气粒两相平面湍射流拟序结构的大涡模拟

采用Eulerian/Lagrangian方法,对空间发展的气粒两相平面湍射流的非定常流动过程进行了数值模拟。以Re数为13000的平面不可压缩湍射流流动为例,气相场采用大涡模拟（large-eddy simulaiton,LES）技术,直接求解大尺度涡运动的Navier-Stokes方程,小尺度涡采用标准Smagorinsky亚格子模式模拟。为了示踪两相射流中气相的运动,同时球 解了标志物的浓度输运方程。颗粒相的运动用Lagrangian方法直接求解。大涡模拟结果表明,在平面射流的过渡区及充分发展区存在丰富的拟序结构及其相互作用。对于稀疏两相射流,不同Stokes数的颗粒运动规律和浓度分布取决于颗粒惯性和气相拟序结构的共同作用。对于Stokes数小于10的两相射流,颗粒相的瞬时浓度场分布与拟序结构密切相关,研究颗粒相的扩散应当考虑拟序结构的影响。     

湍流扩散火焰的大涡模拟及验证

对丙烷/空气射流扩散火焰进行了大涡模拟,并就截面流向速度无量纲分布、速度脉动无量纲分布、温度无量纲分布以及NOx生成与Sandia实验室测量数据进行了对比分析,得出大涡模拟方法可以较好地模拟湍流扩散火焰的结论.在此基础上对拟序结构与化学反应的相互影响以及雷诺数对各尺度的影响进行了研究,结果表明:在流场的起始阶段,燃烧对拟序结构的形成有促进作用;在发展阶段,燃烧对大涡拟序结构有削弱作用,却能使小涡的生成增加,从而增强了化学反应;大涡拟序结构的存在改变了各参数的分布;随着雷诺数的增大,各尺度横向脉动均变大,预测到的NOx减小,产生的最大值前移.     

重力对水平两相湍射流中颗粒行为的影响

研究了重力对空间发展的水平两相平面湍射流中颗粒运动的影响。气相场用 Euler方法求解 ,通过大涡模拟直接求解大尺度涡运动的 Navier- Stokes方程 ,小尺度涡采用 Smagorinsky亚格子模式模拟。颗粒相的运动采用L agrangian方法直接求解。射流 Re数为 1130 0。模拟发现 ,对于 St 1及 St～ 0 (1)的颗粒 ,其在平面射流下游的瞬时分布对重力的影响不敏感。随着颗粒 St数的增大 ,重力对平面射流场中颗粒行为的影响逐渐明显 ,但其作用效果还明显地与两相入射流滑移系数的大小有着直接联系。在小两相入流滑移系数情况下 ,对于 St～ 0 (10 )的颗粒 ,在重力作用下的沉降过程还受到了湍流拟序结构的作用 ,而重力作用导致的更大 St数颗粒的沉降 ,将引起固相粒子在射流下游的非对称分布 ,但它既不是均匀各向同性湍流中颗粒的梯度扩散结果 ,也未呈现出受到湍流拟序结构影响的特征     

基于大涡模拟的发动机缸内湍流流动及拟序结构

应用大涡模拟方法对发动机缸内湍流流场进行了三维瞬态数值分析.主要从湍流脉动、湍动能和缸内拟序结构演变等方面考察了发动机缸内流场特性.计算结果表明:相比雷诺平均模型,大涡模拟方法可以更真实地反映发动机循环过程中缸内气体流动的细节和规律.利用大涡模拟结合Q准则判别法可以较好地识别缸内大尺度湍流拟序结构;拟序结构对于缸内大尺度动能的产生及湍流的维持具有关键的作用.RANS类模型则不具备充分捕获大尺度拟序结构的能力.湍流脉动与活塞平均运行速度接近于成正比.     

气固两相自由射流颗粒弥散实验研究综述

本文为水平浓淡煤粉燃烧技术研发基础工作之一。以双矩形气固两相平行射流为研究背景,作者考察了气固两相平面和圆自由射流的实验研究。综述表明,大涡拟序结构的存在和初始条件对自由射流颗粒弥散的影响是关键的,利用St数可定性估计这种影响,颗粒尺寸和负荷对气固两相流场的湍流结构具有显著影响,大颗粒尾涡脱落显著地影响湍流结构的事实使数值模拟气固两相流动的努力面临着模拟方法选择与指导工程设计的矛盾要求。因此,数值模拟技术需要评估,这是继续的工作。     

大涡模拟二维气固两相平面射流

首先用大涡模拟对二维平面射流进行了数值模拟，并就断面流向速度分布、速度脉动量分布以及雷诺切应力分布与实验值进行对比分析，得出大涡模拟方法可以较好地模拟二雏平面射流。在此基础上对二维气固两相平面射流中直径为30μm(St=2．5)的颗粒扩散特性进行数值模拟。发现由于大尺度涡团的作用，颗粒分布于大尺度涡团的外缘而不是集中于中线附近。这与过去基于时均雷诺方程的数值模拟结果得到的颗粒主要集中于中线的结论有所不同，但与直接模拟二维平面混合层中颗粒运动扩散情况相吻合。     

基于分离涡方法的定日镜风效应研究

选用RANS湍流模型、大涡模拟与分离涡模型等计算流体动力学(CFD)方法,应用由新的湍流脉动流场产生方法DSRFG(discretizing and synthesizing random flow generation)模拟风场实际的湍流边界条件,对定日镜结构的风压分布及流场进行分析。将数值模拟结果与风洞试验数据进行对比,给出风向角为0°工况下定日镜结构的风压系数分布规律及特点,分析5种模型下定日镜周围的流场分布和涡量分布。结果表明,RANS模型虽然能够模拟出结构周围流场的基本规律,但却无法模拟出风场中的紊流和涡旋的分离扩散情况。大涡模拟及分离涡模型对风场中的紊流和涡旋的分离扩散有较好的模拟效果,在风压系数的分布上也与风洞试验数据拟合较好。在定日镜周围流场中脉动风压系数的分布上,分离涡模型的模拟结果较大涡模拟更为接近风洞试验值,且耗时更少。     

绕流圆柱对气固两相流场颗粒聚并效果的研究

采用大涡模拟耦合群平衡模型的方法模拟绕流圆柱流场中颗粒的扩散特性和聚并效果。分析圆柱尾迹的分布,研究流动时间、颗粒St数和颗粒的体积分数对细颗粒聚并的影响。结果发现:随着流场的发展,沿着流动方向涡量值逐渐减小,不同粒径的颗粒受漩涡的影响在流场中的分布不同;增大入口处颗粒的体积分数,可增强颗粒聚并效果,颗粒的体积分数增大到一定值后,颗粒聚并效果变得不明显;当入口颗粒St≤1时,颗粒的St数越小,聚并效果越显著。     

4气门DISI汽油机缸内湍流场POD分析

应用本征正交分解(POD)方法对一台4气门直喷式汽油机(DISI)缸内冷态湍流流场试验测量数据以及大涡模拟计算数据进行了分析,以深入考察该汽油机缸内湍流场拟序结构特性以及湍流场循环变动特性.结果表明,POD方法可以将流场湍流涡团结构按照含能数量进行分解,大尺度拟序结构和小尺度随机脉动涡团可以被有效分离,为研究其各自的形成与演化特征创造了条件.考察各模态的时间系数在不同周期间的统计变化规律,可以了解流场湍流结构的循环变动特性.相比较使用传统相平均方法来研究汽油机循环变动,POD方法可以单独研究不同能量涡团结构的循环变动特点,为进一步深入理解循环变动特性奠定基础.     

入流滑移条件对两相射流特性影响的大涡模拟研究

研究了入流滑移条件对空间发展的气粒两相平面湍射流的非定常流动特性的影响。以Re数13000的平面不可压缩湍射流流动为例,气相场用Euler方法求解,通过大涡模拟(large-eddy　simulation,LES),直接求解大尺度涡运动的Navier-Stokes方程,小尺度涡用标准Smagorinsky亚格子模式模拟。颗粒相的运动用Lagrangian方法直接求解。在不同入流滑移条件下(U     

Two-Way Coupling Vortex Method and Its Application

Because of the success of the discrete vortex method for the simulation of large-scale vortex structure, many researchers extend this method to two-phase flow simulations, especially, to the simulation of particle dispersion in mixing layer, which is characterized by large-scale vortex structure. But the previous work is limited to one-way coupling, which neglects the effect of particles on fluid flow. In this paper, a discrete vortex method involving two-way coupling for two-phase flows is first proposed and then used in numerical simulation of two-dimensional gas-particle mixing layers. The numerical results show that the introduction of particles into the mixing layer has significant effects on the creation, development and merging process of large-scale vortex structures. It makes the mean size of large-scale vortex structure larger and the distance needed for development of large-scale vortex structure shorter.     

Direct numerical simulation of gas-solid two-phase mixing layer

In this paper, the spatially evolving of the higher Reynolds numbers gas-solid mixing layer under compressible conditions was investigated by a new direct numerical simulation technology. A high-resolution solver was performed for the gas-phase flow-field, particles with different Stokes numbers were traced by the Lagrangian approach based on one-way coupling. The processes of the vortex rolling up and pairing in the two-dimensional mixing layer were captured precisely. The large-scale structures developed from the initial inflow are characterized by the counter-rotating vortices. The mean velocity and the fluctuation intensities profiles agree well with the experimental data. Particles with smaller Stokes numbers accumulate at the vortex centers due to the smaller aerodynamic response time; particles with moderate Stokes numbers tend to orbit around individual streamwise vortices and in the periphery of paring vortices; particles with larger Stokes numbers disperse less evenly, showing a concentration distribution in the flow field.     

NUMERICAL INVESTIGATION OF MIXING LAYERS USING THE DISCRETE VORTEX METHOD WITH THE DIFFUSION VELOCITY MODEL

The evolution of large vortex structures and the nondimensional statistical quantities in a two-dimensional mixing layer are investigated by the discrete vortex method, in which viscous diffusion is treated by the diffusion velocity method. Results indicate that continuous amalgamation among the large vortex structures is the primary factor leading to growth of the mixing layer. They agree well with experiments and show a strong capability to discover the self-similar property of mixing layers. Results also indicate that the overestimation of root-mean-square velocity v' is attributed to the effect of three-dimensionality in actual flows, and is neglected by the two-dimensional discrete vortex method.     

Particle mixing in free shear flows

The mixing of particles or droplets in free shear layers is encountered in a variety of combustion systems and industrial applications. Free shear layers are characterized by large scale vortical structures which evolve and interact with time. These vortex structures can play a major role in particle or droplet dispersion. It has recently been postulated that the organized rotating motion of the large-scale structures can enhance the dispersion of intermediate size particles. This paper first reviews the currently-accepted mechanisms and models for particle dispersion in homogenoue, isotropic turbulence and addresses the difference between such flows and free shear layers. The essential features of free shear flows are then described and experiments on particle dispersion in free jets and mixing layers are reviewed. Numerical models which have been developed for particle dispersion in free shear layers, such as plane mixing layers, jets and wakes, are outlined and the results are interpreted in light of the postulated physical model. Both experimental results and numerical simulations strongly imply that particle dispersion in free shear layers is controlled by the motion of large scale vortex structures.     

Investigation of gas-solid two-phase flow across circular cylinders with discrete vortex method

In present paper, a Lagrangian–Lagrangian model is proposed to study gas-solid two-phase flow across single cylinder and two tandem cylinders at high Reynolds number. In this model, the single-phase flow is simulated by discrete vortex method and the particle trajectories are tracked by particle motion equation. A sub-cycle is introduced to adjust time-step in the particle collision model, through which the simulation of two phases is coupled. Validated by the comparison of the particle trajectory in Rankine vortex with literature, this model is used to study single-phase flow across single cylinder and cylinders with different arrangements firstly to get transient flow field and drag coefficients. Then, gas-solid two-phase flow across cylinders with different arrangement is studied and particle distribution is obtained under different Stokes number for horizontal and vertical particle transport cases. The settlement, entrainment and aggregation of solid particles moving with the large-scale coherent vortex structure in the wake of single cylinder and between two cylinders are numerically investigated, and the effects of St number on the distribution of solid particles are obtained.     

Numerical analysis of the effect of boundary layer thickness on vortex structures and heat transfer in the wake behind a hill

This study presents a three‐dimensional numerical analysis of the effect of boundary layer thickness on vortex structures and heat transfer behind a hill mounted in a laminar boundary layer. When the thickness of the velocity boundary layer is comparable to the hill height, a hairpin vortex is formed symmetrically to the center of the spanwise direction in the wake. A secondary vortex is formed between the legs, and horn‐shaped secondary vortices appear under the concave parts of the hairpin vortex. When the boundary layer thickness increases, the legs and horn‐shaped secondary vortices move toward the center of the spanwise direction, and thus heat transport and heat transfer increase there. At this time, high‐turbulence areas generated locally move toward the center of the spanwise direction with an increase in the boundary layer thickness. With a further increase in the boundary layer thickness, steady streamwise vortices are formed downstream of the hill, but the heat transfer decreases. © 2009 Wiley Periodicals, Inc. Heat Trans Asian Res; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/htj.20261     

Computations of flow and heat transfer in a three-dimensional multilouvered fin geometry

This study presents computational results in a complex three-dimensional louver geometry. The three-dimensionality occurs along the height of the fin, where the angled louver transitions to the flat landing and joins with the tube surface. The transition region is characterized by a swept leading edge and decreasing flow area between louvers. The results show that for Re_b=1100, the flow on the angled louver is dominated by spanwise vortex shedding, which is weakly three-dimensional. On the other hand, the flow in the transition region exhibits strong three-dimensionality. A high-energy compact vortex jet forms in the vicinity of the louver junction with the flat landing and is drawn under the louver. The top surface experiences large velocities in the vicinity of the surface and exhibits high heat transfer coefficients. Although the flow slows down at the flat landing, the large induced velocities on the top surface increases the heat transfer coefficient on the tube surface.     

Turbulent open channel flow with heat transfer subjected to the control of a spanwise travelling wave

Turbulent open channel flows with heat transfer subjected to the control of a spanwise travelling wave have been investigated by means of direct numerical simulation (DNS). The three-dimensional Navier–Stokes and energy equations are numerically solved using a fractional-step method. The spanwise travelling wave is induced by a body force that is confined within the viscous layer with its maximum at the bottom wall and decaying exponentially away from it. The objective of this study is to reveal the near-wall turbulence behaviours, the turbulent heat transfer, and thermal structures under the control of the spanwise travelling wave. Three typical frequencies of the spanwise travelling wave, i.e., high-, middle- and low-frequency, corresponding to the exciting periods at T⁺ = 25, 50 and 100, are investigated to reveal the dynamics of turbulent motions and heat transfer. The Prandtl number (Pr) varies from 1 up to 100. To elucidate the behaviours of turbulence statistics and heat transfer, some typical quantities, including the mean velocity, velocity and vorticity fluctuations, temperature and its fluctuation, turbulent heat fluxes, and the structures of the temperature fluctuation, are exhibited and analyzed.