High-fidelity simulation and modeling of turbulent dispersed multiphase flows

Particulate-laden flow refers to a two-phase fluid flow in which one phase is continuous (referred to as the carrier phase) and the other phase consists of fine, insoluble, and typically dilute particles (known as the dispersed or particulate phase). An example of particulate flow is the transport of fine aerosol particles in air, where aerosols represent the dispersed phase and air serves as the carrier phase. Two-phase flow modeling has a wide range of applications in both engineering and scientific disciplines, including atmospheric pollution dispersion, fluidization in combustion processes, and aerosol deposition in pharmaceutical sprays. Accurately predicting the behavior of these flows is essential for optimizing engineering applications and understanding environmental phenomena. This understanding can be achieved through experimental studies and numerical simulations. However, numerical simulations are often hindered by the complexity of turbulent interactions between the carrier phase and the dispersed particles. In this context, the present paper focuses on the development of a novel numerical approach that combines a semi-stochastic model based on the Langevin equation with drift correction and the Wall Modelling Large Eddy Simulation (WM-LES) framework for simulating particle-laden turbulent channel flows in one-way coupling conditions. A comprehensive description of the theoretical framework and its numerical implementation is provided. The paper presents various statistical velocity properties of both the carrier and dispersed phases. The results demonstrate that the proposed method accurately reproduces particle dynamics across a range of Reynolds and Stokes numbers, in comparison with reference data obtained from direct numerical simulations. The model effectively captures key aspects of particle behavior, including velocity fluctuations, the accuracy of particle velocities near the wall, and the expected logarithmic distribution of axial velocities in the logarithmic region. Additionally, the method successfully replicates particle concentration trends both near the fluid wall and within the bulk flow region. The author believes that these findings underscore the potential of the semi-stochastic WM-LES framework to enhance the accuracy and efficiency of particle-laden flow simulations, offering valuable insights into turbulent particle dynamics.