基于混合网格Navier-Stokes方程的并行隐式计算方法研究

针对结构网格很难处理复杂外形和非结构网格无法计算具有边界层的粘性流动的缺点,发展了基于混合网格格点的隐式算法,成功地解决了在工程应用中难于处理的复杂外形粘性流场计算和效率问题。同时针对大规模的工程问题,发展了基于MPI通信技术的染色分层通讯并行计算方法。其中空间离散采用基于Roe格式发展的三阶迎风HLLEW(Harten-Lax-Van Leer-Einfeldt-Wada)或AUSM格式,湍流模型采用k??两方程湍流模型,时间推进考虑到LU-SGS并行等效较困难则采用基于DP-LUR(Data-Parallel Lower-Upper Relaxation)格式的隐式算法,计算CFL数可取到105量级,从2个到128个CPU的并行加速效率都保持在90%以上,大大提高了计算效率。算例对标模M6机翼模型流场进行计算,验证了方法的可靠性;然后对标模DLR-F6翼身组合体进行混合网格粘性与无粘计算结果进行比较,进一步验证混合网格方法;最后计算了DLR-WBNP外挂发动机翼身组合体模型,准确模拟了外挂和超临界机翼的相互干扰流动问题,采用4 CPU 16 CORE到24 CPU 96 CORE,2000步计算时间都不超...     

CFD／CSD紧耦合及新型动网格方法在气动弹性模拟中的应用

研发出一套基于紧耦合的CFD／CSD耦合方法和程序。非定常流场求解采用混合网格有限体积方法,时间离散采用基于LU-SGS隐式格式的双时间步长法。通过求解雷诺平均Navier—Stokes方程模拟了三维机翼的跨音速气动弹性现象。得到了其颤振边界,与风洞实验结果吻合较好,验证了方法和程序的有效性和实用意义。同时将Delau...     

基于贴体网格的VOF方法数模流场研究

提出了一种基于VOF方法的模拟具有复杂边界形状结构物附近流场的新算法,BFC—SIMPLE—VOF算法。采用坐标变换方法实现了任意复杂区域的结构化网格划分,在贴体网格下对二维不可压缩粘性流体的控制方程进行了离散。提出了基于交错网格的修正SIMPLE算法来迭代求解压力一速度场,修正了贴体坐标下的界面跟踪方法（VOF方法）...     

稳定分层二层流室内实验早期格栅湍流结构

采用室内实验混合箱和粒子图像测速技术,本文研究了稳定分层无平均剪切二层流(上层淡水、下层盐水)振动湍流结构。对实验录像进行粒子图像测速技术处理,获得垂向二维流场(垂直于格栅平面)瞬时速度和涡量,并用于计算:①时均速度和时均涡量;②均方根速度;③均匀程度和各向同性程度;④平均流强度;⑤时均泰勒的欧拉积分长度尺度;⑥时均湍动能和时均湍动能垂向通量;⑦水平和垂向速度的欧拉频谱。结果显示:(1)格栅方棒处时间平均速度方向垂直向上,而其两侧的时间平均涡量正负交替,表明格栅附近射流结构占据主要位置且存在反向涡旋对。(2)均方根速度随着距离格栅(水平面)高程的增大而减小,并且满足高程的-1.425(接近-3/2)幂次律,表明格栅湍流均方根速度的垂向变化较为剧烈。(3)靠近混合箱边壁处的均匀程度和各向同性程度都大于1,表明靠近混合箱边壁处存在各向异性湍流。(4)格栅反湍流强度的量级非0,但是,相对较小,表明平均流强度较低,故而,本实验结果仍可与无平均流的情况作对比。(5)时均泰勒的欧拉积分长度尺度随着距离格栅(水平面)高程的增大而线性增大,表明随着湍流向上发展,涡的平均尺度增大。(6)时均湍动能和时均湍动能垂向通量随着距离格栅(水平面)高程的增大而减小,而时均湍动能垂向通量为正值,表明远离格栅时均湍动能衰减,但是,始终向上传递。(7)水平和垂向速度的欧拉频谱随着距离格栅(水平面)高程的增大而减小,幂次律介于ω^-1和ω^-5/3(ω为频率)之间,表明水平和垂向的湍流脉动能量远离格栅均衰减,并受湍流的有限雷诺数效应的影响。     

谱体积方法单元分割与问题单元标记方法的改进研究

谱体积方法是一种本质上解决网格依赖性的高精度CFD计算方法,本文研究了二维Euler方程的谱体积方法,提出一种基于切比雪夫多项式的单元分割方法,建立了基于WENO的变量限制器方法,并发展了结合谱体积和控制体的问题单元标记方法.采用15°超声速压缩拐角和NACA0012跨声速流动两个典型算例进行验证,结果表明,该分区方法具有更好的计算精度,标记方法可有效识别不连续区域,在较少的网格下即可获得与密网格传统有限体积法相当的计算精度.     

Space-filling curves for 2-simplicial meshes created with bisections and reflections

Numerical experiments in J. Maubach: Local bisection refinement and optimal order algebraic multilevel preconditioners, PRISM-97 conference Proceedings, 1977, 121–136 indicated that the refinement with the use of local bisections presented in J. Maubach: Local bisection refinement for n-simplicial grids generated by reflections, SIAM J. Sci. Comput. 16 (1995), 210–227 leads to highly locally refined computational 2-meshes which can be very efficiently load-balanced with the use of a space-filling curve. This paper introduces the construction of this curve which can be produced at almost no costs, proofs that all its properties are invariant under local bisection, and comments on the 3-dimensional case.With the use of a space-filling curve (which passes through all triangular elements), load balancing over several processors is trivial: The load can be distributed over N processors by cutting the curve into N almost equilength parts. Each processor then operates on the triangles which are passed by its part of the curve.     

Numerical simulation of axisymmetric unsteady incompressible flow by a vorticity-velocity method

A new numerical method for solving the axisymmetric unsteady incompressible Navier-Stokes equations using vorticity-velocity variables and a staggered grid is presented. The solution is advanced in time with an explicit two-stage Runge-Kutta method. At each stage a vector Poisson equation for velocity is solved. Some important aspects of staggering of the variable location, divergence-free correction to the velocity field by means of a suitably chosen scalar potential and numerical treatment of the vorticity boundary condition are examined. The axisymmetric spherical Couette flow between two concentric differentially rotating spheres is computed as an initial value problem. Comparison of the computational results using a staggered grid with those using a non-staggered grid shows that the staggered grid is superior to the non-staggered grid. The computed scenario of the transition from zero-vortex to two-vortex flow at moderate Reynolds number agrees with that simulated using a pseudospectral method, thus validating the temporal accuracy of our method.     

Comparison of adjoint‐based and feature‐based grid adaptation for functional outputs

Adjoint‐based and feature‐based grid adaptive strategies are compared for their robustness and effectiveness in improving the accuracy of functional outputs such as lift and drag coefficients. The output‐based adjoint approach strives to improve the adjoint error estimates that relate the local residual errors to the global error in an output function via adjoint variables as weight functions. A conservative adaptive indicator that takes into account the residual errors in both the primal (flow) and dual (adjoint) solutions is implemented for the adjoint approach. The physics‐based feature approach strives to identify and resolve significant features of the flow to improve functional accuracy. Adaptive indicators that represent expansions and compressions in the flow direction and gradients normal to the flow direction are implemented for the feature approach. The adaptive approaches are compared for functional outputs of three‐dimensional arbitrary Mach number flow simulations on mixed‐element unstructured meshes. Grid adaptation is performed with h‐refinement and results are presented for inviscid, laminar and turbulent flows. Copyright © 2006 John Wiley & Sons, Ltd.     

A sliding interface method for unsteady unstructured flow simulations

The objective of this work is to develop a sliding interface method for simulations involving relative grid motion that is fast and efficient and involves no grid deformation, remeshing, or hole cutting. The method is implemented into a parallel, node‐centred finite volume, unstructured viscous flow solver. The rotational motion is accomplished by rigidly rotating the subdomain representing the moving component. At the subdomain interface boundary, the faces along the interface are extruded into the adjacent subdomain to create new volume elements forming a one‐cell overlap. These new volume elements are used to compute a flux across the subdomain interface. An interface flux is computed independently for each subdomain. The values of the solution variables and other quantities for the nodes created by the extrusion process are determined by linear interpolation. The extrusion is done so that the interpolation will maintain information as localized as possible. The grid on the interface surface is arbitrary. The boundary between the two subdomains is completely independent from one another; meaning that they do not have to connect in a one‐to‐one manner and no symmetry or pattern restrictions are placed on the grid. A variety of numerical simulations were performed on model problems and large‐scale applications to examine conservation of the interface flux. Overall solution errors were found to be comparable to that for fully connected and fully conservative simulations. Excellent agreement is obtained with theoretical results and results from other solution methodologies. Copyright © 2006 John Wiley & Sons, Ltd.     

Numerical simulation of fluid flows using an unstructured finite volume method with adaptive tri‐tree grids

A tri‐tree grid generation procedure is developed together with a finite volume method on the unstructured grid for solving the Navier–Stokes equations. A hierarchic numbering system for the data structure is used. The grid is adapted by adding and removing cell elements dependent on the vorticity magnitude. A special treatment is developed to ensure good quality triangular elements around the cylinder boundary. The adopted finite volume method is based on the cell‐centred scheme. The pressure–velocity coupling is treated using the SIMPLE algorithm. A modified QUICK scheme for unstructured grids is derived. The developed method is used to simulate the flow past a single and multiple cylinders at low Reynolds number. The obtained results are in good agreement with the published data. Copyright © 2002 John Wiley & Sons, Ltd.