Numerical investigation of momentum exchange between particles and coherent structures in low Re turbulent channel flow

Abstract

The interaction between particles and coherent structures is studied by using discrete particle simulation combined with direct numerical simulation of gaseous flow in a vertical channel. A conditional sampling scheme is used to examine the modifications of the near-wall quasistreamwise vortices by the momentum exchange between the phases. The particle effect on the fluid flow is modeled by a point-force approximation. The particle diameters are smaller than both the smallest flow length scales and the computational grid spacing. Results are obtained for particle ensembles with four response times ranging from 10 to 200 wall units in numerical simulations with and without gravitational settling in the streamwise direction and interparticle collisions. It is found that the size of the quasistreamwise vortices is increased up to 25% in the presence of particles. The increase is larger for the smallest inertia particles studied, which is partly due to their locally nonuniform spatial distribution. The underlying organized fluid motions induced by the structures are substantially attenuated due to the momentum coupling of the phases. A reduction of 5%–55% is observed in the coherent fluid velocities and vorticities. The size of the coherent structures is additionally augmented by streamwise gravitational settling and interparticle collisions, accompanied by more obvious modifications in the surrounding fluid flow. The latter findings are explained by the stronger direct particle effect in these cases, which is also reflected on the energy redistribution between the fluid velocity components, affecting further the fluid turbulence.

Characteristics of turbulence transport for momentum and heat in particle-laden turbulent vertical channel flows

“…The interaction between the particles and fluid is taken into account, and the effects of particle collisions have been neglected. Figure 1 shows the mean velocity profile and velocity fluctuation intensity in our present DNS are in good agreement with the DNS data of Dristelis and Valchos [14].…”

“…Our DNS results on 128 3 grid points will be compared with the previous datas by Dritselis and Valchos [14] and Lessani and Nakhaei [33] for validating the reliability of our codes. For the DNS study of Dritselis and Valchos [14], dimensions of the channel is 4 π H × 2H × 2 π H in the streamwise, normal, and spanwise directions and it has a 128 3 grid points. The Reynolds number based on channel half-width and the friction velocity is equal to Re τ = 180.…”

“…The computational domain is identical to the dimension of the domain in the study of Dristelis and Valchos [14]. The Reynolds number and Prandtl number are 180 and 0.7, respectively.…”

“…In this section, two cases which have the same parameters with Dritselis and Valchos [14] and Lessani and Nakhaei [33] are performed by DNS, respectively. Our DNS results on 128 3 grid points will be compared with the previous datas by Dritselis and Valchos [14] and Lessani and Nakhaei [33] for validating the reliability of our codes.…”

“…Turbulent momentum transfer in the presence of small particles in a turbulent Couette flow was studied by Richter and Sullivan [13] using direct numerical simulation and point-force model. The modulation of turbulence depends on the particle-fluid density ratio, the volume fraction of particles, the particle Reynolds number, and the particle-turbulence length-scale ratio [14]. The mechanisms of turbulence modulation observed in the presence of inertial particles were reviewed by Balachandar and Eaton [4] and Soldati [15].…”