非定常扩散方程的单元中心型有限体积格式

本文基于调和平均点建立了一种新的单元中心型有限体积格式,用以求解非定常扩散方程.在网格边上离散法向流时,选择该网格边两端点和该边上的一个调和平均点作为辅助插值点,并将这些辅助插值点上的未知量用网格单元中心点的未知量进行替换,最终得到一个只含网格单元中心未知量的有限体积格式.该格式满足线性精确性质和局部守恒性,且适用于任意多边形网格.在六种不同的多边形网格上进行四个数值实验,分别考虑扩散系数是连续的和间断的以及非线性的情况,数值结果表明:本文所构造的格式在六种网格上的L2误差均可达到二阶收敛精度,对于不同类型的扩散系数,该格式保持良好的鲁棒性,并且从编程实现的角度来说,该格式更易于向三维情况推广.     

扩散方程保正的有限体积格式

在大变形网格上数值求解多介质扩散方程时, 如何构造具有保正性的扩散格式一直是人们关注的难题. 本文将简要综述与保正性相关的扩散格式的研究历史, 并为解决这一难题提出新的设计途径,构造出新的具有较高精度的单元中心型守恒保正格式, 它们可兼顾网格几何变形和物理量变化. 本文将给出数值实验结果, 验证新格式在变形的网格上保持非负性.     

三维稳态对流扩散问题的无网格求解算法研究

无网格法是一种不需要生成网格就可模拟复杂形状流场计算的流体力学问题求解算法.为了提高基于Galerkin弱积分形式的无网格方法求解三维稳态对流扩散问题的计算效率,提出了在空间离散上采用基于凸多面体节点影响域的无网格形函数,并通过选取适当节点影响半径因子避免节点搜索问题,同时减少系统刚度矩阵带宽.计算中当节点影响因子为1.01时,无网格方法的形函数近似具有插值特性且本质边界条件的施加与有限元一样简单.三维立方体区域的稳态对流扩散数值算例表明:在保证计算精度的同时,采用凸多面体节点影响域的无网格方法比传统无网格方法最高可节省计算时间42%.因此从计算效率和精度考虑,在运用无网格方法求解三维问题时建议采用凸多面体节点影响域的无网格方法.     

多边形网格上扩散方程新的单调格式

 本文在星形多边形网格上, 构造了扩散方程新的单调有限体积格式.该格式与现有的基于非线性两点流的单调格式的主要区别是, 在网格边的法向流离散模板中包含当前边上的点, 在推导离散法向流的表达式时采用了定义于当前边上的辅助未知量, 这样既可适应网格几何大变形, 同时又兼顾了当前网格边上物理量的变化. 在光滑解情形证明了离散法向流的相容性.对于具有强各向异性、非均匀张量扩散系数的扩散方程, 证明了新格式是单调的, 即格式可以保持解析解的正性. 数值结果表明在扭曲网格上, 所构造的格式是局部守恒和保正的, 对光滑解有高于一阶的精度, 并且, 针对非平衡辐射限流扩散问题, 数值结果验证了新格式在计算效率和守恒精度上优于九点格式.     

扩散方程九点格式中节点未知量的一种新的插值算法

本文研究了二维扩散方程九点格式中节点辅助未知量的插值问题.利用多点通量逼近的边未知量插值算法和一个特殊的极限技巧,获得了节点辅助未知量的一个新的插值算法,并在给定假设下严格分析了该算法中局部线性系统的可解性.新算法满足线性精确准则,具有较高的精度.     

非定常对流扩散方程保正格式解的存在性

本文发展了非定常对流扩散方程的非线性保正格式.该格式为单元中心型有限体积格式,保持局部通量的守恒性,适用于任意星形多边形网格,本文证明了该离散格式解的存在性,并给出数值结果,表明该格式具有二阶精度.     

无网格广义有限差分法模拟三维位势问题

广义有限差分法是一种新型的无网格数值离散方法.该方法基于多元函数泰勒级数展开和加权最小二乘拟合,将控制方程中未知参量的各阶偏导数表示为相邻节点函数值的线性组合,克服了传统有限元等基于网格的方法对网格的依赖性.本文以三维位势问题为例,引入一种新的优化选点技术,克服了传统广义有限差分法在模拟三维复杂几何域问题时遇到的"病态选点问题",极大地提高了该方法的计算精度与数值稳定性.     

三维泊松方程的高精度多重网格解法

利用对称网格点泰勒展开式中各阶导数项明显的对称性,得到了数值求解三维泊松方程的四阶和六阶精度的紧致差分格式,其推导过程简便直接.为了克服传统迭代法在求解高维问题时计算量大、收敛速度慢的缺陷,采用了多重网格加速技术,设计了相应的多重网格算法,求解了三维泊松方程的Dirichlet边值问题.数值实验结果表明,本文所提出的高精度紧致格式达到了期望的精度并且多重网格方法的加速效果是非常显著的.     

Stokes流动的罚-杂交有限元分析*

本文建议了一种用于分析Stokes流动的罚-杂交变分原理,其中,偏应力张量和静水压力事先满足线动量平衡.建立了相应的有限元模型.由此,压力可在列式过程中消去,使得有限元矩阵方程仅以节点速度作为唯一的求解未知量.推导了几种4-节点和8-节点四边形单元.通过数值算例,显示了单元性能.     

射线跟踪技术中物体外形精确建模算法

为了对复杂物体外形进行快速精确建模,利用多面体样条方法进行物体表面的设计及应用研究.通过对样条函数节点进行分类,给出空间中任意拓扑结构网格上样条函数组的具体构造方法.研究曲线曲面控制点与生成样条函数的节点的相互依赖关系,提出在无单位分解条件下构造具有几何不变性曲线曲面的合理化方法.通过构造控制点处多面体样条函数,生成了任意拓扑结构的连续多面体样条曲面.这样的多面体样条曲面在同样性质要求下曲面次数低,只有张量积NURBS曲面的一半;并且与细分曲面相比,所获取的物体外形表面具有坐标计算准确,中间数据较少等优点,可以用于精确构造电磁射线追踪中物体的外形.     

Calculating the vertex unknowns of nine point scheme on quadrilateral meshes for diffusion equation

In the construction of nine point scheme,both vertex unknowns and cell-centered unknowns are introduced,and the vertex unknowns are usually eliminated by using the interpolation of neighboring cell-centered unknowns,which often leads to lose accuracy.Instead of using interpolation,here we propose a different method of calculating the vertex unknowns of nine point scheme,which are solved independently on a new generated mesh.This new mesh is a Vorono¨i mesh based on the vertexes of primary mesh and some additional points on the interface.The advantage of this method is that it is particularly suitable for solving diffusion problems with discontinuous coeffcients on highly distorted meshes,and it leads to a symmetric positive definite matrix.We prove that the method has first-order convergence on distorted meshes.Numerical experiments show that the method obtains nearly second-order accuracy on distorted meshes.     

A Vertex-Centered and Positivity-Preserving Finite Volume Scheme for Two-Dimensional Three-Temperature Radiation Diffusion Equations on General Polygonal Meshes

Two-dimensional three-temperature (2-D 3-T) radiation diffusion equations are widely used to approximately describe the evolution of radiation energy within a multimaterial system and explain the exchange of energy among electrons, ions and photons. In this paper, we suggest a new positivity-preserving finite volume scheme for 2-D 3-T radiation diffusion equations on general polygonal meshes. The vertex unknowns are treated as primary ones for which the finite volume equations are constructed. The edge-midpoint and cell-centered unknowns are used as auxiliary ones and interpolated by the primary unknowns, which makes the final scheme a pure vertex-centered one. By comparison, most existing positivity-preserving finite volume schemes are cell-centered and based on the convex decomposition of the co-normal. Here, the co-normal decomposition is not convex in general, leading to a fixed stencil of the flux approximation and avoiding a certain search algorithm on complex grids. Moreover, the new scheme effectively alleviates the numerical heat-barrier issue suffered by most existing cell-centered or hybrid schemes in solving strongly nonlinear radiation diffusion equations. Numerical experiments demonstrate the second-order accuracy and the positivity of the solution on various distorted grids. For the problem without analytic solution, the contours of the numerical solutions obtained by our scheme on distorted meshes accord with those on smooth quadrilateral meshes.     

A cell edge‐based linearity‐preserving scheme for diffusion problems on two‐dimensional unstructured grids

We propose a new finite volume scheme for 2D anisotropic diffusion problems on general unstructured meshes. The main feature lies in the introduction of two auxiliary unknowns on each cell edge, and then the scheme has both cell‐centered primary unknowns and cell edge‐based auxiliary unknowns. The auxiliary unknowns are interpolated by the multipoint flux approximation technique, which reduces the scheme to a completely cell‐centered one. The derivation of the scheme satisfies the linearity‐preserving criterion that requires that a discretization scheme should be exact on linear solutions. The resulting new scheme is then called as a cell edge‐based linearity‐preserving scheme. The optimal convergence rates are numerically obtained on unstructured grids in case that the diffusion tensor is taken to be anisotropic and/or discontinuous. Copyright © 2017 John Wiley & Sons, Ltd.     

A nine-point scheme with explicit weights for diffusion equations on distorted meshes

In this paper, for the structured quadrilateral mesh we derive a nine-point difference scheme which has five cell-centered unknowns and four vertex unknowns. The vertex unknowns are treated as intermediate ones and are expressed as a linear combination of the neighboring cell-centered unknowns, which reduces the scheme to a cell-centered one with a local stencil involving nine cell-centered unknowns. The coefficients in the linear combination are known as the weights and two types of new weights are proposed. These new weights are neither discontinuity dependent nor mesh topology dependent, have explicit expressions, can reduce to the one-dimensional harmonic-average weights on the nonuniform rectangular meshes, and moreover, are easily extended to the unstructured polygonal meshes and non-matching meshes. Both the derivation of the nine-point scheme and that of new weights satisfy the linearity preserving criterion. Numerical experiments show that, with these new weights, the nine-point difference scheme and its simple extension have a nearly second order accuracy on many highly distorted meshes, including structured quadrilateral meshes, unstructured polygonal meshes and non-matching meshes.     

A vertex‐centered and positivity‐preserving scheme for anisotropic diffusion equations on general polyhedral meshes

We propose a new nonlinear positivity‐preserving finite volume scheme for anisotropic diffusion problems on general polyhedral meshes with possibly nonplanar faces. The scheme is a vertex‐centered one where the edge‐centered, face‐centered, and cell‐centered unknowns are treated as auxiliary ones that can be computed by simple second‐order and positivity‐preserving interpolation algorithms. Different from most existing positivity‐preserving schemes, the presented scheme is based on a special nonlinear two‐point flux approximation that has a fixed stencil and does not require the convex decomposition of the co‐normal. More interesting is that the flux discretization is actually performed on a fixed tetrahedral subcell of the primary cell, which makes the scheme very easy to be implemented on polyhedral meshes with star‐shaped cells. Moreover, it is suitable for polyhedral meshes with nonplanar faces, and it does not suffer the so‐called numerical heat‐barrier issue. The truncation error is analyzed rigorously, while the Picard method and its Anderson acceleration are used for the solution of the resulting nonlinear system. Numerical experiments are also provided to demonstrate the second‐order accuracy and well positivity of the numerical solution for heterogeneous and anisotropic diffusion problems on severely distorted grids.     

A finite volume method on general meshes for a degenerate parabolic convection–reaction–diffusion equation

We propose a finite volume method on general meshes for the discretization of a degenerate parabolic convection–reaction–diffusion equation. Equations of this type arise in many contexts, such as for example the modeling of contaminant transport in porous media. The diffusion term, which can be anisotropic and heterogeneous, is discretized using a recently developed hybrid mimetic mixed framework. We construct a family of discretizations for the convection term, which uses the hybrid interface unknowns. We consider a wide range of unstructured possibly nonmatching polyhedral meshes in arbitrary space dimension. The scheme is fully implicit in time, it is locally conservative and robust with respect to the Péclet number. We obtain a convergence result based upon a priori estimates and the Fréchet–Kolmogorov compactness theorem. We implement the scheme both in two and three space dimensions and compare the numerical results obtained with the upwind and the centered discretizations of the convection term numerically.     

Monotone combined edge finite volume–finite element scheme for Anisotropic Keller–Segel model

In this article, a new numerical scheme for a degenerate Keller–Segel model with heterogeneous anisotropic tensors is treated. It is well‐known that standard finite volume scheme not permit to handle anisotropic diffusion without any restrictions on meshes. Therefore, a combined finite volume‐nonconforming finite element scheme is introduced, developed, and studied. The unknowns of this scheme are the values at the center of cell edges. Convergence of the approximate solution to the continuous solution is proved only supposing the shape regularity condition for the primal mesh. This scheme ensures the validity of the discrete maximum principle under the classical condition that all transmissibilities coefficients are positive. Therefore, a nonlinear technique is presented, as a correction of the diffusive flux, to provide a monotone scheme for general tensors. © 2014 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 30: 1030–1065, 2014     

Post processing of solution and flux for the nodal mimetic finite difference method

We develop and analyze a post processing technique for the family of low‐order mimetic discretizations based on vertex unknowns for the numerical treatment of diffusion problems on unstructured polygonal and polyhedral meshes. The post processing works in two steps. First, from the nodal degrees of freedom, we reconstruct an elemental‐based vector field that approximates the gradient of the exact solution. Second, we solve a local problem for each mesh vertex associated with a scheme degree of freedom to determine a post processed normal flux that is conservative and divergence preserving. Theoretical results and numerical experiments for two‐dimensional (2D) and 3D benchmark problems show optimal convergence rates. © 2014 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 31: 336–363, 2015     

Preconditioning analysis of the one dimensional incremental unknowns method on nonuniform meshes

The condition number of the incremental unknowns matrix on nonuniform meshes associated to the elliptic problem is analyzed. Comparing to the usual nodal unknowns matrix, the condition number of the incremental unknowns matrix is reduced significantly even if the meshes are nonuniform. Furthermore, if a diagonal scaling is used, the condition number of the preconditioned incremental unknowns matrix comes out to be O(1). Numerical experiments are performed respectively on Shishkin mesh and Chebyshev mesh. Computational results with respect to the two particular nonuniform meshes confirm our theoretical analysis.