基于双判定准则的航空发动机叶片缺陷孔洞的边界提取

航空发动机叶片缺陷孔洞的几何形状复杂、位置分布散乱且曲率变化明显,使用单一判定准则难以完整地提取其边界特征点.为实现叶片缺陷孔洞边界的精准提取和孔洞边界线的平滑拟合,提出了一种基于双判定准则的航空发动机叶片缺陷孔洞边界特征点的提取算法.该算法采用K–Dimension tree(K–D树)建立点云的空间拓扑结构,以采样...     

CAARC高层建筑标模动态测力研究

CAARC高层建筑标模的动态测力试验在南京航空航天大学NH－2低速风洞中进行，在均匀流场中以及模拟大气边界层中分别测量了模型的动态气动力和力矩的平均值，均方根值，峰值，试验目的是为检测NH－2风洞的大气层边界模拟技术，动态天平，测试仪器以及动态数据采集和处理系统，试验结果表明，CAARC标模的所有动态气动特性在NH－2中的测量值与国外部报导的结果符合得较好，因此，NH－2风洞的大气边界层模拟技术，动态天平，测试仪器以及动态数据采集与处理系统可以承担高层建筑方面动态测力试验任务。     

非线性晃动问题的ALE边界元方法

利用ＡＬＥ（任意的Ｌａｇｒａｎｇｅ－Ｅｕｌｅｒ）边界元方法数值求解了具有自由液面的非线性晃动问题，即受外力激励下流体的非线性振动问题。把ＡＬＥ有限元方法的思想应用到边界元方法中，得到了ＡＬＥ边界元方法。对于自由液面的非线性动力边界条件，应用Ｇａｌｅｒｋｉｎ加权方法进行了有限元数值离散。为了增加求解精度，对动力边界条件提出了增加误差修正项的数值求解方法。对时间变量采用Ｎｅｗｍａｒｋ方法进行离散。推导了系统非线性方程的预测－多次校正法迭代格式。进行了算例分析与比较，得到了令人比较满意的结果。     

高湍流度对平板边界层传的热影响

在雷诺数为１０６—１０７之间对０．１％—２５％湍流度的自由来流流过平板边界层传热问题进行了实验与数值研究。实验是在吸入式风洞内进行，用自由来流冷却加热平板，自由来流的湍流度可高达３０％。计算模型针对二维定常流，采用代数应力（热流）湍流模型，用ＳＩＭＰＬＥ算法进行计算。结果表明：高湍流度自由来流可使平板的平均Ｓｔａｎｔｏｎ数增加８５％。     

低速大尺寸传热风洞的研制

为进行航空发动机内特别是叶轮机械内的边界层、传热和气膜冷却研究，设计、建造了大尺寸低速回流式风洞。描述了风洞结构和设计中考虑的主要问题。风洞性能测试结果表明，试验段中的气流达到很高的均匀度，不用紊流发生器时可达到很低的紊流度，试验板上的边界层参数和换热系数都和公认的关系式符合良好。     

带跨音结尾激波的二元收-扩管内附面层控制实验研究

本文较深入地研究了由直-曲壁构成的二元收-扩管内收敛段底壁的附画层抽气板形状,在吸除系数为1条件下,吸除附面层对几何喉道附近壁画、对跨音结尾激波与扩压器壁附面层干扰区上游的附面层发展的影响;研究了结尾跨音激波与扩压器底壁附面层干扰区的附面层控制对其下游的流场畸变的影响;文中并对有、无附面层控制下的干扰区下游的动态畸变作了比较.指出,通过对干扰区的附面层抽吸,近壁面的紊流度峰值和平均值大大下降.     

转捩诱导法向力及其对细长尖锥气动特性的影响

边界层转捩时,是否会出现诱导法向气动载荷,这是一个很有意义的问题。本文介绍了在高超音速风洞中完成的静态气动力实验与动态气动力实验,它证实这种诱导法向载荷是存在的,是边界层转捩的不对称性造成的,并对细长锥的气动特性有明显而呈规律性的影响。     

An engineering model of Titan surface winds for Dragonfly landed operations

Winds near the ground on Titan for the Dragonfly landing site (near Selk crater, 10°N) for the mid-2030s (Titan late southern summer, L_s ~ 310°) are estimated for mission design purposes. Prevailing winds due to the global circulation are typically 0.5 m/s, and do not exceed 1 m/s. Local terrain-induced flows such as slope winds appear to be similarly capped at 1 m/s. At various landing sites and times, these two contributions will vectorially combine to yield steady winds (for part of a Titan day, Tsol) of up to 2.0 m/s, but typically less – the slope wind component will be small in the mid-morning. In early afternoon, as on Earth and Mars, solar-driven convection in the planetary boundary layer will cause wind fluctuations of the order of 0.1 m/s, varying with a typical timescale of ~1000 s. Occasionally this convection organizes into coherent ‘dust devil’ vortices: detectable vortices with speeds of 1 m/s are predicted about once per Titan day. We have introduced the convective velocity scale combined with the advection time of PBL cells as a metric to derive the frequency of occurrence of gusts associated with convective vortices (‘dust devils’). Maximum possible vortex winds on Titan of 2.8 m/s may be expected only once per 40 Tsols, and define the maximum wind (4.8 m/s at 10 m height) that Dragonfly must tolerate without damage. The applicability of different wind combinations, scaled to the height of relevant Dragonfly components above the ground (e.g. the maximum corresponds to 3.9 m/s at 1.3 m height) by a logarithmic wind profile, to Dragonfly design and operations are discussed.     

A unifying approach in simulating the shot peening process using a 3D random representative volume finite element model

Using a modified 3D random representative volume (RV) finite element model,the effects of model dimensions (impact region and interval between impact and representative regions),model shapes (rectangular,square,and circular),and peening-induced thermal softening on resultant critical quantities (residual stress,Almen intensity,coverage,and arc height) after shot peening are systematically examined.A new quantity,i.e.,the interval between impact and representative regions,is introduced and its optimal value is first determined to eliminate any boundary effect on shot peening results.Then,model dimensions are respectively assessed for all model shapes to reflect the actual shot peening process,based on which shape-independent critical shot peening quantities are obtained.Further,it is found that thermal softening of the target material due to shot peening leads to variances of the surface residual stress and arc height,demonstrating the necessity of considering the thermal effect in a constitutive material model of shot peeing.Our study clarifies some of the finite element modeling aspects and lays the ground for accurate modeling of the SP process.     

Analysis on capabilities of density-based solvers within OpenFOAM to distinguish aerothermal variables in difusion boundary layer

Open source feld operation and manipulation（OpenFOAM）is one of the most prevalent open source computational fluid dynamics（CFD）software.It is very convenient for researchers to develop their own codes based on the class library toolbox within OpenFOAM.In recent years,several density-based solvers within OpenFOAM for supersonic/hypersonic compressible flow are coming up.Although the capabilities of these solvers to capture shock wave have already been verifed by some researchers,these solvers still need to be validated comprehensively as commercial CFD software.In boundary layer where diffusion is the dominant transportation manner,the convective discrete schemes＇capability to capture aerothermal variables,such as temperature and heat flux,is different from each other due to their own numerical dissipative characteristics and from viewpoint of this capability,these compressible solvers within OpenFOAM can be validated further.In this paper,frstly,the organizational architecture of density-based solvers within OpenFOAM is analyzed.Then,from the viewpoint of the capability to capture aerothermal variables,the numerical results of several typical geometrical felds predicted by these solvers are compared with both the outcome obtained from the commercial software Fastran and the experimental data.During the computing process,the Roe,AUSM＋（Advection Upstream Splitting Method）,and HLLC（Harten-Lax-van Leer-Contact）convective discrete schemes of which the spatial accuracy is 1st and 2nd order are utilized,respectively.The compared results show that the aerothermal variables are in agreement with results generated by Fastran and the experimental data even if the1st order spatial precision is implemented.Overall,the accuracy of these density-based solvers can meet the requirement of engineering and scientifc problems to capture aerothermal variables in diffusion boundary layer.