Document Type

Article

Publication Date

11-1-2016

Abstract

Modeling the response of interacting particles, droplets, or bubbles to subgrid-scale fluctuations in turbulent flows is a long-standing challenge in multiphase flow simulations using the Reynolds-Averaged Navier-Stokes approach. The problem also arises for large-eddy simulation for sufficiently small values of the Kolmogorov-scale particle Stokes number. This paper expands on a recently proposed stochastic vortex structure (SVS) method for modeling of turbulence fluctuations for colliding or otherwise interacting particles. An accelerated version of the SVS method was developed using the fast multipole expansion and local Taylor expansion approach, which reduces computation speed by two orders of magnitude compared to the original SVS method. Detailed comparisons are presented showing close agreement of the energy spectrum and probability density functions of various fields between the SVS computational model, direct numerical simulation (DNS) results, and various theoretical and experimental results found in the literature. Results of the SVS method for particle collision rate and related measures of particle interaction exhibit excellent agreement with DNS predictions for homogeneous turbulent flows. The SVS method was also used with adhesive particles to simulate formation of particle agglomerates with different values of the particle Stokes and adhesion numbers, and various measures of the agglomerate structure are compared to the DNS results.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Rights Information

© 2016 Author(s).

DOI

10.1063/1.4966684

Link to Article at Publisher Website

Share

COinS