Structural dynamic analysis on a parallel computer: The finite element machine

The development of general-purpose finite element computer software systems has provided the capability to analyze a wide range of linear and non-linear structural problems. However, these software systems are severely limited for non-linear response calculations because of the available speed on current sequential computers. Recent and projected advances in parallel multiple instruction multiple data (MIMD) computers provide an opportunity for significant gains in computing speed and for broadening the range of structural problems which may be solved. The key to these gains is the effective selection and implementation of algorithms which exploit parallel computing. This paper documents experiences solving transient response calculations on an experimental MIMD computer, termed the Finite Element Machine. The paper describes the algorithm used, its implementation for parallel computations, and results for representative one- and two-dimensional dynamic response test problems. The results show computation speedups of up to 7.83 for eight processors, and indicate that significant speedups of solution time are possible for non-linear dynamic response calculations through the use of many processors and appropriate parallel integration algorithms. The results are extremely encouraging and suggest that significant speedups in structural computations can be achieved through advances in parallel computers.     

稀疏有限元线性系统的并行算法实现

在对称多处理机系统上,提出了一种求解稀疏对称有限元线性系统的正规化精确并行逆算法。该算法以一种避免数据依赖的反对角运动方法为基础,使用OpenMP编译指导来实现。诸如加速比和效率等数值实验结果的推出,说明在一个对称多处理机系统上,所提出的算法求解方法能更好地提高性能,获得更大的加速。     

Hybrid spectral-element-low-order methods for incompressible flows

In this article we present a new formulation for coupling spectral element discretizations to finite difference and finite element discretizations addressing flow problems in very complicated geometries. A general iterative relaxation procedure (Zanolli patching) is employed that enforcesC ¹ continuity along the patching interface between the two differently discretized subdomains. In fluid flow simulations of transitional and turbulent flows the high-order discretization (spectral element) is used in the outer part of the domain where the Reynolds number is effectively very high. Near rough wall boundaries (where the flow is effectively very viscous) the use of low-order discretizations provides sufficient accuracy and allows for efficient treatment of the complex geometry. An analysis of the patching procedure is presented for elliptic problems, and extensions to incompressible Navier-Stokes equations are implemented using an efficient high-order splitting scheme. Several examples are given for elliptic and flow model problems and performance is measured on both serial and parallel processors.     

A Cascadic Multigrid Algorithm for Semilinear Indefinite Elliptic Problems

We propose a cascadic multigrid algorithm for a semilinear indefinite elliptic problem. We use a standard finite element discretization with piecewise linear finite elements. The arising nonlinear equations are solved by a cascadic organization of Newton's method with frozen derivative on a sequence of nested grids. This gives a simple version of a multigrid method without projections on coarser grids. The cascadic multigrid algorithm starts on a comparatively coarse grid where the number of unknowns is small enough to obtain an approximate solution within sufficiently high precision without substantial computational effort. On each finer grid we perform exactly one Newton step taking the approximate solution from the coarsest grid as initial guess. The linear Newton systems are solved iteratively by a Jacobi-type iteration with special parameters using the approximate solution from the previous grid as initial guess. We prove that for a sufficiently fine initial grid and for a sufficiently good start approximation the algorithm yields an approximate solution within the discretization error on the finest grid and that the method has multigrid complexity with logarithmic multiplier. Received February 1999, revised July 13, 1999     

A high-order Lagrangian-decoupling method for the incompressible Navier-Stokes equations

In this paper we present a high-order Lagrangian-decoupling method for the unsteady convection diffusion and incompressible Navier-Stokes equations. The method is based upon Lagrangian variational forms that reduce the convection-diffusion equation to a symmetric initial value problem, implicit high-order backward-differentiation finite difference schemes for integration along characteristics, finite element or spectral element spatial discretizations and mesh-invariance procedures and high-order explicit time-stepping schemes for deducing function values at convected space-time points. The method improves upon previous finite element characteristic methods through the systematic and efficient extension to high-order accuracy and the introduction of a simple structure-preserving characteristic-foot calculation procedure which is readily implemented on modern architectures. The new method is significantly more efficient than explicit-convection schemes for the Navier-Stokes equations due to the decoupling of the convection and Stokes operators and the attendant increase in temporal stability. Numerous numerical examples are given for the convection-diffusion and Navier-Stokes equations for the particular case of a spectral element spatial discretization.     

A comparison of three kinds of local and parallel finite element algorithms based on two-grid discretizations for the stationary Navier–Stokes equations

Based on two-grid discretizations, three kinds of local and parallel finite element algorithms for the stationary Navier–Stokes equations are introduced and discussed. The main technique is first to use a standard finite element discretization on a coarse grid to approximate low frequencies of the solution, then to apply some linearized discretizations on a fine grid to correct the resulted residual (which contains mostly high frequencies) by some local and parallel procedures. Three approaches to linearization are discussed. Under the uniqueness condition, error estimates of the finite element solution are derived. Numerical results show that among the three kinds of parallel algorithms, the Oseen-linearized algorithm is preferable if we both consider the computational time and the accuracy of the approximate solution.     

A parallel finite element solution method

New parallel computer architectures have revolutionized the design of computer algorithms, and promise to have significant influence on algorithms for structural engineering computations. In this paper, a parallel finite element solution method is presented. The solution method proposed does not require the formation of global system equations, but computes directly the element distortions, as opposed to solving a system of nodal equations. An element or substructure is mapped on to a processor of an MIMD multiprocessing system. Each processor stores only the information relevant to the element or substructure for which the processor represents. The finite element computations can be performed in parallel, in that a processor generates the local stiffness, computes the element distortions and determines the stress-strain characteristics for the element or substructure associated with the processor.     

Parallel analysis of rotationally periodic structures

A parallel finite element solution algorithm for analysing large rotationally periodic structures on MIMD parallel computer systems is described. For a rotationally periodic structure, the global stiffness matrix under the corresponding symmetric coordinate system is periodic, i.e. possesses isomorphic properties, so that the global equation system can be transformed into a number of smaller equation systems which are fully decoupled. These decoupled equation systems then can be solved simultaneously on a multiprocessor parallel computer. The algorithm also generates the decoupled equation systems in parallel, without explicitly assembling the global stiffness matrix of the structure. A prototype implementation of the algorithm on an array of transputers is presented, and the efficiency of the program is also studied in this paper. Finally, a numerical example is given to demonstrate the speedup of the program.     

An incomplete nested dissection algorithm for parallel direct solution of finite element discretizations of partial differential equations

A multilevel algorithm is presented for direct, parallel factorization of the large sparse matrices that arise from finite element and spectral element discretization of elliptic partial differential equations. Incomplete nested dissection and domain decomposition are used to distribute the domain among the processors and to organize the matrix into sections in which pivoting is applied to stabilize the factorization of indefinite equation sets. The algorithm is highly parallel and memory efficient; the efficient use of sparsity in the matrix allows the solution of larger problems as the number of processors is increased, and minimizes computations as well as the number and volume of communications among the processors. The number of messages and the total volume of messages passed during factorization, which are used as measures of algorithm efficiency, are reduced significantly compared to other algorithms. Factorization times are low and speedups high for implementation on an Intel iPSC/860 hypercube computer. Furthermore, the timings for forward and back substitutions are more than an order-of-magnitude smaller than the matrix decomposition times.     

Newton Iterative Parallel Finite Element Algorithm for the Steady Navier-Stokes Equations

A combination method of the Newton iteration and parallel finite element algorithm is applied for solving the steady Navier-Stokes equations under the strong uniqueness condition. This algorithm is motivated by applying the Newton iterations of m times for a nonlinear problem on a coarse grid in domain Ω and computing a linear problem on a fine grid in some subdomains Ω _j ⊂Ω with j=1,…,M in a parallel environment. Then, the error estimation of the Newton iterative parallel finite element solution to the solution of the steady Navier-Stokes equations is analyzed for the large m and small H and h≪H. Finally, some numerical tests are made to demonstrate the the effectiveness of this algorithm.     

The scattered decomposition for finite elements

The solution of finite element problems with irregular geometries on a parallel computer of the hypercube type (MIMD, distributed memory) is considered. The technique of scattering the decomposition is found to be easy to implement and to effectively load balance the computation.     

A parallel monotone iterative method for the numerical solution of multi-dimensional semiconductor Poisson equation

Various self-consistent semiconductor device simulation approaches require the solution of Poisson equation that describes the potential distribution for a specified doping profile (or charge density). In this paper, we solve the multi-dimensional semiconductor nonlinear Poisson equation numerically with the finite volume method and the monotone iterative method on a Linux-cluster. Based on the nonlinear property of the Poisson equation, the proposed method converges monotonically for arbitrary initial guesses. Compared with the Newton's iterative method, it is easy implementing, relatively robust and fast with much less computation time, and its algorithm is inherently parallel in large-scale computing. The presented method has been successfully implemented; the developed parallel nonlinear Poisson solver tested on a variety of devices shows it has good efficiency and robustness. Benchmarks are also included to demonstrate the excellent parallel performance of the method.     

A multilevel hybrid Newton–Krylov–Schwarz method for the Bidomain model of electrocardiology

A multilevel hybrid Newton–Krylov–Schwarz (NKS) method is constructed and studied numerically for implicit time discretizations of the Bidomain reaction–diffusion system in three dimensions. This model describes the bioelectrical activity of the heart by coupling two degenerate parabolic equations with a stiff system of ordinary differential equations. The NKS Bidomain solver employs an outer inexact Newton iteration to solve the nonlinear finite element system originating at each time step of the implicit discretization. The Jacobian update during the Newton iteration is solved by a Krylov method employing a multilevel hybrid overlapping Schwarz preconditioner, additive within the levels and multiplicative among the levels. Several parallel tests on Linux clusters are performed, showing that the convergence of the method is independent of the number of subdomains (scalability), the discretization parameters and the number of levels (optimality).     

OpenMP based parallel normalized direct methods for sparse finite element linear systems

A new parallel normalized exact inverse algorithm is presented for solving sparse symmetric finite element linear systems on symmetric multiprocessor systems (SMP), based upon an antidiagonal motion approach (“wave”-like pattern) for overcoming the data dependencies. The proposed algorithm was implemented using OpenMP directives. Numerical results, such as speedups and efficiency, are presented illustrating the efficient performance on a symmetric multiprocessor computer system, where the proposed algorithmic solution method achieves good speedups.

Implementation of an element-by-element solution algorithm for the finite element method on a coarse-grained parallel computer

An element-by-element solution algorithm for systems of equations arising in applying the finite element method in solid mechanics was implemented on the loosely coupled array of processors (lCAP) parallel computer located at IBM Kingston. The element-by-element algorithm has previously been shown to be advantageous over direct solution algorithms for large problems on sequential computers. It also has the advantage that it can be implemented in parallel on machines such as the lCAP in a relatively straightforward manner. The results show that solution speedup efficiencies of approximately 95% can be readily achieved with this method, with no indication that the speed-up efficiency drops off as more processors are added. The implementation used is applicable to other coarse-grained parallel architectures in addition to the lCAP computer.     

Parallel processing, neural networks and genetic algorithms

In an earlier paper[1] some recent developments in computational technology to structural engineering were described. The developments included: parallel and distributed computing; neural networks; and genetic algorithms. In this paper, the authors concentrate on parallel implementations of neural networks and genetic algorithms. In the final section of the paper the authors show how a parallel finite element analysis may be undertaken in an efficient manner by preprocessing of the finite element model using a genetic algorithm utilizing a neural network predictor. This preprocessing is the partitioning of the finite element mesh into sub-domains to ensure load balancing and minimum interprocessor communication during the parallel finite element analysis on a MIMD distributed memory computer. © 1998 Published by Elsevier Science Limited. All rights reserved.     

Multipoint flux mixed finite element methods for slightly compressible flow in porous media

In this paper, we consider multipoint flux mixed finite element discretizations for slightly compressible Darcy flow in porous media. The methods are formulated on general meshes composed of triangles, quadrilaterals, tetrahedra or hexahedra. An inexact Newton method that allows for local velocity elimination is proposed for the solution of the nonlinear fully discrete scheme. We derive optimal error estimates for both the scalar and vector unknowns in the semidiscrete formulation. Numerical examples illustrate the convergence behavior of the methods, and their performance on test problems including permeability coefficients with increasing heterogeneity.     

A communications system for irregular local interaction problems on a concurrent computer

The solution of a large class of problems requires the repeated evaluation of matrix vector products: y = Ax. An appropriate data decomposition and communications system to exchange x components among processors is necessary for efficient evaluation of these vector products on an MIMD concurrent computer. A communications system is presented for the case of a sparse matrix A that arises from a finite element or finite difference discretization of a partial differential equation on an irregular region, or from some kind of finite range interaction between particles. The method presented here uses a domain decomposition of the physical space to distribute A and x among processors. A packed form of the matrix is used which turns out to be very convenient to set up the data structures necessary to send and receive the extra x components. The resulting communications scheme has been used in a multigrid solver for finite element static elasticity problems and in a program which solves an eigenvalue problem. Speed up factors were determined on a 32 processor Caltech/JPL Mark II hypercube with good results. The communications system is not hypercube specific and can easily be implemented on other types of MIMD parallel computers.     

A MIMD implementation of a parallel Euler solver for unstructured grids

A mesh-vertex finite volume scheme for solving the Euler equations on triangular unstructured meshes is implemented on a MIMD (multiple instruction/multiple data stream) parallel computer. Three partitioning strategies for distributing the work load onto the processors are discussed. Issues pertaining to the communication costs are also addressed. We find that the spectral bisection strategy yields the best performance. The performance of this unstructured computation on the Intel iPSC/860 compares very favorably with that on a one-processor CRAY Y-MP/1 and an earlier implementation on the Connection Machine.The authors are employees of Computer Sciences Corporation. This work was funded under contract NAS 2-12961     

Parallel EFG algorithm for heat transfer problems

At present, meshless element free Galerkin (EFG) method is being successfully applied in the areas such as solid mechanics, fracture mechanics and thermal. Being a meshless method, it has many advantages over finite element method. One big hurdle with the wide implementation of this method is its computational cost. Therefore, in this paper, a parallel algorithm is proposed for the EFG method. The parallel code has been written in FORTRAN language using MPI message passing library and executed on a four node (eight processors) MIMD type, distributed memory ‘PARAM 10000’ parallel computer. The total time, communication time, speedup and efficiency have been estimated for a three-dimensional heat transfer problem to validate the proposed algorithm. For eight processors, the speedup and efficiency are obtained to be 4.66 and 58.22%, respectively, for a data size of 1320 nodes.