U-RANS/PDF方法的并行计算研究

本文对非结构网格中U-RANS/PDF混合方法的并行计算进行了研究,通过对方柱绕流的数值模拟,讨论了并行程序的有效性和计算效率.非结构网格上U-RANS/PDF混合方法的并行计算,提高了PDF方法的计算效率和处理复杂几何结构的能力,使得PDF方法能够方便地应用于实际燃烧问题.     

八叉树网格到非结构混合网格的转换

在数值模拟中, 非结构网格的优势是可以采用相同的数值格式统一处理任意复杂的计算区域, 但在网格生成过程中难度大, 并且不容易控制网格质量。树结构网格可以认为是介于结构网格和非结构网格之间的一种网格, 目前已经有相对成熟的方法快速在复杂区域内生成二维四叉树网格和三维八叉树网格。在实际应用中, 数值方法往往需要在连接协调的非结构网格上做离散, 树结构网格中不同尺寸的网格之间连接不是协调的, 在应用上会受到很多限制。文章实现了树结构网格到非结构混合网格的转换, 这种转换在二维情况下就是将四叉树网格转换为非结构三角形和四边形的混合网格, 三维情况下则将八叉树网格转换为非结构混合网格。这一转换过程的难点在于需要考虑数千种不同的八叉树单元, 并给出能实现连接协调的非结构混合网格划分。可以出现的网格单元包括六面体、三棱柱、金字塔和四面体这4种不同情况。通过特别的分类, 实现了程序的自动生成, 这种程序自动生成技术一方面可以避免人工编写大量程序时的失误, 另一方面也使得对数以千计的不同情况的处理成为可能。通过对几个简单网格的测试, 对网格数据转换方法做了初步的验证。      

离心叶轮及无叶扩器内湍流的非结构化网格数值解法

利用非结构化网格上的有限体积法对三维湍流进行了数值研究,提出了不用求解单元顶点处变量值的二阶混合差分格式,既避免了复杂耗时的单元顶点处变量值的计算,又避免了普通差分格式在非结构化网格上易于引起网格界面方向相关性问题。最后利用非结构化网格计算程序,数值求解了一台离心风机内叶轮及无叶扩压器通道内三维相对定常湍流,计算结果与实验值的比较表明速度的计算值与实验值吻合较好。     

一种可扩展的三维半导体器件并行数值模拟算法

在基于漂移-扩散模型的三维半导体器件数值模拟中,通过有限体积法进行数值离散,采用完全耦合的牛顿迭代求解非线性代数方程组,并使用基于代数多重网格预条件子的GMRES方法求解牛顿迭代中的线性方程组,构造一种稳健且高度可扩展的非结构四面体网格上求解半导体方程的并行算法.基于PHG平台实现该算法的并行计算程序,并对PN结和MOS场效应晶体管等问题进行了最大网格规模达到5亿单元、最大并行规模达到1 024进程的大规模数值模拟实验,结果表明,该算法计算效率高,可扩展性好.     

非结构网格上大涡模拟数值方法研究

随着计算能力的飞速发展,在非结构网格上进行大涡模拟是研究叶轮机械真实几何结构下复杂流动问题的有效途径。本文针对非结构网格上数值格式人工黏性过大的缺点,利用湍流中动能级联特性,发展了适用于大涡模拟的低耗散数值格式,并结合亚格子模型,考察了其在基础湍流算例中的有效性。通过与传统迎风格式的对比,说明本文发展的低耗散格式可以显著提高非结构网格上大涡模拟的精度和可靠性。     

非结构网格上大涡模拟数值方法研究EI北大核心CSCD

随着计算能力的飞速发展,在非结构网格上进行大涡模拟是研究叶轮机械真实几何结构下复杂流动问题的有效途径。本文针对非结构网格上数值格式人工黏性过大的缺点,利用湍流中动能级联特性,发展了适用于大涡模拟的低耗散数值格式,并结合亚格子模型,考察了其在基础湍流算例中的有效性。通过与传统迎风格式的对比,说明本文发展的低耗散格式可以显著提高非结构网格上大涡模拟的精度和可靠性。     

多网格谱元法及其在高性能计算中的应用*

针对传统谱元法在每个单元内只能存在单一均匀介质,应用在复杂非均匀介质的波传播模拟中可能造成极大计算规模的问题,发展了多网格谱元法。该方法在谱元法单元内引入独立的辅助网格,用于精细描述单元内的介质和外力分布变化,在较稀疏的主网格上进行波场的求解。基于声波和弹性波方程推导了多网格谱元法公式,并对几种典型模型进行了波场的数值模拟。与传统谱元法的对比结果表明,此算法在复杂非均匀介质的弹性波传播模拟中可以利用较少的网格点数达到不低于传统算法的精度。此外,实现了并行化的多网格谱元法,获得了较好的并行效率。     

基于网格片的氧碘化学激光器多块并行数值模拟

为实现氧碘化学激光器(COIL)喷管流场大规模数值模拟,采用VICON程序中的基本方程与数值算法,应用JASMIN框架中的基本数据结构——网格片,以及多块结构网格拼接并行算法,发展了三维多块COIL并行模拟程序。数值实验结果展示了该并行模拟程序的正确性及可扩展性,并在2048个处理器核上模拟450万网格单元算例,加速比超过420。     

600 MW汽轮机高压联合进汽阀内部流场的数值模拟

本文采用非结构网格对某600MW汽轮机高压联合进汽阀不同阀门升程下内部复杂流场进行了数值模拟,详细分析了其流场结构.数值计算结果与模型实验进行了对比,并对实验中的一些现象给出了合理的解释.     

方柱绕流大涡模拟

采用有限体/有限元混合格式、非结构网格和大涡模拟方法求解可压缩的N-S方程,对Re=22 000的方柱绕流进行数值模拟,并对不同的边界条件进行详细的分析比较.通过对以往研究经验的总结和利用精细的边界条件,使得采用二阶精度的数值格式和较稀疏的网格仍然得到了令人满意的计算结果,甚至优于以往采用密网格的模拟结果.     

Lattice Boltzmann computational fluid dynamics in three dimensions

The recent development of the lattice gas method and its extension to the lattice Boltzmann method have provided new computational schemes for fluid dynamics. Both methods are fully paralleled and can easily model many different physical problems, including flows with complicated boundary conditions. In this paper, basic principles of a lattice Boltzmann computational method are described and applied to several three-dimensional benchmark problems. In most previous lattice gas and lattice Boltzmann methods, a face-centered-hyper-cubic lattice in four-dimensional space was used to obtain an isotropic stress tensor. To conserve computer memory, we develop a model which requires 14 moving directions instead of the usual 24 directions. Lattice Boltzmann models, describing two-phase fluid flows and magnetohydrodynamics, can be developed based on this simpler 14-directional lattice. Comparisons between three-dimensional spectral code results and results using our method are given for simple periodic geometries. An important property of the lattice Boltzmann method is that simulations for flow in simple and complex geometries have the same speed and efficiency, while all other methods, including the spectral method, are unable to model complicated geometries efficiently.     

A High-Order Discontinuous Galerkin Method for the Two-Dimensional Time-Domain Maxwell's Equations on Curved Mesh

In this paper, a DG (Discontinuous Galerkin) method which has been widely employed in CFD (Computational Fluid Dynamics) is used to solve the two-dimensional time-domain Maxwell's equations for complex geometries on unstructured mesh. The element interfaces on solid boundary are treated in both curved way and straight way. Numerical tests are performed for both benchmark problems and complex cases with varying orders on a series of grids, where the high-order convergence in accuracy can be observed. Both the curved and the straight solid boundary implementation can give accurate RCS (Radar Cross-Section) results with sufficiently small mesh size, but the curved solid boundary implementation can significantly improve the accuracy when using relatively large mesh size. More importantly, this CFD-based high-order DG method for the Maxwell's equations is very suitable for complex geometries.     

High-order compact schemes for incompressible flows: A simple and efficient method with quasi-spectral accuracy

In this paper, a finite difference code for Direct and Large Eddy Simulation (DNS/LES) of incompressible flows is presented. This code is an intermediate tool between fully spectral Navier–Stokes solvers (limited to academic geometry through Fourier or Chebyshev representation) and more versatile codes based on standard numerical schemes (typically only second-order accurate). The interest of high-order schemes is discussed in terms of implementation easiness, computational efficiency and accuracy improvement considered through simplified benchmark problems and practical calculations. The equivalence rules between operations in physical and spectral spaces are efficiently used to solve the Poisson equation introduced by the projection method. It is shown that for the pressure treatment, an accurate Fourier representation can be used for more flexible boundary conditions than periodicity or free-slip. Using the concept of the modified wave number, the incompressibility can be enforced up to the machine accuracy. The benefit offered by this alternative method is found to be very satisfactory, even when a formal second-order error is introduced locally by boundary conditions that are neither periodic nor symmetric. The usefulness of high-order schemes combined with an immersed boundary method (IBM) is also demonstrated despite the second-order accuracy introduced by this wall modelling strategy. In particular, the interest of a partially staggered mesh is exhibited in this specific context. Three-dimensional calculations of transitional and turbulent channel flows emphasize the ability of present high-order schemes to reduce the computational cost for a given accuracy. The main conclusion of this paper is that finite difference schemes with quasi-spectral accuracy can be very efficient for DNS/LES of incompressible flows, while allowing flexibility for the boundary conditions and easiness in the code development. Therefore, this compromise fits particularly well for very high-resolution simulations of turbulent flows with relatively complex geometries without requiring heavy numerical developments.     

KIVA程序网格生成算法的改进及雷诺应力模型计算

针对KIVA程序中网格生成不能适用于复杂区域的缺点,对其提出了一种改进算法,该算法采用Laplace椭圆方程,生成计算区域的网格,具有收敛快、适用于复杂区域的特点。该方法只要给定边界点,就可以自动生成内部网格,无需差值形成内部网格。针对几种部同的内燃机燃烧室采用雷诺应力(DSM)模型,计算了压缩、膨胀、进气和排气过程的内流场,计算结果正确给出了缸内的湍流各向异性的特点。     

Prediction of radiative heat transfer in 2D irregular geometries using the collocation spectral method based on body-fitted coordinates

Recently, an efficient numerical method, which is called the collocation spectral method (CSM), for radiative heat transfer problems, has been proposed by the present authors. In this numerical method there exists the exponential convergence rate, which can obtain a very high accuracy even using a small number of grids. In this article, the CSM based on body-fitted coordinates (BFC) is extended to simulate radiative heat transfer problems in participating medium confined in 2D complex geometries. This numerical method makes simultaneously the use of the merits of both the CSM and BFC. In this numerical approach, the radiative transfer equation (RTE) in orthogonal Cartesian coordinates should be transformed into the equation in body-fitted nonorthogonal curvilinear coordinates. In order to test the efficiency of the developed method, several 2D complex irregular enclosures with curved boundaries and containing an absorbing, emitting and scattering medium are examined. The results obtained by the CSM are assessed by comparing the predictions with those in references. These comparisons indicate that the CSM based on BFC can be recommended as a good option to solve radiative heat transfer problems in complex geometries.     

用非结构网格与欧拉方程计算复杂区域的二维流动

提出用Delaunay三角化方法生成非结构网格的一种过程。所生成的网格可用于复杂多连通域内的可压流计算。采用Euler方程和格心有限体积法,研制出程序,给出了算例。     

An algebraic multigrid solver for transonic flow problems

This article presents the latest developments of an algebraic multigrid (AMG) based on full potential equation (FPE) solver for transonic flow problems with emphasis on advanced applications. The mathematical difficulties of the problem are associated with the fact that the governing equation changes its type from elliptic (subsonic flow) to hyperbolic (supersonic flow). The flow solver is capable of dealing with flows from subsonic to transonic and supersonic conditions and is based on structured body-fitted grids approach for treating complex geometries. The computational method was demonstrated on a variety of problems to be capable of predicting the shock formation and achieving residual reduction of roughly an order of magnitude per cycle both for elliptic and hyperbolic problems, through the entire range of flow regimes, independent of the problem size (resolution).     

Proper orthogonal decomposition of atomistic flow simulations

An adaptive proper orthogonal decomposition based on time windows (WPOD) for analysis of velocity fields from atomistic simulations is presented. The method effectively separates the field into ensemble average and fluctuation components, and can be applied to both stationary and non-stationary flows in simple and complex geometries. The criteria to distinguish POD modes representing ensemble average and fluctuation components are based on adaptive examination of eigenvalue decomposition, specifically on the rate of decay of POD eigenvalues and analysis of POD eigenvectors. The WPOD method is efficient and its superior accuracy leads to smooth field gradients that can be used effectively in multiscale formulations. The method has been applied in molecular dynamics and dissipative particle dynamics simulations. We demonstrate the method for several cases including steady and unsteady flow and red blood cell flow, but the same approach can be used in atomistic simulations of materials.     

Application of the high-resolution Godunov method to the multi-fluid flow calculations

In this paper, we have numerically solved the multi-fluid problems using an operator-split two-step high-resolution Godunov PPM (parabolic piecewise method) for the flow in complex geometries. By using the front capturing method,the PPM integrator captures the interface in the solution process. The basic multi-fluid integrator is coupled to a Cartesian grid algorithm where a VOF (volume of fluid) representation of the fluid interface is also used. As an application of this method, we test the 2D interracial advection example and simulate an experimental hypervelocity launcher model from Sandia National Laboratories. The computational design of the hypervelocity launcher is also given in the paper.     

Lattice Boltzmann method on quadtree grids for simulating fluid flow through porous media: A new automatic algorithm

During the past two decades, the lattice Boltzmann (LB) method has been introduced as a class of computational fluid dynamic methods for fluid flow simulations. In this method, instead of solving the Navier Stocks equation, the Boltzmann equation is solved to simulate the flow of a fluid. This method was originally developed based on uniform grids. However, in order to model complex geometries such as porous media, it can be very slow in comparison with other techniques such as finite differences and finite elements. To eliminate this limitation, a number of studies have aimed to formulate the lattice Boltzmann on the unstructured grids. This paper deals with simulating fluid flow through a synthetic porous medium using the LB method and on the quadtree grid structure. To this end, the LB method was used on nonuniform grids coupled with a technique for image reconstruction which resulted in the quadtree grids for simulation of fluid flow through porous media. Accuracy and efficiency of this algorithm is compared against the conventional LB method based on uniform grids. While the decrease in computational time in the proposed LB method on nonuniform grids is found to be significant regarding the size of the initial and reconstructed images, the same level of accuracy is obtained when compared with the conventional LB method on uniform grids.