Stochastic subgrid-scale modeling for particles dispersion and deposition for a filtered DNS channel flow field
Farid Rousta,Bamdad Lessani,Goodarz Ahmadi
DOI: https://doi.org/10.1016/j.ijmultiphaseflow.2024.104801
IF: 4.044
2024-03-21
International Journal of Multiphase Flow
Abstract:The effect of subgrid-scale velocity fluctuations on the deposition and dispersion of solid particles in a filtered DNS turbulent channel flow at Reτ=180 , and for particles with Stokes numbers of St = 1, 2, 5, 24 was studied. First, the turbulent flow field was simulated using the DNS on a high-resolution grid. Then, the filtered DNS flow fields were generated using several low-resolution grids. Particle tracking was performed for the DNS and filtered DNS channel flow fields, and the effects of filtering on particle deposition and concentration were studied. It was shown that filtering significantly decreases the particle deposition velocities at lower Stokes numbers (St = 1, 2, 5). However, at higher Stokes numbers (St = 24), the deposition velocity was less affected. Next, the ability of a stochastic model to generate the correct particle deposition velocities and concentration profiles was studied. A Langevin stochastic equation for modeling the subgrid-scale (SGS) fluid velocity fluctuations as seen by particles was developed and tested. The two input parameters of the stochastic model are the variance of SGS fluctuations in the wall-normal direction and the Lagrangian SGS time-scale. It was seen that the model prediction is sensitive to the SGS time scale used in the Langevin equation. The model was capable of predicting the correct deposition velocities of and concentration profiles of lower inertia particles when the proper time-scale was used. For higher inertia particles and when a larger amount of energy was removed by filtering, the model was unable to predict the deposition velocities.
mechanics