High-order discontinuous Galerkin solver on hybrid anisotropic meshes for laminar and turbulent simulations

Efficient and robust solution strategies are developed for discontinuous Galerkin （DG） discretization of the Navier-Stokes （NS） and Reynolds-averaged NS （RANS） equations on structured/unstructured hybrid meshes. A novel line-implicit scheme is devised and implemented to reduce the memory gain and improve the computational eificiency for highly anisotropic meshes. A simple and effective technique to use the mod- ified Baldwin-Lomax （BL） model on the unstructured meshes for the DC methods is proposed. The compact Hermite weighted essentially non-oscillatory （HWENO） limiters are also investigated for the hybrid meshes to treat solution discontinuities. A variety of compressible viscous flows are performed to examine the capability of the present high- order DG solver. Numerical results indicate that the designed line-implicit algorithms exhibit weak dependence on the cell aspect-ratio as well as the discretization order. The accuracy and robustness of the proposed approaches are demonstrated by capturing com- plex flow structures and giving reliable predictions of benchmark turbulent problems.     

基于非结构/混合网格模拟黏性流的高阶精度DDG/FV混合方法

间断Galerkin有限元方法(discontinuous Galerkin method, DGM) 因具有计算精度高、模板紧致、易于并行等优点, 近年来已成为非结构/混合网格上广泛研究的高阶精度数值方法. 但其计算量和内存需求量巨大, 特别是对于网格规模达到百万甚至数千万的大型三维实际复杂外形问题, 其计算量和存储量对计算资源的消耗是难以承受的. 基于“混合重构”的DG/FV 格式可以有效降低DGM 的计算量和存储量. 本文将DDG 黏性项离散方法推广应用于DG/FV 混合算法, 得到新的DDG/FV混合格式, 以进一步提高DG/FV混合算法对于黏性流动模拟的计算效率. 通过Couette流动、层流平板边界层、定常圆柱绕流, 非定常圆柱绕流和NACA0012 翼型绕流等二维黏性流算例, 优化了DDG 通量公式中的参数选择, 验证了DDG/FV 混合格式对定常和非定常黏性流模拟的精度和计算效率, 并与广泛使用的BR2-DG 格式的计算结果和效率进行对比研究. 一系列数值实验结果表明, 本文构造的DDG/FV混合格式在二维非结构/混合网格的Navier-Stokes 方程求解中, 在达到相同的数值精度阶的前提下, 相比BR2-DG格式, 对于隐式时间离散的定常问题计算效率提高了2 倍以上, 对于显式时间离散的非定常问题计算效率提高1.6 倍, 并且在一些算例中, 混合格式具有更优良的计算稳定性. DDG/FV 混合格式提升了计算效率和稳定性, 具有良好的应用前景.      

A comparison of stability computational methods for periodic solution of nonlinear problems with application to rotordynamics

In this paper a comparative study of five different stability computational methods based on the Floquet theory is presented. These methods are compared in terms of accuracy and CPU performance. Tests are performed on a set of nonlinear problems relevant to rotating machinery with rotor-to-stator contact and a variable number of degrees of freedom, whose periodic solutions are computed with the Harmonic Balance Method (HBM).     

ADER discontinuous Galerkin schemes for aeroacoustics

In this paper we apply the ADER approach to the Discontinuous Galerkin (DG) framework for the two-dimensional linearized Euler equations. The result is an efficient high order accurate single-step scheme in time which uses less storage than Runge–Kutta DG schemes, especially for very high order of accuracy. The aim is to obtain an arbitrarily accurate scheme in space and time on unstructured grids for accurate noise propagation in the time domain in very complex geometries. We will present numerical convergence rates for ADER-DG methods up to 10th order of accuracy in space and time on structured and unstructured meshes. To cite this article: M. Dumbser, C.-D. Munz, C. R. Mecanique 333 (2005).     

弹性动力学高阶核无关快速多极边界元法

基于核无关的快速多极方法, 发展了一种弹性动力学问题的快速、高精度边界元分析方法. 采用基于二次曲面单元的Nystr?m 离散, 将边界积分方程转化为求和形式, 可以方便地进行加速计算;由于采用二次元, 边界元分析精度很高. 将一种新型快速多极方法用于Nystr?m 边界元法的加速计算, 该方法的数值实现简便、不依赖于积分方程基本解的表达式, 因此通用性很好;该方法还具有最优的计算量和存储量、精度高且可以控制. 结合Nystr?m 边界元系数矩阵和快速多极方法转换矩阵的特点, 提出一种大幅度降低边界元内存消耗的策略. 数值结果表明, 该方法无论在分析精度, 还是计算速度和内存消耗上, 都大大优于同类方法, 是一种快速、通用的工程弹性动力学问题大规模数值分析方法.      

A HIGH ORDER KERNEL INDEPENDENT FAST MULTIPOLE BOUNDARY ELEMENT METHOD FOR ELASTODYNAMICS

基于核无关的快速多极方法, 发展了一种弹性动力学问题的快速、高精度边界元分析方法. 采用基于二次曲面单元的Nyström 离散, 将边界积分方程转化为求和形式, 可以方便地进行加速计算;由于采用二次元, 边界元分析精度很高. 将一种新型快速多极方法用于Nyström 边界元法的加速计算, 该方法的数值实现简便、不依赖于积分方程基本解的表达式, 因此通用性很好;该方法还具有最优的计算量和存储量、精度高且可以控制. 结合Nyström 边界元系数矩阵和快速多极方法转换矩阵的特点, 提出一种大幅度降低边界元内存消耗的策略. 数值结果表明, 该方法无论在分析精度, 还是计算速度和内存消耗上, 都大大优于同类方法, 是一种快速、通用的工程弹性动力学问题大规模数值分析方法.     

Parallelization of the ILU(0) preconditioner for CFD problems on shared‐memory computers

The use of ILU(0) factorization as a preconditioner is quite frequent when solving linear systems of CFD computations. This is because of its efficiency and moderate memory requirements. For a small number of processors, this preconditioner, parallelized through coloring methods, shows little savings when compared with a sequential one using adequate reordering of the unknowns. Level scheduling techniques are applied to obtain the same preconditioning efficiency as in a sequential case, while taking advantage of parallelism through block algorithms. Numerical results obtained from the parallel solution of the compressible Navier–Stokes equations show that this technique gives interesting savings in computational times on a small number of processors of shared‐memory computers. In addition, it does this while keeping all the benefits of an ILU(0) factorization with an adequate reordering of the unknowns, and without the loss of efficiency of factorization associated with a more scalable coloring strategy. Copyright © 1999 John Wiley & Sons, Ltd.     

A divergence-conforming hybridized discontinuous Galerkin method for the incompressible Reynolds-averaged Navier-Stokes equations

We present a hybridized discontinuous Galerkin (HDG) method for the incompressible Reynolds-averaged Navier-Stokes equations coupled with the Spalart-Allmaras one-equation turbulence model. The method extends upon an HDG method recently introduced by Rhebergen and Wells for the incompressible Navier-Stokes equations. With a special choice of velocity and pressure spaces for both element and trace degrees of freedom (DOFs), the method returns pointwise divergence-free mean velocity fields and properly balances momentum and energy. We further examine the use of different polynomial degrees and meshes to see how the order of the scalar eddy viscosity affects the convergence of the mean velocity and pressure fields, specifically for the method of manufactured solutions. As is standard with HDG methods, static condensation can be employed to remove the element DOFs and thus dramatically reduce the global number of DOFs. Numerical results illustrate the effectiveness of the proposed methodology.     

基于静动态混合重构的DG/FV混合格式

通过比较紧致格式和间断Galerkin(DG)格式, 提出了``静态重构'和``动态重构'的概念,对有限体积方法和DG有限元方法进行统一的表述. 借鉴有限体积的思想, 发展了基于``混合重构'技术的一类新的DG格式, 称之为间断Galerkin有限元/有限体积混合格式(DG/FV格式). 该类混合格式通过适当地扩展模板(拓展至紧邻单元)重构单元内的高阶多项式分布, 在提高精度的同时, 减少了传统DG格式的计算量和存储量. 通过典型一维和二维标量方程的计算发现新的混合格式在有些情况下具有超收敛(superconvergence)性质.      

Sparse element for shallow water wave equations on triangle meshes

Within the mixed FEM, the mini‐element that uses a bubble shape function for the solution of the shallow water wave equations on triangle meshes is simplified to a sparse element formulation. The new formulation has linear shape functions for water levels and constant shape functions for velocities inside each element. The suppression of decoupled spurious solutions is excellent with the new scheme. The linear dispersion relation of the new element has similar advantages as that of the wave equation scheme (generalised wave continuity scheme) proposed by Lynch and Gray. It is shown that the relation is monotonic over all wave numbers. In this paper, the time stepping scheme is included in the dispersion analysis. In case of a combined space–time staggering, the dispersion relation can be improved for the shortest waves. The sparse element is applied in the flow model Bubble that conserves mass exactly. At the same time, because of the limited number of degrees of freedom, the computational efficiency is high. The scheme is not restricted to orthogonal triangular meshes. Three test cases demonstrate the very good accuracy of the proposed scheme. The examples are the classical quarter annulus test case for the linearised shallow water equations, the hydraulic jump and the tide in the Elbe river mouth. Copyright © 2013 John Wiley & Sons, Ltd.     

Evaluation of discontinuous Galerkin and spectral volume methods for scalar and system conservation laws on unstructured grids

The discontinuous Galerkin (DG) and spectral volume (SV) methods are two recently developed high‐order methods for hyperbolic conservation laws capable of handling unstructured grids. In this paper, their overall performance in terms of efficiency, accuracy and memory requirement is evaluated using a 2D scalar conservation laws and the 2D Euler equations. To measure their accuracy, problems with analytical solutions are used. Both methods are also used to solve problems with strong discontinuities to test their ability in discontinuity capturing. Both the DG and SV methods are capable of achieving their formal order of accuracy while the DG method has a lower error magnitude and takes more memory. They are also similar in efficiency. The SV method appears to have a higher resolution for discontinuities because the data limiting can be done at the sub‐element level. Copyright © 2004 John Wiley & Sons, Ltd.     

High-order Finite Difference and Finite Volume WENO Schemes and Discontinuous Galerkin Methods for CFD

In recent years, high order numerical methods have been widely used in computational fluid dynamics (CFD), to effectively resolve complex flow features using meshes which are reasonable for today's computers. In this paper, we review and compare three types of high order methods being used in CFD, namely the weighted essentially non-oscillatory (WENO) finite difference methods, the WENO finite volume methods, and the discontinuous Galerkin (DG) finite element methods. We summarize the main features of these methods, from a practical user's point of view, indicate their applicability and relative strength, and show a few selected numerical examples to demonstrate their performance on illustrative model CFD problems.     

基于图像四叉树的改进型比例边界有限元法研究

为提高数值计算的精度,断裂力学问题的数值模拟需要在裂纹扩展的局部区域采用较密的网格,而远离裂纹扩展的区域可采用较疏的网格,且对于裂纹扩展问题的数值模拟,大多数数值方法又存在局部网格重剖分的问题.论文提出了一种基于图像四叉树的改进型比例边界有限元法用于模拟裂纹扩展问题,该方法可根据结构域几何外边界的图像全自动进行四叉树网格剖分,无需任何人工干预,网格剖分效率极高,由于比例边界有限元法本身的优势,四叉树网格的悬挂节点可以直接地视为新的节点,无需任何特殊处理.通过引入虚节点的思想,将裂纹与四叉树单元边界交叉点作为虚节点,虚节点的自由度作为附加自由度处理,并采用水平集函数表征材料内部的裂纹面,含不连续裂纹面的子域可通过节点水平集函数识别,使得裂纹扩展时无需进行网格重剖分,界面的几何特征通过比例边界有限元子域的附加自由度表征.最后,通过若干算例验证了该方法的性能,建议的改进型比例边界有限元法在求解复合型应力强度因子和模拟材料内部裂纹扩展路径时均具有较高的精度.     

加筋路径驱动的板壳自适应等几何屈曲分析

加筋薄壁结构常被用于航空航天结构的轻量化设计.随着结构尺寸和几何特征的增加,需要更加精细的网格来满足分析精度的要求.传统的等几何方法采用NURBS张量积形式的拓扑结构,使得在分析过程中难以实现局部细化,而全局细化则会增加不必要的自由度.为了提升加筋板壳结构的数值分析精度和效率,提出一种基于RPHT (rational polynomial splines over hierarchical T-meshes)样条的加筋板壳自适应等几何屈曲分析方法.样条网格可以沿着加筋路径进行自适应的局部细化,有效提升低自由度下加筋板壳结构等几何屈曲分析的精度.首先,蒙皮和筋条分别采用RPHT样条曲面和NURBS样条曲线进行建模,几何建模与数值仿真采用统一的几何语言,实现建模与分析的一体化.其次,采用几何投影算法和样条插值算法实现筋条与蒙皮之间的高效高精度强耦合,并建立基于加筋路径驱动自适应网格细化方法.最后,曲线加筋板和网格加筋壳两个算例验证本方法的高效性和鲁棒性,通过与基于NURBS的等几何分析进行对比,本方法能够明显降低分析模型的总自由度.     

Higher‐order and adaptive discontinuous Galerkin methods with shock‐capturing applied to transonic turbulent delta wing flow

Discontinuous Galerkin (DG) methods allow high‐order flow solutions on unstructured or locally refined meshes by increasing the polynomial degree and using curved instead of straight‐sided elements. DG discretizations with higher polynomial degrees must, however, be stabilized in the vicinity of discontinuities of flow solutions such as shocks. In this article, we device a consistent shock‐capturing method for the Reynolds‐averaged Navier–Stokes and k‐ ω turbulence model equations based on an artificial viscosity term that depends on element residual terms. Furthermore, the DG method is combined with a residual‐based adaptation algorithm that targets at resolving all flow features. The higher‐order and adaptive DG method is applied to a fully turbulent transonic flow around the second Vortex Flow Experiment (VFE‐2) configuration with a good resolution of the vortex system.Copyright © 2013 John Wiley & Sons, Ltd.     

基于等几何分析的边界元法求解二维Laplace方程

针对二维Laplace问题,提出了基于非均匀有理B样条的等几何边界单元法(IGABEM),并利用径向积分法来处理奇异积分。该方法实现了几何与求解域的无缝融合,不仅实现了求解域与几何的完美匹配,而且节约了前处理时间。该方法可以很容易地实现模型的细分,并且在仅增加少量自由度的情况下获得更高的精度。数值算例表明,该方法能够有效地求解二维Laplace方程,且具有非常好的计算精度。     

Conforming finite element computations applied to a spatially periodic,harmonic Navier-Stokes problem

In this paper computations in the two dimensional case of a harmonic Navier-Stokes problem with periodic boundary conditions are presented. This study of an incompressible viscous fluid leads to a non-symmetric linear problem (very low Reynolds number). Moreover unknown functions have complex values (monochromatic dynamic behaviour). Numerical treatment of the incompressibility condition is a generalization of the classical treatment of Stokes problem. A mixed formulation, where discrete pressure plays the role of Lagrange multipliers is used (Uzawa algorithm). Two conforming finite element methods are tested on different meshes. The second one uses a classical refinement in the shape function: the so-called bulb function. All computational tests show that the use of a bulb function on each element gives better results than refinement in the mesh without introducing too many degrees of freedom. Finally numerical results are compared to experimental data.     

A solution-adaptive mesh procedure for predicting confined explosions

Explosion hazards constitute a significant practical problem for industry. In response to the need for better-resolved predictions for confined explosions, and particularly with a view to advancing safety cases for offshore oil and gas rigs, an existing unstructured, adaptive mesh, finite volume Reynolds-averaged Navier–Stokes computational fluid dynamics code (originally developed to handle non-combusting turbomachinery flows) has been modified to include a one-equation, eddy break-up combustion model. Two benefits accrue from the use of unstructured, solution-adaptive meshes: first, great geometrical flexibility is possible; second, automatic mesh adaptation allows computational effort to be focused on important or interesting areas of the flow by enhancing mesh resolution only where it is required. In the work reported here, the mesh was adaptively refined to achieve flame front capture, and it is shown that this results in a 10%–33% CPU saving for two-dimensional calculations and a saving of between 57% and 70% for three-dimensional calculations. The geometry of the three-dimensional calculations was relatively simple, and it may be expected that the use of unstructured meshes for truly complex geometries will result in CPU savings sufficient to allow an order-of-magnitude increase in either complexity or resolution. © 1998 John Wiley & Sons, Ltd.     

A high‐order discontinuous Galerkin method for all‐speed flows

In this article, we present a discontinuous Galerkin (DG) method designed to improve the accuracy and efficiency of steady solutions of the compressible fully coupled Reynolds‐averaged Navier–Stokes and k ? ω turbulence model equations for solving all‐speed flows. The system of equations is iterated to steady state by means of an implicit scheme. The DG solution is extended to the incompressible limit by implementing a low Mach number preconditioning technique. A full preconditioning approach is adopted, which modifies both the unsteady terms of the governing equations and the dissipative term of the numerical flux function by means of a new preconditioner, on the basis of a modified version of Turkel's preconditioning matrix. At sonic speed the preconditioner reduces to the identity matrix thus recovering the non‐preconditioned DG discretization. An artificial viscosity term is added to the DG discretized equations to stabilize the solution in the presence of shocks when piecewise approximations of order of accuracy higher than 1 are used. Moreover, several rescaling techniques are implemented in order to overcome ill‐conditioning problems that, in addition to the low Mach number stiffness, can limit the performance of the flow solver. These approaches, through a proper manipulation of the governing equations, reduce unbalances between residuals as a result of the dependence on the size of elements in the computational mesh and because of the inherent differences between turbulent and mean‐flow variables, influencing both the evolution of the Courant Friedrichs Lewy (CFL) number and the inexact solution of the linear systems. The performance of the method is demonstrated by solving three turbulent aerodynamic test cases: the flat plate, the L1T2 high‐lift configuration and the RAE2822 airfoil (Case 9). The computations are performed at different Mach numbers using various degrees of polynomial approximations to analyze the influence of the proposed numerical strategies on the accuracy, efficiency and robustness of a high‐order DG solver at different flow regimes. Copyright © 2014 John Wiley & Sons, Ltd.