On the role of oxide layer thickness in iron particle cloud ignition: A carrier-phase DNS study

Carrier-phase direct numerical simulations (CP-DNS) of a three-dimensional turbulent mixing layer are performed to investigate how the ignition characteristics of iron particles affect the overall combustion. The particle heat-up rate and the oxide layer thickness are shown to be key factors controlling ignition success, highlighting the importance of thermal coupling between particles and the surrounding hot gases. Slow particle heat-up does not initiate ignition immediately, but causes a build-up of oxide layer thickness, increasing the temperature required for ignition and leading to particle ignition failure. Thereby, the ignition temperature of particles was found to increase from nearly 1200 K to just above 1600 K, far above the temperature of the hot gas. This increase in ignition temperature to values higher than the hot gas temperature hinders the further ignition of particles and the formation of iron flames. A particle-driven gas-phase flame is observed only under preheated condition that promotes particle ignition and the formation of a diffusion-like flame front. These findings provide new insights into the processes that determine flame stabilization in iron particle combustion and provide a reference for the further development of iron combustion modeling.
... mehr
Novelty and significance statement: The formation of a self-sustained iron flame fundamentally relies on the successful ignition of individual particles. The ignition behavior of iron particles is governed by solid-state oxidation kinetics and can be influenced by the oxide layer thickness (Mi et al., 2022). Although significant progress has been made in studying single-particle ignition, limited data exist on how the ignition dynamics are affected in a turbulent flow. We present a comprehensive study on iron particle ignition dynamics in turbulent flows affected by oxide layer thickness by means of three-dimensional direct numerical simulations of a turbulent mixing layer. Additionally, the high-fidelity DNS datasets generated here provide valuable reference data for future model development.