用于厚大ADI铸件的分级淬火工艺

为消除厚大球铁件等温淬火时出现的珠光体组织,试验了"预淬火+等温淬火"的分级淬火工艺,预淬火冷却介质为过饱和硝酸盐水溶液.预淬火时间为28 s,使厚壁部位在650～500℃高温区加速冷却.试验结果表明:工件的基体金相组织为针状铁素体+高碳块状奥氏体,力学性能达到了技术要求.     

ADI铁路机车制动器耐磨件的试制

用牌号为QTD 1400-1的等淬球铁(ADI)试制了东风4B内燃机车制动装置用衬套等铸件,其球化率为2~3级、石墨球大小5~7级、珠光体体积分数大于80%,等淬热处理工艺为:880~920℃×1~2 h+240~280℃×1~2 h,热处理后金相组织为针状铁素体+富C奥氏体+球状石墨,力学性能检测结果为:抗拉强度1 460~1 510 MPa,伸长率1.6%~4%,硬度46.0~48.5 HRC。通过30万km行车试验证明,ADI衬套等耐磨件可以满足内燃机车制动装置的服役使用要求,在相同工况条件下的磨耗是原来钢衬套的1/3~1/5。     

Fast alternating direction implicit method for efficient transient thermal simulation of integrated circuits

This paper presents fast alternating direction implicit (FADI) method for efficient transient thermal simulation of integrated circuits. The FADI method is formulated from Peaceman–Rachford's ADI and Douglas–Gunn's ADI methods. The update procedure of the proposed method has basic implicit form that features derivative‐free right‐hand side and hence, better efficiency and conciseness. Subsequently, through the basic implicit form of FADI method, the relationship between classical Peaceman–Rachford's and Douglas–Gunn's ADI methods can be clarified and elucidated in detail. A unified boundary condition that can cater to common kinds of boundary conditions in thermal simulation is also introduced. To further accelerate FADI method, the graphics processing unit is also utilized through Compute Unified Device Architecture implementation. It is shown that high efficiency gain can be achieved using the proposed FADI method through large time step size and data parallelism, while maintaining stability and good accuracy. As numerical illustration, an integrated circuit structure with microchannel cooling is demonstrated. Numerical results further ascertain the cooling effect of the microchannels. Copyright © 2015 John Wiley & Sons, Ltd.     

交替方向隐式时域有限差分法中的Berenger理想匹配层

提出一种新方案．用于离散Berenger场分裂形式Maxwell方程．将理想匹配层吸收边界条件由通常的时域有限差分法推广到交替方向隐式时域有限差分法．和已有的离散方案比较，采用文中提出的离散方案可使理想匹配层吸收边界的反射误差成数量级地减小．     

Pentadiagonal alternating-direction-implicit finite-difference time-domain method for two-dimensional Schrödinger equation

In this paper, we have proposed a pentadiagonal alternating-direction-implicit (Penta-ADI) finite-difference time-domain (FDTD) method for the two-dimensional Schrödinger equation. Through the separation of complex wave function into real and imaginary parts, a pentadiagonal system of equations for the ADI method is obtained, which results in our Penta-ADI method. The Penta-ADI method is further simplified into pentadiagonal fundamental ADI (Penta-FADI) method, which has matrix-operator-free right-hand-sides (RHS), leading to the simplest and most concise update equations. As the Penta-FADI method involves five stencils in the left-hand-sides (LHS) of the pentadiagonal update equations, special treatments that are required for the implementation of the Dirichlet’s boundary conditions will be discussed. Using the Penta-FADI method, a significantly higher efficiency gain can be achieved over the conventional Tri-ADI method, which involves a tridiagonal system of equations.     

预条件的平方Smith法求解大型Sylvester矩阵方程

提出了一种预条件的平方Smith算法求解大型连续Sylvester矩阵方程,该算法利用交替方向隐式迭代(ADI)来构造预条件算子,将原方程转换为非对称Stein方程,并在Krylov子空间中应用平方Smith法迭代产生低秩逼近解。数值实验表明,与已知的Jacobi迭代法等算法相比,该算法有更好的迭代效率和收敛精度。     

Implicit methods for fluid dynamics

This paper represents the contributions to the development of implicit procedures for solving the equations of fluid dynamics made by Briley and McDonald (1975)[1], Beam and Warming (1976, 1978) [2] and [3] and Lombard et al. (1983) [4]. The contributions of Briley and McDonald and Beam and Warming are well known, but Lombard has not been fully recognized for his innovative contributions to flux vector splitting and use of the DDADI (Diagonally Dominate Alternating Direction Implicit) algorithm. Their contributions are presented herein.Fully implicit algorithms are applied to two complex flow problems of current interest, (1) hypersonic non-equilibrium flow about a blunt nosed body and (2) flow within an MFD (magneto-fluid dynamics) accelerator. These two applications would be exceeding costly without the use of fully implicit methods.     

自适应SW-ADI方法解反应扩散方程的并行实现

根据解反应扩散方程的自适应样条小波-交替方向(SW-ADI)方法,使用MPI、OpenMP两种并行编程模式,对串行程序进行了直接并行化,并在上海大学的高性能计算机自强2000上分别用MPI和OpenMP实现了对方程的求解。对运算结果进行了分析并给出了与串行程序相比较的并行加速比。     

Modified alternating direction-implicit iteration method for linear systems from the incompressible Navier–Stokes equations

In order to solve the large sparse systems of linear equations arising from numerical solutions of two-dimensional steady incompressible viscous flow problems in primitive variable formulation, Ran and Yuan [On modified block SSOR iteration methods for linear systems from steady incompressible viscous flow problems, Appl. Math. Comput. 217 (2010), pp. 3050–3068] presented the block symmetric successive over-relaxation (BSSOR) and the modified BSSOR iteration methods based on the special structures of the coefficient matrices. In this study, we present the modified alternating direction-implicit (MADI) iteration method for solving the linear systems. Under suitable conditions, we establish convergence theorems for the MADI iteration method. In addition, the optimal parameter involved in the MADI iteration method is estimated in detail. Numerical experiments show that the MADI iteration method is a feasible and effective iterative solver.     

A parallel and scalable algorithm for ADI method with pre-propagation and message vectorization

It is hard to implement the ADI method in an efficient way on distributed-memory parallel computers. We propose “P-scheme” which parallelizes a tridiagonal linear system of equations for the ADI method, but its effectiveness is limited to the cases where the problem size is large enough mainly because of the communication cost of the propagation phase of the scheme.

In order to overcome this difficulty, we propose an improved version of the P-scheme with “message vectorization” which aggregates several communication messages into one and alleviates the communication cost. Also we evaluate the effectiveness of message vectorization for the ADI method and show that the improved version of the P-scheme works well even for smaller problems and linear and super-linear speedups can be achieved for 8194 × 8194 and 16,386 × 16,386 problems, respectively.