CFD modelling of nanoparticles dispersion in a dust explosion apparatus
Résumé
A computational fluid dynamics simulation (CFD) has been developed in order to describe the dispersion of a confined gas-solid flow in a typical test designed for the determination of dusts cloud flammability (modified Hartmann apparatus). For this purpose, an analysis of the biphasic flow development inside such an explosion tube has been performed. The equipment considered in this study consists of a vertical tube in which a flammable dust is dispersed by the injection of air at high pressure to be ignited after a time period in which the mixture can be considered as homogeneous (commonly 60 ms). This analysis has been focused on the assessment of flow conditions pertaining to the agglomeration and fragmentation of the disperse phase. Previous studies performed by Bouillard et al (2010) established significant variations in the combustion process for solid materials notably with a submicronic size distribution. For this reason, the present study has been intended to determine the zones where turbulence causes variations on the particles size distribution (agglomeration/fragmentation dynamics) and therefore on their mechanism of combustion. In this order, a CFD simulation has been developed by using an Euler- Lagrange approach. Based on previous studies (Bouillard et al., 2010) and on industrial considerations, aluminium powders were chosen for this case study. Different Reynolds Averaged Navier-Stokes (RANS) turbulence models were compared to analyze flow variables associated to computational cost and description of the dispersion phenomena. A transonic flow condition is reached at zones near the injection. The model was adjusted by inclusion of specific characteristics that described the mutual interactions between the two phases and other effects such as Brownian motion for non-turbulent regions. After adjusting the biphasic CFD simulation to the design specifications of the apparatus, this analysis has led to the identification of a unsteady behaviour with high segregation levels during the first 30 ms of dispersion. This fact is associated to the high number of collisions among the aggregates and the variations in their size distribution during a typical characterization test. Nevertheless, it has been evidenced that the homogeneity assumption in the mixture can be accepted because the minimum explosive concentration (0.03-0.14 kg/m3) is accomplished at ignition time in the regions where the electrodes are located. This fact has confirmed the importance of turbulence on solids segregation which causes fluctuations on determined parameters according to ignition delay. Due to the strong influence of fragmentation phenomenon on sub-micronic particles, this study was mainly focused on nanoparticles suspensions. From an industrial point of view, this choice is relevant as it will allow the quantification of nanopowders dispersion risk, both for toxicological and explosion considerations.