国家数值风洞(NNW)工程中的黏性自适应笛卡尔网格方法研究进展

相比传统的结构和非结构网格,笛卡尔网格具有自动化和高质量兼顾的优势,是未来网格技术发展的重要方向。面向笛卡尔网格生成软件开发需求,依托国家数值风洞（NNW）工程,介绍了黏性自适应笛卡尔网格技术相关课题的研究工作进展,重点从网格生成技术、自适应方法、黏性物面边界处理方法等方面展开。网格生成技术方面,基于全线程树数据结构的思想,开展了叉树数据结构的优化;针对物面离散单元信息的快速检索,构建了稳定、平衡、高效的叉树式物面网格数据结构;对于网格单元类型的判断,发展了染色方法和包围盒方法以提高效率;针对大规模网格量,发展了高性能并行计算技术。自适应技术方面,发展了基于几何特征和流场解特征的自适应方法,建立了流场结构捕捉判据,并开展了三维复杂构型的应用;为了降低全局网格规模,发展了各向异性自适应方法,构建了各向异性特征识别的综合判据;发展了边界层法向射线加密技术,以减少边界层内的网格数目。物面边界处理方面,发展了重叠笛卡尔网格方法,构建了重叠网格尺度匹配技术和数据交互方法;开展了壁面函数方法在笛卡尔网格框架下的适应性应用,以放宽壁面网格尺度限制、降低边界层网格规模,并通过典型算例验证了方法的可靠性。     

网格技术

网格计算是一种新兴的解决大规模复杂问题的应用技术，网格是Internet的发展方向。文章阐述了网格的定义、体系结构、以及实现网格计算的关键技术，介绍了国内外网格技术的研究现状、应用前景和发展趋势。     

基于非结构网格流场计算的网格重排序

提出基于非结构网格流场计算时的网格重排序,优化了网格信息在内存中的存贮位置,使得计算过程中调用数据的处理速度加快,有效提高了计算效率。以M6机翼及DLR-F4翼身组合体的跨音速无粘流场为算例,分别通过显式四步龙格-库塔推进格式和隐式LU-SGS推进格式来验证网格重排序后对计算效率的提高作用。原始网格均用Delaunay方法生长,并分别使用了两种重排序方法进行比较。计算结果表明:重排序后的计算效率可以提高到未排序时的3到5倍,很具实用价值。     

基于图剖分的多块结构网格负载平衡方法

负载平衡是影响并行计算性能的重要因素。针对多块结构网格,给出了一种改进的多层次图剖分负载平衡方法。该方法设计了新的网格剖分算法,采用改进的子块分裂方法与图剖分算法的循环调用实现结构对接网格剖分,并通过建立不同物体重叠网格间的连接关系,实现了结构重叠网格的负载平衡。采用2个典型算例对方法进行了对比验证,数值结果表明,子块分裂方法对剖分结果具有重要影响,采用循环调用算法及改进的子块分裂方法能有效地实现计算负载均衡及通信量优化,同时显著减少了网格块数及因虚网格导致的内存需求,有利于提高并行效率。该负载平衡方法与网格拓扑无关,适用于多块结构对接网格及重叠网格,且整体型剖分方式对于多块结构重叠网格具有更好的剖分效果。     

非结构化网格生成中背景网格对源的响应

背景网格对源的响应方法对非结构网格的生成具有重要意义,如果背景网格参数能恰当地反映出源的变化,则所生成的非结构网格就能满足网格设计的要求。将背景网格的三角形单元分为7个子域,根据源在单元中的位置对网格进行局部加密或变形,使每一个源点都与一个背景网格结点相对应,再通过求解热传导方程,使背景网格参数分布光滑,基本反映网格设计的要求,从而提高了非结构化网格的生成质量。     

E 级计算给 CFD 带来的机遇与挑战

E 级(Exascale)计算机有望在2020年前后投入使用,E 级计算将给计算科学和科学研究带来革命性的变化。计算流体力学(CFD)作为超大规模计算机应用的重要领域之一,将迎来前所未有的发展机遇,同时也将面临极其严峻的技术挑战。本文对当前国内外超大规模 CFD 计算的现状进行了概述,探讨了未来 CFD 的发展趋势,并对 E 级计算给 CFD 带来的机遇与挑战进行了分析,最后提出了适应未来 E 级计算的 CFD 发展思路与建议。为了实现 E 级计算的宏伟目标,CFD 与计算机科学、应用数学等学科的“协同设计”势在必行。我们期待本文的分析能对我国的高性能计算应用、CFD 发展有一定的启示作用。     

气体运动论BGK格式的翼型绕流数值模拟

以气体运动论BGK(Bhatnagaar-Gross-Krook)格式为基础,通过改进非均匀网格下的通量计算方法,把气体运动论BGK方法应用到二维贴体网格上模拟了可压缩的无黏和黏性流动。在重构阶段,以交界面周围4个网格单元为一组,先把每个网格单元的宏观量(密度、动量和总能量)转换在相对于交界面的坐标系下,然后使用van Leer限制器插值求出交界面的宏观量,在求解阶段再使用BGK理论求出流过交界面的通量。通过上述方法使贴体网格和均匀网格,在计算格式上一致,与其他求解二维贴体网格的气体运动论BGK方法相比,该方法相对简单,计算量减小。结合贴体网格通量计算和气体运动论BGK格式方法理论,分别对亚、跨、超声速下的NACA0012翼型绕流进行了数值模拟和对比研究,所得可压缩无黏流动的数值结果与北大西洋公约组织航天研究与发展咨询组(AGARD)的结果吻合良好。本文还模拟了低雷诺数状态下的可压缩黏性流动,所得结果与文献中的数值计算结果吻合良好,证明了该方法在数值模拟及激波捕捉方面具有鲁棒性,为在复杂流动中的应用提供了一种简单的途径。     

非结构弹性体动网格方法应用策略及改进

针对3种代表性的经典弹性体动网格方法,从网格最大变形能力、质量评价指标和空间分布趋势3个方面综合分析了各方法的网格变形特性,重点总结了各方法的适用范围.结合工程实际,研究了特定问题的弹性体动网格应用策略.最后,整合各经典弹性体动网格方法的优势提出了改进方法,并通过NACA 0006翼型的超声速俯仰振荡加以验证.研究结果表明:基于面积的弹性体法适用于加密型网格,而基于距离的弹性体法适用于均匀型网格,该应用策略可为工程问题提供指导.改进方法效果明显,在大变形时相对基于距离的弹性体法的各项网格质量增幅均在10%以上,相对基于面积的弹性体法的最小网格质量增幅几乎达到4倍,并且在数值模拟中具有较高的精度,可显著增强弹性体法的鲁棒性.      

Flight Operations in the New Millennium

New approaches are being studied for real-time interaction, and related supporting processes, with spacecraft and instruments in deep space. Spacecraft are evolving, improving in many ways, and generally becoming more robust. Operations is changing also, and will be more automated in the future. However, there is a challenge. Deep space missions are not all alike. The operations phases of discovery and exploration are an extension of the research that creates the mission; they are the time of obtaining results. This examines the historical role of flight operations and its evolving processes to develop an understanding of the operational methods that will be effective in the future. It takes people, equipment, software, space, and connectivity for operations success. A balance has to be struck between improving technology, gaining knowledge, automation, and realistic expectations. Finally, the recommended methods to gain efficiency in operations are system-wide services and shared resources. These common processes will meet the challenge of varied missions.     

基于非结构网格有限差分法的扎染算法

应用有限差分法时遵循贴体坐标系下离散等价方程的离散准则,格式中仅用到当地网格点的坐标变换系数,据此可以构建新的非结构网格有限差分法,数值算例表明在空间二维情况下用连接离散点的3条网格线构造出来的一阶迎风格式可以稳定收敛。把这种非结构网格有限差分法推广到三维,在4条网格线基础上进行离散计算,然后提出一种局部区域多重网格、重复计算的新算法。首先在包含物体的整个区域采用直角坐标系下的均匀网格进行计算,然后消除物体内部点对物体外部流场区域的影响,最后在物面和直角网格之间很小的局部区域填补非结构网格进行计算,计算过程类似于中国传统的扎染工艺。处理流场内存在曲面物体边界问题时,相较于常规结构网格差分算法,扎染算法尽管在计算中包含需要特殊处理的无用网格点,也存在重复计算的局部区域,但是以这些计算能力的浪费为代价,换取了编程简便、内存量小、网格生成快捷、易于扩展网格规模等优势。定性分析表明,这种扎染算法非常适合研制大规模并行计算,建立的计算流体力学应用软件可以充分发挥数十万至百万处理器核心的超级计算机的效能。     

PVM平台下通量守恒的嵌入自适应网格解激波和底部流

运用嵌入式自适应网格技术来提高结构网格求解复杂流场的分辨率。该方法根据流场的形成,将密集网格嵌入到物理量剧烈变化的区域,使求解精度得以提高的同时,避免较大的计算开销,从而达到求解效率和求解精度之间的平衡。 Ｐ Ｖ Ｍ 平台使得被嵌入的网格可以在网络中不同的机器上完成,大幅度降低了墙上时间。原始网格与被嵌入网格之间数值通量守恒的实现增强了该方法在捕捉间断时的有效性和可靠性。     

一种利用栅格光源特征编码的室内定位技术

目前,几种主流的室内定位技术分别存在定位精度低、设备成本高、易受环境干扰等一种或多种缺点,基于可见光通信的室内定位技术成为了研究热点。不同于主流的室内定位技术,提出了一种利用栅格光源特征编码的室内定位方案,先使光源照射区被划分为栅格,再用光源闪烁传递栅格特征信息,用户端接收闪烁光信号,对其解码实现定位。利用的编码方式结构巧妙,可以有效避免误读,同时相邻栅格编码相似度高,有利于跨栅格定位。在实验室中,以普通投影仪生成栅格光源,以市售安卓智能手机为用户终端,在较理想条件下的测试表明,该方案在动态和静态情况下均可确定用户位置,定位精度达到半个栅格,定位准确率达到100%。     

运动激波自适应网格算法中权函数问题的研究

变分原理的自适应网格技术被应用到运动激波问题的求解上,在解的大梯度区自动加密网格,从而非常成功地算出了激波。通过分析发现权函数选择和取值对自适应网格技术至关重要     

Computational intelligence and soft computing for spaceapplications

Systems using computational intelligence and soft computing have been successfully developed for many industrial and space applications. These systems seek to emulate the type of reasoning that humans perform when solving complex tasks. The field of soft computing, as defined by Zadeh-the inventor of fuzzy logic-encompasses fuzzy logic as well as other methodologies such as neural networks, genetic algorithms, and uncertainty management. It is expected that soft-computing techniques will eventually become as common and prevalent as traditional methods of computer science. This paper presents an overview of applications of fuzzy logic and soft computing to space projects. The role of fuzzy systems that can learn from experience to improve their performance is discussed. We present a report on applications of these adaptive systems to NASA space projects such as the orbital operations of the Space Shuttle, which include attitude control and rendezvous/docking operations. We also provide insights on the future of computational intelligence and soft computing and of their vast potential in industrial applications     

新型重叠网格洞面优化方法及其应用

高度自动化的洞面优化方法是重叠网格关键技术之一。通过对割补法的分析,针对其洞面切割后洞边界位置不确定性及不可预知性问题,提出将其与物面距离优化准则相结合的新型混合洞面优化方法。该方法针对各物体间体网格,采用物面距离优化准则,将洞边界位置控制在物体中间的位置,使重叠区域更合理、可预测;针对背景网格,将其物面距离设置为较大值,采用物面距离优化准则将其落入物体网格内部的单元一并挖去;针对物面处重叠的网格,采用割补法对其进行切割和填补,保证物面及近壁面网格得以有效的重叠。该方法在保证高适用性的前提下提高了网格重叠质量,自动化程度高,无需人工干预。3个典型复杂流动算例计算结果与实验结果吻合良好,网格重叠区域流场变量传递正确,等值线过渡光滑,流场刻画准确,证明该方法准确可靠。     

Future of Venus Research and Exploration

Despite the tremendous progress that has been made since the publication of the Venus II book in 1997, many fundamental questions remain concerning Venus’ history, evolution and current geologic and atmospheric processes. The international science community has taken several approaches to prioritizing these questions, either through formal processes like the Planetary Decadal Survey in the United States and the Cosmic Vision in Europe, or informally through science definition teams utilized by Japan, Russia, and India. These questions are left to future investigators to address through a broad range of research approaches that include Earth-based observations, laboratory and modeling studies that are based on existing data, and new space flight missions. Many of the highest priority questions for Venus can be answered with new measurements acquired by orbiting or in situ missions that use current technologies, and several plausible implementation concepts have been studied and proposed for flight. However, observations needed to address some science questions pose substantial technological challenges, for example, long term survival on the surface of Venus and missions that require surface or controlled aerial mobility. Missions enabled by investments in these technologies will open the door to completely new ways of exploring Venus to provide unique insights into Venus’s past and the processes at work today.     

A fast and automatic full-potential finite volume solver on Cartesian grids for unconventional configurations

To meet the requirements of fast and automatic computation of subsonic and transonic aerodynamics in aircraft conceptual design,a novel finite volume solver for full potential flows on adaptive Cartesian grids is developed in this paper.Cartesian grids with geometric adaptation are firstly generated automatically with boundary cells processed by cell-cutting and cell-merging algo rithms.The nonlinear full potential equation is discretized by a finite volume scheme on these Cartesian grids and iteratively solved in an implicit fashion with a generalized minimum residual (GMRES) algorithm.During computation,solution-based mesh adaptation is also applied so as to capture flow features more accurately.An improved ghost-cell method is proposed to implement the non-penetration wall boundary condition where the velocity-potential of a ghost cell is modified by an analytic method instead.According to the characteristics of the Cartesian grids,the Kutta condition is applied by specially computing the gradients on Kutta-faces without directly assigning the potential jump to cells adjacent wake faces,which can significantly improve the solution con verging speed.The feasibility and accuracy of the proposed method are validated by several typical cases of sub/transonic flows around an ONERA M6 wing,a DLR-F4 wing-body,and an unconventional figuration of a blended wing body (BWB).The validation cases demonstrate a fast convergence with fully automatic grid treatment and computation,and the results suggest its capacity in application for aircraft conceptual design.     

结构/非结构混合网格数值模拟栅格翼

为了详细了解栅格翼内部的流动特性,设计满足飞行器性能要求的栅格翼,本文开展了多种不同单栅格外形的气动特性数值计算,对单个栅格的外形进行了初步选型研究;然后采用结构网格、结构与非结构混合网格求解NS方程,对带有两片栅格翼的导弹进行了数值模拟,结构网格计算结果与结构非结构混合网格计算结果一致,并与国外文献的计算结果和风洞试验结果进行了对比,三者的结果吻合较好.采用混合网格比结构网格的网格规模小许多,可以节省大量计算机内存和计算时间,是一种比较好的栅格翼数值模拟方法.     

强制偏流喷管的三维网络生成

运用泊松方程实强了制偏流喷管内的三维网络生成，方程中的源项采用了与边界形状和边界网格点分布无关的数学形式，源项内的自由参数由横向网格线与边界的交角及横向网格线在边界处的局部曲率两个约束条件来确定，实现了由边界上的网格分布直接控制内部的网格点，使生成的网格具有良好的正交性和合理性。     

基于伴随方程的网格自适应及误差修正

基于流场方程的离散伴随优化理论和三维非结构网格,建立了网格自适应技术和目标函数误差修正方法。详细研究了用流动伴随变量进行目标函数的误差估计和修正技术,构造了适用于格心格式有限体积法的流场变量插值技术和网格单元剖分判据,初步实现了网格物面投影和空间单元优化,发展了适用于有限体积法的整套网格自适应方法。对NACA0012翼型和ONERA-M6机翼绕流进行了自适应数值模拟,并对升、阻力等目标函数进行了误差修正。数值结果表明,本文自适应方法能正确地捕捉到影响目标函数计算精度的敏感区域,网格自适应和误差修正两项技术显著提高了升、阻力等目标函数的计算精度。