted by the particle models, which is shown to strongly influence the reaction front characteristics. Overall, the present work illustrates the complex relationship between characteristics of single particle ignition and combustion, polydispersity, and properties of reaction fronts in iron-air suspensions.