Uniform Treatment of Numerical Time-Integrations of the Maxwell Equations R. Horváth TU/e, Philips Research Eindhoven TU/e Eindhoven, 27th June, 2002 Scientific.

Slides:



Advertisements
Similar presentations
Courant and all that Consistency, Convergence Stability Numerical Dispersion Computational grids and numerical anisotropy The goal of this lecture is to.
Advertisements

Finite Difference Discretization of Hyperbolic Equations: Linear Problems Lectures 8, 9 and 10.
4 th order Embedded Boundary FDTD algorithm for Maxwell Equations Lingling Wu, Stony Brook University Roman Samulyak, BNL Tianshi Lu, BNL Application collaborators:
Chapter 2 Solutions of Systems of Linear Equations / Matrix Inversion
Chapter 8 Elliptic Equation.
Algebraic, transcendental (i.e., involving trigonometric and exponential functions), ordinary differential equations, or partial differential equations...
Point-wise Discretization Errors in Boundary Element Method for Elasticity Problem Bart F. Zalewski Case Western Reserve University Robert L. Mullen Case.
5/4/2015rew Accuracy increase in FDTD using two sets of staggered grids E. Shcherbakov May 9, 2006.
Jan Verwer Convergence and Component Splitting for the Crank-Nicolson Leap-Frog Scheme Hairer-60 Conference, Geneva, June 2009 TexPoint fonts used in EMF.
1 L-BFGS and Delayed Dynamical Systems Approach for Unconstrained Optimization Xiaohui XIE Supervisor: Dr. Hon Wah TAM.
Computer-Aided Analysis on Energy and Thermofluid Sciences Y.C. Shih Fall 2011 Chapter 6: Basics of Finite Difference Chapter 6 Basics of Finite Difference.
LMS Durham Symposium Dynamical Systems and Statistical Mechanics 3 – 13 July 2006 Locating periodic orbits in high-dimensional systems by stabilising transformations.
1 Lecture 11: Unsteady Conduction Error Analysis.
Introduction to Numerical Methods I
Preference Analysis Joachim Giesen and Eva Schuberth May 24, 2006.
Ordinary Differential Equations (ODEs) 1Daniel Baur / Numerical Methods for Chemical Engineers / Implicit ODE Solvers Daniel Baur ETH Zurich, Institut.
Computational Fluid Dynamics I PIW Numerical Methods for Parabolic Equations Instructor: Hong G. Im University of Michigan Fall 2005.
Math 3120 Differential Equations with Boundary Value Problems
Finite Difference Time Domain Method (FDTD)
Finite Difference Methods to Solve the Wave Equation To develop the governing equation, Sum the Forces The Wave Equation Equations of Motion.
Error estimates for degenerate parabolic equation Yabin Fan CASA Seminar,
Chapter 16 Integration of Ordinary Differential Equations.
1Chapter 2. 2 Example 3Chapter 2 4 EXAMPLE 5Chapter 2.
1 Part 1: Ordinary Differential Equations Ch1: First-Order Differential Equations Ch2: Second-Order Differential Equations Ch3: The Laplace Transform Ch4:
Lecture 35 Numerical Analysis. Chapter 7 Ordinary Differential Equations.
Time integration methods for the Heat Equation Solution u for t= Tobias Köppl TUM.
Advanced Computer Graphics Spring 2014 K. H. Ko School of Mechatronics Gwangju Institute of Science and Technology.
Section 1.1 Introduction to Systems of Linear Equations.
Introduction to Numerical Methods for ODEs and PDEs Methods of Approximation Lecture 3: finite differences Lecture 4: finite elements.
Algorithms for a large sparse nonlinear eigenvalue problem Yusaku Yamamoto Dept. of Computational Science & Engineering Nagoya University.
Introduction to Numerical Analysis I MATH/CMPSC 455 PA=LU.
Lecture 3.
Lecture Objectives: Analyze the unsteady-state heat transfer Conduction Introduce numerical calculation methods Explicit – Implicit methods.
Efficient Integration of Large Stiff Systems of ODEs Using Exponential Integrators M. Tokman, M. Tokman, University of California, Merced 2 hrs 1.5 hrs.
Discontinuous Galerkin Methods Li, Yang FerienAkademie 2008.
The swiss-carpet preconditioner: a simple parallel preconditioner of Dirichlet-Neumann type A. Quarteroni (Lausanne and Milan) M. Sala (Lausanne) A. Valli.
7. Introduction to the numerical integration of PDE. As an example, we consider the following PDE with one variable; Finite difference method is one of.
Mathe III Lecture 4 Mathe III Lecture 4 Mathe III Lecture 4 Mathe III Lecture 4.
1 Complex Images k’k’ k”k” k0k0 -k0-k0 branch cut   k 0 pole C1C1 C0C0 from the Sommerfeld identity, the complex exponentials must be a function.
Rieben IMA Poster, 05/11/ UC Davis /LLNL/ ISCR High Order Symplectic Integration Methods for Finite Element Solutions to Time Dependent Maxwell Equations.
16.920J/SMA 5212 Numerical Methods for PDEs Lecture 5 Finite Differences: Parabolic Problems B. C. Khoo Thanks to Franklin Tan SMA-HPC ©2002 NUS.
AUTOMATIC CONTROL THEORY II Slovak University of Technology Faculty of Material Science and Technology in Trnava.
Numerical Analysis – Differential Equation
CIS/ME 794Y A Case Study in Computational Science & Engineering 2-D conservation of momentum (contd.) Or, in cartesian tensor notation, Where repeated.
The first pair of Maxwell’s equations Section 26.
1 Chapter 7 Numerical Methods for the Solution of Systems of Equations.
1 EEE 431 Computational Methods in Electrodynamics Lecture 7 By Dr. Rasime Uyguroglu
2.2 The Inverse of a Matrix. Example: REVIEW Invertible (Nonsingular)
 Recall that when you wanted to solve a system of equations, you used to use two different methods.  Substitution Method  Addition Method.
F. Fairag, H Tawfiq and M. Al-Shahrani Department of Math & Stat Department of Mathematics and Statistics, KFUPM. Nov 6, 2013 Preconditioning Technique.
The story so far… ATM 562 Fovell Fall, Convergence We will deploy finite difference (FD) approximations to our model partial differential equations.
1 LECTURE 6 Stability of Parabolic PDEs. 2 Aim of Lecture Last week we discussed Parabolic PDEs –Looked at Explicit and Implicit Methods –Advantages and.
1 Spring 2003 Prof. Tim Warburton MA557/MA578/CS557 Lecture 28.
First-order Differential Equations Chapter 2. Overview II. Linear equations Chapter 1 : Introduction to Differential Equations I. Separable variables.
Lecture 9 Numerical Analysis. Solution of Linear System of Equations Chapter 3.
Modern Control Systems (MCS) Dr. Imtiaz Hussain URL :
EEE 431 Computational Methods in Electrodynamics
ECE 1304 Introduction to Electrical and Computer Engineering
Introduction to Numerical Methods I
Chapter 30.
Boundary Element Analysis of Systems Using Interval Methods
First order non linear pde’s
Chapter 26.
Scientific Computing Partial Differential Equations Implicit Solution of Heat Equation.
Introduction: A review on static electric and magnetic fields
MATH 175: Numerical Analysis II
Numerical Analysis Lecture11.
Computer Animation Algorithms and Techniques
Generalized Finite Element Methods
Presentation transcript:

Uniform Treatment of Numerical Time-Integrations of the Maxwell Equations R. Horváth TU/e, Philips Research Eindhoven TU/e Eindhoven, 27th June, 2002 Scientific Computing in Electrical Engineering Eindhoven, The Netherlands June, 23rd - 28th, 2002

Outline TU/e Eindhoven, 27th June, 2002 Introduction, Yee-method, problems Operator splitting methods Splitting of the semi-discretized Maxwell equations (Yee, NZCZ, KFR) Proof of the unconditional stability of the NZCZ-method Comparison of the methods

TU/e Eindhoven, 27th June, 2002 Maxwell Equations

TU/e Eindhoven, 27th June, 2002 Second order spatial discretization using staggered grids. Space Discretization in the Yee-method

TU/e Eindhoven, 27th June, 2002 Time Integration in the Yee-method So-called leapfrog scheme is used: E 0 H 1/2 E 1 H 3/2 E 2 H 5/2 E 3 H 7/2... Stability condition:

TU/e Eindhoven, 27th June, 2002 Operator Splitting Methods I. Sequential splitting (S-splitting):

TU/e Eindhoven, 27th June, 2002 Operator Splitting Methods II. Splitting error: For the S-splitting method (1st order): Lemma: s: number of the sub-systems

TU/e Eindhoven, 27th June, 2002 Operator Splitting Methods III. S-splitting: Other splittings: Strang-splitting (second order): Fourth order splitting:

TU/e Eindhoven, 27th June, 2002 Semi-discretization of the Maxwell Equations Applying the staggered grid spatial discretization, the Maxwell equations can be written in the form: :skew-symmetric, sparse matrix, at most four elements per row. The elements have the form. : consists of the field components in the form

TU/e Eindhoven, 27th June, 2002 Splitting I. (Yee-scheme, 1966) We zero the rows of the magnetic field variables We zero the rows of the electric field variables Lemma:

TU/e Eindhoven, 27th June, D example for splitting  H z x E y H 9

TU/e Eindhoven, 27th June, D example - Yee-method

TU/e Eindhoven, 27th June, 2002 Splitting II. (KFR-scheme, 2001) The matrices are skew-symmetric matrices, where the exponentials exp(A.K ) can be computed easily using the identity: Stability: the KFR-method is unconditionally stable by construction.

TU/e Eindhoven, 27th June, D example - KFR-method

TU/e Eindhoven, 27th June, 2002 Splitting III. (NZCZ-scheme, 2000) Discretization of first terms in the curl operator Discretization of second terms in the curl operator skew-symmetric The sub-systems with and cannot be solved exactly. Numerical time integrations are needed.

TU/e Eindhoven, 27th June, 2002 Splitting III. (NZCZ-scheme, 2000) Solving the systems by the explicit, implicit, explicit and implicit Euler-methods, respectively, we obtain the iteration Theorem: the NZCZ-method is unconditionally stable. Proof: We show that if is fixed ( ), then the relation is true for all k with a constant K independent of k.

TU/e Eindhoven, 27th June, 2002 Splitting III. (NZCZ-scheme, 2000)

TU/e Eindhoven, 27th June, D example - NZCZ-method

TU/e Eindhoven, 27th June, D example - Real Schur Decomposition No splitting error!

TU/e Eindhoven, 27th June, D examples - Comparison

TU/e Eindhoven, 27th June, 2002 Comparison of the methods

TU/e Eindhoven, 27th June, 2002 Comparison of the methods

TU/e Eindhoven, 27th June, 2002 Comparison of the methods What is the reason of the difference between the Yee and the KFR- method? Compare Yee and KFR1: Yee: sufficiently accurate, strict stability condition KFR: unconditionally stable, but an accurate solution requires small time-step. In the long run it is slower than the Yee-method. NZCZ: unconditionally stable. With a suitable choice of the time- step the method is faster than the others (with acceptable error).

TU/e Eindhoven, 27th June, 2002 Thank You for the Attention