A Nonlinear Sub-grid Scale Model for Compressible Turbulent Flow

本文采用直接过滤的Navier-Stokes方程组作为可压缩湍流大涡模拟控制方程组。方程组中因过滤产生的高阶相关项用Taylor级数展开近似,但仅保留级数的一阶导数项。这样产生的误差相当于丢失了模型的部分耗散作用,本文用一种动力学非线性亚格子模型来补偿丢失的耗散影响。本文根据Caylay-Hamilton定理导出了一个非线性亚格子模型,模型中包含的系数由动力学模式确定。与传统的动力学Smagorinsky模式相比,这种动力学非线性模型(DNM)稳定性更好,且不需要在统计均匀方向进行统计平均来计算模型系数,因此减少了计算开销。本文用这种非线性亚格子模型对绕双椭球高超声速湍流进行了模拟,并将所得结果与实验值、计算值及理论值进行了对比。结果表明,本文模型可以有效地模拟可压缩湍流流场。     

Y形进气道流场数值研究

Y形进气道在两侧进气布局的单发轻型战斗机和教练机上应用较为普遍。由于流动的复杂．不易对其内部的流动进行实测．为此本文基于Favre平均的N-S方程和B／L代数湍流模型．采用了Jameson格式和矩阵人工粘性耗散，对某型亚声速飞机的Y形进气道流场进行了数值模拟。提出了一种解决网格块之间插值问题的处理方法．研究了进气道两侧管道气流的掺混作用及二次涡．得出了进气道内外流场的马赫数、压力和速度分布，以及进气道的性能随反压和攻角的变化曲线。研究结果可为进气道的设计提供一定的参考。     

建筑物风环境三维定常数值仿真

针对建筑物的风环境问题，采用计算流体动力学CFD软件对两幢建筑物进行了数值模拟，得出了建筑物的风场分布，以及风载的分布规律，结果表明，气流在来流方向第一幢建筑物的顶部、两侧的前沿产生分离，建筑物之间产生强烈的漩涡，顶部的局部负压最大。     

基于模型降阶的随机最优输出反馈阵设计

本文研究了线性离散随机系统的模型降阶和最优输出反馈矩阵的设计问题。本文提出的设计方法不需要对原系统进行Anderson-Moore的迭代计算,直接利用求得的降阶模型解出近似的最优输出反馈阵。文中举例说明了设计方法的应用并进行了计算机仿真,仿真结果是令人满意的。     

Gravity wave momentum flux generation close to mid-latitude Andes in mesoscale simulations of late 20th and 21st centuries

Adequate representations of diverse dynamical processes in general circulation models (GCM) are necessary to obtain reliable simulations of the present and the future. The parameterization of orographic gravity wave drag (GWD) is one of the critical components of GCM. It is therefore convenient to evaluate whether standard orographic GWD parameterizations are appropriate. One alternative is to study the generation of gravity waves (GW) with horizontal resolutions that are higher than those used in current GCM simulations. Here we assess the seasonal pattern of topographic GW momentum flux (GWMF) generation for the late 20th and 21st centuries in a downscaling using the Rossby Centre regional atmospheric model under the Intergovernmental Panel on Climate Change A1B emission conditions. We focus on one of the world’s strongest extra-tropical GW zones, the Andes Mountains at mid-latitudes in the Southern Hemisphere. The presence of two GCM sub-grid scale structures locally contributing to GWMF (one positive and one negative) is found to the East of the mountains. For the late 21st century the strength of these structures during the GW high season increases around 23% with respect to the late 20th century, but the GWMF average over GCM grid cell scales remains negative and nearly constant around −0.015 Pa. This constitutes a steady significant contribution during GW high season, which is not related to the GWMF released by individual sporadic strong GW events. This characteristic agrees with the fact that no statistically significant variation in GWMF at source level has been observed in recent GCM simulations of atmospheric change induced by increases in greenhouse gases.     

倾转四旋翼飞行器垂直飞行状态气动特性

综合采用基于滑移网格技术的计算流体力学(CFD)方法与悬停状态气动干扰试验方法,对倾转四旋翼(QTR)飞行器垂直飞行状态的流场进行模拟与试验,研究飞行器垂直飞行状态气动特性以及部分参数对气动特性的影响.结果表明:倾转四旋翼飞行器在垂直飞行状态,前后旋翼之间干扰不明显,但旋翼与机翼的干扰明显;旋翼旋向对旋翼与机翼的干扰不...     

一种基于求解椭圆型方程的结构动网格生成方法

基于椭圆型网格生成法,实现了一种简单高效的贴体结构动网格生成方法,可用于具有移动边界问题的非定常流动数值模拟。该方法提出,在网格变形过程中,Poisson方程需要的控制网格间距和正交性的源项可以通过提取已知的静态网格源项直接得到,并在整个动网格生成过程中保持不变。因此,在椭圆型网格生成中需要通过外迭代确定源项的过程可以得到省略,而且该方法不需要人工指定参数。这使得方法具有高效和易于嵌入到已有程序中的特点。数值模拟结果证明,采用这种方法获得的网格能够较好地保持静态网格原有的正交性和光滑性,在相同迭代步数约束下,网格求解效率低于传统弹簧模拟法,但鲁棒性优于弹簧模拟法。     

旋翼桨尖涡生成及演化机理的高精度数值研究

为了细致捕捉直升机旋翼桨尖涡的生成和演化过程,建立了一个基于高精度网格和高阶通量计算格式的旋翼桨尖涡计算流体力学(CFD)求解方法。在该方法中,流场求解选取旋转坐标系下的Navier-Stokes方程为控制方程;空间离散采用迎风Roe格式,并采用低耗散的5阶WENO(Weighted Essentially Non-Osciltatory)格式进行对流通量的计算;时间推进则采用双时间法,在伪时间步上使用隐式LU-SGS(Lower-Upper Symmetric Gauss-Seidel)推进格式;应用嵌套网格方法实现桨叶网格和背景网格的数据交换。应用所建立的方法对悬停状态的旋翼桨尖涡流场进行了高精度模拟,在桨叶网格上精细地捕捉到了桨尖涡的具体生成过程,在背景网格上捕捉到了更多圈数的桨尖涡尾迹,并对桨尖涡的演化机理进行了深入研究。结果表明:建立的高精度数值方法能够有效地对旋翼桨尖涡的生成和演化过程进行细致模拟;悬停状态下旋翼桨尖气流在上下表面压力梯度的作用下经历了边界层增厚、逐渐卷起形成涡核以及最终脱离桨叶形成桨尖涡的过程。     

进气道堵盖打开过程数值模拟

为获得整体式固冲发动机转级过程中进气道堵盖(包括入口堵盖与出口堵盖)打开过程的流动形态,建立了进气道二维模型,利用Fluent动网格技术和UDF方法,开展了进气道堵盖打开过程非稳态流场研究。结果表明,在入口堵盖打开前,进气道前端形成强烈的弓形激波;在入口打开、出口未开的过程中,沿进气道轴向各监测点压强呈现周期性变化,振荡频率为100 Hz左右,出口堵盖位置压强振荡幅值为0.53 MPa;在出口打开后,补燃室残余热量形成的压强峰导致进气道在短时间内无法起动,随着背压降低至小于进气道再起动反压,进气道完全起动。     

Numerical Simulation of Rotor Flow Field Based on Overset Grids and Several Spatial and Temporal Discretization Schemes

A numerical method based on solutions of Euler/Navier-Stokes (N-S) equations is developed for calculating the flow field over a rotor in hover. Jameson central scheme, van Leer flux-vector splitting scheme, advection upwind splitting method (AUSM) scheme, upwind AUSM/van Leer scheme, AUSM+ scheme and AUSMDV scheme are implemented for spatial discretization, and van Albada limiter is also applied. For temporal discretization, both explicit Runge-Kutta method and implicit lower-upper symmetric Gauss-Seidel (LU-SGS) method are attempted. Simultaneously, overset grid technique is adopted. In detail, hole-map method is utilized to identify intergrid boundary points (IGBPs). Furthermore, aimed at identification issue of donor elements, inverse-map method is implemented. Eventually, blade surface pressure distributions derived from numerical simulation are validated compared with experimental data, showing that all the schemes mentioned above have the capability to predict the rotor flow field accurately. At the same time, vorticity contours are illustrated for analysis, and other characteristics are also analyzed.