Search

The generation of nanostructured particles in high-temperature flames is important both for the control of emissions from combustion devices and for the synthesis of high-value chemicals for a variety of applications. The physiochemical processes that lead to the production of fine particles in turbulent flames are highly sensitive to the flow physics and, in particular, the history of thermochemical compositions and turbulent features they encounter. Consequently, it is possible to change the characteristic size, structure, composition, and yield of the fine particles by altering the flow configuration. This review describes the complex multiscale interactions among turbulent fluid flow, gas-phase chemical reactions, and solid-phase particle evolution. The focus is on modeling the generation of soot particles, an unwanted pollutant from automobile and aircraft engines, as well as metal oxides, a class of high-value chemicals sought for specialized applications, including emissions control. Issues arising due to the numerical methods used to approximate the particle number density function, the modeling of turbulence-chemistry interactions, and model validation are also discussed. [ABSTRACT FROM AUTHOR]

Multiphase flows occurring in nature and in technological applications are often turbulent. The large range of length scales and timescales in turbulent multiphase flows makes direct numerical simulation of the microscale governing equations intractable for many applications. In this article we review a systematic approach for developing large-eddy-simulation (LES) tools for dispersed multiphase flows starting from the microscale model. A key intermediate step is the mesoscopic model for the dispersed phase, formulated in terms of kinetic equations, that contains the physical models for the flow. Owing to the phase-space variables, direct solution of the mesoscopic model is usually intractable, and additional mathematical approximations are introduced to arrive at a macroscopic model. We show that self-conditioned LES models can be derived for both the mesoscopic and macroscopic models, but the former is preferred to ensure consistency and physical accuracy. The principal difficulties and open challenges in closing the LES model equations are highlighted. [ABSTRACT FROM AUTHOR]