A large-scale parallel hybrid grid generation technique for realistic complex geometry

High-Performance Computing (HPC) systems and Computational Fluid Dynamics (CFD) have made significant progress in recent years; however, as the basis of the large-scale parallel computing, the massive grid generation of billions of cells has become a bottleneck problem. In this study, a parallel grid generation technique is proposed to generate large-scale mixed grids with arbitrary cell types and scales. The basic idea of our method is analogous to the global mesh refinement technique. An initial coarse grid with arbitrary cell types is regarded as a background mesh which is partitioned into subzones, and subzones are assigned onto different CPU cores. After the cells and faces in each subzone are split, the inserted new points of the solid wall are projected onto the original CAD entities to preserve the geometry accurately. Finally, the tangled cells caused by the projection in the boundary layer are untangled by a local Radial Basis Function mesh deformation technique. Furthermore, a parallel partition approach and an efficient wall distance computing technique for massive grids are developed also to shorten the preprocessing time. The tests show that the preprocessing efficiency has been increased by two or three orders compared with traditional methods. Billions of grids are generated for the AIAA JSM high-lift model and the Chinese CHN-T1 transport model to test the ability of the parallel grid generation technique. The maximum scale up to 19 billion mixed elements is generated using 16 384 CPU cores in parallel, and the mesh quality is acceptable for CFD simulations.     

Vortex amplifier internal geometry and its effect on performance

Consideration of the internal geometry of a vortex amplifier reveals eight geometrical ratios that describe the device. The effects of each of these ratios on the performance of a vortex amplifier are considered and experimental results presented. The most significant ratios when designing for vortex amplifier performance are identified     

Effect of glycerol addition on phase inversion in horizontal dispersed oil-water pipe flows

The effect of interfacial tension on the phase inversion process during horizontal pipe flow of an oil-aqueous solution was investigated. Interfacial tension was varied by adding small amounts of glycerol in the water phase. At these glycerol concentrations the density and viscosity of the aqueous phase changed by 1% or less. Exxsol™ D140 (5.5 mPa s, 828 kg m⁻³) was used as the oil phase. The experiments were carried out in a 38 mm ID acrylic test pipe. The phase continuity and appearance of phase inversion were investigated using conductivity (wire and ring) probes and an Electrical Resistance Tomographic (ERT) system. The ERT also provided diagrams of the phase distribution in a pipe cross section. Drop size distribution was monitored using a dual impedance probe. It was found that starting from a water continuous flow with increasing oil fraction at constant mixture velocity the mixture inverted initially in the middle of the pipe (measured at 19 mm from the top pipe wall) while a higher oil fraction was required for inversion at the top (measured at 4 mm from the top pipe wall) and finally the rest of the pipe. The addition of glycerol did not affect the phase fraction where the initial inversion occurred but caused an increase in the oil fraction needed to complete the inversion. The drop size measurements were used to explain this behaviour. Pressure drop was found to decrease with increasing oil fraction but this trend reversed when inversion spread to the pipe wall and the oil continuous phase came in contact with it.     

The effects of confinement and inertia on the production of droplets

Recent experiments of Sibillo et al., Phys. Rev. Lett. 97:054502, (2006) investigate the effect of walls on flow-induced drop deformation for Stokes flow. The drop and the fluid in which it is suspended have the same viscosities. The capillary numbers vary from 0.4 to 0.46. They find that complex start-up transients are observed with overshoots and undershoots in drop deformation. Drop breakup is inhibited by lowering the gap. The ratio of initial drop radius to wall separation is fixed at 0.34. We show that inertia can enhance elongation to break the drop by examining Reynolds numbers in the range of 1 to 10. The volumes of the daughter drops can be larger than in the unbounded case, and even result in the production of monodisperse droplets.      

Impact of tetrahedralization on parallel conforming octree mesh generation

This work discusses the performance impact from the tetrahedralization of non‐conforming meshes generated by a parallel octree method capable of handling arbitrary immersed geometries. First, we review conforming techniques for meshes created from octrees. Then we implement a tetrahedralization scheme based on templates in a parallel meshing algorithm that uses a linear octree with 2:1 balancing constraint. Besides, we also propose a change on the partitioning strategy for the same meshing algorithm to improve the octree refinement load balancing. Scalability analyses show that the chosen tetrahedralization technique preserves algorithm performance. Copyright © 2014 John Wiley & Sons, Ltd.     

基于结构化网格的导管螺旋桨敞水性能计算和艇桨干扰分析

本文采用结构化网格建模求解RANS方程,建立了导管桨和艇桨组合体的水动力性能数值计算方法.计算中以JD75简易导管与Ka4-70螺旋桨组合的导管螺旋桨及其与SUBOFF艇体模型的组合体为研究对象,利用RNG k-ε湍流模型封闭RANS方程得到控制方程.通过滑移网格方法计算了导管桨敞水性能,然后将该桨置于全附体SUBOFF潜艇模型后,计算螺旋桨与主艇体间的干扰特性,文中对流场等结果进行了分析比较.     

Effect of primary jet geometry on ejector performance: A cold-flow investigation

The following cold-flow study examines the interaction of the diffracted shock wave pattern and the resulting vortex loop emitted from a shock tube of various geometries, with an ejector having a round bell-shaped inlet. The focus of the study is to examine the performance of the ejector when using different jet geometries (primary flow) to entrain secondary flow through the ejector. These include two circular nozzles with internal diameters of 15 mm and 30 mm, two elliptical nozzles with minor to major axis ratios of a/b = 0.4 and 0.6 with b = 30 mm, a square nozzle with side lengths of 30 mm, and two exotic nozzles resembling a pair of lips with axis ratios of a/b = 0.2 and 0.5 with b = 30 mm. Shock tube driver pressures of P₄ = 4, 8, and 12 bar were studied, with the pressure of the shock tube driven section P₁ being atmospheric. High-speed schlieren photography using the Shimadzu Hypervision camera along with detailed pressure measurements along the ejector and the impulse created by the ejector were conducted.     

Design and mechanical characteristics analysis of a new viscous damper for piping system

In this paper, the design of a new viscous damper is presented and its mechanical characteristics are investigated experimentally. The motion equation of a system consisting of a drop machine and the damper is set up. By numerically simulating this equation, the curve of the damper cavity generatrix is obtained on the assumption that the resisting force is constant. Then the new damper with big capacity and high-energy dissipation rate is designed. Drop tests using this damper and a Pro225-damper bought in the market are performed, respectively. On one hand, the experimental resisting forces of the new damper approximate constants, which illustrates that the simulation is viable. On the other hand, some advantages of the new damper over the Pro225-damper are found.     

Numerical study of flow and heat transfer during a high-speed micro-drop impact on thin liquid films

Numerical simulation of high-speed micro-droplet impingement on thin liquid film covering a heated solid surface has been carried out. Effect of droplet Weber number and liquid film thickness on the characteristics of flow and heat transfer has been investigated using the coupled level set and volume of fluid method. The code is validated against both the experimental and numerical results from the literature. Results show that the crown dynamics is mostly affected by variations in the initial film thickness but is weakly influenced by changes in the Weber number. The liquid within the film can be categorized as three regions based on the heat transfer distribution: the static film region, the transition region, and the impact region. The transient local wall temperature shows three stages: first stage when the temperature decreases rapidly, followed by a second stage in which the temperature starts to rise and then becomes almost constant in the third stage. After drop impact, the local Nusselt number continuously increases until reaching a maximum value, and then decreases approaching the initial impact stage. Our analysis of the change in Weber number shows that larger Weber number contributes to intense temperature variation at the crater core relative to other radial locations. Lastly, the results reveal that the thinner liquid film leads to lower wall temperature and hence, higher average Nusselt number.     

AUTOMATIC MESH GENERATION OF 3—D GEOMETRIC MODELS

In this paper the presentation of the ball-packing method is reviewed, and a scheme to generate mesh for complex 3-D geometric models is given, which consists of 4 steps: (1) create nodes in 3-D models by ball-packing method, (2) connect nodes to generate mesh by 3-D Delaunay triangulation, (3) retrieve the boundary of the model after Delaunay triangulation, (4) improve the mesh.     

Study of skin model and geometry effects on thermal performance of thermal protective fabrics

Thermal protective clothing has steadily improved over the years as new materials and improved designs have reached the market. A significant method that has brought these improvements to the fire service is the NFPA 1971 standard on structural fire fighters’ protective clothing. However, this testing often neglects the effects of cylindrical geometry on heat transmission in flame resistant fabrics. This paper deals with methods to develop cylindrical geometry testing apparatus incorporating novel skin bioheat transfer model to test flame resistant fabrics used in firefighting. Results show that fabrics which shrink during the test can have reduced thermal protective performance compared with the qualities measured with a planar geometry tester. Results of temperature differences between skin simulant sensors of planar and cylindrical tester are also compared. This test method provides a new technique to accurately and precisely characterize the thermal performance of thermal protective fabrics.     

通气空泡生成和溃灭特性试验研究

为了研究通气流量对超空泡生成和溃灭的影响,在西北工业大学高速水洞实验室进行了系列通气超空泡生成和渍灭特征的试验.结果表明:空泡生成过程中,空泡长度和直径的变化特性与空泡溃灭时的变化特性相反;通气流量对空泡生成和溃灭均有较大影响,数据显示通气量对空泡生成的影响大于对空泡溃灭的影响.此研究为进一步研究超空泡的快速稳定生成提供了实验基础.     

Experimental investigation of the performance of a thermoelectric generator based on Peltier cells

An experimental investigation is carried out to characterize the performance of thermoelectric modules used for electric power generation over a range of different resistance loads. The performance of a Peltier cell used as a thermoelectric generator is evaluated in terms of power output and conversion efficiency. The results show that a thermoelectric module is a promising device for waste heat recovery.     

基于AFT-Delaunay的二维解耦并行网格生成算法

面向平面任意几何区域网格生成,提出了一种将波前法AFT(Advancing Front Technique)与Delaunay法相结合的解耦并行网格生成算法。算法主要思想是沿着求解几何区域惯性轴,采用扩展的AFT-Delaunay算法生成高质量三角形网格墙,递归地将几何区域动态划分成多个彼此解耦的子区域;采用OpenMP多线程并行技术,将子区域分配给多个CPU并行生成子区域网格;子区域内部的网格生成复用AFT-Delaunay算法,保证了生成网格的质量、效率和一致性要求。本算法优先生成几何边界与交界面网格,有利于提高有限元计算精度;各个子区域的网格生成彼此完全解耦,因此并行网格生成过程无需通信。该方法克服了并行交界面网格质量恶化难题,且具有良好的并行加速比,能够全自动、高效率地并行生成高质量的三角网格。     

Enhancing control of grid distribution in algebraic grid generation

Three techniques are presented to enhance the control of grid-point distribution for a class of algebraic grid generation methods known as the two-, four- and six-boundary methods. First, multidimensional stretching functions are presented, and a technique is devised to construct them based on the desired distribution of grid points along certain boundaries. Second, a normalization procedure is proposed which allows more effective control over orthogonality of grid lines at boundaries and curvature of grid lines near boundaries. And third, interpolating functions based on tension splines are introduced to control curvature of grid lines in the interior of the spatial domain. In addition to these three techniques, consistency conditions are derived which must be satisfied by all user-specified data employed in the grid generation process to control grid-point distribution. The usefulness of the techniques developed in this study was demonstrated by using them in conjunction with the two- and four-boundary methods to generate several grid systems, including a three-dimensional grid system in the coolant passage of a radial turbine blade with serpentine channels and pin fins.     

基于网格生成的随机凹凸型混凝土骨料细观建模方法

基于目前混凝土细观数值分析多采用二维模型,提出了一种基于网格预先生成的二维随机凹凸型骨料的混凝土细观建模方法。该方法利用已生成的网格状态进行骨料投放,有效解决了在细观三相介质中的网格质量问题。采用随机凹凸型骨料使骨料模型更加接近真实形态,避免了该方法对骨料形貌描述不准确的问题。该方法可保证投放的随机性以及级配分布的合理性,二维投放骨料含量可达75%。最后,采用混凝土损伤塑性模型模拟了混凝土保护层锈胀开裂的细观断裂破坏过程,分析了混凝土裂纹萌生、扩展过程及破坏形态,数值结果与实验结果吻合良好。     

基于转换模板的三维实体全六面体网格生成方法

以四面体一六面体基本转换模板为基础,提出一系列具有伸缩性的扩展转换模板．可根据需要选择不同模板及其组合,将四面体分解为不同数量、不同密度过渡形式的六面体单元。这样,初始四面体网格不需要划分得很细,生成的六面体单元数量也可以通过采用不同规格的扩展转换模板而得到控制。提出了基于CAD几何造型的边界节点坐标修正方法．从而使边界网格更好地拟合几何模型边界。     

脆性材料破坏模拟的网格生成算法

脆性材料的破坏过程具有随机性,当前的网格生成算法没有充分考虑脆性材料破坏时裂纹扩展和碎块生成的随机性。在Persson网格生成算法与Delaunay随机网格剖分理论基础上,提出了一种可根据模拟需要动态控制网格品质的网格生成算法。通过对随机分布点的Delauna三角化,生成初始网格,然后将网格体系比拟为桁架结构,网格节点即为桁架节点。桁架节点在虚拟力作用下可动态调整位置,并最终达到整个体系受力平衡。对Persson 算法中的尺寸分布函数和收敛条件进行了修正,从而提高了收敛速度,并适用于任意形状对象的网格剖分。 基于VC++平台开发了算法程序。通过实例对算法进行了验证,表明算法能够满足脆性材料破碎模拟的需要。     

Numerical modeling of the initial stage of the generation of unsteady vortices from sharp corner in plane compressible flow

ＮＵＭＥＲＩＣＡＬＭＯＤＥＬＩＮＧＯＦＴＨＥＩＮＩＴＩＡＬＳＴＡＧＥＯＦＴＨＥＧＥＮＥＲＡＴＩＯＮＯＦＵＮＳＴＥＡＤＹＶＯＲＴＩＣＥＳＦＲＯＭＳＨＡＲＰＣＯＲＮＥＲＩＮＰＬＡＮＥＣＯＭＰＲＥＳＳＩＢＬＥＦＬＯＷＨｕａｎｇＤｕｎ（黄敦）（Ｄｅｐａｒｔ．...