A numerical study on discrete combustion of polydisperse magnesium aero-suspensions

In this numerical study, discrete combustion of polydisperse magnesium dust clouds was investigated. A numerical model accounting for the effects of ignition, thermal conduction, and radiation was formulated to simulate the spatiotemporal distribution of temperature. Three distribution models, i.e., Dagum, log-normal, and Beta prime, were used to describe the magnesium particle -size polydispersity. The numerical model was first validated by comparison against experimental data on discrete combustion of both mono-sized and polydisperse magnesium aero-suspensions. Subsequently, the flame propagation characteristics of mono-sized and log-normally polydisperse cases at two different mean magnesium particle sizes were compared. The comparison shows that polydisperse magnesium dust clouds have higher flame propagation speeds than their mono-sized counterparts. Finally, the differences among the polydisperse cases with different size distributions were compared, revealing that magnesium powders with a higher percentage of small particles give rise to higher flame propagation speeds. Furthermore, results show that in comparison with the Dagum and Beta prime distributions, the log-normal distribution results in a lower flame propagation speed and a higher minimum ignition energy. As either the particle size decreases or the dust-cloud concentration increases, the flame propagation speed increases and the minimum ignition energy decreases. (C) 2019 Elsevier Ltd. All rights reserved.