On the use of EBRSM turbulence model to improve continuous random walk model prediction in inhomogeneous turbulent flows

Abstract

In this paper the suitability of the normalized Langevin stochastic equation, coupled with the RMS fluctuations velocity predicted by EBRSM (Elliptic Blending Reynolds Stress Model) second order turbulence model, in inhomogeneous turbulent flows was studied. The gas-particle flow was numerically investigated by using OpenFOAM V9 CFD toolkit for two configurations. First, in a two-dimensional duct flow in which $2\times 10^5$ randomly distributed particles, with diameters ranging from 10 nm to 23 $\mu$m, were introduced in the channel and followed by solving the particle equation of motion including the drag and Brownian forces under the one-way coupling assumption. The performance of the Continuous Random Walk (CRW) model with the EBRSM prediction of the Reynolds stress tensor for predicting the behavior of fluid-tracer and inertial particles in an inhomogeneous turbulent flow was examined as well as their deposition velocities. In addition, the particle laden flow inside an obstructed three dimensional channel was studied to evaluate the characteristics of particle deposition according to the ratio of rough-element spacing to its height e/H. Thus, the deposition rate on the different surfaces and the particles deposition profiles on the cavities between the rough-elements and on windward rib surfaces were evaluated for e/H = 4, 7 and 10. By exploring the concentration profiles and deposition velocities of particles, it was concluded that the Normalized-CRW model including EBRSM flow prediction leads to accurate and satisfying results, compared to the use of correlations for RMS fluctuations velocity values, and can be applied in more complex flows (such as in industrial configurations).