• 제목/요약/키워드: discretized iterations

검색결과 4건 처리시간 0.015초

A MESH INDEPENDENCE PRINCIPLE FOR PERTURBED NEWTON-LIKE METHODS AND THEIR DISCRETIZATIONS

  • Argyros, Ioannis K.
    • Journal of applied mathematics & informatics
    • /
    • 제7권1호
    • /
    • pp.139-159
    • /
    • 2000
  • In this manuscript we study perturbed Newton-like methods for the solution of nonlinear operator equations in a Banach space and their discretized versions in connection with the mesh independence principle. This principle asserts that the behavior of the discretized process is asymptotically the same as that for the original iteration and consequently, the number of steps required by the two processes to converge to within a given tolerance is essentially the same. So far this result has been proved by others using Newton's method for certain classes of boundary value problems and even more generally by considering a Lipschitz uniform discretization. In some of our earlierpapers we extend these results to include Newton-like methods under more general conditions. However, all previous results assume that the iterates can be computed exactly. This is mot true in general. That in why we use perturbed Newton-like methods and even more general conditions. Our results, on the one hand, extend, and on the other hand, make more practical and applicable all previous results.

3차원 정상상태의 드리프트-확산 방정식의 해석 프로그램 개발 (A development of the 3-dimensional stationary drift-diffusion equation solver)

  • 윤현민;김태한;김대영;김철성
    • 전자공학회논문지D
    • /
    • 제34D권8호
    • /
    • pp.41-51
    • /
    • 1997
  • The device simulator (BANDIS) which can analyze efficiently the electrical characteristics of the semiconductor devices under the three dimensional stationary conditions on the IBM PC was developed. Poisson, electon and hole continuity equations are discretized y te galerkin method using a tetrahedron as af finite element. The frontal solver which has exquisite data structures and advanced input/output functions is dused for the matrix solver which needs the highest cost in the three dimensional device simulation. The discretization method of the continuity equations used in BANDIS are compared with that of the scharfetter-gummel method used in the commercial three-dimensional device. To verify an accuracy and the efficiency of the discretization method, the simulation results of the PN junction diode and the BJT from BANDIS are compared with those of the commercial three-dimensiional device simulator such as DAVINCI. The maximum relative error within 2% and the average number of iterations needed for the convergence is decreased by more than 20%. The total simulation time of the BJT with 25542 nodes is decreased to about 60% compared with that of DAVINCI.

  • PDF

Numerical Solution of Nonlinear Diffusion in One Dimensional Porous Medium Using Hybrid SOR Method

  • Jackel Vui Lung, Chew;Elayaraja, Aruchunan;Andang, Sunarto;Jumat, Sulaiman
    • Kyungpook Mathematical Journal
    • /
    • 제62권4호
    • /
    • pp.699-713
    • /
    • 2022
  • This paper proposes a hybrid successive over-relaxation iterative method for the numerical solution of a nonlinear diffusion in a one-dimensional porous medium. The considered mathematical model is discretized using a computational complexity reduction scheme called half-sweep finite differences. The local truncation error and the analysis of the stability of the scheme are discussed. The proposed iterative method, which uses explicit group technique and modified successive over-relaxation, is formulated systematically. This method improves the efficiency of obtaining the solution in terms of total iterations and program elapsed time. The accuracy of the proposed method, which is measured using the magnitude of absolute errors, is promising. Numerical convergence tests of the proposed method are also provided. Some numerical experiments are delivered using initial-boundary value problems to show the superiority of the proposed method against some existing numerical methods.

Iterative coupling of precise integration FEM and TD-BEM for elastodynamic analysis

  • Lei, Weidong;Liu, Chun;Qin, Xiaofei;Chen, Rui
    • Structural Engineering and Mechanics
    • /
    • 제67권4호
    • /
    • pp.317-326
    • /
    • 2018
  • The iterative decomposition coupling formulation of the precise integration finite element method (FEM) and the time domain boundary element method (TD-BEM) is presented for elstodynamic problems. In the formulation, the FEM node and the BEM node are not required to be coincident on the common interface between FEM and BEM sub-domains, therefore, the FEM and BEM are independently discretized. The force and displacement converting matrices are used to transfer data between FEM and BEM nodes on the common interface between the FEM and BEM sub-domains, to renew the nodal variables in the process of the iterations for the un-coincident FEM node and BEM node. The iterative coupling formulation for elastodynamics in current paper is of high modeling accuracy, due to the semi-analytical solution incorporated in the precise integration finite element method. The decomposition coupling formulation for elastodynamics is verified by examples of a cantilever bar under a Heaviside-type force and a harmonic load.