基于SYCL的多相流LBM模拟跨平台异构并行计算研究

异构并行体系结构是当前高性能计算的重要技术趋势。由于各种异构平台通常支持不同的编程模型,跨平台性能可移植异构并行应用开发非常困难。SYCL是一个基于C++语言的单源跨平台并行编程开放标准。目前针对SYCL的研究主要集中于与其他并行编程模型的性能比较,对SYCL中提供的不同并行内核实现及其性能优化研究得较少。针对这一现状,基于SYCL编程模型对开源多相流数值模拟软件openLBMmflow实现跨平台异构并行模拟,通过对比基础并行版本、细粒度调优的ND-range并行版本以及计算到工作项多对一映射方法,系统总结了SYCL并行应用的性能优化方法。测试结果表明,在Intel Xeon Platinum 9242 CPU以及NVIDIA Tesla V100 GPU上,相比优化后的OpenMP并行实现,在不需要额外调优的情况下,基础并行版本在CPU上获得了2.91的加速比,表明了SYCL的开箱即用性能具备一定优势。以基础并行版本为基准,ND-range并行版本通过改变工作组大小及形状,在CPU与GPU上分别取得了最高1.45以及2.23的加速比。通过优化计算到工作项的多对一映射改变每个工作项处理...     

基于3D-LBM的多相流快速模拟

对于流体和其他物体的交互, 提出了一种基于Lattice Boltzmann的建模和绘制方法。针对固、液交互提出了外力叠加机制, 考虑了障碍物对流体的单向作用; 在液、液交互时考虑了两种液体的相互作用力。采用GPU硬件加速技术对LBM算法进行了加速, 并采用基于屏幕空间的绘制技术对流体表面进行了绘制。实现了两种不相溶液体交互, 以及液体与固体交互场景的模拟。     

基于神威加速计算架构的LBM多级并行计算

格子玻尔兹曼方法(lattice Boltzmann method, LBM)是一种基于分子运动理论计算流体力学(computational fluid dynamics, CFD)的方法, 提高LBM的并行计算能力是高性能计算领域的一项重要的研究内容. 本文基于SW26010Pro处理器, 通过区域分解、数据重构、双缓冲、向量化等优化方法, 实现了LBM的多级并行. 基于以上优化方案, 测试了5 600万网格规模, 实现结果显示, 相比于MPI进行级并行, 碰撞过程的平均加速倍数达到61.737、迁移过程的平均加速倍数达到17.3, 同时对方腔流案例做了强扩展测试, 网格规模为1200×1200×1200, 以6.2万计算核心为基准, 百万核心的并行效率超过60.5%.     

面向超大规模并行模拟的LBM计算流体力学软件

格子玻尔兹曼方法(Lattice Boltzmann Method,LBM)是一种基于介观模拟尺度的计算流体力学方法,已被广泛用于理论研究和工程领域。提高LBM计算流体软件的并行模拟能力,是高性能计算及应用研究中的一项重要内容。该研究基于"神威·太湖之光"超级计算系统,设计并实现了一套高效扩展的LBM计算流体力学软件。针对国产众核处理器SW26010的架构,文中设计了以下几种提高SWLBM方针速度和可扩展性的多级并行技术,包括面向19点stencil的数据复用、碰撞过程向量化、主从异步并行通信计算隐藏等。基于以上并行优化方案,文中测试了高达56 000亿网格的数值模拟,SWLBM软件持续浮点计算性能达到4.7 PFlops,软件模拟速度提高了172倍。相比百万核心10 000*10 000*5 000网格风场模拟,SWLBM整机千万核心的并行效率可达87%。测试结果表明,SWLBM有能力为工业应用提供实用的大规模并行模拟解决方案。     

多相流搅拌器流场数值模拟软件

为对搅拌器多相流体间的最终乳化结果进行模拟,结合FLUENT与ANSYS软件,以VB为前台程序开发搅拌器流场数值模拟软件.用VB开发用户交互界面,通过后台调用ANSYS的APDL完成前处理模块,实现搅拌器叶片、筒体的建模,网格划分和组元创建;通过后台调用FLUENT运行日志文件实现流体流场仿真运算与结果后处理,并以图形...     

基于高性能并行计算的旋转网球空气动力学模拟

基于曙光5000A高性能并行集群架构和软硬件环境,采用计算流体力学方法,针对旋转物体特有的马格努斯力作用,应用Realizable k-ε湍流模型,研究网球在不同球速和转速条件下周围流场压力分布及网球表面压力分布特性,并提出一种旋转网球空气动力学模拟方法。通过增加网球旋转加大上下表面压力差,增强马格努斯效应使网球急坠,球速增加造成周围气体流态紊乱,使得网球可控性变差,阻力和升力系数分别随网球球速和旋转速度增加而增加。实验结果表明,多核并行计算可有效提高计算精度和运算效率。     

面向多相流模拟的体积通量无散度SPH方法

针对高密度比多相流体模拟中存在的相间密度计算误差问题及产生的不合理对流运动模拟效果,提出一种基于体积通量无散度的隐式流体压强求解方法.首先,分析传统多相流模拟方法产生密度近似误差的原因;其次,提出“体积-压缩率”的关联计算方式,构建流体压缩率与压强间的线性关系;再次,分别设计恒定体积求解器和体积通量无散度求解器,以实现多相流模拟过程中流体体积的不可压缩性和速度场的无散度特性.为验证所提方法性能,以流体模拟方法 DFSPH为对比对象,分别以模拟效果合理性、数值计算稳定性与收敛性为定性和定量评估指标,依次开展两相溃坝、热对流等多相流交互实验.结果表明,该方法能够实现高效、稳定的多相流交互模拟视觉效果,在同等多相流条件下较DFSPH方法耗费更少计算时间实现收敛,在各种复杂模拟场景中均具有良好的健壮性、有效性和可扩展性,尤其适用于高密度比流体交互模拟.     

3D打印氧化锆复合浆体的MRT LBM流动分析

陶瓷3D打印的成功应用有助于改善陶瓷材料成形难的问题.针对挤出式的浆料直写工艺,提出一种单螺杆挤出结构代替原有的挤出结构,为了分析陶瓷浆体在挤出结构中的流动情况,将多松弛参数的格子玻尔兹曼方法(MRT LBM)引入作为流体动力学方法,首先利用流变测试仪对氧化锆复合陶瓷浆体的流变方程进行了测试拟合,将其代人MRT LBM中,获取浆体在流道中的流线图,以及速度分布图,并与单松弛参数的格子玻尔兹曼方法(SlRT LBM)进行了对比,在碰撞步中应用多个松弛时间可进一步改善精度和稳定性,对比结果验证了MRT LBM可有效应用于复杂流体的流动分析中.     

气固两相流全分辨率直接数值模拟的并行算法

用沉浸边界法对气固两相流进行全分辨率直接数值模拟,介绍并行算法及其并行效率。考察球形颗粒的空间分辨率等对计算精度的影响,对颗粒雷诺数Rep=1~150,颗粒直径与计算网格比大于20时,该并行算法获得的计算精度较高。     

基于组态软件MCGS的多相流实验控制系统

基于MCGS开发的多相流实验控制系统,能够实现油、气、水单相、多相流程的控制,自动采集并记录温度、压力、流量等参数,避免了人工数据记录、整理和烦琐的数据处理工作,提高了实验精度与工作效率。     

基于Lattice Boltzmann模型的液-液混合流模拟

引入了一种二元Lattice Boltzmann Model（LBM）,实现丁两种液体组成的混合流的模拟．不同于其它的类似模型,它区分考虑了流体的粘性和扩散特性,可以很容易地模拟各种互溶或者不互溶的混合流现象．此外,由于LBM的运算大都是线性的局部运算,这使得它很容易在可编程图形处理器（Graphics Process Unit,GPU）上进行加速,从而进行实时模拟．给出了若干二元混合流的模拟结果．     

Accelerated Lattice Boltzmann Schemes for Steady-State Flow Simulations

A matrix formulation of the steady Lattice Boltzmann equation is presented. It is shown that the strict steady-state formulation, combined with preconditioned iterative solvers, leads to significant computational savings as compared to the standard explicit LBE scheme.     

Optimization Strategies for the Entropic Lattice Boltzmann Method

The entropic formulation of the lattice Boltzmann method (LBM) features enhanced numerical stability due to its compliance with the Boltzmann H-theorem. This stability comes at the price of some computational overhead, associated with the need of adjusting the local relaxation time of the standard LBM in such a way as to secure compliance with the H-theorem. In this paper, we discuss a number of possible optimization strategies to reduce the computational overhead of entropic LBMs.     

Lattice Boltzmann simulation of lid-driven flow in trapezoidal cavities

In this paper, an incompressible lattice Bhatnagar–Gross–Krook (LBGK) model proposed by Guo et al. is used to simulate lid-driven flow in a two-dimensional isosceles trapezoidal cavity. Due to the complex boundary of the trapezoidal cavity, here the extrapolation scheme proposed by Guo et al. is used to treat curved boundary. In our numerical simulations, the effects of the Reynolds number (Re) and the top angle θ on the strength, center position and number of vortices in the isosceles trapezoidal cavities are studied. Re is varied from 100 to 15,000, and the top angle θ ranges from 50 to 90. Numerical results show that, as Re increases, the phenomena in the cavity become more and more complex, and the number of the vortexes increases. We also found that the vortex near the bottom wall breaks up into two smaller vortices as θ increases up to a critical value. Furthermore, as Re is increased, the flow in the cavity undergoes a complex transition (from steady to the periodic flow, and finally to the chaotic flow). At last, the scope of critical Re for flow transition from steady to periodic state, and from periodic to chaotic state is presented for different top angles θ.     

Topology optimization of flow domains using the lattice Boltzmann method

We consider the optimal design of two- (2D) and three-dimensional (3D) flow domains using the lattice Boltzmann method (LBM) as an approximation of Navier-Stokes (NS) flows. The problem is solved by a topology optimization approach varying the effective porosity of a fictitious material. The boundaries of the flow domain are represented by potentially discontinuous material distributions. NS flows are traditionally approximated by finite element and finite volume methods. These schemes, while well established as high-fidelity simulation tools using body-fitted meshes, are effected in their accuracy and robustness when regular meshes with zero-velocity constraints along the surface and in the interior of obstacles are used, as is common in topology optimization. Therefore, we study the potential of the LBM for approximating low Mach number incompressible viscous flows for topology optimization. In the LBM the geometry of flow domains is defined in a discontinuous manner, similar to the approach used in material-based topology optimization. In addition, this non-traditional discretization method features parallel scalability and allows for high-resolution, regular fluid meshes. In this paper, we show how the variation of the porosity can be used in conjunction with the LBM for the optimal design of fluid domains, making the LBM an interesting alternative to NS solvers for topology optimization problems. The potential of our topology optimization approach will be illustrated by 2D and 3D numerical examples.     

Implementing lattice Boltzmann computation on graphics hardware

The Lattice Boltzmann Model (LBM) is a physically-based approach that simulates the microscopic movement of fluid particles by simple, identical, and local rules. We accelerate the computation of the LBM on general-purpose graphics hardware, by grouping particle packets into 2D textures and mapping the Boltzmann equations completely to the rasterization and frame buffer operations. We apply stitching and packing to further improve the performance. In addition, we propose techniques, namely range scaling and range separation, that systematically transform variables into the range required by the graphics hardware and thus prevent overflow. Our approach can be extended to acceleration of the computation of any cellular automata model.     

格子波尔兹曼方法的医学图像同步去噪增强算法

格子波尔兹曼的理论基础为分子动理学和统计力学,具有算法简单高效,执行速度快,易于并行处理等优点,近年来成功地应用于图像处理领域.本文针对医学图像反差较低且包含噪声污染的特点,通过设计分段线性拉伸函数作为格子波尔兹曼方程的外力项,实现医学图像的同步去噪和反差增强.同时构建基于TV下降流的PDE反差增强模型与本文的方法进行对比,实验表明,本文的方法处理效果优于TV下降流对应的效果.     

Application of lattice Boltzmann method in free surface flow simulation of micro injection molding

The filling flow in micro injection molding was simulated by using the lattice Boltzmann method (LBM). A tracking algorithm for free surface to handle the complex interaction between gas and liquid phases in LBM was used for the free surface advancement. The temperature field in the filling flow is also analyzed by combining the thermal lattice Boltzmann model and the free surface method. To simulate the fluid flow of polymer melt with a high Prandtl number and high viscosity, a modified lattice Boltzmann scheme was adopted by introducing a free parameter in the thermal diffusion equation to overcome the restriction of the thermal relaxation time. The filling flow simulation of micro injection molding was successfully performed in the study.     

Three-dimensional lattice Boltzmann simulations of high density ratio two-phase flows in porous media

A three-dimensional multiphase lattice Boltzmann model is implemented to study the spontaneous phase transport in complex porous media. The model is validated against the analytical solution of Young’s and Laplace’s laws. Afterward, three-dimensional porous layers are randomly generated to investigate droplet penetration into a substrate, liquid transport in a porous channel as well as extraction of a droplet from a porous medium. Effects of several geometrical and flow parameters such as porosity, density ratio, Reynolds number, Weber number, Froude number and contact angle are considered. A parametric study of the influence of main non-dimensional parameters upon the impact of liquid drops on permeable surface is performed. Results show that while increasing Froude number causes spreading of the droplet on the surface, increasing Reynolds number, Weber number, porosity and liquid-air density ratio will enhance the penetration rate into the surface. Furthermore, increasing the contact angle decreases both the spreading and the penetration rate at the same time. In the same way, for the liquid transport in a porous channel, it is found that increasing the porosity and Reynolds number will result in increasing penetration rate in the channel. For the extraction of a droplet from a porous medium, it is shown that by increasing the gravitational force and/or porosity the droplet extracts faster from the substrate.