Large-Eddy simulation of iron oxide nanoparticle synthesis in turbulent spray flames using a Lagrangian FDF approach with different mixing models

The formation of iron oxide nanoparticles within a turbulent spray flame was investigated using Large-Eddy Simulation (LES) coupled with a transported Lagrangian Filtered Density Function (FDF) method. In the SpraySyn system, iron nitrate is dissolved in a solvent mixture of ethanol and 2-ethylhexanoic acid, then injected and atomized through an air-blast nozzle surrounded by a pilot flame. For the first time, a transported Lagrangian FDF method embedded within LES has been integrated with a sectional model to study iron oxide nanoparticle synthesis in spray flames. This study compares the diameter of average volume and the particle size distribution obtained from two classical mixing models – the Interaction by Exchange with the Mean (IEM) and the modified Curl model – with different mixing constants against the traditional Eulerian Simulations and experiments. The Lagrangian FDF simulations, utilizing an optimized mixing constant, produced significantly broader nanoparticle size distributions compared to traditional Eulerian simulations that overlook sub-filter-scale effects on particle formation. These results demonstrate that the current approach can capture subgrid-scale heterogeneities in LES, enabling more accurate modeling of localized particle collisions and growth dynamics.