On predicting particle-laden turbulent flows

The paper provides an overview of the challenges and progress associated with the task of numerically predicting particle-laden turbulent flows. The review covers the mathematical methods based on turbulence closure models as well as direct numerical simulation (DNS). In addition, the statistical (pdf) approach in deriving the dispersed-phase transport equations is discussed. The review is restricted to incompressible, isothermal flows without phase change or particle-particle collision. Suggestions are made for improving closure modelling of some important correlations.Lecture presented at a workshop on turbulence in particulate multiphase flow, Fluid Dynamics Laboratory, Battelle Pacific Northwest Laboratory, Richland, WA, March 22–23, 1993.     

Dispersion of particles in wall-bounded particle-laden turbulent flows with high wall permeability

A series of numerical simulations were performed to investigate the distribution and deposition properties of particles in turbulent flows bounded by permeable walls using the Large Eddy Simulation (LES) with a Lagrangian trajectory approach. The wall permeation speeds were taken from 10⁻⁴ to 10⁻² of the bulk velocity. The directions of the permeation speed were the same at both walls, and they were inward on one wall but outward on the other wall to reserve the fluid mass. Particles with Stokes number (respecting viscous time scale) around 0.1, 1 and 10 were released in the fully developed turbulent channel flow. The particle–particle interaction and the retroaction from particles to the fluid were neglected. The fluid-phase turbulence statistical properties and particle's transport characteristics by vortexes were then analyzed in details. If the wall permeation exists, the turbulence intensities will be depressed close to the outward permeable wall but increased near the inward permeable wall. Not influenced by the wall permeation, the suspended particles with St⁺ ∼O(1) tend to accumulate in the less vortical zones away from the wall, while those particles in the flow regions near the outward permeable wall will distribute disregarding of the vorticity. The turbulence structures near the outward permeable wall are found to exert promotional effects on the particle deposition rate, but such effects are different for particles with various Stokes number. A distribution tendency of streamwise streaks for the deposited particles is also found on the wall imposed by the high outward permeation speed and the clustering deposition pattern is more obvious with increasing particle Stokes number.     

Spectral-based simulations of particle-laden turbulent flows

In this paper, we discuss the application of spectral-based methods to simulation of particle-laden turbulent flows. The primary focus of the article is on the past and ongoing works by the authors. The particles are tracked in Lagrangian framework, while direct numerical simulation (DNS) or large-eddy simulation (LES) is used to describe the carrier-phase flow field. Two different spectral methods are considered, namely Fourier pseudo-spectral method and Chebyshev multidomain spectral method. The pseudo-spectral method is used for the simulation of homogeneous turbulence. DNS of both incompressible and compressible flows with one- and two-way couplings are reported. For LES of particle-laden flows, two new models, developed by the authors, account for the effect of sub-grid fluctuations on the dispersed phase. The Chebyshev multidomain method is employed for the works on inhomogeneous flows. A number of canonical flows are discussed, including flow past a square cylinder, channel flow and flow over backward-facing step. Ongoing research on particle-laden LES of inhomogeneous flows is briefly reported.     

Two-phase micro- and macro-time scales in particle-laden turbulent channel flows

The micro-and macro-time scales in two-phaseturbulent channel flows are investigated using the direct numerical simulation and the Lagrangian particle trajectorymethods for the fluid-and the particle-phases,respectively.Lagrangian and Eulerian time scales of both phases are calculated using velocity correlation functions.Due to flowanisotropy,micro-time scales are not the same with the theoretical estimations in large Reynolds number(isotropic) turbulence.Lagrangian macro-time scales of particle-phaseand of fluid-phase seen by particles are both dependent onparticle Stokes number.The fluid-phase Lagrangian integral time scales increase with distance from the wall,longerthan those time scales seen by particles.The Eulerian integral macro-time scales increase in near-wall regions but decrease in out-layer regions.The moving Eulerian time scalesare also investigated and compared with Lagrangian integraltime scales,and in good agreement with previous measurements and numerical predictions.For the fluid particles themicro Eulerian time scales are longer than the Lagrangianones in the near wall regions,while away from the walls themicro Lagrangian time scales are longer.The Lagrangianintegral time scales are longer than the Eulerian ones.Theresults are useful for further understanding two-phase flowphysics and especially for constructing accurate predictionmodels of inertial particle dispersion.     

Measurement of turbulent spectra of particle-laden flows in a horizontal channel

Frequency spectra of air turbulence of particle-laden flows were investigated by use of a laser-Doppler velocimeter to discover the eddy-length scales that are influenced by the transported particles. The influence of glass and steel particles of 100–1,000 μm diameter was measured in a horizontal channel and a horizontal pipe for the streamwise and transverse components of the velocity vector. Particles that were small compared with the integral length scale of the particle-laden flow decrease the turbulent power density of the greatest eddies in varying degrees, depending on mass loading and distance from the wall. All fractions create turbulence in their wakes, the size of which depends on loading and slip velocity. These results support the hypothesis that the particles consume energy by following the large eddies that are much greater than the particle diameters, and in so doing, turbulence is created by this energy. Received: 28 September 2000/Accepted: 9 April 2001     

Modelling turbulent and dispersion heat fluxes in turbulent porous medium flows using the resolved LES data

Resolved large eddy simulations (LESs) of turbulent conjugate heat transfer in porous media are performed by the lattice Boltzmann method (LBM) for modelling turbulent and dispersion heat flux terms of the double-averaged energy equation. The considered porous structures are square rod arrays, staggered cube arrays and body centred cubic foam. In the LBM, the double-distribution function method which solves the distribution functions for the velocity and the internal energy is used. For the velocity and thermal fields, the D3Q27 multiple-relaxation-time method and the regularized D3Q19 single-relaxation-time method are applied, respectively. A priori tests using the LES data suggest that the trends of the sum of the dispersion and volume-averaged turbulent heat fluxes can be well captured by the second order gradient diffusion model.     

LES prediction of space-time correlations in turbulent shear flows

We compare the space-time correlations calculated from direct numerical simulation(DNS) and large-eddy simulation(LES) of turbulent channel flows.It is found from the comparisons that the LES with an eddy-viscosity subgrid scale(SGS) model over-predicts the space-time correlations than the DNS.The overpredictions are further quantified by the integral scales of directional correlations and convection velocities.A physical argument for the overprediction is provided that the eddy-viscosity SGS model alone does not includes the backscatter effects although it correctly represents the energy dissipations of SGS motions.This argument is confirmed by the recently developed elliptic model for space-time correlations in turbulent shear flows.It suggests that enstrophy is crucial to the LES prediction of spacetime correlations.The random forcing models and stochastic SGS models are proposed to overcome the overpredictions on space-time correlations.     

Accelerating fully resolved simulation of particle-laden flows on heterogeneous computer architectures

An efficient computing framework, namely PFlows, for fully resolved-direct numerical simulations of particle-laden flows was accelerated on NVIDIA General Processing Units (GPUs) and GPU-like accelerator (DCU) cards. The framework is featured as coupling the lattice Boltzmann method for fluid flow with the immersed boundary method for fluid-particle interaction, and the discrete element method for particle collision, using two fixed Eulerian meshes and one moved Lagrangian point mesh, respectively. All the parts are accelerated by a fine-grained parallelism technique using CUDA on GPUs, and further using HIP on DCU cards, i.e., the calculation on each fluid grid, each immersed boundary point, each particle motion, and each pair-particle collision is responsible by one computer thread, respectively. Coalesced memory accesses to LBM distribution functions with the data layout of Structure of Arrays are used to maximize utilization of hardware bandwidth. Parallel reduction with shared memory for data of immersed boundary points is adopted for the sake of reducing access to global memory when integrate particle hydrodynamic force. MPI computing is further used for computing on heterogeneous architectures with multiple CPUs-GPUs/DCUs. The communications between adjacent processors are hidden by overlapping with calculations. Two benchmark cases were conducted for code validation, including a pure fluid flow and a particle-laden flow. The performances on a single accelerator show that a GPU V100 can achieve 7.1–11.1 times speed up, while a single DCU can achieve 5.6–8.8 times speed up compared to a single Xeon CPU chip (32 cores). The performances on multi-accelerators show that parallel efficiency is 0.5–0.8 for weak scaling and 0.68–0.9 for strong scaling on up to 64 DCU cards even for the dense flow (φ = 20%). The peak performance reaches 179 giga lattice updates per second (GLUPS) on 256 DCU cards by using 1 billion grids and 1 million particles. At last, a large-scale simulation of a gas-solid flow with 1.6 billion grids and 1.6 million particles was conducted using only 32 DCU cards. This simulation shows that the present framework is prospective for simulations of large-scale particle-laden flows in the upcoming exascale computing era.     

Direct numerical simulation of wall turbulent flows with microbubbles

The marker‐density‐function (MDF) method has been developed to conduct direct numerical simulation (DNS) for bubbly flows. The method is applied to turbulent bubbly channel flows to elucidate the interaction between bubbles and wall turbulence. The simulation is designed to clarify the structure of the turbulent boundary layer containing microbubbles and the mechanism of frictional drag reduction. It is deduced from the numerical tests that the interaction between bubbles and wall turbulence depends on the Weber and Froude numbers. The reduction of the frictional resistance on the wall is attained and its mechanism is explained from the modulation of the three‐dimensional structure of the turbulent flow. Copyright © 2001 John Wiley & Sons, Ltd.     

Direct numerical simulation of a particle-laden low Reynolds number turbulent round jet

Direct numerical simulation method is used for the investigating of particle-laden turbulent flows in a spatially evolution of low Reynolds number axisymmetric jet, and the Eulerian–Lagrangian point-particle approach is employed in the simulation. The simulation uses an explicit coupling scheme between particles and the fluid, which considers two-way coupling between the particle and the fluid. The DNS results are compared well with experimental data with equal Reynolds number (R_e = 1700). Our objects are: (i) to investigate the correlation between the particle number density and the fluctuating of fluid streamwise velocity; (ii) to examine whether the three-dimensional vortex structures in the particle-laden jet are the same as that in the free-air jet and how the particles modulate the thee-dimensional vortex structures and turbulence properties with different Stokes number particles; (iii) to discover the particle circumferential dispersion with different Stokes number particles. Our findings: (i) all the particles, regardless of their particle size, tend to preferentially accumulate in the region with large-than-mean fluid streamwise velocity; (ii) the small Stokes number particles take an important part in the modulation of three-dimensional vortex structures, but for the intermediate and larger sized particles, this modulation effect seems not so apparent; (iii) the particle circumferential dispersion is more effective for the smaller and intermediate sized particles, especially for the intermediate sized particles.     

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

The dynamic and thermal performance of particle-laden turbulent flow is investigated via direction numerical simulation combined with the Lagrangian point-particle tracking under the condition of two-way coupling, with a focus on the contributions of particle feedback effect to momentum and heat transfer of turbulence. We take into account the effects of particles on flow drag and Nusselt number and explore the possibility of drag reduction in con-junction with heat transfer enhancement in particle-laden turbulent flows.The effects of particles on momentum and heat transfer are analyzed,and the possibility of drag reduc-tion in conjunction with heat transfer enhancement for the prototypical case of particle-laden turbulent channel flows is addressed.We present results of turbulence modification and heat transfer in turbulent particle-laden channel flow,which shows the heat transfer reduction when large inertial parti-cles with low specific heat capacity are added to the flow. However,we also found an enhancement of the heat transfer and a small reduction of the flow drag when particles with high specific heat capacity are involved.The present results show that particles,which are active agents,interact not only with the velocity field,but also the temperature field and can cause a dissimilarity in momentum and heat transport.This demonstrates that the possibility to increase heat transfer and suppress friction drag can be achieved with addition of par-ticles with different thermal properties.     

Direct numerical simulation of compressible turbulent flows

This paper reviews the authors＇ recent studies on compressible turbulence by using direct numerical simulation （DNS）,including DNS of isotropic（decaying） turbulence, turbulent mixing-layer,turbulent boundary-layer and shock/boundary-layer interaction.Turbulence statistics, compressibility effects,turbulent kinetic energy budget and coherent structures are studied based on the DNS data.The mechanism of sound source in turbulent flows is also analyzed. It shows that DNS is a powerful tool for the mechanistic study of compressible turbulence.     

Wall-resolved large-eddy simulation of turbulent channel flows with rough walls

Turbulent flows over rough surfaces widely exist in nature and industry. Investigating its mechanism is of theoretical and practical significance. In this work we simulate the turbulent channel flow with rough walls using large-eddy simulation with rough elements resolved using the curvilinear immersed boundary method and compare the results obtained in this work with those in the paper by Yuan and Piomelli( J. Fluid Mech., vol. 760, pp. R1, 2014), where the volume of fluid method was employed for modeling rough elements. The mean streamwise velocity profiles predicted by the two methods agree well with each other. Differences in Reynolds stresses and dispersive stresses are observed, which are attributed to the different approaches in dealing with the complex geometry of the rough surface.     

非球形颗粒两相流的数值模拟研究进展

非球形颗粒两相流是多相流的重要研究方向之一, 常见于自然界及工业生产过程中. 不同于球形颗粒, 由于非球形颗粒形状的各向异性, 除了颗粒平动行为, 还需要考虑颗粒的转动与取向行为, 颗粒的取向与转动行为会影响颗粒所受的力和力矩. 为了准确模拟非球形颗粒的运动行为, 目前非球形颗粒两相流的数值模拟研究主要基于欧拉?拉格朗日的求解框架展开, 常见的非球形颗粒两相流数值模拟方法主要包括点颗粒法与全分辨颗粒法. 本文将对这两类方法进行介绍, 同时会全面介绍非球形颗粒两相流研究的基础理论模型, 并系统总结非球形颗粒在简单基本流和复杂湍流中的研究进展, 包括对于非球形颗粒在湍流中的取向与转动行为机理, 以及颗粒对湍流减阻调制作用的研究. 最后, 本文提出了非球形颗粒两相流研究存在的问题及未来研究方向.      

环形通道内湍流旋流流动的数值模拟

本文应用一种考虑湍流－旋流相互作用及湍流脉动各向异性的新的代数Ｒｅｙｎｏｌｄｓ应力模型,对环形通道内的湍流旋流流动进行了数值模拟,研究了改主为旋流流数,进口轴向速度及半径比等参数对环形通道内湍流流动的影响,以及对强化环形通道内传热的作用。     

A sharp interface Cartesian grid method for viscous simulation of shocked particle-laden flows

A Cartesian grid-based sharp interface method is presented for viscous simulations of shocked particle-laden flows. The moving solid–fluid interfaces are represented using level sets. A moving least-squares reconstruction is developed to apply the no-slip boundary condition at solid–fluid interfaces and to supply viscous stresses to the fluid. The algorithms developed in this paper are benchmarked against similarity solutions for the boundary layer over a fixed flat plate and against numerical solutions for moving interface problems such as shock-induced lift-off of a cylinder in a channel. The framework is extended to 3D and applied to calculate low Reynolds number steady supersonic flow over a sphere. Viscous simulation of the interaction of a particle cloud with an incident planar shock is demonstrated; the average drag on the particles and the vorticity field in the cloud are compared to the inviscid case to elucidate the effects of viscosity on momentum transfer between the particle and fluid phases. The methods developed will be useful for obtaining accurate momentum and heat transfer closure models for macro-scale shocked particulate flow applications such as blast waves and dust explosions.     

Stochastic simulations of particle-laden isotropic turbulent flow

Numerical simulations are performed of dispersion and polydispersity of particles in isotropic incompressible turbulence. The mass loading of the particles is assumed to be small; thus the effects of particles on turbulence is neglected (one-way coupling). A stochastic model is employed to simulate the carrier phase. The results of the simulations are compared with direct numerical simulation (DNS) data and theoretical results. The stochastic model predicts most of the trends as portrayed by DNS and theory. However, the continuity effect associated with the crossing trajectories effect is not captured. Also, the peaking in the variation of the particle asymptotic diffusivity coefficient with the particle time constant is not observed. For evaporating particles, the stochastic model predicts thinner probability density functions (pdfs) for the particle diameter as compared with DNS generated pdfs. The model is implemented to investigate the effects of gravity on evaporation. It is shown that the depletion rate increases with increase of the drift velocity at short and intermediate times, but an opposite trend is observed at long times. The standard deviation and skewness of the particle diameter indicate peak values in their variations with the drift velocity. Dispersion of evaporating particles decreases with respect to that of non-evaporating particles at small drift velocities; an opposite trend is observed at large drift velocities. The effects of the initial evaporation rate and the particle Schmidt number on the evaporation in the gravity environment are also studied.     

Efficient simulation of nonlinear viscoelastic fluid flows

This paper presents a new and efficient method for computing the flow of a non-Newtonian fluid. The approach is based on two independent concepts:Time-dependent simulation of viscoelastic flow: A new decoupled algorithm, presented in P. Saramito, Simulation numérique d'ecoulements de fluides viscoélastiques par éléments finis incompressibles et une méthode de directions alternées; applications, Thèse de l'Institut National Polytechnique de Grenoble, 1990 and P. Saramito, Numerical simulation of viscoelastic fluid flows using incompressible finite element method and a θ-method, Math. Modelling Num. Anal., 35 (1994) 1–35, enables us to split the major difficulties of this problem, and to solve it more efficiently. Moreover, this scheme is of order two in time, and can be used to obtain stationary flows in an efficient way.Conservative finite element method: this method combines the incompressible Raviart Thomas element, the discontinuous Lesaint-Raviart element, and a finite volume element method. It satisfies exactly the mass conservation law, and leads to an optimal size for the nonlinear system in terms of the total degree of freedom versus the mesh size.We apply our numerical procedure to the Phan-Thien-Tanner model with a classical benchmark: the four to one abrupt contraction. The numerical solutions exhibit good behavior, especially near the singularity, in the vicinity of the re-entrant corner. The numerical experiments present the main features of such flows: vortex development and overshooting of the velocity profile along the axis of symmetry in the entry region.